Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Prepublication copy - subject to further editorial correction 14-26 /
14 Findings, Conclusions, and Future Needs 14.1 FRAMING THE ISSUE The Internet has enormous potential to contribute to public welfare and private well-being. One dimension of that potential involves the use of the Internet to enhance and transform education for the nation's youth, and many public policy decisions have been taken to provide Internet access for educational purposes. Easy access to the Internet (and related online services) has many advantages for children access to educational materials; collaborative projects, publications, online friendships, and pen pals; access to subject matter experts; recreation, hobby, and sports infor- mation; and so on. While such potential for contributing to the nation's welfare in gen- eral and to the education of its children in particular is recognized, the Internet also presents to the public a wide variety of concerns. This fact in itself should not be surprising few powerful and widely deployed tech- nologies have been used solely for socially beneficial purposes. But the Internet poses many challenges for which there are no precedents, and much of the controversy about inappropriate sexually explicit material on the Internet arises because of these differences. 14.1.1 Social Dimensions What is the issue to be addressed? Although the nominal title of the project was "Tools and Strategies to Protect Kids from Pornography on 357
358 YOUTH, PORNOGRAPHY, AND THE INTERNET the Internet and Other Inappropriate Material," a key fact is that "pornog- raphy" is a term whose meaning is not well specified. People neverthe- less use the term as though it did have a well-specified meaning, and they often fail to recognize that what one may consider pornographic, another may not. For this reason, the committee chose the term "inappropriate sexually explicit material" when in common parlance it might have used the term "pornography." Using the former term keeps in the foreground the question of "inappropriate according to whose standards?" Internet exposure of children to sexually explicit material is only one dimension of exposure, albeit important, because sexually explicit mate- rial and other sexual content exist in a wide variety of other commonly accessible media such as video cassettes, magazines, and cable television. Further, concerns over obscenity may well be a proxy for the desire to suppress access to other sexually explicit or sexually oriented content that would not be judged legally obscene. Internet exposure of children to inappropriate sexually explicit mate- rial is also only one dimension of inappropriate or potentially dangerous activities in which youth may engage. The Internet is also a medium that can facilitate face-to-face meetings between people who do not know each other prior to their Internet contact, and when there is a great disparity of experience and age between these parties, the younger less-experienced person could be more subject to exploitation and physical danger. Other types of material may also be judged by various parties to be inappropri- ate for children. Some of the approaches to protection from sexually explicit material may be applicable to such other material. The views of people about "pornography" on the Internet and what to do about it reflect a broad range of values and moral commitments. What is pornographic to some people may be simply mainstream advertising to others; what is morally wrong to some may be entirely acceptable to others; what is legal to show to minors in one community may be regarded as wholly inappropriate by those in another community; and what counts as responsible choice according to one set of values may be irresponsible be- havior according to a different set of values. Approaches taken to protect children should be flexible enough to honor that diversity. 14.1.2 Developmental Dimensions Children from birth to the age of legal majority pass through a wide range of developmental stages as they mature into adults (and further- more the age of legal majority is not statutorily uniform). The impact of any given piece of sexually explicit material is likely to vary widely with age or, more importantly, level of maturity, and the approaches taken to
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 359 protect children of a given maturity level should take into account the characteristics of their level of maturity. Moreover, the experiences of individuals can influence how sexual content affects them, especially con- sidering that increasing numbers of adolescents, who are still legally mi- nors, are sexually active. Finally, age usually affects the extent to which children can understand dangers and engage in safe behavior. The information needs of children that the Internet can and should meet also change with the developmental stage of the child in question. For example, juniors and seniors in high school have a much broader range of information needs (i.e., for doing research related to their educa- tion) than do those in the third grade or in junior high school. This, in turn, leads to the question of how to provide older children with access to a broader range of material while preventing younger ones from access- ing material that is not deemed appropriate given their developmental level. 14.1.3 Legal Dimensions As a matter of law, sexually explicit material that is "obscene with respect to minors" must be made available to adults without restriction, though it can be restricted for minors. Certain other sexually explicit materials (obscenity, child pornography) enjoy no First Amendment pro- tection at all. Material that is determined to be obscene or obscene with respect to minors must pass certain tests, including tests related to com- munity standards. For both classes of material, the community standards for making such determinations likely change over time, and in recent years, mores about sex and consumption of sexually explicit material may have changed in such a way as to reduce (but not to eliminate) the scope of both categories. Thus, there is in practice considerable ambiguity about what should fall into these categories, and the fact that community standards are inte- gral to the application of the law in this area means that material cannot be determined to be obscene or obscene with respect to minors solely on the basis of the material itself. Over the past decade, the number of federal obscenity prosecutions has been very small compared with those in previous years, thus complicating to a significant degree the concept of "community standards." The First Amendment is relevant regarding the extent to which and circumstances under which public institutions of various types can restrict access to particular types of information. Finally, in the public policy domain, U.S. regulation of sexually ex- plicit material is most likely to have an effect on commercial sources inside the United States, and far less effect on sources located abroad.
360 YOUTH, PORNOGRAPHY, AND THE INTERNET 14.1.4 Technical Dimensions Making some material available to adults but not to children requires that providers have a reasonably reliable way of differentiating between them. In the physical world, such differentiation can often be accom- plished with reasonable ease (e.g., by checking a driver's license or other identification). But in the Internet context, rules based on age differentia- tion are highly problematic and technically difficult to enforce. Content providers must also have a clear understanding of the difference between material that is and is not inappropriate for children. Although many of the issues concerning Internet access to various types of material that may be regarded as inappropriate arise for other media as well, the Internet changes significantly the convenience and anonymity of access, thus reducing certain constraints that may be opera- tive in other media. For example, online chat rooms and instant messages (IMs) have few analogs in the physical world, and these are channels through which a great deal of communication between strangers can oc- cur. For this reason, special attention to the Internet dimensions of the issue may be warranted. The adult online industry notwithstanding, inappropriate sexually explicit material is available from many non-commercial online sources. Thus, approaches that focus primarily on access to inappropriate sexually explicit material provided by the adult online industry (widely seen as the crux of today's problem) are likely to have limited relevance to prob- lems arising from non-commercial sources. For a great deal of inappropriate sexually explicit material (specifi- cally, material accessible through Web sites), a reduction of the number of Web sites containing such material, in and of itself, is not likely to reduce the exposure of children to such material. The reason is that a primary method for obtaining access to such material is through search engines, and the likelihood that a search will find some inappropriate material for a given set of search parameters is essentially independent of the number of Web pages represented in that search. That said, if the number of such Web sites is small enough that no Web site operator can flout the rules of respon- sible behavior with impunity,] regulation of their behavior (through public tin this context, responsible behavior refers to actions taken to reduce the likelihood that children will obtain access to inappropriate sexually explicit material. To illustrate, one method of inducing Web site operators to act responsibly is to establish codes of behavior to which they must adhere under pain of government enforcement actions (whether civil or criminal). By definition, such an approach requires government action, and with a plethora of operators, the likelihood of being the target of government action is very small hence the number of operators must be reduced to a "sufficiently small" number. A second illustration of inducing Web site operators to act responsibly is to create disincentives for
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 361 policy and/or self-regulatory approaches) becomes significantly easier, and enforceable codes of responsible behavior can have a significant im- pact on the extent to which operators of Web sites that contain adult- oriented, sexually explicit material make their products and services ac- cessible to children. 14.1.5 Economic Dimensions The adult online industry is one of the primary sources of sexually explicit images (e.g., on "teaser" home pages) that are accessible without any attempt to differentiate between adults and children. Such teaser pages allow potential customers to sample what would be available with payment, but children have easy access to the free content. The sexually explicit material provided by the adult online industry is available to children through a variety of routes, including mistyped Web site ad- dresses, links returned by search engines in response to search terms with sexual connotations, and spam containing links to adult Web sites. The revenue models of the adult online industry suggest that broad exposure is needed to attract potential customers, and so the industry engages in tactics that seek to generate the broadest possible audience. Moreover, these tactics to gain exposure cannot be used at low cost if they are to differentiate between adults and children. The result is that chil- dren can be "swept up" in the industry's reach for larger audiences of potentially paying customers. The adult online industry is only one component of supply. The low cost of creating and maintaining a Web site means that the production of sexually explicit material is now within the financial reach of almost any- one. For example, Web cameras can be purchased for under $100, en- abling anyone so inclined to produce a video stream of sexually explicit material. In the Internet environment, an astronomically large volume of mate- rial is available for free, including art, literature, science, advertising, and irresponsible behavior. In this method, the key is to associate disincentives with a large number of parties so that irresponsible Web site operators will feel the pressure of those disincentives, for example, by establishing causes of action allowing those affected by irre- sponsible behavior to take action against such operators. (One example in a different do- main is the establishment of liability (and an associated bounty) for junk faxes, an action that dramatically reduced the number of such faxes.) Whether or not these or other actions can in fact reduce the number of Web sites to a "sufficiently small" number is an open question, especially in a context in which U.S. actions are unlikely to affect Web sites oper- ated by foreigners. Note that these illustrations are just that illustrations and their inclu- sion in the report is not intended to signal endorsement or rejection by the committee.
362 YOUTH, PORNOGRAPHY, AND THE INTERNET government information, as well as sexually explicit material of every vari- ety. Restricting what any individual may access (or protecting him or her from certain kinds of material) will inevitably impose additional costs on users. Such costs may include denial of access to useful information and loss of privacy for those wishing to access certain kinds of information. 14.2 ON THE IMPACT ON CHILDREN OF EXPOSURE TO SEXUALLY EXPLICIT MATERIAL AND EXPERIENCES As described in Chapter 6, factors such as certain ethical and legal considerations, an increasing conservatism of university review boards that approve research studies involving human subjects (institutional re- view boards), and a lack of research funding have contributed to a pau- city of research regarding the impact on children of exposure to sexually explicit material. Furthermore, the extant scientific literature does not support a scientific consensus on a claim that exposure to sexually explicit material does or does not have a negative impact on children, and there is no adequate research base for understanding the impact of sexu- ally explicit material of various kinds and how different approaches to protection may vary in effectiveness and outcome. It is important to consider why many young people search for adult- oriented sexually explicit material in the first place. Adolescents go to these sites for many of the same reasons that adults do. Human beings are sexual. Sexuality is a part of identity, and a facet of identity that is a focus during adolescence when youth come of reproductive age. It is not surprising that many children especially preadolescents and older are curious about sex, and adolescents who are sexually mature are looking for information about sex and are making choices in this arena. In other earlier eras, they might well be married, but today in Western culture marriage among those in their early and mid-teens is frowned upon. To the extent that adults (parents and families, schools, libraries) provide accurate information and guidance about sexuality in its biological, psy- chological, emotional, and social dimensions and information and guid- ance that is responsive to the situations that their children are facing it can be argued that young people will be less drawn to searching for adult- oriented sexually explicit material. This is not to say that parents are wrong to be concerned about their children's exposure to sexually explicit material. There is no reason to suppose that all negative impacts from exposure are necessarily shown or manifested in science-based research studies. The moral and ethical val- ues of parents whether or not religious in orientation and a desire to be involved in providing context and guidance for a child exposed to such material are important and understandable drivers of such concerns.
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 363 The committee believes that it would not be difficult to come to a consensus on the undesirability of some set of sexually explicit material involving depictions of extreme sexual behavior. That is, such a set could be developed by construction image after image could be shown to a group of individuals drawn from a broad cross section of the community. (In some ways, the committee consists of just such a group.) Under this procedure, the images that everyone on the committee deemed inappro- priate for children would constitute the set and the set would be sub- stantial in size. Such a consensus would not be based so much on scien- tific grounds (as the committee knows of no reliable scientific studies that address this point) as much on a sense that such exposure would offend its collective moral and ethical sensibilities. Furthermore, the committee believes that a significant fraction of this set would likely be deemed obscene if prosecuted. Yet, the fact that such a set could be defined by construction does not mean that it is possible to craft unambiguous rules that define this set without capturing material that would either be protected speech under existing First Amendment precedents or unobjectionable to some number of group members. And, in the absence of such rules, disagreement is inevitable over what else other than "similar" material should be cap- tured in any definition. The story is quite different for child pornography. In contrast to the diversity of views about what material must count as obscenity or ob- scene with respect to minors (and hence a diversity of views on what harm might result to children from being exposed to such material), there is a much broader social consensus that child pornography results in harm to the children depicted in such images and that child pornography is morally wrong as well.2 Over the past decade, the incidence of child pornography has risen as new communications channels such as the In- ternet have facilitated the exchange of child pornography. A similar argument applies to sexual predation. By design, the Inter- net facilitates contact between people who do not know each other. While much that is good and valuable and safe can come from interactions with strangers, parents rightly have some concern when their children talk to strangers in an unsupervised manner. These concerns arise in the physi- cal world, and they are magnified in the online environment where the range of personality types and intentions is both less known and less 2The social consensus is strongest when children are used to create sexual imagery. How- ever, the breadth of the legal definition of child pornography has also led to the attempted prosecution of works of art that involve children in various states of nudity (e.g., the works of Jock Sturges), and it is fair to say that there is less of a social consensus around such material.
364 YOUTH, PORNOGRAPHY, AND THE INTERNET controllable. Further, the Internet has enabled potential predators to seek out a wider range of vulnerable children. The committee believes that the issue of face-to-face meetings be- tween children and their Internet acquaintances is very different from that of being exposed to inappropriate material on the Internet because the potential dangers that face-to-face meetings entail are much greater. Furthermore, while the majority of children report that they brush off aggressive solicitation encounters or treat them as a relatively minor an- noyance, a significant minority do report being upset or disturbed by them (see Section 5.4.3~. In addition, even when children were distressed by such encounters, a large fraction of them did not report the incident to parents or other authorities. Finally, the committee believes that there is a consensus regarding involuntary exposure to sexually explicit material. Regardless of one's views on the impact of voluntary exposure to sexually explicit material, the committee believes that there is a reasonably strong consensus in- deed, one reflected in its own deliberations that involuntary Internet exposure to sexually explicit material is inappropriate and undesirable and should not be occurring, and it is particularly inappropriate and undesirable in the context of minors being exposed to such material.3 14.3 ON APPROACHES TO PROTECTION Much of the debate about "pornography on the Internet" focuses on the advantages and disadvantages of technical and public policy solu- tions.4 Technology solutions seem to offer quick and inexpensive fixes that allow adult caregivers to believe that the problem has been addressed, and it is tempting to believe that the use of technology can drastically reduce or even eliminate the need for human supervision. Public policy approaches promise to eliminate sources of the problem. In the committee's view, this focus is misguided: neither technology nor public policy alone can provide a complete or even a nearly com- plete solution. As a rule, public policy aimed at eliminating sources of sexually explicit material can affect only indigenous domestic sources, 3Are there any circumstances under which involuntary exposure might be beneficial? Perhaps. Consider a situation in which discussions about sex made a child uncomfortable. It might still be a reasonable thing for a concerned parent to have a conversation about sex with his or her child. Needless to say, this kind of situation does not occur frequently in the context of Internet media. 4The discussion in this section "section 14.3' is complementary to the findings and gen- eral observations in Chapters 8 through 13, but does not repeat them systematically. Read- ers are urged to consult those chapters for more specific findings especially about technol- ogy-based tools such as filters and monitoring programs.
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 365 and a substantial fraction of such material originates overseas. Nor is technology a substitute for education, responsible adult supervision, and ethical Internet use. For these reasons, the most important finding of the committee is that developing in children and youth an ethic of responsible choice and skills for appropriate behavior is foundational for all efforts to protect them- with respect to inappropriate sexually explicit material on the Internet as well as many other dangers on the Internet and in the physical world. Social and educational strategies are central to such development, but technology and public policy are important as well and the three can act together to reinforce each other's value. Social and educational strategies are a primary focus of the committee because most children are likely to be confronted, on occasion, with mate- rial that they or their parents regard as inappropriate, or find them- selves in online situations that are potentially dangerous. Parents must balance their concerns about exposure to harmful things on the Internet against the benefits gained from exposure to positive things on the Inter- net, and the question of how children can learn to handle and defend themselves becomes the primary issue. Social and educational strategies that promote and teach responsible decision making are at the core of such defense. Social and educational strategies are also important for teaching chil- dren how to recognize and avoid situations that might expose them to inappropriate material or experiences. Though technology has a role to play here as well, developing "street smarts" about how to avoid trouble is likely to be a far more reliable and robust approach to protection. In short, a child who responsibly chooses appropriate materials to access and appropriate things to do on the Internet and who knows what do to about inappropriate materials and experiences should he or she come across them is much safer than a child whose parents and school teachers rely primarily on technology and public policy to solve the prob- lem for them. Moreover, social and educational strategies to promote and teach responsible choice have applicability far beyond the limited ques- tion of "protecting kids from porn on the Internet," because they are relevant to teaching children to think critically about media messages, to conduct effective Internet searches for information and to navigate with confidence, and to evaluate the credibility of the information they receive. Social and educational strategies are not quick or inexpensive, and they require tending and implementation. Adults must be trained to teach children how to make good choices on the Internet. They must be willing to engage in sometimes-difficult conversations. And, because social and educational strategies place control in the hands of the youth targeted, children may make mistakes as they learn to internalize the
366 YOUTH, PORNOGRAPHY, AND THE INTERNET object of these lessons. However, by understanding why certain actions were mistakes, children will more effectively learn the lessons that par- ents and other adults hope that they will learn. Virtually all of the high school students to whom the committee spoke said that their "Internet savvy" came from experience, and they simply learned to cope with cer- tain unpleasant Internet experiences. They also spoke of passing their newfound expertise down to younger siblings, hence becoming the new de facto educators for younger kids in the "second wave of digital children." Technology-based tools, such as filters, can provide parents and other responsible adults with additional choices as to how best to fulfill their responsibilities. Though even the most enthusiastic technology vendors acknowledge that their technologies are not perfect and that supervision and education are necessary when technology fails, tools need not be perfect to be helpful and used properly (an important caveat), they can be an important aspect of comprehensive programs for Internet safety. First, technology can help to create a child-rearing environment that parents can moderate and shape according to their values and the matu- rity of their children. In an Internet context, a controlled and moderated environment does not mean that a child's every keystroke and mouse click are preprogrammed forever. But it does mean that the child can exercise choices within safe limits set by parents or other responsible adults and as a child learns more and develops greater maturity, those limits can be expanded in ways that allow greater freedom of choice and that may at some point entail greater risk taking as well. An example of a technology-moderated environment for Internet ac- cess might call for the provision of Internet access in young childhood that is explicitly limited to child-friendly content (i.e., access based on "white lists"), strongly filtered Internet access in middle childhood (i.e., access based on extensive "black lists"), less filtered Internet access in preadolescence and early adolescence (i.e., access based on a slow reduc- tion in the number of categories deemed inappropriate), and monitored Internet access in middle to late adolescence (i.e., unfettered Internet ac- cess but accompanied by warnings about inappropriate material). Par- ents wishing to provide a more risk-free environment might delay the introduction of less restrictive measures; those wishing to promote a more free flow of information to their children might accelerate the introduc- tion of less restrictive measures. Second, technology can help to keep parents and other responsible adults informed about what their children are doing online. Of course, the circumstances of obtaining such information matter quite a bit. An intent to provide guidance that helps the child make informed and re- sponsible choices relies on the presence of openness about the presence of
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 367 monitoring. On the other hand, an intent to "catch" the child in doing something wrong is likely to result in behavior that simply reduces the flow of information to the parent, such as the child obtaining Internet access in a venue that is not monitored. Third, technology offers many desirable benefits: speed, scalability, standardization, and reduced cost. Because non-technical approaches take valuable time from parents, teachers, and others that is badly needed to address many other issues related to raising responsible young people, it will often be the case that a mix of non-technical strategies and technol- ogy-based tools provides the most cost-effective way to protect children on the Internet. Choosing the right combination of social and educational strategies and technology-based tools depends a great deal on the nature of the problem that parents, teachers, and librarians are trying to solve. For example, recall from Chapter 8 that deliberate access and inadvertent exposure to inappropriate sexually explicit materials pose different pro- tection problems. Recognizing that responsible adults have to deal with both problems, the fact remains that certain tools and certain strategies are more appropriate for the former than the latter, and vice versa. Tools that warn of impending exposure to inappropriate material rather than blocking are better suited to dealing with the problem of inadvertent exposure. One important point arising from the relative preponderance of males relative to females as consumers of adult-oriented sexually explicit material (Chapter 3) is that adolescent males are probably much more likely to seek out such material deliberately. If so, social and educational strategies that aim at reducing the desire of such individuals to seek out such material may well be more relevant to male children than to female children. As for the role that public policy can play in protecting children, regulation may help to shape the environment in which these strategies and tools are used by reducing at least to some extent the availability of inappropriate sexually explicit material on the Internet. Public policy can help to influence the adult online industry to take actions that better deny children's access to their material and/or influence others in the private sector to support self-regulatory or self-help measures. Furthermore, through Prosecution of violators of existing laws that prohibit the trans- tJ 1 mission of obscene material, public policy can help to some extent to reduce the number of providers of such materials. Successful law enforcement depends on many factors, including ap- propriately formulated statutes, adequate resources, and a willingness to enforce existing law and regulation. In the Internet safety arena, the participation of citizens (e.g., reporting illegal activity) is also an essential element of law enforcement. Proper training of law enforcement person-
368 YOUTH, PORNOGRAPHY, AND THE INTERNET net including those who take complaints, those who investigate com- plaints, and those who prosecute cases at federal, state, and local levels remains critical to effective law enforcement. Finally, public policy (or the threat of regulation) can also encourage and promote self-regulatory efforts that contribute to certain public policy goals, such as Internet ser- vice providers taking down materials posted in violation of the terms of service to which users agree as a condition of use. Coping with non-commercial sources presents different issues to law enforcement authorities. In order to attract customers, commercial sources must draw attention to themselves, which means that their activities are hard to conceal from law enforcement authorities. But non-commercial sources such as peer-to-peer file-sharing networks may present no easily accessible target for legal action (and under many circumstances can oper- ate invisibly to law enforcement). To address non-commercial sources, law enforcement officials must conduct what amount to "sting" operations.5 Such operations are controversial, are personnel-intensive, and may not offer large leverage, as non-commercial sources are likely to have signifi- cantly smaller audiences than commercial sources. Table 14.1 contains an illustration of how social and educational strat- egies, technology tools, and public policy can work together. 14.4 TRADE-OFFS AND COMPLEXITY After more than a year of intensive study of the issue, the committee was struck by its extraordinary complexity. Such complexity manifests itself in many ways, but nowhere more prominently than in understand- ing the trade-offs involved in the development of any comprehensive approach to protecting children on the Internet from inappropriate mate- rials and experiences. 50ne example of a "sting" operation is an exercise in which a law enforcement official assumes an online identity corresponding to that of a minor, and engages potential preda- tors seeking to entice a minor for the purpose of initiating a sexual encounter. A second example is the use of customer lists belonging to an online service that provided child pornography, lists confiscated after the service itself was convicted of child pornography charges. In this latter case, known as Operation Avalanche, law enforcement officials con- tinued to operate the online service's Web site and sent e-mail to subscribers offering them the opportunity to purchase child pornography. Those who responded received controlled deliveries of child pornography made by investigators, and search warrants were executed on the residences of those customers immediately after the deliveries were made (see AsC News, 2001, "An Avalanche of Child Porn: Investigators Use Subscription List to Track Down Pedophiles," November 14, available online at <http://abcnews.go.com/sections/ business/TechTV/TechTV_Avalanche_Porn_011 114.html>~.
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS TABLE 14.1 An Illustration of Mutual Reinforcement 369 Impact on Suppliers of Inappropriate Material Impact on Potential Consumers of Inappropriate Material Social/ Business decisions and ethics Teach children to make choices educational (e.g., ISPs can choose to refrain to stay away and to target strategies from carrying USENET searches more precisely to newsgroups with a large reduce chances of inadvertent amount of child pornography) exposure; teach parents how to educate their children about these strategies and to monitor what their children are doing online Technology- Labeling material so that Help guide children to based tools parents and others can easily appropriate sites and help to ascertain the appropriateness prevent access to inappropriate of that material for their sites, consistent with parental children values and preferences Public policy Shape the environment by Provide support for social and reducing deceptive practices educational strategies (e.g., (e.g., mouse/rapping, spam, new standards of learning for capturing of misspelled web K-12, outreach to parents to sites) educate them about the Internet) The nature of trade-offs is such that doing "better" with respect to one goal implies doing less well with respect to some other goal or goals. In the present instance, any approach that improves protection for children and youth from inappropriate Internet material or experiences is almost certain to have a negative impact on other values or goals that most par- ents or communities would generally find appropriate and desirable. Note that these latter values and goals may also be associated with chil- dren and the Internet. Two points must be made about the existence of trade-offs. First, the existence of such trade-offs is not an argument, per se, against attempts to "do better" at protecting children and youth from inappropriate Internet material or experiences. Second, the fact that trade-offs exist for any given method suggests that a mix of methods may well be more effective than exclusive or primary reliance on any one method. To illustrate these trade-offs, the next few sections discuss trade-offs that decision makers must address in considering the use of social and educational strategies, technology-based tools, and public policy actions.
370 YOUTH, PORNOGRAPHY, AND THE INTERNET 14.4.1 Social and Educational Trade-offs The committee has identified social and educational strategies to teach children and youth how to make good decisions about using the Internet as foundational for any approach to protection. But socialization and education are inherently processes that operate over a long time scale- thus, they cannot be expected to demonstrate immediate results. Further- more, they are not simple to implement, and they require forethought, planning, and extensive follow-through. They can be costly, both in terms of dollars and in terms of time. Perhaps the most important trade-off associated with using social and educational strategies is that they may conflict with other pressing social and educational needs. For example, most K-12 curricula are al- ready overloaded, and information and media literacy curricula must compete for time in the schedule with physical education, art, music, sex education, consumer literacy, and a variety of other pressures on the curriculum. Education in these areas is also important, and passionate advocates can be found for all of them. Because the amount of time in a curriculum is more or less fixed, there are only three possibilities. One is that something must be removed if something else is added, and the elimination of any given subject area is always controversial, if only because the importance of one subject area must be weighed against the importance of another. A second possibility is that by increasing the efficiency of education in the existing areas of study (so that the same ground can be covered in shorter amounts of time), time can be made available to add information and media literacy. But increasing efficiency is an enormously difficult problem, and in prac- tical terms, it is not clear how to do so. The third logical possibility is to obtain the needed time for information and media literacy by trimming the instructional time devoted to existing subject areas. The risk in this approach is that coverage of those latter areas may become inadequate as a result. The dilemma is no easier to resolve in the family context, where fam- ily time together is at a high premium in many families. Parental efforts to supervise children and youth using the Internet must compete with making sure that children clean their rooms, do their homework, get to the soccer or basketball game on time, avoid unhealthy use of drugs and alcohol, work part-time jobs, and so on. The difference in knowledge about technology and its uses between many parents and their children further inhibits informal candid discussion about issues related to the Internet. Siblings and friends charged with providing peer assistance may also play tricks on children to get them into trouble.
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 371 Finally, educational strategies to teach children and youth to refrain from seeking out inappropriate sexually explicit materials face a power- ful challenge in that many adolescents, especially boys, are highly moti- vated to seek out such materials. Some children do let educational efforts "roll off their backs" and are not influenced by them, and although educa- tion and socialization are the only approaches that have the chance of reaching all or most children or that have any chance of reducing a child's desire for such materials, the expectations for such education and social- ization should not be unrealistic. 14.4.2 Technology Trade-offs One technology trade-off is illustrated in the balance that users of technology-based tools must strike between two concerns shielding the child from inappropriate material and experiences against enabling child- centered control over the flow of information to him or her. One can increase shielding but only at the expense of reducing a child's discre- tionary control over and access to information. A related trade-off is the issue of false positives and false negatives. While false positives and false negatives in general trade off against each other, even when human beings are actively involved in making judg- ments about the appropriateness of material to be shown to children, the trade-off is most stark when technology is used to assess appropriateness, and a fundamental reality of technology is that it does not provide an accurate and inexpensive way to assess content. As a general rule, in- creasing the probability that a device or system will identify inappropri- ate sexually explicit material as such also increases the probability that some not-inappropriate material will also be improperly identified as inappropriate. If false positives are generally tolerable (i.e., one is gener- ally willing to pass up useful information as the price of protecting against inappropriate material, as might be the case for many risk-averse par- ents~, then the automated assessment of content does have significant utility. Still another technological trade-off arises because of tensions between flexibility and ease of use. Products that are flexible can be customized to the needs of individual users, and most people say that they want flexible products. Yet, a highly customizable product involves many decisions for a user to make, and many users find the actual exploitation of a product's flexibility to be a chore. As a result, the most common use of any technology tool is in its default "out of the box" configuration. Thus, for practical purposes, it is fair that any assessment of a tool place great weight on what the tool does out of the box. Against that standard, many
372 YOUTH, PORNOGRAPHY, AND THE INTERNET tools for protecting children from inappropriate Internet material and experiences place a far greater emphasis on labeling Internet material and experiences as inappropriate than on carefully selecting whether appro- priate material might be labeled as inappropriate. That is, if it might be inappropriate, it is labeled and flagged as such. Filters thus favor over- blocking rather than underblocking, while monitors are likely to flag more material as questionable rather than less. This systematic bias in favor of overcaution provokes anxiety in many people. For example, to the extent that technology-based tools are pre- ferred instruments of government policy, concerns arise that government may be implicitly endorsing a lesser degree of information flow to the public and to children than would be the case in the absence of tech- nology-based tools. It is also an empirical observation about many tools on the market identifying certain kinds of information as inappropriate for children that their supporters and advocates are perceived as support- ing an underlying political agenda. Finally, political pressures and man- dates to deploy technology-based tools distort the market, in the sense that they create artificial demand for solutions to problems that commu- nities do not perceive to exist to a significant degree (if they did, they would deploy these tools without such a mandate). Even the protection itself that technology offers with respect to pro- tecting children and youth from inappropriate material involves a poten- tial trade-off. To the extent that technology-imposed limits on choice work as intended (i.e., to block rather than discourage access to material that may be inappropriate), children and youth lose an opportunity to exercise responsible choice, and hence an associated chance to learn how to make responsible decisions. For younger children who tend to be less good at decision making than older children the consequences of bad choices may be more serious, because of their relative inexperience about life and their greater impressionability; for such children, opportu- nities for unconstrained decision making and choice are not appropriate under most circumstances. On the other hand, children who develop internal standards of ap- propriateness can be safer in those situations in which they or their parents and guardians cannot rely on technology to protect them. This is important given the increasing ubiquity of Internet access points in many venues. When responsible and respected adults and mentors talk with these older children about responsible decision making and estab- lish sanctions for inappropriate choices, they create an environment that encourages and supports responsible choice, which in turn is likely to be conducive to the development of positive habits. Such habits are most important as adolescents reach the age of majority, when they will have
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 373 the rights of adults and no constraints relative to Internet access. On this point, parents must decide how to proceed. 14.4.3 Public Policy Trade-offs As noted in Chapter 9, the viability of many public policy proposals depends on how policy makers make certain trade-offs between the goal of helping to shield children and youth from inappropriate sexually ex- plicit material on the Internet and other desirable societal goals. For example, the committee believes that spam containing material that is obscene for minors should not be sent to children. But laws banning such e-mail to minors are potentially problematic in an online environment in which it is very difficult to differentiate between adults and minors. (In- deed, regulation that depends on regulating a certain type of content (namely, sexually explicit speech) is inherently more suspect on First Amendment grounds than are proposals that regulate speech indepen- dent of content.) On the other hand, a ban of all spam regardless of content may be seen as too broad because it affects many other interests- for example, those parties with a commercial interest in using e-mail channels to advertise non-sexual goods and services. The committee also believes that it would be desirable for adult Web site operators who exhibit material that is obscene for minors to use age verification systems so that children would not be able to access such material. However, in an online environment in which it is very difficult to differentiate between adults and minors, it is not clear whether deny- ing access based on age can be achieved in a way that does not unduly constrain the viewing rights of adults. Thus, as one illustrative example from Section 9.3.2, the government might offer a grant of immunity from prosecution under obscenity laws to Web site operators that use age veri- fication systems to prevent minors from accessing such material.6 In this instance, the trade-off is helping to protect children from exposure to certain kinds of inappropriate sexually explicit material (such a mea- sure would help to reduce the inadvertent discovery of such material from commercial Web sites) in return for limitations on possible obscen- ity prosecutions. Aggressive enforcement of obscenity laws also presents trade-offs. Increased prosecution of obscenity would likely require increased re- 6More specifically, immunity for adult Web site operators from prosecution under ob- scenity laws would be granted if (a) they take actions to prevent their content from being indexed by search engines, and (b) they provide a plain-text front page that warns users that the site they are about to enter has adult-oriented content and an "exit" button to a child-friendly site if the user acknowledges being under 18.
374 YOUTH, PORNOGRAPHY, AND THE INTERNET sources, and those resources must be taken from some other activity. If, as is likely, the other activity represents prosecutions of other crimes, policy makers must make the judgment that it would be wise to pursue more obscenity prosecutions rather than other criminal prosecutions, or that more prosecutions for obscenity would necessarily be the best use of additional resources if such resources are available. Such judgments are complex and require a careful weighing of many competing factors well beyond the scope of this report. 14.5 TAKE-AWAY MESSAGES FOR DIFFERENT PARTIES In the committee's judgment, the bottom line on reducing the expo- sure of children to inappropriate material and experiences on the Internet is that those who rely exclusively, or even primarily, on technology and public policy will find that the resulting protection will rest on uncertain and shifting ground and is likely to fail their children when exposure to inappropriate material or dangerous situations occurs. If one installs tools and/or passes legislation in the hope that they will "take care of the problem," and that in doing so one's responsibilities will thus be ad- equately discharged, children are highly likely eventually to encoun- ter inappropriate material or experiences on the Internet in some venue, whether by accident or on purpose. And such an encounter will come as a disturbing surprise to the parent, teacher, librarian, or public policy maker who feels that he or she has done all that needed to be done. In the end, responsible choice which is foundational for safe Internet use by children is closely tied to the values that parents and communities wish to impart to their children, and that influence judgments about the proper mix of education, technology, and public policy to adopt. Box 14.1 describes some of the behavioral aspects of Internet safety for children that families, schools, and libraries might wish to teach. 14.5.1 Parents Parents have a primary responsibility for guiding children into matu- rity. They have responsibilities to their children in the physical world, and they have corresponding responsibilities in cyberspace. With respect to Internet usage, developing maturity implies the ability to make safe, responsible, and morally appropriate choices about what to do and what to see on the Internet and a facility for coping constructively with inap- propriate or objectionable experiences. Box 14.2 describes one possible "best practices" scenario focused on the home. In addition, parents might wish to keep the following points in mind.
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 375
376 YOUTH, PORNOGRAPHY, AND THE INTERNET · When children are young, a wider and more extensive range of precautions and a limited range of choices are appropriate. However, as children mature, precautions can generally be relaxed and the range of choices expanded. If technology is used to limit access, consider the age- appropriateness of the limits you wish to impose. · Interactive dialog (whether through e-mail, instant messages, voice and/or video, chat rooms) with strangers is a much greater potential threat to a child's well-being than other things that he or she is likely to encounter on the Internet. Account for the fact that children often overes- timate their ability to make good judgments about the intentions of people whom they do not know. · An environment in which your child will feel comfortable talking to you is more likely to result in conversation about potentially serious
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 377 issues. If they are afraid of your reaction to what they tell you, they will be less likely to tell you when they encounter things that might make you uncomfortable. Thus, think in advance about what you will say and do when your child is exposed to material that you regard as inappropriate. · For most people (children included), instant messages and chat dialogs are no different than telephone conversations, mail, or diary en- tries. Reading logs of chat room or IM dialogs is likely to be seen in the same way as listening in on a child's telephone conversations or reading a child's diary. For some parents, the cost of violating a child's privacy is worth the gain of learning what is on a child's mind. For others, the cost of violating a child's privacy is too high a price to pay under most ordi- nary circumstances. A related question is that of the extent to which a child's rights to information go beyond what a parent is willing to grant. Some parents are willing to state explicitly that their children should be able to obtain information on certain subjects, even if it would make them (the parents) uncomfortable, while others find such a notion to be unac- ceptable. Thus, consider your own philosophy regarding privacy and information rights for your children. · The policies of your child's school, library, and friends' families regarding Internet use and access have an effect on your child's Internet experience. Coordinate the Internet experiences available at home and at school, the library, or at the homes of friends (if you approve of those policies), or provide an alternative for your child (if you disapprove of them). · Children often copy what their parents say verbally and actually do behaviorally. If you believe that the viewing of sexually explicit mate- rial is inappropriate for your children, explain to your children why such behavior is inappropriate for them and set a good example in your own household. · Illegal or inappropriate behavior or material can be reported to the responsible authorities. For example, the National Center for Missing and Exploited Children provides one-stop reporting to law enforcement authorities of child pornography found online by the general public (http://www.cybertipline.com or 1-800-843-5678~. Local law enforcement agencies (e.g., state or town police) are often recipients of complaints related to online solicitations of children for sexual purposes, and the Federal Bureau of Investigation and the U.S Customs Service have au- thority in this matter as well. Certain online services provide mecha- nisms for reporting inappropriate activity. For example, AOL provides a way for users to notify AOL's "Community Action Team," which has the responsibility of enforcing AOL's terms of service. Among other things, AOL's terms of service forbid vulgar language and disruptive behavior in chat rooms or instant messages and e-mail and instant messages that
378 YOUTH, PORNOGRAPHY, AND THE INTERNET contain requests from strangers for passwords, credit card numbers, or other personal information. Finally, the committee believes that a parental lack of knowledge about Internet culture and the diversity of possible Internet experiences has been and continues to be a source of both complacency (because some do not know what on the Internet may be of concern to them and their children) and excessive fear (because they do not know enough to be able to place these dangers in proper perspective). For this reason, paren- tal education about the Internet continues to be an important part of a comprehensive program of Internet safety education for children. 14.5.2 Teachers and Librarians Teachers have a responsibility for educating students and for provid- ing a safe environment in which learning can occur. Libraries have a responsibility for providing the communities they serve with a broad range of useful materials appropriate for their needs. Box 14.3 and Box 14.4 describe possible "best practices" scenarios for schools and libraries, respectively. In addition, teachers, school administrators, and libraries might wish to keep the following points in mind. · The educational and informational needs of young people at vari- ous ages from kindergarten to the seniors in high school vary enor- mously, and the Internet content that is made available to students at
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 379 these various stages of development must be matched to their needs, skills, and maturity. Furthermore, teachers and librarians have an impor- tant role in educating parents about the Internet its benefits and its dan- gers and they are sources of advice about how to cope with the dangers that the Internet may pose. · Transparency is a virtue, both for adult supervisors and child In- ternet users. Transparency can include an understanding of why a given site is or is not regarded as inappropriate, access to detailed information about what is deemed inappropriate, knowledge of instances when ac- tions are being monitored or influenced, and the ability for on-site adult supervisors to override non-local decisions about inappropriateness. · The total cost of technology solutions, rather than just the initial deployment costs, must be evaluated. Specifically, technology generally entails continuing costs of maintenance and upgrade, as well as human staff members to ensure that the technology is being used appropriately and is serving its intended functions. For example, staff may be needed to process requests for filtering overrides (if filtering is in place) or to pro- cess parental complaints (if filtering is not in place). · Efforts to provide educational outreach to parents must account for busy family schedules as well as, perhaps, a certain resistance among some to deal with technology-based issues with which they are unfamiliar. For example, outreach activities requiring parents to go far out of their way
380 YOUTH, PORNOGRAPHY, AND THE INTERNET tend to be poorly attended; conversely, parents would be more inclined to attend programs that are conveniently located and scheduled consistently with the community's work hours. One program described to the commit- tee during the first workshop involved taking computers to venues that had little to do with technology such as a gardening class or music festival. By demonstrating how the technology could have useful applications even in such venues, a context arises naturally in which it is possible to raise issues related to young people, the Internet, and parenting. Also, coordina- tion with volunteer organizations outside schools and libraries may make more resources available for these outreach efforts. PTA organizations, for example, are well suited to provide such outreach efforts. 14.5.3 Industry Various components of industry can make a major contribution to the Internet safety of children. The segments of industry relevant to the issue include ISPs and online service providers, makers of access devices such as personal computers, software vendors, content providers, and the adult online industry. · ISPs and online service providers could: Provide easily understood and implemented parental controls. As noted in Chapter 11, flexibility is often not used because presenting a range of options is confusing to the parent trying to configure a system. On the other hand, a single "one-size-fits-all" approach does not take into account the needs of individual children. One approach to simplifying parental controls, implemented by some service providers, is to set de- fault measures depending on the age of a child, while giving the parent the ability to adjust these defaults appropriately. Design and provide educational and child-friendly areas. Chapter 10 discusses a number of means to concentrate child-friendly and compel- ling content that would attract children's online attention. To achieve critical mass, such content would be designed to be broadly appealing over a wide variety of topics, including information on relationships, sexual health, and other topics of interest to adolescents. Provide a uniform channel for user complaints about child por- nography and/or obscene material. Such a channel (which could be as simple as a link to the CyberTipline) could easily be placed on the com- plaints or customer service page of the service provider. Refrain from carrying material that they believe to be illegal, such as suspected child pornography. (As discussed in Chapter 4, the "Good Samaritan" provisions of the Communications Decency Act remain in force and largely immunize providers exercising such control over con-
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 381 tent from liability for failures in such control.) For example, on their own authority, ISPs could refrain from carrying Usenet newsgroups known to carry large amounts of child pornography. · Makers of access devices such as personal computers could: Provide in-the-box Internet safety tips and best practices. Many such brochures published by reputable organizations are available for free, and "out-of-the-box" information relevant to Internet safety could be a valuable route to greater publicity for such efforts. Provide configuration options for children in a household. At greater cost, the initial machine-setup configuration process could be modified to ask the device owner or administrator if children are ex- pected to use the device. If so, the process could guide the owner or administrator through a setup process for specifying parental controls for the use of the device. (For example, the setup process might specify certain limits on Internet access if a child logs into the device.) · Software vendors could: Develop software to prevent mouse/rapping. Recall that mouse- trapping (discussed in Chapter 3) refers to the phenomenon in which a user who tries to leave a sexually explicit site is automatically forwarded to another such site. Some programs are available today to block "pop- up" advertisements, and these work to prevent mousetrapping as well. However, these programs must be invoked before the pop-up ad or mousetrapping occurs, and they also disable some useful features of Web sites. A "panic button" could be installed that is always available on the user's screen, and if a user is mousetrapped, clicking on the "panic but- ton" should close all current browser windows. Develop software to help configure computers to be child-friendly. Such software would operate in lieu of the child-oriented setup routines described above. Integrate label-based filtering options into Web browsers (see below). Include in software regularly used by adults tips for their children's Internet safety. For example, upon initial installation of a software prod- uct, the software setup program could ask if the user wanted to view a screen of Internet safety tips for children, and if so, could display such tips or direct him or her to an appropriate Web site. Provide content creation tools that have been adapted to speed the process of content developers creating labels that can be used by PICS- based filtering schemes. · Content providers could: Participate in labeling schemes. As discussed in Chapter 12, the
382 YOUTH, PORNOGRAPHY, AND THE INTERNET success of a labeling scheme depends primarily on its widespread adop- tion. In October 2001, AOL Time Warner, Yahoo, and Microsoft Network announced their adoption of the content labeling system of the Internet Content Rating Association. Integrate educational and entertainment value into content devel- oped for children. If children are to use content voluntarily, such integra- tion is likely to enhance its appeal for both children and parents. Add links to age-appropriate sexual and emotional health content to Web sites visited by older youth. As noted in Section 14.2, many older youth are interested in information that relates to their sexuality. Given this fact, it makes some sense to provide reliable and appropriate infor- mation to meet this need, rather than leaving them on their own to find unveiled information that may be of questionable value or to seek out adult-oriented, sexually explicit content that depicts sexual behavior in ways that are demeaning or disturbing. · The adult online industry could: Take more effective steps to keep children from accessing their products. For example, operators of adult Web sites could set up their home pages without sexually explicit material (i.e., the cyber-equivalent of a brown paper wrapper around an adult magazine), and use the robot.txt protocol (described in Chapter 2, Box 2.3) to prevent indexing of pages (even free teaser pages) that contain such material. Thus, children searching the Web would never find the sexually explicit material directly (though they might be directed to an adult site's home page). Stop the practice of involuntary mouse/rapping. For example, it would be simple to offer a user seeking to leave a Web site a choice about whether or not to be redirected. Such a simple step could do much to reduce the perception of irresponsible behavior on the part of the adult online industry. Use contracts to require more responsible behavior among affili- ates that use content provided by commercial sources of adult-oriented, sexually explicit imagery. For example, a contract between a content provider and an affiliate might require that the affiliate would have to put its content behind plain brown wrappers and so on, and the firms that supply content would be in a position to penalize them if they did not (by cutting off a content source). Finally, the information technology industry should not be discour- aged from undertaking serious technology-based efforts to help parents and other responsible adults to improve and enhance the safety of their children's Internet experiences and to reduce the amount of inappropri- ate material to which they may be exposed. Indeed, as a primary benefi-
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 383 ciary of the Internet age (as well as being instrumental in creating it), the information technology industry arguably has a special responsibility to help safeguard children on the Internet. Developing more discriminating filters, enabling parental controls, supporting research of the type de- scribed in Section 14.6, and adopting and promoting labeling schemes are a few of the ways that the IT industry has sought to discharge its respon- sibilities in this area, and a further enhancement and strengthening of these types of effort can only expand the range of options that parents and other responsible adults can exercise. 14.5.4 Makers of Public Policy Public policy at the local, state, and federal levels helps to shape the environment in which Internet access occurs. But because the scope of public policy actions are by definition pervasive throughout the com- munity to which those actions are relevant, public policy makers must proceed judiciously. Public policy actions are most effective when they are based on reli- able science rather than anecdote, and when they reflect a strong social, ethical, and moral consensus. For example, the sentiment that child por- nography and sexual molestation of children are wrong is shared by people among a very broad spectrum of political views. Thus, it is rea- sonable to say that these are serious national problems, and addressing them continues to be an important task for the nation. Furthermore, the scale of these problems already large is increasing, and in any event outstrips the resources available to deal with them. By contrast, public policy makers should tread lightly when it comes to other areas in which a consensus is not so apparent. For example, the committee heard from parents who did not trust the federal government to take actions to reduce children's Internet exposure to inappropriate materials. The striking aspect of this sentiment was that it was expressed by both conservative and liberal parents. Public policy makers should also be wary of cheap, easy, or quick solu- tions. As the discussion in Chapter 12 on filtering demonstrates, such "solu- tions" may not fix the problem that they seek to solve at least not to the extent that they would enable resources and attention to be turned else- where. It is true that the cost of social and educational strategies tends at first blush to be considerably larger than the costs of protective technologies but the benefits that accrue are also correspondingly higher. Students will be more able to avoid problematic experiences and material of their own voli- tion, and will be better able to cope with them when they occur. Finally, it is necessary to underscore the fact that public policy can go far beyond the creation of statutory punishment for violating some ap-
384 YOUTH, PORNOGRAPHY, AND THE INTERNET proved canon of behavior. Certainly, legal sanctions are one possible pub- lic policy option, and such sanctions act both to punish those who behave in a way contrary to law and to deter others from conducting themselves in a similar way. Options such as more vigorous prosecution of existing obscenity laws are discussed at length in Chapter 9. But public policy can be used much more broadly and can shape the Internet environment in many ways. For example, public policy can be used to: · Reduce uncertainty in the regulatory environment. Uncertainty in the regulatory environment is often inhibiting to business plans. For example, prior to the enactment of the "Good Samaritan" section of the Communications Decency Act (discussed in Section 4.2.3), at least one court case (Stratton Oakmont v. Prodigy) had suggested that an ISP exercis- ing editorial control over messages posted on its bulletin boards in accor- dance with its acceptable use policy was subject to liability as a publisher for content available through the ISP that was inconsistent with its accept- able use policy. This precedent gave ISPs incentives to refrain from exer- cising editorial control, and the Good Samaritan provisions of the Com- munications Decency Act never overturned in the courts eliminate such liability. · Promote media literacy and Internet safety education. Promotion can include: Funding the development of model curricula for media literacy and Internet safety. There are political sensitivities related to the federal government's role in education, which is a local responsibility, but devel- oping models and giving grants are both well-accepted federal practices in the area of education. Encouraging and supporting professional development for teach- ers on Internet safety and media literacy. Professional development for teachers seeking to use the Internet for pedagogical purposes is sparse compared with the need,7 but Internet safety and media literacy do not account for more than a very small fraction of the sparse support that is available. Supporting outreach to educate parents, teachers, librarians, and other responsible adults about Internet safety education issues. Grants could also be made available to non-profit and community organizations to run Internet safety programs. In addition, given the gap in knowledge between adults and their children, a media-based educational outreach campaign has some potential for reducing this gap. For example, public 7u.s. Department of Education. 2001. Teachers' Toolsfor the 21st Century: A Report on Teachers' Use of Technology. Available online at <http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid= 2000102>.
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 385 service announcement campaigns could help to inform parents of online issues related to their children (TV ads, enclosures in phone and/or Internet bills). Another part of media outreach might be the creation of a national "children's safety day" a day in which the media would high- light all that can be done to improve children's safety and well-being, and adults would take concrete steps to review safety issues with children. Internet safety education could well be one aspect of such a day. · Support the development of and access to high-quality Internet material that is educational and attractive to children in an age-appropri- ate manner. Public support for the development of PBS programming has a long tradition of providing such material in the TV medium, and with the dearth of comparable material for children on the Internet, such support could have a substantial impact. In addition, educational portals that organize existing safe and educationally appropriate content (see comments under Section 14.5.3) would improve the accessibility of that content to diverse audiences. · Support self-regulatory efforts by private parties. For example, public policy can provide financial or legal incentives for ISPs and content providers to behave in responsible ways. It can also coordinate and facili- tate private efforts to self-police the Internet environment (analogous to the rating efforts of the music and game industry). Such self-policing relies on the public at large to report possibly illegal material or behavior, ISPs to take actions consistent with their terms-of-service provisions, and information exchange mechanisms to ensure that all parties have the in- formation needed to take appropriate action. · Support research in areas that are relevant to the issue of Internet safety. Some of the relevant areas are the impact of exposure to sexually explicit material on children at various ages; the Internet use patterns of children; and the in-practice effectiveness of various social and educa- tional strategies, technology-based tools, and public policy at federal, state, and local levels at increasing the safety of children's Internet experi- ences. Of special interest is the possibility of research and development into technologies specifically designed to accurately identify sexually ex- plicit material and thereby enhance and improve the effectiveness of tools that can help to reduce the exposure of children to inappropriate Internet materials and experiences.8 As discussed in Chapter 2, the general infor- mation-retrieval problem is a very difficult one to solve. However, re- search intended to focus on the identification of sexually explicit material may progress more rapidly than work on the more general problem. The ~Non-industrial support for such research may be justified on the grounds that, as dis- cussed in Chapter 12, there is little market incentive for more accurate methods for identify- ing sexually explicit materials (and hence for more accurate filtering).
386 YOUTH, PORNOGRAPHY, AND THE INTERNET reason is that in seeking to identify sexually explicit material (rather than arbitrarily selected material), one might be able to make use of heuristic methods and clues signaling the presence of sexually explicit content that are not available for use in the general case.9 · Support efforts to enable families and other private parties to exer- cise greater latitude of choice in the Internet experiences of the children for whom they are responsible. For example, parents and others lack a good evaluative guide to technology-based tools and must assess the claims of vendors on their own. This report helps to provide a framework for understanding how to think about such tools but does not provide specific product guidance. Publicly available assessments of specific tools rated multidimensionally according to common criteria could help par- ents and others select tools that are appropriate for their own situations. Finally, makers of public policy must keep in mind the international dimensions of the Internet. This does not mean that U.S. actions should not be undertaken, or that they will be wholly ineffective, but expecta- tions for the impact of such actions must necessarily be moderated com- pared to the case in which the United States is the only significant actor. 14.6 RESEARCH NEEDS As the length of this report suggests, the problem of protecting chil- dren from inappropriate material and experiences on the Internet is com- plex. Reliable information in a number of areas is needed. Indeed, throughout its work, the committee was concerned about the lack of reli- able and valid science-based information for many dimensions of the problem it was addressing. Such information would have helped to strengthen committee deliberations. · The effectiveness of technology-based tools and social and educa- tional strategies in practice should be examined and characterized. Chap- ter 12 discusses one aspect of evaluating the performance of filters, based on a "head-to-head" comparison of how filters performed in blocking inappropriate materials. But protection of children is a holistic enterprise that must account for the totality of their Internet experience which sug- gests the need for an examination of all of the tools in all of the venues in which children use the Internet. The same is true for understanding how social and educational strategies affect the behavior of children "on the ground." 9Note that more effective technology in this area could be used for the benefit of those who want to block such material and of those who want to search more precisely for it.
FINDINGS, CONCLUSIONS, AND FUTURE NEEDS 387 · Given the inadequate scientific basis for understanding how media exposure to sexually explicit material affects children, more research in this area is needed. Systematic research on U.S. children may not be possible, for all of the reasons described in Chapter 6, but cross-national longitudinal studies in countries similar to the United States may shed some light on these issues. In addition, it may be possible to conduct research to study exposure to such material that occurs in the course of children's use of the Internet (which is likely to include both accidental and deliberate exposure). A further dimension of media research is the impact of media exposure on the development of personal character in youth. · It seems reasonable that providing high-quality, age-appropriate information about sex and sexual health addressing the physical, emo- tional, social, and psychological issues on the minds of children would have a dampening effect on the urge of many adolescents to search for inappropriate sexually explicit materials. To test this proposition, longi- tudinal studies of children and adolescents who receive such information would be helpful. · There is a need for research and development directed toward a comprehensive curriculum that provides in detail what parents need to know in order to handle the issue of Internet safety with their children. Such a curriculum would provide specifics on what a parent might do operationally. · A better understanding of the population of online sexual preda- tors might help government efforts to prevent sexual predation. If, for example, a large percentage of online sexual predators were already- convicted child molesters, that fact might argue for longer prison terms for such felons, closer supervision of their activities after release, and/or prohibitions on their use of online resources. 14.7 CONCLUSION The Internet offers enormous potential to enhance the intellectual, educational, social, and personal development of children. Those who take actions to address the concerns described above must bear in mind the potential benefits that the Internet offers. Thus, any "appropriate" mix of actions should be seen as balancing competing goals and values rather than endorsing the absolute supremacy of any one goal or value. Furthermore, evolution with respect to technology and the e-business environment, as well as possible changes in community standards gov- erning obscenity, means that there are no foreseeable technological "sil- ver bullets" or single permanent solutions to be crafted. Rather, any approach adopted to protect children must adapt to changing circum-
388 YOUTH, PORNOGRAPHY, AND THE INTERNET stances. While technology and public policy have important roles to play, social and educational strategies that impart to children the character and values to exercise responsible choices about Internet use and the knowl- edge about how to cope with inappropriate material and experiences is central to promoting children's safe Internet use.