Terry Fisher

First some background (from Chapter 4 of my book): �Roughly speaking, price discrimination refers to the practice of charging different consumers different prices for access to the same good or service. Somewhat more precisely, it means charging different consumers different prices when the variation cannot be explained by differences in the costs of the versions of the good or service that is supplied to them. A traditional example of price discrimination is the practice of airlines to offer discounts to customers who �stay over� at their destinations for a Saturday night. The cost to the airline of supplying the service in question–the various expenditures necessary to carry a person in a plane from, say, Boston to San Francisco and back–is the same regardless of whether the passenger stays in San Francisco for a day or a week. The only reason for varying the price is to extract higher fares from people who are able and willing to pay more (in this case, business travelers, most of whom have expense accounts), without pricing out of the market people who are able and willing to pay less (in this case, tourists).

�Price discrimination increases a seller�s profits. Why, then, don�t all sellers engage in it? The main reason is that, with rare exceptions, price discrimination is only feasible when the seller both has some degree of market power and is able to discourage �arbitrage.� Market power exists when there are no readily available, equally satisfactory substitutes for the good or service that the seller is offering. Arbitrage occurs when a customer who purchases a good or service at a low price is able to resell it (typically at a profit) to someone who otherwise would be obliged to pay a high price for it�.�

Record companies, film studios, and pharmaceutical companies all enjoy some degree of market power. Their ability to engage in price discrimination in the past has been limited primarily by the difficulty they have had in curtailing arbitrage. (The �first sale� doctrine in copyright law affirmatively protects arbitrage, and drug companies have had trouble preventing purchasers of drugs at low prices from reselling them to others at higher prices.) This has not prevented the companies from engaging in price discrimination altogether, but has forced them to engage in relatively crude versions. The best known is the so-called �windowing� marketing system in the film industry, in which the studios divide up their markets temporally — first milking the market of especially eager or wealthy viewers by charging high license fees to first-run theatres (which, in turn, charge substantial fees for admission), then gradually releasing the film at diminishing prices in a series of secondary markets. (More detail on this practice can be found in Chapter 2 of my book.) The pharmaceutical companies, by contract, have practiced geographic price discrimination. Except when prevented by �exhaustion� doctrines (e.g., within the EU), they price their products differently in different countries, depending on the residents� ability and willingness to pay.

In the near future, I suggest, we are likely to see much more precise and ubiquitous forms of price discrimination � certainly in the entertainment industry and perhaps in the drug industry. With respect to the former, the main reason is that (assuming that an ACS is not adopted in the near term), the record companies and film studios are likely to intensify their efforts to develop digital rights management systems (perhaps with support from governments, in the form of enhanced versions of the Digital Millennium Copyright Act and its ilk). DRM systems are designed, among other purposes, to frustrate arbitrage � i.e., to prevent the first purchaser or licensee of a recording from transferring it to others. The net result: the record companies and films studios, if their DRM systems work, will begin to divide the universe of people who wish access to their works into much thinner slices and to set different prices for each subgroup. In the old days, everyone paid the same price for a CD. Soon, the price you are charged for access to a recording will vary with such things as your age, your zip code (a proxy for your income), what sort of device you will play it on, etc. Whether we will see a similar trend in the drug industry is less certain, but the overall trend toward the curtailment of international exhaustion rules, combined with intensified pressure on drug companies to lower their prices in poor regions, may enable and induce the pharmaceutical houses to further refine their price discrimination schemes.

Should we applaud or lament these trends? My contention is that there�s no blanket answer to that question � that price discrimination has both good and bad effects, and that its overall social desirability varies by context.

Its benefits include:
(a) It enables the entertainment and drug companies to make more money. If (a big �if�) we believe that augmentation of their profits will increase their incentives to engage in socially beneficial innovation, that�s good.
(b) It increases the ability of the poor to gain access to copyrighted or patented material. When the material in question consists of a life-saving drug, that means that price discrimination saves lives.
(c) Usually (not invariably), it redistributes wealth from the rich to the poor � and thus functions like a progressive income tax.

Its disadvantages include:
(d) The time and effort necessary to devise and implement price-discrimination schemes represent a social waste. Some forms of the practice (most importantly, so-called �second degree� price discrimination � in which the seller of a product induces potential buyers to reveal their resources or preferences by offering them different versions [such as business-class and coach-class plane tickets]) � are especially wasteful.
(e) In the view of at least some people, it is immoral. �Charging whatever each submarket market will bear,� �gouging,� seems improper. All people ought to have equal access (i.e., access on the same terms and at the same price) to a particular good or service.

My own view is that, with respect to drugs, argument (d) and (e) are trumped by the compelling interest in getting life-saving drugs into more hands (argument (b)). With respect to recorded entertainment, where increasing access to the products as issue is nowhere near so important, the right way to weigh these competing considerations is much less clear. Moreover, in context of entertainment, price discrimination has some additional troubling features, summarized in the following excerpt from Chapter 4:

�Sadly, [the] benefits [of enhanced opportunities to engage in price discrimination] would pale in the face of three other pernicious effects. The first is cultural. As Wendy Gordon has suggested, [a world characterized by pervasive price discrimination] would have a distressingly granular feel. Each time you listened to a song or watched a movie, you would know that a tiny meter, attached to your credit card or bank account, was whirring. Each time you bought a lobster in an upscale grocery store, you would know that your �consumer profile� was being adjusted–and that, next week, the prices you paid for all entertainment products would be one notch higher. Awareness of these effects would likely make you more calculating. Also, perhaps, less altruistic. Conscious that you are paying for each bit of entertainment you consume, you might be less inclined to �give back� freely to the culture the fruits of your own imagination. To most people, such a world seems unattractive.

�Second, creative and critical uses of entertainment products would likely suffer in this new environment�. [S]uppliers would most likely charge higher prices to people who wished to modify their products or incorporate them in other works. Cover artists, rap artists (who rely on digital sampling), and ordinary consumers who just like to �play around� with recordings, would all pay substantially higher fees. This price rise, plus inevitable imperfections in the rate-setting mechanisms, would sometimes place access to digital products beyond the financial means of such second-tier creators. Society at large, consequently, would lose the benefit of their derivative products, and opportunities for semiotic democracy would be curtailed�.

�Finally, when added to the many other opportunities to make money considered in this chapter, the ability to engage in precise price discrimination would radically increase the incomes of the creators of music and movies. The result would be a massive transfer of wealth from consumers to producers–much more than necessary to offset the losses that the producers are currently suffering or are likely to suffer in the future. That effect seems troubling in its own right. In addition, it would exacerbate the problem discussed at the end of Chapter 2 — in which the large and highly visible incomes of �star� performers draw inefficiently large numbers of aspirants into the music and movie businesses. If that problem is bad now, it would become worse in the altered legal and business environment.�

Bottom line: we should adjust legal rules to assist drug companies to engage in price discrimination, but, other things being equal, we should legal reforms that afford the owners of copyrights in songs and films similar opportunities.

Reactions?

Terry Fisher

Background. To understand the various reform proposals, it is crucial to keep in mind both the strengths and the weaknesses of the patent system when applied to drugs. Here are the key points: The argument that patent protection is essential to stimulate research and development, although quite shaky with respect to many fields of technology, is very strong with respect to pharmaceutical products. Why? Because (a) developing and securing approval for new drugs is extremely expensive and risky; (b) once produced, drugs can be reverse-engineered easily; and (c) the marginal cost of manufacturing them is very low. This combination of circumstances means that, if the companies that developed new drugs were not protected by the patent system against competition from generic manufacturers, they would be unable to make a profit and thus would soon stop producing the vaccines, treatments, and cures from which we all benefit. Unfortunately, the high prices that the pharmaceutical companies, armed with patents, charge for their products has the effect of denying access to crucial drugs many poor, uninsured people who could have afforded them had they been sold at or near their marginal cost. (The companies could if they wanted make the drugs available at reduced prices to poor patients, but they are reluctant to do so, primarily because they fear that the drugs would be resold in other markets, thereby eroding the companies� primary sources of revenue.) Bottom line: patents are especially socially valuable in the context of medicines, but they also have especially severe social side-effects.

Against this backdrop, how might we get adjust the legal system to more life-saving drugs into the hands of residents of developing countries? Nine options:

(1) Reduce the levels of patent protection for drugs in developing countries, allowing for the production or importation of inexpensive generic versions. To be sure, adoption of this approach would diminish the profits of the pharmaceutical companies. But remember, from the preceding post, that the developing-markets provide the companies only a tiny proportion of their overall revenues. Forcing them to forfeit that source of revenue, so the argument goes, would not materially impair incentives for innovation � and would save millions of lives. The principal impediment to this approach is the TRIPS Agreement, � a treaty binding on all of the member countries of the World Trade Organization � which, even as clarified by the recent �Doha Declaration,� sets sharp limits on the ability of developing countries to either deny or limit patent rights. Some commentators urge that the TRIPS treaty be modified to provide developing countries more flexibility in suspending patents rights so as to facilitate distribution of cheap drugs to their citizens � for example by liberalizing the currently cumbersome requirements that countries must satisfy before imposing compulsory licenses. An especially sweeping recommendation of this general sort is Tim Hubbard�s and Jamie Love�s proposal for an alternative treaty that, in brief, would give developing countries (indeed, all countries) the option of supporting pharmaceutical research either by establishing and enforcing patent protections for drugs or through some other mechanism (several of which are discussed below) that would be more efficient and thus less costly to administer. The adoption of such a treaty would give developing countries a way of providing their citizens access to cheap generic drugs (by eliminating or curtailing patent protection for pharmaceutical products) without shirking their responsibility to help finance R&D. However, persuading the United States or many other developed countries to acquiesce in such a substantial modification of the TRIPS Agreement would be very difficult.

(2) Facilitate Differential Pricing of Drugs. The basic idea here is that the drug companies would be much more likely to make cheap versions of their products available to poor patients if they were confident that those versions would not be resold to rich patients who otherwise could and would purchase the drugs at market (i.e., monopoly) rates. At the international level, the simplest way of providing the companies that assurance would be to strengthen the already substantial impediments to the resale of drugs first distributed in a poor country to consumers in a wealthy country. (Notice that such a ban on �parallel imports� runs counter to the dominant rhetoric of both of the two major political parties in the U.S. today � a rhetoric that depicts any differences in the prices at which drugs are sold in different countries as at least presumptively unfair.)

(3) Regulate the Pharmaceutical Companies. European countries have long regulated drug prices. The United States has traditionally been loathe to do so. Developing countries could force drug prices down simply by limiting what the drug companies could charge for them. The obvious hazard of this strategy is that it might prompt the drug companies to pull out of the developing-country markets altogether. A much more complex approach that Talha Syed and I are just beginning to explore would regulate the drug companies in a fashion analogous to the way we regulate car manufacturers. The so-called CAFE (Corporate Average Fuel Economy) standards oblige carmakers to ensure that the average fuel economy of the fleets of automobiles and light trucks they produce exceed specified levels. This system, in force since 1976, has been credited with inducing carmakers both to invest billions of dollars into research on mechanisms for improving fuel efficiency and to sell more small, high-mileage cars. Both of these effects have been beneficial to the environment. Suppose we imposed an analogous regime on drug manufacturers � specifically, by requiring them to ensure that the average ratio of health benefits to prices of all of the drugs they sell worldwide exceeds a specified level. (The health benefits of a particular drug would be measured by the Disability Adjusted Life Years � another index developed by the World Health Organization � it saved.) Under such a regime, if a pharmaceutical firm sold many high-priced �lifestyle� drugs in developed countries, it would be obliged, in order to meet its statutory obligations, to sell many lifesaving drugs cheaply somewhere � presumably in developing countries. A system of transferable credits might enable companies with low ratios to pay other companies to distribute more cheap drugs or vaccines. One of the advantages of such a system is that it would not attempt to micromanage the R&D or sales decisions of the drug companies (always a hazardous undertaking) but would let them figure out how most efficiently to meet the overall standards.

(4) Donor Leverage. Much of the research that ultimately enables pharmaceutical companies to develop new drugs is done in either government labs or university labs funded by the government. In exchange for making their research results available to the companies, the government or the universities might insist that any products produced in part through the use thereof be provided (either by the pharmaceutical companies themselves or by generic manufacturers) at low prices to the residents of developing countries. Statutory authority for the imposition of such conditions already exists in the United States, in the form of Bayh-Dole �march-in� rights, but that authority has almost never been exercised. Efforts to persuade the universities to exercise their powerful leverage in this fashion are currently being made by Yochai Benkler at Yale Law School. (See his recent essay in Science Magazine.)

(5) Tax Incentives. Developed-country governments might offer tax breaks to pharmaceutical companies that distribute drugs at low prices in developing countries.

(6) Governmental �Push� Progams. Governments in developed countries might conduct or fund research into tropical diseases as a form of foreign aid. If a treaty of the sort advocated by Hubbard and Love were in place, governments in developing countries might conduct or fund such research as a way of fulfilling their treaty obligations to contribute a certain percentage of their GNP to health-related research.

(7) Governmental �Pull� Programs Alternatively, governments (for either of the two reasons just suggested) might offer prizes (like those now being contemplated by NASA as a way of stimulating space research) to pharmaceutical firms that succeeded in producing vaccines or cures for specified, currently neglected diseases. A variation on the same theme: governments (perhaps with the assistance of private foundations and NGOs) might make advance commitments to purchase specified amounts at specified prices of newly developed vaccines or drugs that address particular health needs. (Kremer; Sachs)

(8) An Alternative Compensation System. In the same general vein, some economists (e.g., Calandrillo; Shavell & Ypersele; Kremer) have proposed that governments might substitute a �reward� system for the current patent regime as the primary mechanism by which we stimulate the development of new pharmaceutical products. Such a system would resemble the ACS we discussed earlier this week as a possible solution to the current crisis in the entertainment industry. A pharmaceutical company that created a therapeutically useful drug would be paid by the government. Thereafter, anyone (not just the company that created it) would be free to manufacture the drug and sell it at any price. Competition would soon drive the price down close to the marginal cost of production (just what we want). A few details: Where would the money necessary to pay such rewards come from? Probably from an income tax. Would the rewards be paid all at once or periodically, over time? Probably the latter. How would the magnitude of the rewards be determined? By a government agency, using a combination of sales figures (i.e., the total number of pills or injections sold during the relevant time period by all manufacturers of the drug in question, not just the original developer) plus a measure of the drug�s therapeutic value to estimate its total social value. Would all drugs be covered? Perhaps, but a less controversial (and expensive) variant of the idea would encompass only drugs that addressed currently neglected diseases.

(9) Harness �Traditional Knowledge.� The members of groups indigenous to tropical countries often are aware of the medicinal powers of the plants that grow in the region. Pharmaceutical companies sometimes use that knowledge to identify materials from which they might develop drugs � typically, by employing �ethnobotanists� to interview the members of such groups and then, relying on the information they provide, to collect plant samples, sometimes in violation of local �natural resources� laws. This system is not used as often or as effectively as it might be, in part because the members of the indigenous group ordinarily receive little or no compensation. (Most importantly, they do not share in the patent rights in the drugs that are developed in this way.) More effective (and fair) use of this research technique might be fostered if the TRIPS Agreement (mentioned above) were modified to require all member countries of the World Trade Organization to recognize as a form a �inequitable conduct� under their patent laws the fact that either materials or knowledge used to develop a patented pharmaceutical product was extracted from a country in violation of its laws. Such a reform would dramatically strengthen the hands of developing countries in negotiating deals with drug companies � and might ultimately lead to development of more drugs in this fashion.

I have surely not done justice to the complexity of most of these proposals. But I hope to have provided enough information in these quick sketches to foster a discussion.

The phrase, �healthy life expectancy� (HALE) is used by the World Health Organization to denote �the number of years in full health that a newborn can expect to live based on current rates of ill-health and mortality.� The HALE of the residents of Japan is currently 75 years. The HALE of the residents of Canada and most countries in Western Europe is between 70 and 75. That of residents of the United States and the remainder of the countries in Western Europe is between 65 and 70. By contrast, the HALE of the residents of most countries in sub-Saharan Africa is between 28 and 45. (Click here to see a map of the world, showing the HALEs of all countries.)

Why the huge disparity? Why do people living in developing countries die so much younger than people living in the developed countries? As one might expect, many factors are at work. The most important are: poor nutrition, sanitation, and water; climates amenable to mosquitoes and other sources of disease; inadequate education concerning methods for preventing and treating diseases; poor health-care services; and, last but not least; lack of access to appropriate medicines.

It�s the last of these factors that we�ll focus on here � partly because, as Michael Kremer has shown, it represents an ever increasing source of the problem and partly because it may be the easiest of the factors to remedy through adjustments in the relevant legal rules.

The inaccessibility of appropriate medicines in developing countries has two causes or facets. First, second, existing drugs are priced at levels that place them out of the reach of most residents of those countries. The most dramatic example involves AIDS drugs. 93% of the residents of Southeast Asian countries who are infected with HIV cannot afford the anti-retroviral drugs that would save their lives. 99% of the 25,000,000 residents of sub-Saharan Africa who are infected with HIV cannot afford the anti-retroviral drugs. The figures for other diseases are not quite so appalling, but are still very grim.

Second, too little effort and money is being devoted by American, European, and Japanese pharmaceutical companies into the development of vaccines or treatments for the kinds of diseases that disproportionately affect developing countries � e.g., malaria, measles, pertussis, and diarrhoel diseases. One especially dramatic indicator: Between 1975 and 1997, 1233 new drugs were licensed in the world for the treatment of human diseases. Of that number, only 13 addressed the aforementioned �tropical diseases.� Of the 13, five were byproducts of veterinary research. (Sources: Pecoul; Kremer) Why this bias? The primary explanation is that the pharmaceutical companies derive the overwhelming majority of their revenue from countries where tropical diseases are not common. (For example, 94% of the 2003 revenues of American pharmaceutical companies came from North America, Western Europe, Australia, and Japan. Only 6% came from the rest of the world, which bore over 99% of the burdens of the tropical diseases mentioned above.) It�s neither surprising nor immoral that the companies would focus their research and development on diseases prevalent in regions where they make money and neglect diseases prevalent elsewhere. But the net effect is tragic: millions of people are needlessly suffering and dying.

So that�s a very brief sketch of the essential features of the problem. In the next post, I�ll identify a few potential solutions.

Valuation Problems Excellent question by Cory, echoed by Erik: �It doesn�t seem that this system addresses variation of value to the consumer. The Economist, for example, can charge a significantly higher yearly subscription fee than Entertainment Weekly, because its relative value to its (I suspect) smaller subscriber base is much higher. How does this system support niche items of high value to their niche?� It�s quite right that my proposal contains a mechanism for incorporating only one of the many variables that give rise to differences among recordings in terms of their value to consumers � namely, duration. (Click here for the relevant portion of Chapter 6.) But that seems to me acceptable with respect to music and film, where differences in value are not very great � as reflected in the fact that, in the current, market-based system, CDs and DVDs of all types sell for very similar prices, and the cost of admission to theatres varies little with the content of what�s shown on the screen. The same cannot be said (as Cory�s example notes) for print media, software, or games � which partly explains why I haven�t proposed incorporating such materials in my plan.

Why not rely on voluntary contributions? Ian�s post, and the FairShare proposal to which he directs our attention, presents this common argument in a novel and unusually attractive form. My response: Your efforts to reinforce, by creating opportunities to invest in promising artists, the willingness of music consumers to donate money to their favorite musicians is commendable, but I see two problems. The first is the notorious reluctance of consumers to make voluntary contributions to creators. People tip waiters partly because it�s a well-established social custom and partly because they are in face-to-face contact with the waiters and would feel ashamed to snub them. Neither constraint operates on the Internet. The unfortunate result is the failure of PayPal etc. to generate significant revenues for artists. The second worry is that, if current trends continue, it will be less and less feasible for musicians to make money through sales of their recordings. (Some observers � including, for example, Jens in her thoughtful post � celebrate or at least accept that outcome; others lament it. But it seems hard to deny.) The result is that, in the absence of an ACS or some other substantial reform initiative [more on this below], over time fewer and fewer investors in your system will recover any money, which will make it resemble ever more closely a pure busking regime, which hasn�t worked thus far.

Big Government Several posts emphasize the hazards of letting a government set up and run such a system. I agree that the dangers are serious � and are discussed in some detail in Chapter 6. Awareness of those risks partly underlies the proposal made at the end of the chapter for a voluntary Entertainment Coop, which would resemble a government-run ACS, but would rely upon subscriptions, rather than taxes. (If you�re curious concerning how such a system might be constructed and what might prompt people to sign up for it, check out this summary.) But, as by now should be apparent, I am less despairing concerning the ability of a responsible government agency to manage such a regime than are several people who have participated in this last round of discussion. Government-run collecting societies in Europe and judicially supervised private collecting organizations in the U.S. (ASCAP and BMI) are far from perfect, but they are not disastrous either. Certainly, composers are better off in their presence that they would be in their absence. I�ve tried, in my book, to identify their defects and suggest ways in which they could be corrected. In the end, I find efforts of that sort more promising than any of the alternatives.

Other Options Speaking of alternatives, if we don�t move toward an ACS or an Entertainment Coop, what is likely to happen? Most likely, one of three things:

1. Unauthorized copying continues to increase, and consumers increasingly rely up materials obtained (free) online for their entertainment needs. The film industry, in its current form, collapses � perhaps replaced by small, independent studios, financed by donations from corporations, foundations, and government agencies. Musicians continue to make recordings (inexpensively, using the rapidly improving digital recording technologies) but don�t earn any money from them, treating them instead as advertising for their performances.

2. The record companies and film studios, dismayed by the prospect of #1, persuade Congress to reinforce the copyright system substantially � for example, by adopting the INDUCE Act and sharply increasing criminal penalties for unauthorized reproduction of digital recordings. We see a protracted �war on piracy� very similar to the longstanding �war on drugs.�

3. Alternatively, the record companies and film studios persuade Congress to adopt some version of the Consumer Broadband and Digital Television Promotion Act � which requires the manufacturers of all consumer electronic equipment to embed in their products technology that recognizes and respects watermarks, and to remove from their devices all analog ports.

I�ve already explained in prior posts why I think #1 is an unattractive outcome � though plainly I have not persuaded everyone. That said, #1 is the least probable of the scenarios. The record and film industries are sufficiently powerful, and the majority of Congressmen are sufficiently sympathetic to them, that, if the fundamental transformation contemplated by #1 seems imminent, we will see legislation of type #2 or type #3. The merits and demerits of those routes are explored in Chapter 4 of my book. Before addressing them, I�d be curious concerning whether anyone has a different forecast.

(1) Porn. Lots of people currently pay for access to pornographic films. If an ACS were instituted, some of the money raised through taxes would also end up in the pockets of the creators of pornographic films distributed for free online. Should we be troubled by this outcome? In my view, no. Arguably, we should be troubled by the fact that porn is so popular. But channeling money to pornographers through an ACS doesn�t seem any worse than channeling money to pornographers through the market.

A possible retort: �I may not be entitled to object if my neighbor buys porn, but I certainly don�t want any of my tax dollars going to porn.� The response: If you consume an average amount of recorded music and film, and if the sampling system used to manage an ACS accurately estimated the frequency with which specific recordings were consumed, then your tax dollars would go only to the artists you patronize. In other words, you would help pay for porn only if you watched porn. (For the same reason, your tax dollars will go to Britney only if you listen to Britney.) Admittedly, for this argument to hold, it would be crucial for the sampling system to work well, but we�ve already addressed this issue.

A final (and more powerful) retort: �The problem is not that my tax dollars would go to pornographers, it�s that legislators would refuse to establish a system in which any tax dollars went to pornographers. In other words, when instituting an ACS, they would insist on excluding from its coverage all pornography. We would thus start down a very dangerous road � in which government officials made judgments concerning which artistic works are meritorious and which are not.� I agree that this is a serious hazard. It�s not inherent in an ACS. But it highlights how an ACS might be distorted in the course of legislative implementation.

(2) Cross-subsidies. In an ACS, people who consume little or no recorded entertainment would pay the same amount as people who consume a great deal. It�s frequently argued that that�s both unfair and will lead to various economic distortions. This argument has a good deal of bite; such cross-subsidies definitely constitute one of the drawbacks of the system. The problem is mitigated by the fact that (under my version of the idea, anyway) only broadband accounts are taxed. (Thus someone who truly uses the Internet only for email and to check the weather can avoid paying the most substantial of the taxes.) The problem could be mitigated further if we used �Ramsey pricing� to set the tax rates (an issue discussed in more detail in Chapter 6). Finally, it could be reduced sharply if, as �.ant� suggest in his or her post) the ISPs upon whom the tax is levied charged high-volume subscribers more than low-volume subscribers. But the problem cannot be made to disappear altogether. Is it a fatal objection? In my view, no. The amounts of the taxes are modest. (The largest of them is the (indirect) tax on broadband subscriptions � which, in the first year in which the system were in operation, would be roughly $5 per month.) We tolerate vastly larger degrees of inequality in many other domains. (Think, for example, of public schooling, which is funded in the U.S. primarily by property taxes, borne in equal amounts by homeowners with no children and homeowners with many.) Bottom line: the overall savings of the system seem sufficient to justify a modest degree of inequality in the distribution of its burdens.

(3) Derivative Works. One of the greatest advantages of an ACS, in my view, is that it creates a sensible mechanism for dealing with composite recordings � rap songs that include samples of other recordings, modified or expurgated films, mash-ups. In sharp contrast to copyright law, an ACS regime would authorize persons to rework existing recordings, refashioning them into new products � provided they complied with two conditions: (a) they identified the owners of the copyrights in the original works when they registered their new works; and (b) they gave credit where credit is due. When a modified work were downloaded or streamed, both the creator of the original work and the modifier would get a share of the resultant revenue. (For considerably more detail concerning how such a system would work in practice, see Chapter 6.) The net result would be to liberate the transformative cultural activities the potential for which is perhaps the most important of the many benefits of the Internet, while continuing to compensate fairly creators of various sorts.

(4) Who Gets Paid? There are very good reasons, suggested by Kristin in her post, for requiring that at least some portion of the payments from an ACS be distributed directly to artists, rather than paid to the intermediaries (e.g., record companies) to whom the artists have assigned their copyrights. (Those reasons are explored in more detail in Chapter 5.) But there are two related catches: moving in this direction further decreases the already small probability that the major intermediaries would support the plan; and it would also increase the likelihood that the system would be challenged successfully as an unconstitutional �taking� of private property without just compensation.

For yet more objections to ACSs � and responses to those objections � check out the pertinent portions of Andrew Orlowski�s recent speech at the �In the City� Convention � available at http://www.theregister.co.uk/2004/09/23/orlowski_interactive_keynote/page7.html.

That�s probably enough on this issue. Tomorrow, I�ll venture some comments on an equally divisive topic: price discrimination. Thursday or Friday, if all goes well, I hope to veer in a different direction, discussing the current crisis involving the distribution of drugs in developing countries.

Thanks to everyone for your comments thus far.

Terry Fisher

First of all, it�s important to emphasize that I�m not the only person who has proposed an ACS. The pioneering treatment is Neil Netanel�s. Among academics, Jessica Litman, Raymond Ku, and Glynn Lunney have all contributed importantly to the project. The general idea has been in the air for at least a century. Certainly, several of the people who have gone before me would disagree with some of the comments set forth below. With that caveat, here�s the summary of my own version:

�The owner of the copyright in an audio or video recording who wished to be compensated when it was used by others would register it with the Copyright Office and would receive, in return, a unique file name, which then would be used to track its distribution, consumption, and modification. The government would raise the money necessary to compensate copyright owners through a tax � most likely, a tax on the devices and services that consumers use to gain access to digital entertainment. Using techniques pioneered by television rating services and performing rights organizations, a government agency would estimate the frequency with which each song and film was listened to or watched. The tax revenues would then be distributed to copyright owners in proportion to the rates with which their registered works were being consumed. Once this alternative regime were in place, copyright law would be reformed to eliminate most of the current prohibitions on the unauthorized reproduction and use of published recorded music and films. The social advantages of such a system, we will see, would be large: consumer convenience; radical expansion of the set of creators who could earn a livelihood from making their work available directly to the public; reduced transaction costs and associated cost savings; elimination of the economic inefficiency and social harms that result when intellectual products are priced above the costs of replicating them; reversal of the concentration of the entertainment industries; and a boost to consumer creativity caused by the abandonment of encryption. The system would certainly not be perfect. Some artists would try to manipulate it to their advantage, it would cause some distortions in consumer behavior, and the officials who administer it might abuse their power. But, on balance, it is the most promising solution [to the intensifying crisis in the entertainment industry].�

Here are five worries/objections often raised by skeptics of such a system � followed by efforts to address each:

(1) Does such a system �scale� internationally (see John Allsopp�s post)? This is definitely one of the weaknesses of an ACS. If adopted in only one country (say, the U.S.), it would �leak� across national boundaries � in the sense that French artists whose creations are downloaded in the U.S. would be paid portions of the tax revenues collected by the American government from American consumers, while French consumers would gain free access through the Internet to the creations of American artists, and American artists would not collect anything from French taxpayers. To be sure, American artists would be no worse off under such a regime than they are at present. Nevertheless, awareness of this aspect of the system would contribute to the already substantial resistance of Americans to the adoption of such a regime.

Is there any escape from this bind? Modification of the pertinent international treaties (most likely, the TRIPS Agreement) to force other nations to adopt similar regimes is extremely unlikely in the near future. Harmonization thus would have to occur through voluntary adoption of ACSs by other countries. Chapter 6 of my book explains: �The success of the system might prompt countries other than the United States to institute similar systems. Each would impose taxes on its own residents� ISP subscriptions and purchases of electronic equipment. Each would establish a registration system, permitting copyright owners from every country to register audio and video recordings. (Ultimately, these separate national offices might be superseded or supplemented by a global registry for digital works.) Using schemes like those outlined above, each country would estimate the relative frequency with which those recordings were consumed by its residents � and would then distribute its tax revenues accordingly, to both domestic and foreign registrants. An interlocking set of national regimes of this sort would cure the third of the three major disadvantages of a tax-and-royalty system noted in the previous section � namely its tendency to leak across national boundaries. All of the national regimes would continue to leak, of course. But the leaks would occur in both directions � and would fairly reflect the extent to which consumers within one country were relying for their entertainment on works created by artists in other countries.�

(2) Would artists really be compensated in proportion to the frequency with which their creations are consumed? Skepticism on this score takes two very different forms. Some people (like Jeremy Williams of Warner Bros.) worry that too much money would be channeled to minor contributors to the stew of entertainment products, leaving the studios with too little income to support high-cost modern productions. Others (like John Allsopp in his post) have the opposite concern � that the enormous number of artists whose works are downloaded in small numbers will not get their fair shares. The ability of the system to assuage both concerns depends on the quality of the sampling system used to count consumption rates. In Chapter 6, I devote a fair amount of space to a discussion of how one might design an effective sampling system. In brief, here are its essential features:

�[T]he Copyright Office [w]ould randomly select a set of entertainment consumers who were willing to allow the Office to monitor what they actually listen to and watch. The imperfections of the Nielsen model could be avoided (or at least mitigated) through the following [three] related adjustments. First, the process of gathering data concerning consumers� habits could and should be automated. Software � distributed as �plugins� for playback devices or bundled with peer-to-peer file-sharing applications � would automatically record the registration numbers of the songs and films that sample members heard and watched (all the way through) and periodically transmit that information to the Office. Sample members would thus experience no inconvenience and would have few opportunities to misreport their choices. Next, the size of the sample employed by the Copyright Office would have to be vastly larger than the sizes of the samples used by Nielsen. This would be essential to enable reasonably accurate estimates of the frequency with which each member of an enormous array of songs and films were being consumed. It would be feasible because of the low cost of the automated reporting system. [Third and finally, to] persuade a representative set of households to permit their consumption patterns to be monitored, one would have to provide them credible assurances of privacy. In other words, they would have to be persuaded that the data the Copyright Office gathered concerning the frequency with which they watched particular films or listened to particular songs would be aggregated when determining the amounts of money paid to artists, would be discarded after each monthly accounting, and would not be made available to any other public or private entity�.�

But the proof concerning the efficacy of such precautions is in the pudding. At the Berkman Center, we�re hoping to build a sampling system along these lines � and then to try it out in conjunction with a voluntary Entertainment Coop. One of our major goals is to address the legitimate anxiety that, under an ACS, artists of different sorts would be either underpaid or overpaid.

(3) Won�t unscrupulous artists and third parties �game� the system, artificially inflating the number of times their works are supposedly �consumed� and thus depriving deserving artists of their fair shares of the ACS fund? The primary answer to this serious source of concern is that, once again, a great deal depends upon the quality of the sampling system used to estimate consumption rates. A secondary response is that the �gaming� problem is most serious with respect to downloads and is much less worrisome with respect to streamed works. As we shift, increasingly, from a world in which people create permanent collections of audio and video recordings to a world in which people listen to �streamed� songs and watch �streamed� films, the �ballot-stuffing� problem will diminish.

(4) Shouldn�t ISPs, rather than individual consumers, pay the taxes necessary to run such a system? Yes, and/but they would undoubtedly pass a substantial portion of the tax on to their subscribers.

(5) Wouldn�t consumers end up paying just as much for access to entertainment under an ACS as they currently do under the copyright regime? If so, we will have accomplished nothing more than substitution of a creaky government bureaucracy for a creaky copyright system. This, in my view, is the objection most easily met. Even if the distribution of digital recordings over the Internet fully displaced the current mechanisms by which recordings are distributed, consumers would end up paying much less under an ACS than they currently do. Specifically, even using highly conservative assumptions, the average American household would end up paying no more than $250, roughly half of the $470 the average household currently pays for access to recorded entertainment, and would receive, in return unlimited amounts of ad-free music and movies. That seems a gain dramatic enough to warrant seeking solutions to the hazards and complications discussed above.

I�m hopeful that other aspects of the system will emerge in the course of the discussion, but those seem like enough for now.

I thought I�d begin by briefly summarizing the argument of the first chapter, and then ask whether, particularly in light of some recent articles and developments, the argument holds up.

Here�s the argument in a nutshell: In combination, three technological developments � digital recording and storage systems, compression/decompression systems, and the Internet � have created dramatically new ways of making, keeping, sharing, and enjoying audio and video recordings. Full exploitation of those new techniques would have many social and economic benefits: large costs savings (enabling consumers to get more entertainment for less or artists to be paid more); greater convenience and precision in the ways that recordings are delivered to consumers; a sharp increase in the number of musicians and filmmakers who can reach global audiences and make decent livings; enhancement of the diversity of materials available to the public; and a dramatic increase in the number of persons who participate in the making of culture (a trend for which I use the term [not of my own invention] �semiotic democracy�). The same developments, however, pose three serious dangers: corrosion of the systems by which artists and intermediaries have traditionally made money from their recordings; threats to artists� �moral rights�; and destabilization of the cultural icons in reference to which we partially define ourselves. The chapter concludes with the claim that we ought to strive, through a restructuring of the legal system and associated business models, to capitalize on the opportunities created by the new technologies, while minimizing the concomitant hazards. (The rest of the book then goes on (a) to argue that we�ve failed to achieve that goal thus far and (b) to sketch some ways in which we might.)

Before addressing the particular ways in which I and others have tried to solve the crisis, it might be helpful to consider whether my characterization of the crisis is fair and balanced. One potential line of criticism would point to the recent paper by Oberholzer and Strumpf, �The Effect of File Sharing on Music Sales (http://www.unc.edu/~cigar/papers/FileSharing_March2004.pdf) (which appeared after my manuscript was set) as evidence that I have seriously exaggerated the extent to which the new technologies (in this case, P2P services) have, at least thus far, threatened traditional business models. Another potential line of criticism would argue that moral rights (artists� interest in the integrity of their creations and in being given appropriate credit for their creations) are overblown � either in general or with respect to digital recordings, of which unlimited copies may be made. I�d be curious to hear reactions on these or any other pertinent issues.

Two unrelated procedural points: First, unfortunately, on Monday I will be on planes from roughly 1:00 am to noon Eastern time � and thus, unable, during that period, to respond to comments. But thereafter, I�ll be back online. Second, it may be impossible, given the importance and imminence of the election, to keep a conversation going this week about online entertainment. If so, I�ll have to adapt in some way.

Terry Fisher