SOME PUZZLING BEHAVIOR BY OWNERS OF INTELLECTUAL
PRODUCTS: AN ANALYSIS
S. J. LIEBOWITZ*
Creators of intellectual properties frequently seem to prefer “rental,” or sale with restricted rights, over outright sale of the intellectual properties. Surprisingly, these individuals also seem to prefer legislative provisions forcing them to restrict purchasers’ rights rather than allowing this result to be achieved through voluntary negotiations. In this paper, I examine two examples of this behavior: the sale of syndication rights for television programs and the separation of broadcast rights from synchronization rights for music in television broadcasts. The desire to restrict purchasers’ rights is explained in two ways. First, I appeal to recent developments in utility theory, namely, regret theory. Second, I show that sales of intangible goods such as intellectual properties are less closely related to use than are sales of tangible goods, and that restricting rights is an attempt to enhance such a linkage. I give several explanations for using legislative solutions, although I acknowledge that further research is needed on both questions before definitive answers can be given.
When a private good is exchanged, the original owner usually loses his rights over that good. It is unusual for the government to insist on restricting the original owner’s ability to trade away his rights over the good when it is “sold” in the marketplace. Intellectual properties, unlike most private goods, provide battlegrounds for governmental actions intended to restrict the rights associated with exchange. Owners of intellectual products often favor governmental actions restricting the property rights of those who purchase intellectual properties. This paper investigates the motivations of the owners of intellectual properties who seek to restrict their own ability to trade away their rights over such properties in the marketplace.
Although most private goods transactions do not involve the seller’s restricting the buyer’s rights when the good is sold, it is not unusual for the seller to restrict the buyer’s actions when the good is rented. Rentals, as opposed to outright sales, are for a limited duration and often provide very restricted
rights. Because some elements of rentals are similar to the restrictions presently desired by owners of intellectual properties, I use rentals as an analogy.
It is assumed here that two types of exchange exist. In the first case, that of unrestricted rights, the buyer receives the good and can do whatever he wishes with it. The original seller loses all rights over the good, as is usually associated with sales. In the second case, that of restricted rights, the seller imposes significant restrictions on the buyer s use of the good, as is often associated with rentals. The exchange of goods with restricted rights imposes additional costs because the buyer must be monitored to make sure that use of the good is restricted as per the contract.
In this paper, I examine two specific cases that illustrate this phenomenon. The first is the struggle between television networks and television program producers regarding the networks’ rights to buy shows. Producers of programs seek to restrict the networks’ rights to own programs sold to the networks. In particular, producers seek to retain ownership of the programs when reruns are sold in the syndication market. The second case is the disagreement between authors of copyrighted music and television broadcasters regarding the sale of music rights contained in television programming. Television broadcasters seek to be allowed to purchase television programs that include the rights to retransmit music contained in the programming. Under the present practice, broadcasters pay the owners of the music a separate broadcast fee to broadcast the programs. Copyright owners oppose such a change.
In both of these issues, the owners of intellectual products seek to induce the government to restrict the property rights of purchasers of the products. These are not the only examples of this type of behavior. For example, droit de suite, which allows artists to derive income every time one of their works is resold, could have been used. However, the two issues examined here are current and newsworthy controversies.
In these two examples, the transactions seem to take place in relatively free markets where presumably any type of deal could be worked out—including one not allowing the purchaser unrestricted rights. One immediate question, then, is why creators of these properties seek legislative solutions when it appears that they could achieve their goal with mere contractual solutions. This is in addition to the main question of why owners of intellectual properties seek to restrict the use of these properties more so than the owners of physical goods seek to restrict the use of their goods.
Intellectual products differ from most private goods in that intellectual products are not themselves tangible, and can be perceived only when physical embodiments are produced. Thus, one cannot partake of an author’s creation without either reading the story in book form or seeing it performed on stage, in a movie theater, or on television. Technology now allows multitudinous outlets for many intellectual properties. One challenge of current public policy
is constructing the proper set of rights for purchasers of the intellectual properties in any outlet, particularly when the outlets overlap with each other.
The enlightened economic purpose for providing any property rights system is quite straightforward: to provide incentive for the efficient production of products. The stated intention of intellectual property laws is to provide greater incentive for producing intellectual properties. Producers of intellectual products receive this incentive by being granted copyright protection (or other forms of intellectual property protection such as patents) which tends to allow increased revenues to copyright owners, who have the sole right to duplicate their works. Of course, the value of any additional works induced by copyright legislation must be balanced against the social harm caused by any unnecessary (monopoly) restriction in the number of authorized copies made from intellectual products that would have been created without such stringent restrictions. Also, strengthening copyright protection will not always lead to an increase in the incentives to produce such works (see Liebowitz, 1985).
Copyright laws attempt to prevent unauthorized duplication. Sometimes the act of duplication is simple to define, as in the case of making physical copies of a book. In other instances it is more difficult, as in the case of a group of musicians performing music written by someone else. In this latter case, for example, the musicians usually will have paid for the (sheet) music to be able to play it. Society, through its laws, indicates that purchasing the sheet music is not by itself considered sufficient payment when the music is to be used in concert, on records, on television, or on radio. Therefore, additional rights have been granted to copyright owners for the performance, broadcast, and recording of their music, in the hope that copyright owners will generate greater revenues.
Providing these latter rights can be viewed as allowing the copyright owner to price discriminate among different types of music purchasers more fully than would be possible without these rights. Those individuals planning to perform a piece of music in concert, for example, are likely to place a higher value on the music than are most other users. But this higher value is hard to capture in the price of the sheet music since it is difficult to identify and charge a higher fee at the time of sale to purchasers planning to use the sheet music for these ends. These individuals, therefore, must pay an additional fee when they perform the music in concert, with the fee itself being a function of the concert size.
Once a set of rights exists, different payment mechanisms are possible. It is possible to collect an ex ante flat fee based on expected use, as in a sale, or to collect an ex post fee based on (imperfectly) measured actual use of the copyrighted material, as in a rental. The policy problems associated with these payment mechanisms are illustrated below.
In 1970, the Federal Communications Commission (FCC) approved the financial interest and syndication rule. This rule forbids television networks from taking a financial interest in most of the programs they broadcast. Under this rule, networks pay a license fee entitling them to broadcast each episode of a television program twice a year. The rights over the show then revert to the program producers who, if the show is successful, can sell (syndicate) the program’s episodes to local stations. In essence, this rule amounts to forcing program producers to rent their programs to networks for one season.
Syndication fees for successful programs can be very large, and have been the center of a rather heated controversy. In 1983, the New York-based networks apparently had convinced the FCC to abandon the financial interest and syndication rule before a fierce lobbying effort by Hollywood’s motion picture studios, star actors, independent producers, writers, and directors derailed the momentum that the networks had built up (see Landro and Saddler, 1983).
On the surface, there appears little reason for the bitter controversy. One would think that if program producers preferred renting their programs instead of selling them outright, they could choose that form of contract with or without the financial interest and syndication rule. Clearly, the price of renting a program for one year would be less than the purchase price, although statements made by the participants indicate little such awareness. One might expect that the difference between the two prices would be the expected value of the program’s syndication rights. Therefore, it would appear that if program producers sold their programs voluntarily, the price would go up and they should be no worse off than they would be under the syndication rule. That is because program producers would sell the programs only if the value of ownership were greater for the networks than that for the program producers. One would also expect that the actors, writers, producers, etc., would be made better off by an arrangement allowing the greatest flexibility in negotiations between networks and producers.
Individual producers, however, may be made better off or worse off by these rules. Under the current system, those programs not successful in the syndication market receive no payment in that market, while those programs that are syndicated receive all of the syndication revenues. If networks were allowed to buy programs and did so, they would include in their reservation bids the expected value of syndication revenues. Because some uncertainty over a program’s future syndication revenues always exists, some programs that did worse than expected would be given undeserved syndication revenues
while other programs that did better than expected would be given syndication revenues below those derived from their actual performance.
Two major differences between renting and purchasing programs would seem dependent on who has the advantage in predicting the future and on who has the advantage in bearing the risk of a show’s making it into syndication. Casual empiricism would indicate that networks likely are better able to hear the risk since the large number of programs (portfolio) that they air would allow a greater risk reduction than would the smaller portfolios of producers. One also would think that networks would be better able to predict a show’s ability to make it into syndication, first, because they can influence the likelihood of success with their programming decisions and, second, because they can specialize in estimating and monitoring the appeal of the numerous shows that they broadcast.
Voluntary negotiations apparently would let the more efficient prognosticators and risk bearers take on these burdens, to the benefit of both networks and program creators. Yet this obviously is not the perception of the producers, writers, actors, etc. This analysis, therefore, must be incomplete, or else the Hollywood community must be expending resources on an effort which, if successful, will make it poorer.
When television programs are broadcast, the rights to broadcast the music and lyrics contained within them are owned primarily by two performing rights societies, the American Society of Composers, Authors and Publishers (ASCAP) and Broadcast Music Inc. (BMI). A performing rights society, whose members are authors (music and lyrics) and music publishers, grants a license to a broadcaster for the society’s entire repertoire of music for a yearly fee based on the revenues of the station. The society then pays back the revenues to the composers and publishers, based on a measurement of usage after deducting the overhead cost of running the organization.
Performing rights organizations presently are at odds with two recently proposed pieces of legislation (HR. 3521 in the House and S. 1980 in the Senate). These bills would require that program producers obtain the performing rights at the time of initial production, and that these rights be delivered along with the program at the time of purchase. The broadcast rights still would be purchased, but by the creator of the program and not by the broadcasters.
Copyright law allows separate rights for both recording (synchronization
rights) and performing (broadcast rights). This means that a musician writing a song for a television program may receive up to three payments by the time the song is broadcast. First, he will receive payment for creating the song if it is commissioned by the television producers. Second, he will receive payment when the song is recorded on videotape (synchronization payment). Third, he will receive payment when the program is broadcast (broadcast right). The bill would merge these second and third payments into one payment.
Broadcasters and performing rights organizations are expending considerable resources to influence the outcome of these bills. Yet the differences to broadcasters would appear to be due mainly to shifting risk to them from authors and music publishers. After all, under the proposed system, the upfront payment should be higher by an amount presumably equal to the expected value of the performing rights. Those copyright owners of music contained in programs that are more successful than expected would receive less remuneration under a system based on upfront payment 5 and those copyright owners of music associated with less successful programs would receive greater compensation.
In this section, I address two issues. First, economic analysis should tie able to predict why creators of artistic works seek to restrict the property rights associated with their goods while owners of private goods seem content without such restrictions on their goods. Second, creators of artistic works should be able to restrict property rights of the goods that they sell through contractual means. Why, then, are these issues being fought in the legislature, with copyright owners supporting proposals that would seem to restrict flexibility?
I have two explanations for why creators of artistic works might seek to restrict the use of their intellectual products to a greater degree than do sellers of most other products. The first concerns the utility functions of copyright owners, which I believe are likely examples of a relatively new theoretical extension of utility theory known as ‘regret theory.” The second explanation concerns the incomplete knowledge of market conditions possessed hy creators of intellectual products and methods for them to transcend this incompleteness.
Role of risk and regret. Most of the above analysis is concerned with risk bearing. It is possible that neither the owner of an intellectual product nor anyone else knows how successful it will be. Yet the copyright owner must exchange the product in the marketplace. If the copyright owner sells the product with no restrictions on use, the buyer then bears the risk of the product’s future success. The buyer should be willing to pay a price up to his expected value of the product. The buyer reaps the rewards if the product is a big hit and suffers the losses if the product is a flop. If the copyright owner sells the good with a very restricted set of property rights, then he retains some or most of the risk himself.
Since those who purchase intellectual properties usually are major purchasers of artistic creations—such as television networks—they would seem better able to form portfolios of these items and therefore better able to bear the risk. Yet, an explanation exists for why the artists might wish to bear the risk. The previous discussion of risk was couched in terms of expected utility theory. An alternative approach still in its infancy—regret theory—allows for different conclusions. Under this theory (see Sugden, 1986), consumers take into account the “regret” felt from not gambling on a bet that eventually pays off, as well as the ‘rejoice” felt if a gamble not taken does not payoff. When they maximize their utility, they do not merely maximize expected value of the payoff function. It seems reasonable that artists would regret more acutely the loss of riches from not sticking with their own creations than would most people in not taking a bet with the same odds. Under this theory, then, artists will accept risks shunned by most other individuals when those risks concern the artists’ own creations. This could explain why artists oppose letting networks bear the risk of programs’ success in the syndication market.
Transcending incomplete knowledge of use. Private goods differ from intellectual products in several ways. Most private goods are tangible in the sense that they can be touched, used, and used up. Physical goods provide services over some period of time, but eventually they deteriorate into an unusable form. Cars, machines, and buildings all wear out and eventually become incapable of providing any more service. Intellectual products, on the other hand, are public goods. This means that one person’s consumption doesn’t reduce anyone else’s possible consumption of the good. Also, intellectual properties cannot be “used up” in the ordinary sense. The story contained in a book can continue to be “used” forever and can be used an infinite number of times, although any particular copy of a book will wear out with use.
Because physical goods wear out, and because the rate at which they wear out tends to be related to use of the good, a positive correlation will exist between the use of a physical good and the frequency of the good’s purchase. This means that those who use a physical good more intensively will pay more for their use since they will use more units of the good. For example, an automobile used as a taxicab will have a shorter lifespan than will one used only as a private vehicle. Thus, a taxicab company keeping one automobile on the road will pay more over 30 years than will a family keeping one automobile on the road for the same 30-year period because the cab company will have to purchase more automobiles over the 30 years. (A more detailed examination of these issues can be found in Liebowitz, 1983.)
This positive association between payment and use helps the producer of a physical good to maximize revenues for a given demand for services of the
physical good when he sells it with no restrictions. The producer need not place restrictions on the purchaser’s use of the good to create a positive association between use and payment since this association occurs naturally and since the monitoring costs implied by these restrictions might be uneconomical.
Unfortunately for producers, the positive association between use and payment, and between value and payment, does not exist for intellectual products. An individual’s more frequent uses of an intellectual product need not translate into more frequent purchases. To maximize revenues from a given demand for services, the seller of an intellectual product would seek to restrict the property rights of purchasers so that those who use the good more intensively pay a higher price. A rental agreement with monitoring of use, or sales with incomplete property rights, would better allow the producer to price discriminate and to maximize profits.
One necessary condition for the seller to he able to restrict the purchasers’ property rights is that the producer have some degree of monopoly power. If competition exists among owners of intellectual properties, the price will he driven down toward cost. If customers preferred unrestricted rights upon purchase and if production were profitable, the products would be sold with no restrictions. The government, however, provides a legal monopoly to the creators of intellectual products. Since there exist very few, if any, intellectual products that are perfect substitutes for each other, this protection confers real monopoly power for at least some properties so covered. Therefore, owners of intellectual properties can engage in activities dependent on monopoly power to a far greater degree than can ordinary producers, and thus the former have a greater preponderance of these restrictive contracts.
Voluntary negotiations seemingly would allow the most flexibility and the greatest welfare for all participants. After all, if artists want to bear the risk of the future value of their creations, they should be willing to bid higher amounts for the future rights of their creations and would retain these rights under voluntary negotiations. Yet, a reason must exist for why the legislative route has been taken. The political arena is difficult to analyze, and my conclusions here are more tentative than those dealing with the proclivity of copyright owners to sell intellectual properties with restricted rights.
Negotiations will allow the rights to be granted to the market participant with the highest value. Why, then, do both parties not agree to voluntary negotiations? Several hypotheses are common to both examples:
(1) Unequal bargaining power causes the weaker party to forgo voluntary
negotiations. (a) Syndication rights: Perhaps producers are competitive and networks are monopsonistic. (A monopsonist is the sole buyer of a resource.) Of course, the networks’ first-year rental payments still are negotiated under monopsonistic conditions, but the syndication rights are negotiated on more equal terms, with the purchase of syndication rights taking place in the more competitive local broadcaster market. This allows the program producers to generate greater revenues in syndication than the networks would have paid them. (b) Music broadcast rights: Perhaps performing rights organizations are monopolistic. If this is the case, the performing rights organizations can act with a higher degree of monopoly power than could individual composers. Thus, the organization can raise prices and generate more revenue for composers. Of course, this assumes that composers do not compete these rents away in other markets such as the synchronization rights market.
(2) One party thinks it can generate short-run quasi-rents from the other party. (a) Syndication rights: Perhaps the networks believe that they can usurp ownership over programs already produced for which they have not paid any indirect syndication rights in the negotiated fees. (b) Music broadcast rights: Composers authored their works with the expectation that they would have a performing rights organization collecting revenues. If the performing rights are negotiated individually, competition among composers would drive down the price to a level below that expected and required for future creation of music (the long-run marginal cost of creating music) since the cost of writing the music is sunk.
(3) One party or both parties are irrational and think they can legislate a repeal of supply and demand. For example, perhaps artistic individuals believe that government intervention will allow them to receive both higher prices and higher quantities. This explanation is empty, in the sense that no empirical evidence can be adduced to refute it. Yet, one should not rule out the possibility that this may in fact be the correct explanation.
Some hypotheses apply to one of the following examples but not the other:
(4) Program producers can shift the costs of monitoring and enforcement to the FCC. One problem with this explanation is that although both parties to the contract would benefit from the reduced enforcement costs, the networks object to the government monitoring.
(5) Program producers fear that the networks will produce their own shows, thus increasing competition and decreasing profits in the production of television programs. This could explain the behavior of program producers but not that of the actors, writers, and directors, who presumably would benefit by more competition for their talents, unless it was believed that networks would drive the independent producers out of business and monopsonize the market for creative inputs.
(6) An organization originally created to monitor radio transmissions may not wish to relinquish these activities, even if it would be better for the presumed beneficiaries of the performing rights organizations to include the broadcast rights together with the synchronization rights.
Only future empirical work can determine which of these explanations, if any, is correct. My own guess, based on casual evidence gathered while observing these controversies, is that the syndication rights battle is due to factor (5): outside producers’ fear of network competition, and actors’, writers’, and directors’ fear of any change, particularly one that might increase their employers’ monopsony power. Since the potential monopsuny power envisioned by actors and writers probably is exaggerated, allowing the financial interest and syndication rule to expire probably would be in the public’s best interest. The music broadcast rights controversy probably is a mixture of factors (1) and (2), where the performing rights organizations seek to prevent individual authors from driving the prices into the ground through competition. I feel that the evidence here requires that further work be undertaken before even preliminary policy conclusions can be drawn.
Public policy with respect to intellectual property has been increasing in importance as these properties have been increasing in value. 1 have attempted to examine some puzzling behavior of intellectual property owners, while remaining aware of the differences between intellectual properties and physical properties. I have examined two interesting policy problems in copyright law. I hope that some foundation has been laid and that some interest has been created to generate future work on these and related questions.
Landro, L., and J. Saddler, “Network, Film Moguls Blitz Capital in Battle for TV Rerun Profits,” Wall Street Journal, November 8, 1983, p. 1.
Liebowitz, S. J. “Tie-in Sales and Price Discrimination, Economic Inquiry, luly 1983, 387-399. Copying and Indirect Appropriability Photocopying of Joumals,” Journal of Political Economy. October 1983, 945—957.
Sugden. R. “New Developments in the Theory of Choice under Uncertainty, Bulletin of Economic Research, ]anuary 1986, 1—24.
* Department of Economics and Business, North Carolina State University. I thank Yoram Barzel, Jack Hirshlelfer, and Craig Newmark for their helpful comments and the John Olin Foundation for supporting me as a fellow at tlse University of Chicago’s Law School while this paper was l)ciug written. An earlier version of this palser was presented at the 61st Annual Western Economic Association International (>snferenee San Francisco, Calif., July 1986.
 Specifically, the networks were limited in the number of hours that they could produce in house. In 1980, they were allowed to produce 2.5 hours of prime-time programming per week which increases to 5 hours per week for the 1988—1989 season.
 Presently, producers of television shows take into account the expected syndication fees when they produce their programs. The networks’ payments for programs consistently are lower than the costs of producing the episodes, and the producers are able to break even or make a profit only if the show is sold in syndication..
 Stations have the option of paying for the music contained in individual programs—as opposed to paying for an entire year of programming—and of dealing directly with the authors. Tlsis option virtually is never used, however.
 I thank Jack Hirshleifer for informing me of this line of research.
 Of course, the value of an intellectual product may deteriorate over time due to changes in the tastes and ronditions of those who consume the intellectual product. But the intellectual product itself does not deteriorate.
 It should be noted that this logic applies not only to copyrighted and patented goods, but to any goods (read services) that are intellectual in nature and are noncorporeal. such as advice on topics in medicine, law, or accounting. These other goods tend to be related more uniquely to particular uses by particular individuals, and so tlse problem discussed here is somewhat less acute. Instances no donist occur, however. where sellers try to restrict their clients’ use of the advice.