Edited Version Originally Published in 45 Duke L. J. 1133 (1996).
Copyright 1998 Jack M. Balkin. All Rights Reserved.
MEDIA FILTERS AND THE V-CHIP
Introduction–To V or Not to V
One of the most controversial features of the Telecommunications Act of 1996(2) is its intervention in longstanding disputes about violence and indecency in the media. Due in part to the urging of President Clinton and his Democratic allies, the new Act requires that all television sets over thirteen inches include a “V- chip,” a device that would allow parents to block violent and indecent television programming.(3)
Despite its name, the V-chip is not a single chip at all, but a combination of different technologies. All television programs currently have the capacity to carry extra information–like closed captioning–as well as sound and pictures. An electronic circuit in a television or cable box can be designed to block programs by reading a numerical code broadcast along the same band used for closed captioning. Viewers then use a remote control device to select from a menu of choices as to how much violence, bad language, sex, and nudity they wish to tolerate. An experimental rating system developed in Canada, for example, features a five-number scale, with three separate categories for sex, profanity and violence. Higher numbers signify higher levels of each category. When the V-chip circuitry reads a rating equal to or higher than the consumer’s preselected standards, the picture is replaced by a large black box. A V-chip system can also be designed to recall previous settings and block all unrated programs. However, in order to prevent bad language from being transmitted, it must be able to block sound as well as pictures.(4)
In response to the passage of the Telecommunications Act, members of the American entertainment industry met with President Clinton on February 29, 1996, and promised the delivery of an industry-sponsored ratings system for the V-Chip within a year. On December 19, 1996, an industry committee, led by Jack Valenti, president of the Motion Picture Association of America (MPAA), unveiled a ratings system quite different from the Canadian model described above. In its original formulation, this system did not offer separate categories for violence, sexual content, or bad language; instead it considered all three together and rated programming based on appropriateness to age, much like the MPAA ratings system. The current age-based ratings categories are TV-G (General Audience– suitable for children of all ages), TV-PG (Parental Guidance Suggested– material that some parents would find unsuitable for younger children), TV-14 (Parents Strongly Cautioned– material that many parents would find unsuitable for children under 14), and TV-M (Mature Audiences Only– specifically designed for adults and unsuitable for children under 17). There are also two special categories for programs designed specially for children: TV-Y (All Children– designed to be appropriate for all children), and TV-Y7 (Directed to Older Children– designed for children age 7 and above).(5) In response, Canadian broadcasters announced their own age-based system on May 5, 1997; it primarily rates violence as opposed to sexual content or bad language.(6) On July 10, 1997, the major American networks (excluding NBC) agreed to supplement their age-based ratings with codes for violence (V), sexual situations (S), coarse language (L) or suggestive dialogue (D). Children’s programming rated TV-Y7 may carry additional codes for “fantasy violence” (FV).(7) Although the V-Chip system has yet to be fully implemented in mass market televisions as of the date of this writing, broadcast and cable programmers have already begun displaying age-based ratings at the beginning of many different television shows; they intend to begin display of supplemental content-based ratings sometime in the fall of 1997.
Critics charge that the V-chip raises serious First Amendment problems. This essay explores a few of them. But my more important goal is to use the debate over the V-chip to rethink the foundations of broadcast regulation. The federal courts, including the Supreme Court, have justified content-based restrictions on broadcast indecency partly on the grounds of the special nature of the mass media. Yet their justifications for special treatment have been, on the whole, unconvincing. I will argue that the real issues have little to do with traditional justifications of scarcity, public interest, and pervasiveness. They have to do with how different media permit the filtering of information. Different communication technologies are better adapted to different kinds of informational filters. For example, broadcast media permit different and more limited filters than print media. The V-chip promises to change all that by creating a new system for filtering broadcast information. But this new technology raises many new and unexpected problems. In particular, it raises the possibility that in the Information Age, control of filters may be one of the most important forms of power over human thought and human expression. In the Information Age, the informational filter, not information itself, is king.
I. The Difference Broadcasting Makes
The constitutional status of the V-Chip is inextricably linked to the special constituitonal treatment of broadcasting in American constitutional law. For many years, broadcast media have been subject to much greater content-based regulation than print media. For example, in FCC v. Pacifica Foundation, the Supreme Court upheld the constitutionality of FCC restrictions on indecency as applied to a radio broadcast of George Carlin’s “Filthy Words” monologue.(8) More recently, the D.C. Circuit upheld “safe harbor” provisions that permit indecent speech on broadcast television only from 10 p.m. to 6 a.m.(9)
First Amendment scholars are divided as to whether this special treatment is constitutional. They have good reason to be concerned. “Indecency,” like violence, is an unclear and wavering category. By definition, it includes sexually explicit speech that could not be regulated as obscene. This is a much larger category than many people imagine. It includes, for example, not only expression expressly designed for sexual stimulation, but also expression that is offensive to some but not obscene because it has genuine literary, artistic, political or scientific value. Thus, indecent expression can include not only the more salacious contents of the Playboy Channel, but also political speeches laced with four-letter words and serious discussions of AIDS and homosexuality.
Similar problems hound the regulation of violence.(10) It is not always clear what kinds of violence do the most harm to children. Is the violence in cartoons worse than the violence in live action programs? Does unrealistic violence do more harm than depictions that bring home the horrors of war and death? Does the violence reported on the local and national news contribute to the problem, and if so, should it also be restricted in the interests of our children?
In assessing the constitutionality of restrictions on violence and indecency, it is important to remember that the programming at issue here would be constitutionally protected if it appeared in print media, in a movie theater, or on a videocassette. There must be some special justification for abandoning general First Amendment principles in broadcast regulation.
Traditionally, content-based regulations of the broadcast media have been justified on two basic grounds: the scarcity of the airwaves and the pervasiveness of the medium. Other explanations–the fact that broadcasters hold licenses from the government, and the importance of empowering democracy–tend to be parasitic on the scarcity rationale. Unfortunately, each of these justifications becomes problematic when applied to questions of violence and indecency.
The most common argument for special content-based regulations of the media is based on the scarcity of the airwaves. The word “scarcity” is poorly chosen. All valuable resources are scarce. The scarcity problem in broadcasting stems from the fact that no two broadcasters can use the same frequency at the same time in the same geographical area, or they will block each other out.(11)But this problem can be dealt with by creating a system of property rights dividing up the airwaves according to frequency, time, place, and broadcasting power; it does not require a system of government licenses.(12) Moreover, the existing system has actually created an artificial scarcity in broadcast television. Many VHF and UHF channels go unused in many localities.(13)
The spread of cable television has increasingly made the scarcity argument implausible. More than half of all American homes now receive cable,(14) and cable television wiring passes by most of the rest.(15) If the government is really interested in reducing scarcity and increasing choices, it should simply subsidize cheap cable television for the remaining households instead of artificially limiting access through the award of broadcasting licenses.
In any case, scarcity is a particularly badly suited justification for content-based regulation of violence and indecency. At best, scarcity provides a reason to put things on the air, not to keep things off. Because airtime is limited, governments may require that stations broadcast certain kinds of public interest programming, like local news or children’s programming; it may also require that candidates for public office have the opportunity to purchase airtime and respond to personal attacks. But limited resources do not justify keeping particular programming off the air if there is otherwise sufficient room for it. To be sure, requiring that some things be on the air will necessarily require broadcasters to leave other things off. But the scarcity rationale does not by itself give the government any right to choose what that forgone programming will be, unless it thinks that scarcity entitles it to dictate the whole of the broadcaster’s day. The justification for keeping indecency off the air cannot be to make room for the presidential debates; it must lie elsewhere.(16)
The other major justification usually offered for special treatment of the broadcast media is that these media are uniquely “pervasive.” Like the term “scarcity,” the term “pervasiveness” is also badly chosen. In fact, courts seem to use the term “pervasive” to stand for a conglomeration of five different sorts of justifications about broadcasting, often not fully distinguished. The broadcast media are pervasive first, because they are the most powerful medium of communication, and second, because they are ubiquitous. Yet the fact that a mode of communication is particularly powerful or ubiquitous is not necessarily a reason for regulating it. That would suggest that the only speech that escapes regulation is that which doesn’t do its job very well.
A third interpretation is that television is “pervasive” because it is constitutive of our culture. We now live in a television culture, in which an increasing number of our cultural allusions are drawn from television. If we are what we eat, then perhaps we are also what we watch. Hence, many people secretly and not so secretly worry that whoever controls television controls culture, and they want to make sure that our culture is not thereby debased. But when the matter is put so starkly, the desire to use government to control culture by controlling what people watch on television cannot be a constitutional justification for the regulation of free expression.
A fourth meaning of “pervasiveness” is a restatement of the captive audience doctrine: government may protect audiences when their privacy is invaded in an intolerable manner by offensive speech they cannot escape. Television is “pervasive” in the sense that there are significant cultural pressures to have a television set and keep it in one’s home. Once television is in the home, it is difficult to protect unwilling listeners from encountering programs they don’t want to watch other than by keeping the television turned off at all times. The captive audience doctrine, it is said, has special force in the home because expectations of privacy are higher there.(17) Although television can be watched outside of the home–in a sports bar, for example–the fact that the overwhelming majority of Americans watch it at home is said to justify content-based regulation that would be impermissible if applied to the print media. However, when applied to adults, the captive audience rationale tends to prove too much. One can also accidentally come across printed material in the home. A newspaper or magazine might have offensive language buried on page fifteen; a videocassette might have offensive language or pictures in a “coming attractions” segment. But the fact that such a “sneak attack” might occur in the home does not justify content-based regulation of print media or videocassettes.
The fifth and final interpretation of “pervasive” is, to my mind, the most important, and the only one that really justifies special content-based regulation for the broadcast media. It is a concern about parental control of children’s viewing habits.(18) Television is pervasive because it is difficult to keep it away from children and children away from it. Once television is in the home, parents must continually supervise what children watch, which is difficult and time-consuming. Many households now own multiple television sets, so that children can watch in the privacy of their own room, away from parental supervision. It is always possible for parents to remove television completely from the home. However, because of television’s cultural importance, many parents do not feel able or willing to deny their children the right to watch television at home, especially when the children can watch it at their friends’ houses.
Although concerns about children make the most sense doctrinally, it’s important to remember that they have little to do with scarcity. Even if there were 500 channels, the problem of parental supervision would still exist, and might even be enhanced. Nor does this justification for regulation turn on the fact that broadcast television is an especially powerful medium of communication, or that it is conveyed in the easily assimilable form of pictures. Parents can watch rented movies on a VCR that are every bit as unacceptable for children as anything one might watch on television. But these movies cannot be regulated in the same way that television broadcasting can.(19)
This final rationale for broadcast regulation is often described as the protection of children, but the real issue is parental control. The two are not necessarily the same. We generally assume that parents love their children and discipline them in ways that are, on the whole, best for them. But parents do not always do so, and we do not second-guess their decisions except in extreme cases. Parents are currently free to bring home R-rated videos full of violence and nudity and let their children watch them. They can subscribe to premium cable channels showing these movies and leave their cable lock boxes unused. If violence and indecency really are bad for children, and we think protection of children is paramount, we should take steps to criminalize such behavior, whether or not parents misguidedly believe such exposure is harmless. Yet I suspect that such proposals would be severely criticized, and not merely by civil libertarians. Most parents do not want the government deciding what is best for their children when the decisions are contrary to their wishes; they want the government to assist them in controlling their children in ways they think appropriate. (20)
In short, behind the slogans of “scarcity” and “pervasiveness” lurks the real issue of parental control. This explains, I think, how current calls for media regulation are tied to the underlying anxieties of the moment. Calls for censorship (which exist at all times) arise most heatedly in moments of great cultural change and uncertainty. After all, where cultural mores are relatively stable, censorship can be achieved informally and without the constraints of law. But we now live in a time of cultural upheaval, caused by significant economic and technological changes as well as changes in mores. Not surprisingly, many people are especially anxious about these changes; they see the world they once knew slipping away. Like the drunk who searches for his keys near the lamppost because the light is better there, people tend to fix upon the mass media as the likely cause of cultural ills and regulation of the mass media as a likely solution.
The First Amendment prohibits relatively direct control over what adults can be exposed to. Hence the focus naturally turns to control of children, who are under their parents’ authority and whom parents see as the natural inheritors and perpetuators of their cultural values. The desire to preserve culture in the face of widespread cultural change (and, in particular, economic and technological change) leads to anxieties over children and the desire to reassert parental control over them.
The problem we face today, however, is that new forms of technology increasingly upset established patterns of parental control. Children can operate VCRs and computers better than their parents. They spend more time in front of the television than at the family dinner table. Technology threatens to render parents’ means of cultural reproduction ineffectual. It is no wonder, then, that new forms of communication technology, whether they be movies, records, radio, television, or the Internet, produce new cultural anxieties and new calls for censorship and control.
II. Thinking about Media in Terms of Filters
It might be best to start over again and think about where the real differences between broadcast and other media lie. I believe that the answer to this question must begin with features that all communications media share in common. All media, I shall argue, whether voice, print, or broadcast, share two features in differing degrees. The first is the ability of the recipient to exclude information; the second is the presence or absence of filtering mechanisms. Filtering and excludability are related to each other, because filtering information usually depends on the present or potential ability to exclude it. (21)
Print media lend themselves easily to filtering precisely because print media are easy to exclude. If I want to avoid the information contained in a newspaper, I can simply avoid buying it. If I go into a bookstore, I can buy the book I want without buying other books. I can take the books I want home and then lock them up so that my children cannot see them. Print media are also easy to select and organize. Because my books are discrete units, I can organize them alphabetically. I can read them when I want and in the order I want.
Filtering mechanisms fall into three basic types or functions–they can organize information (for example, by classifying it), they can select information, or they can block information. Within the last category, one can block information for one’s self or for others (for example, one’s children). All of these functions have important relationships to excludability. Blocking information clearly involves exclusion, but so do selection and organization. To select information, I must be able to take it and not other information. To organize information, I must be able to create categories into which that information (and not other information) falls, and through which that information could (in theory) be selected. (22)
Although I have divided up filters into blocking filters, selecting filters, and organizing filters, these functions substantially overlap. The V-chip is a good example. The V-chip is a blocking filter for children, but it also is a selecting filter for their parents. It lets adults choose whether or not to view violent or indecent material. Equally importantly, the V-chip is an organizing filter, because it creates two types of programming–programming that is blocked by the V-chip and programming that is not. Or, if the V-chip has multiple settings, it creates multiple categories of programming.
Installing an informational filter simultaneously raises and lowers pre-existing costs of searching for, blocking and receiving information. A V-chip raises costs of receiving information substantially to children, and moderately to adults. But it also lowers costs as well, because it subdivides the body of programming and makes certain types of choices easier.
Throughout this essay I shall speak in terms of “informational” filters. Nevertheless, because of its computer-age connotations, the term “information” is likely to be understood much too narrowly. Many people associate information with statements of fact, or with strings of ones and zeros that can be read by a computer. Yet much of what people want to and do filter out is not information in that limited sense. They want to filter out dirty language, violence, and nudity. They also want to filter out dangerous ideas and views they do not agree with or expressions that offend and anger them. “Information,” in the broader sense I am concerned with, is cultural information. It does not consist merely of statements of fact, but includes anything that can be understood by someone in a culture, and have a corresponding effect on their reason, emotions, or behavior. Cultural information is the counterpart of cultural understanding. Hence it is involved in not only the production of knowledge or ignorance, but also persuasion or offense, refinement or coarsening, ennoblement or corruption.
Because there is too much information in the world, all communications media produce attempts at filtering by their audiences.(23) The desire to filter is not, however, always matched by available methods of filtering. Each medium offers different means of exclusion, and different costs of exclusion. Filtering is an effective strategy precisely to the extent that excludability is possible and cost-effective. Each medium’s ability to permit exclusion determines and limits the kinds of filtering that are available to it. Even when filtering is used to organize information or facilitate selection, it is still limited by the possibilities of exclusion characteristic of the medium. An example from the print media may illustrate this point. Suppose that Time magazine started publishing lots of four-letter words and sexually suggestive pictures. Eventually Time magazine would get a reputation as the sort of magazine that does that sort of thing. Parents would, after a time, discover this. Some of them would cancel their subscriptions to Time magazine; others would not keep it lying around the house. Advertisers would also notice the change. They would discover that the demographics of the readership had changed and would shift their money accordingly. A magazine’s reputation can act as a kind of filtering device, although it is social rather than technological. It signals the likely content of the magazine. Many filters work by offering signals to the audience. Examples are titles of books and headers in the delivery of e-mail. Nevertheless, Time’s new reputation would have little practical effect as a filtering device unless parents could exclude it by refusing to buy it or by not bringing it into their homes. If they could not act on Time magazine’s new reputation by excluding it, the use of reputation as an informational filter would do little good, other than perhaps to warn parents to discount what they read in the magazine (which is itself a filter of a different sort). (24)
III. Filtering and Choice
At first glance, filtering seems to overlap with a much more familiar concept–choice. But the two ideas are distinct in important ways. Not all informational filtering involves conscious or deliberate choice by an audience. Indeed, the importance of filters consists precisely in the ways that they obviate or skew choice even as they enable it. “Choice” is a word with largely positive connotations of personal responsibility, respect for individual intelligence and protection of personal autonomy. Filtering, on the other hand, is morally ambiguous; it may make little demands on individual intelligence, may involve considerable surrender of personal responsibility and may actually undermine personal autonomy. Filtering, especially in the Information Age, increasingly involves delegation of choice to another party. Thus, it is very important not to collapse filtering into choice, thereby absorbing the latter’s positive moral connotations.
Let me give an example drawn from my own experience as a legal academic. Currently there is more literature being published in law and related academic fields than any person can possibly keep up with. Hence I and many other legal academics make use of filters. One is a periodicals list. Another is searching on data- bases like LEXIS and Westlaw. The periodicals list gives me the titles of articles in different law reviews. This filter is widely distributed in identical form to many academics; I do not receive a version tailored to my specific needs. By contrast, an informational filter like a LEXIS or Westlaw database is partially modifiable by the use of search terms.
Both the periodicals list and the computer databases already filter out publications even before they offer me possibilities for choice. For example, I do not have a choice about what law-related journals to include or exclude. In LEXIS and Westlaw databases, I am limited to the journals that are currently on-line and the databases’ selected periods of coverage. They will show me nothing published before 1982, for example. Recently a fellow law professor wrote me asking for a cite to an article I wrote in 1990. She could not find the article on LEXIS because the article was published in Cardozo Law Review and LEXIS’ coverage of that review begins in 1994. Any articles written before that time do not appear on the database. My colleague depended heavily on LEXIS because it was easy to use; looking outside of it took considerable effort.
The LEXIS database has an interesting effect on the cost of obtaining and filtering information. It lowers the costs of searching for materials if one uses the database, while the costs of more traditional hard-copy searches remain constant, at least in the short run. (25)Furthermore, if one shifts to LEXIS as a primary research tool, certain types of filtering choices (i.e., searches) become easier and less expensive to make than others, even if the latter choices would be easier or less expensive using a different filtering system.(26)
This example demonstrates one of the important side effects of informational filters. If everyone uses LEXIS to do basic legal research, articles that do not meet LEXIS’ selection criteria will increasingly disappear from view, because the filter changes the differential costs of searching for and receiving certain kinds of information. And this example suggests a larger point. The structure and content of public communication can be and often is shaped by the informational filters people most commonly use and depend on.
I use filters like the periodicals list and LEXIS and Westlaw because they have definite advantages. I do exercise choice in using them. Yet my choice is at the same time limited. In using a particular filter, I have delegated choice to some other entity–in this case, the people who put together the periodicals list, and the people who run LEXIS and Westlaw. I hope that they know what they are doing, and that, over time, they will include most of the journals I might want to read. But they might not, and, as a result, my choices may be limited or skewed without my even knowing it.
Literary critics have always known about filters. They call them canons and anthologies. Canons and anthologies are special kinds of filters that involve special forms of delegation. People who construct canons and anthologies decide what is important to read, and, by implication, what is less important to read. Canons and anthologies can be very helpful filters. They introduce people to the works most worth reading or most often discussed in the academic literature. In this way they can enable not only choice, but also the search for truth. But as repeated debates over canonicity have shown, canons and anthologies can also skew or inhibit these values.
More generally, the marketplace produces any number of informational filters. Book publishers screen manuscripts to determine which ones are most likely to be worth reading. Bookstores stock, classify, and sell books by category and likely readership interest. Magazines specialize in particular kinds of stories and particular political approaches, and the public can use their reputations as informational filters.
All of these examples involve different filters that work in different ways, but each filter involves some form of delegation. When many people need to filter the same body of information, there are considerable efficiencies in delegating that task to someone else. The need for filtering gives rise to people who provide that service, either through market demand, through social custom, or through governmental regulation. Filtering and delegation thus go hand in hand. And because increasing amounts of information inevitably lead to the need for filtering, they inevitably lead to the need for delegation. This gives people to whom we delegate the construction of informational filters an important degree of power. It is a necessary power caused by the limited space in our minds and limited time available for absorption, as well as the positive need to block harmful, useless or offensive information. The power of delegation is, if anything, enhanced in an age of exploding information. We must–and do–trust and rely on delegations to filterers to give a relatively appropriate picture of the world. For it is the picture of the world we get through informational filters that will largely determine whether we think that the people we have delegated this power to are, in fact, doing their jobs properly. There is something ironic about this. In the Information Age, we were told, information would be power. It is turning out to be quite the opposite. In the Information Age, it seems, power does not rest with those who have access to information. It rests with those who filter it.
1. Lafayette S. Foster Professor, Yale Law School. My thanks to Owen Fiss and Scot Powe for their comments on a previous draft, and to Michael Adler for his research assistance. A slightly different version of this essay was originally published as “Media Filters, The V-Chip, and the Foundations of Broadcast Regulation, 45 Duke L. J. 1131 (1996). It appears with the permission of the Duke Law Journal.
3. section 551, 110 Stat. at 139-42. As its name implies, the V-chip technology has been touted primarily as a means of controlling television violence. But its uses are not limited to that category. The Telecommunications Act of 1996 specifically lists its concerns as “sexual, violent, or other indecent material.” section 551(b)(1), 110 Stat. at 140.
5. See TV Parental Guidelines, http://www.tvguidelines.org/#Children.
6. See Canadians Propose 7-Point TV Ratings System, Communications Daily, May 6, 1997; Etan Vlessing, Canada Aims at TV Violence: Nation will go light on language, sex in its ratings system, The Hollywood Reporter, May 6, 1997.
7. Paige Albiniak, Ratings get revamped: networks, except for NBC, agree to add content labels; includes related articles on the rating system and the V-chip, Broadcasting and Cable Vol. 127 (July 14, 1997), at 4.
9. Action for Children’s Television v. FCC (ACT III), 58 F.3d 654 (D.C. Cir. 1995) (en banc), cert. denied, 116 S. Ct. 701 (1996). The relevant law actually permitted stations to broadcast indecent programming from midnight to 6 a.m., but it also permitted public television stations that go off the air before midnight to broadcast the same programming starting at 10 p.m. Public Telecommunications Act of 1992, Pub. L. No. 102-356, section 16(a), 106 Stat. 949, 954 (codified at 47 U.S.C. section 303 note (Supp. V 1993) (Broadcasting of Indecent Programming; FCC Regulations). Because the court found the exception for public broadcasting to undermine the purposes of the legislation, it remanded the case to the FCC with instructions to limit the ban on indecent programming to the period from 6 a.m. to 10 p.m. ACT III, 58 F.3d at 669-670.
12. Critiques of the scarcity rationale are by now legion. For a sampling, see Lucas A. Powe, Jr., American Broadcasting and the First Amendment 197-209 (1987); Matthew L. Spitzer, Seven Dirty Words and Six Other Stories 1013-20 (1986); Ronald H. Coase, The Federal Communications Commission, J.L. & Econ., Oct. 1959, at 1, 12-27.
14. See H.R. Conf. Rep. No. 862, 102d Cong., 2d Sess. 56 (1992) (stating that near- ly 56 million households and more than 60% of all households with televisions are cable subscribers); Robert S. Tanner, Note, The Data Transfer Industry: Communications Regu- lation for the Next Century, 17 Hastings Comm. & Ent. L.J. 917, 922-23 (1995) (citing Chesapeake & Potomac Tel. Co. v. United States, 830 F. Supp. 909, 915 (E.D. Va. 1993)).
15. Leland L. Johnson, Toward Competition in Cable Television 179 (1992) (“Cable systems have become accessible to more than 95 percent of the nation’s homes.”); see also US West, Inc. v. United States, 855 F. Supp. 1184, 1192 (W.D. Wash. 1994) (“Cable service is now available … at 96% of all U.S. homes … .”), aff’d, 48 F.3d 1092 (9th Cir. 1994), cert. granted and judgment vacated, 116 S. Ct. 1037 (1996).
16. Many other justifications for regulation of violence or indecency often tend to be parasitic on the scarcity rationale. For example, in Red Lion Broadcasting Co. v. FCC, 395 U.S. 367 (1969), the Supreme Court suggested that content-based regulation was permissible because broadcasters do not own the airwaves outright. Id. at 394. They hold licenses from the government, and therefore the government can impose conditions on that license. By itself, this argument tends to prove too much. The government’s conditions may be unconstitutional conditions. The government does not license the airwaves as an act of governmental largesse–the usual means of justifying conditions on licenses. See, e.g., Rust v. Sullivan, 500 U.S. 173, 199 n.5 (1991) (stating that restrictions on abortion counseling by recipients of Title X subsidies do not violate the First Amendment because the subsidy may be declined). Rather, the licensing scheme exists because the government decided to take complete control of the airwaves and parcel out licenses instead of auctioning off rights to broadcast at certain times in certain locations and with certain degrees of broadcast strength. The government does not license the manufacture and distribution of paper or printing presses. Even if it did so, it could not constitutionally justify imposing content-based conditions on their use. Thus, the conditionson-licensing justification ultimately rests on the prior justifications for licensing, which depend in turn upon the scarcity rationale.
FCC Chairman Reed Hundt has suggested that restrictions on violence and indecency on television may be justified by the fact that the First Amendment is designed to protect democracy. Reed E. Hundt, The Public’s Airwaves: What Does the Public Interest Require of Television Broadcasters?, 45 Duke L.J. 1089, 1097 (1996). He argues that violent and indecent programs were not what James Madison had in mind when he wrote the First Amendment; they do nothing to promote discussion of public issues. See id. at 1126. This seems to conflate an argument that some speech is of lesser constitutional value with an argument that the broadcast media are special. Assuming that Chairman Hundt’s claim about degrees of constitutional value is sound, it applies equally well to violent and indecent depictions in movies and the print media. It cannot by itself distinguish broadcast media from other media.
In any case, the argument tends to prove too much: A great deal of non-indecent and nonviolent programming on television has only the faintest relationship to promoting democracy. Yet it does not follow that this programming is subject to content-based regulation because it is also of low constitutional value.
In fact, the argument from democracy is best viewed as an adjunct to the scarcity argument. Because airtime is scarce, television must make room for programming that enhances democratic values. But again, this argument does not justify keeping indecency or violence off the air; rather, it justifies keeping public interest programming on.
19. With a few exceptions–zoning regulations for adult movie theaters and procedures for pre-screening obscene films–the regulation of movies is largely along the lines of the print model. See, e.g., City of Renton v. Playtime Theatres, 475 U.S. 41 (1986) (upholding zoning requirements for adult movie theaters), reh’g denied, 475 U.S. 1132 (1986); Young v. American Mini Theatres, 427 U.S. 50 (1976) (same), reh’g denied, 429 U.S. 873 (1976); Freedman v. Maryland, 380 U.S. 51 (1965) (establishing procedures for reviewing obscene films).
20. The D.C. Circuit recognized that parental control and protection of children were separate interests, but it did not acknowledge the degree to which they might conflict in practice. ACT III, 58 F.3d at 660-61. In fact, safe harbor rules do not perfectly mesh with the goals of enhancing parental control. As Chief Judge Harry Edwards has pointed out, safe harbor rules actually preempt some parental choice, because children cannot watch certain programming (for example, a documentary on AIDS prevention) even if parents want them to. Alliance for Community Media v. FCC, 56 F.3d 105, 145-46 (D.C. Cir.) (en banc) (Edwards, C.J., dissenting), aff’d in part and reversed in part, sub nom. Denver Area Education Telecom. Consortium v. FCC, 116 S.Ct. 2374 (1996); ACT III, 58 F.3d at 670 (Edwards, C.J., dissenting).
21. When we think of excludability, we think of captive audiences, and when we think of captive audiences, we think of justified expectations of privacy. Although excludability and privacy are related concepts, they are related in complicated ways. Practical excludability does not by itself determine whether we have a justified expectation of privacy. On the one hand, expectations of privacy seem to be based in part on the prac- tical possibility of exclusion. Because one can be assaulted by billboards or by voices in the street, one’s justified expectations of privacy are lower there. On the other hand, expectations of privacy are sometimes thought justifiable whether or not excludability is practically possible. Even though we may not be able to prevent electronic eavesdrop- ping, we have a justified expectation that others will not eavesdrop.
22. Thus, in alphabetizing a list, I make selection by first letters possible. Placing books in the Library of Congress cataloguing system enables readers to find books on some subjects without having to look at others.
Some forms of organization have a more attenuated relationship to excludability. Suppose I merely imagine a system for cataloguing library books. Although this system organizes information, the organization is not implemented. The imaginary system merely gives directions for implementation, which if enacted, would require selection, and hence the possibility of exclusion. Finally, purely mental organization of information does not involve physical selection and exclusion, but it does involve mental selection and exclusion.
23. The desire by producers of information to gain an audience and the contrary desire by recipients to block out unwanted information creates an arms race between filters and ways of getting around filters; hence the use of flashier graphics, louder music, and increasingly hyperbolic claims to attract people and gain a slice of their increasingly valuable and limited attention spans. Audiences use remote controls to flip through channels during commercials; advertisers respond by varying the content, length, and timing of their advertisements, and broadcasters respond by varying the length of time between shows. Marcia Mogelonsky, Coping With Channel Surfers, Am. Demographics, Dec. 1995, at 13. Occasionally methods of filtering and methods of evading filters start to merge. For example, best seller lists are both means for weeding out books that might be interesting to read and means by which publishers promote books as worth reading. Kudos on the backs of books also have this dual character.
24. This example also suggests that filtering is a heuristic device for dealing with information, and as a heuristic device, it is usually imprecise. For example, I may use Time magazine’s new reputation as a reason not to bring it into my home. There may be lots of things in the new Time magazine that are just like the old Time maga- zine–useful news stories, for example. But the overall reputation, and not the precise content of the stories, will determine whether parents subscribe and what kinds of adver- tising revenues the magazine generates.
26. Law librarians have long understood that the shift to computer databases alters the way legal research is conducted. See, e.g., James A. Sprowl, A Manual For Computer-Assisted Legal Research 14-15 (1976).