Professor Frederick Schauer

October 2, 2002

My name is Frederick Schauer, and I am Frank Stanton Professor of the First Amendment at the Kennedy School of Government, Harvard University. I have previously served as Professor of Law at he University of Michigan, and have served as Visiting Professor of Law at the law schools of Harvard University, the University of Chicago, and the University of Virginia. In 1985-96 I served as a Commissioner of the Attorney General's Commission on Pornography, and was the principal author of the Commission's findings, analysis, and conclusions, including the findings, analysis, and conclusions on the subject of child pornography. I am the author of The Law of Obscenity (BNA, 1976), Free Speech: A Philosophical Enquiry (Cambridge, 1982), The First Amendment: A Reader (West, 1992, 1996), and numerous articles on the law of obscenity and pornography, on freedom of speech and press, and on constitutional law generally. Among my publications are an article in the 1982 Supreme Court Review analyzing the child pornography case of New York v. Ferber, 458 U.S. 747 (1982), an article in the 1995 Supreme Court Review analyzing the child pornography case of United States v. X-Citement Video, Inc., 115 S. Ct. 464 (1994), and other articles specifically focused on obscenity and pornography law in the American Bar Foundation Research Journal, West Virginia Law Review, North Carolina Law Review, Hastings Law Journal, and Georgetown Law Journal.

I appear before the Committee not as a representative of the Kennedy School of Government nor of Harvard University. Nor do I appear on behalf of any other person, corporation, or organization, and I have no political, financial, organizational, or other connections with anyone interested in one way or another in the proposed legislation. As a political independent, I have not been a member of a political party for over twenty-five years, and I do not represent or consult for clients, or their lawyers, directly or indirectly. I appear today at the unsolicited request of the Committee on the Judiciary, as I did in 1996 when S.2520's predecessor, the Child Pornography Prevention Act of 1996, was considered by the Congress.

The bill before the Committee, the Prosecutorial Remedies and Tools Against the Exploitation of Children Today ("PROTECT") Act of 2002, is a proposed congressional response to the Supreme Court's decision in Ashcroft v. Free Speech Coalition, 122 S. Ct. 1389 (April 16, 2002), in which the Supreme Court invalidated on First Amendment grounds two different provisions of the Child Pornography Prevention Act of 1996. Both of these provisions were aimed at stemming the increasing proliferation of virtual child pornography, a genre of child pornography produced with modern computer technology to resemble closely the genre of unlawful child pornography, but which, unlike "traditional" child pornography, does not employ real children in its production. To the Supreme Court in Free Speech Coalition, the distinction between actual child pornography, using real children, and virtual child pornography, in which actual children are not employed in the process of production, was the crucial factor. Although the Supreme Court reaffirmed that legally obscene virtual child pornography could be proscribed without affronting the First Amendment, the Court also reaffirmed that the key to allowing the proscription of child pornography that was not legally obscene was the presence of an actual child in the production process. Without the exploitation of a real child in the production of the material, said the Court, the requirements of existing obscenity law could not be circumvented, nor could the requirements of child pornography law as established by Ferber be satisfied.

The Supreme Court's opinion in Free Speech Coalition is hardly above criticism. But whether open to academic or congressional criticism, Justice Kennedy's opinion for a 7-2 Court still represents the definitive and authoritative interpretation of the First Amendment in the child pornography context, and thus represents the law. Legislation inconsistent with Free Speech Coalition would not only be inconsistent with current constitutional law, therefore, and not only be certain to fail in light of this very recent 7-2 decision of the Supreme Court, but would also represent a tactical mistake in the attempt to combat the horror of child pornography. As the six-year course of litigation under the previous Act so well demonstrates, constitutionally suspect legislation under existing Supreme Court interpretations of the First Amendment, whatever we may think of the wisdom and accuracy of those interpretations, puts the process of prosecuting the creators of child pornography on hold while the appellate courts proceed at their own slow place. There is room in our legislative world for legislation that is largely symbolic, but for Congress to enact symbolic but likely unconstitutional legislation would have the principal effect of postponing for conceivably six more years the ability to prosecute those creators of child pornography whose prosecution is consistent with the Supreme Court's view of the First Amendment.

Child pornography damages real children in at least four ways. One is by requiring the sexual abuse of children in its production. A second is by creating a permanent and mortifying record of that abuse that will stay with the children for the remainder of their lives. Both of these harms provided the support for the Supreme Court's willingness in Ferber to allow the prosecution of even non-obscene child pornography. In addition, child pornography is often used as a way of convincing real children to engage in sexual acts with those who would abuse them, and is often used to fuel and reinforce the predilections of the abusers. Significantly, neither of these last two justifications require the use of real children in the production process, but the harm to real children still exists.

Despite the fact that real children are harmed when virtual child pornography is used by child abusers both to reinforce their own desires and to assist in the abuse of real children, the Supreme Court in Free Speech Coalition held both of these justifications to be constitutionally impermissible. Because a wide variety of non-pornographic and plainly constitutionally protected material is often used to persuade children to engage in sexual activity, and because the First Amendment equally plainly protects material advocating or encouraging or approving of otherwise illegal activity, Justice Kennedy's opinion for the Supreme Court essentially ruled that the third and fourth of the above-mentioned justifications could not be used to support constitutionally acceptable virtual child pornography legislation.

Against that background, S.2520 needs to be evaluated, and if necessary rewritten, to ensure that it neither conflicts with the Supreme Court's holdings in Free Speech Coalition nor relies on justifications that the Court has so recently rendered constitutionally illegitimate, no matter how compelling and empirically well-supported they may be. In what follows, I will divide my comments in the various themes implicated by S.2520.

1. Pandering. In Ginzburg v. United States, 383 U.S. 463 (1966), the Supreme Court upheld the obscenity conviction of the publisher of a magazine called Eros, relying largely on the way in which the magazine had been "pandered," which Justice Brennan defined as the "commercial exploitation of erotica solely for the sake of their prurient appeal." S.2520 adds a new §18 U.S.C. 2252A(a)(3)(B), which would make criminal the pandering, including by computer, of legally obscene child pornography.

The new provision is limited to the legally obscene, and is likely for that reason to be constitutionally permissible. Ginzburg did not create or recognize pandering as an independent offense, and this understanding of pandering as being largely about the evidence available to prove appeal to the prurient interest, the first prong of the obscenity standard in Miller v. California, 413 U.S. 15 (1973), has been reaffirmed by both Hamling v. United States, 418 U.S. 87 (1973), and more recently in Free Speech Coalition. 122 S. Ct. At 1405-06. But although the pandering of otherwise constitutionally protected material cannot be made unlawful consistent with the First Amendment, the pandering of otherwise obscene material is tantamount to the advertising of an unlawful transaction. Because obscenity is unlawful conduct, and because the advertising (or, therefore, pandering) of unlawful products and services is not itself protected by the First Amendment, see Virginia Pharmacy Board v. Virginia Citizens Consumer Council, 425 U.S. 748 (1976), the limitation of this provision to legally obscene material renders it constitutional even though the solicitation, pandering, or promotion is not itself obscene. Were it not for this limitation, the pandering provision would likely be unconstitutional, because the Supreme Court has never indicated that pandering can be an independent offense. Thus, in order to reinforce the constitutional basis for §2252A(a)(3)(b), it might be wise for Congress to make clear that it views such advertising and solicitation as the advertisement of an otherwise illegal product, thus establishing the "commercial advertisement of an unlawful product" foundation for this section.

By contrast to S.2520, Section 4 of H.R.4623 treats pandering as an independent offense without the necessity of a showing that the material pandered is in fact legally obscene or is in fact child pornography made with the use of a real child. In the absence of such a showing, the "advertising for an unlawful transaction" rationale disappears, and the pandering provision appears instead as a prohibition on the advertising of an immoral or unhealthy but lawful product, plainly protected by the First Amendment under recent court rulings. See Rubin v. Coors Brewing Co., 514 U.S. 476 (1995). Before Free Speech Coalition, there existed plausible arguments to justify such an expansion of the pandering idea in the context of virtual child pornography. After Free Speech Coalition, however, such arguments, and the statutes they would support, are highly likely to be exercises in symbolism alone.

2. Defining Child Pornography. S.2520 would make two changes in the definition of child pornography, both in response to the Supreme Court decision in Free Speech Coalition. The bill requires that material coming under 18 U.S.C.§2256(8)(B) not only be or appear to be of a minor engaged in sexually explicit conduct, but be legally obscene as well. With this modification, §2256(8)(B) is plainly constitutional, but the question remains about its effectiveness. Although the vast majority of child pornography is in fact legally obscene, the numerous procedural and substantive hurdles to proving obscenity make the difference between what is theoretically obscene and what can actually be proved beyond a reasonable doubt to be obscene an important distinction. It is to this issue that §2256(8)(D) appears to be directed.

In Free Speech Coalition, the Supreme Court struck down the existing §2256(8)(D), concluding that the First Amendment prohibited the pandering of non-obscene and otherwise lawful material. S.2520 appears to address this issue in two ways. First, it requires that the depiction be of a minor, or a person who appears to be a minor, engaged in actual and specified sexual activities. And, second, it requires the prosecution to prove that the material lacks serious literary, artistic, political, or scientific value, just as in an obscenity case.

The new §2256(8)(D) in fact comes quite close to the legal definition of obscenity found in Miller. The new §256(8)(D) would contain exactly the third prong - lacking in serious literary, artistic, scientific, or political value - and §2256(8)(D)'s specification of the types of depictions that are impermissible tracks both the letter and the spirit of Miller. Although the United States and most states have included the "patently offensive" language in their post-Miller obscenity laws, a close reading of Miller makes it clear that what Miller requires is not necessarily the ritual inclusion of the "patently offensive" language, but rather a specification of the type of material that the legislature finds to be patently offensive. Because the proposed new §2256(8)(D) does this, and because there has never been any indication that the activities specified are not within the range that a legislature may constitutionally find to be patently offensive, the second prong of the Miller standard appears to be satisfied as ell.

This leaves only the first prong - appeal to the prurient interest, which is not included.
Although a showing of appeal to the prurient interest is part of the constitutionally required definition of obscenity under existing law, it is certainly plausible to believe that it may not be necessary in this particular context. That is, it is plausible to believe that obscenity in the virtual or actual child pornography context would not necessarily require proof of an appal to the prurient interest, especially given the inclusion here of the "serious literary, artistic, political, or scientific value" aspect, the aspect that carries almost all of the First Amendment concerns, as part of the prosecution's burden of proof. As a result, it is reasonable to argue that the new §2256(8)(D) comes not under Ferber (at least not where no real child is involved), but under Miller as slightly, and only slightly, modified in the particular context where children are involved. This is consistent with the spirit of Ginsberg v. New York, 390 U.S. 629 (1968), in which the Supreme Court allowed modification of the obscenity standard (although, to be sure, not the explicit words of the standard) when material is directed to children, and much the same approach, based on Ginsberg, would seem to apply to material directed to children in the particular way in which child pornography is used.

S.2520 could be strengthened slightly in this regard if the new §2256(8)(D) also contained some elements of the pandering section, since the Supreme Court from Ginzburg to Free Speech Coalition has emphasized that pandering is best seen as proof of an appeal to the prurient interest. Without the addition of any reference to pandering, the proposed §2256(8)(D) is close enough to Miller, especially in its most important First Amendment dimensions, and far enough away from the provision invalidated in Free Speech Coalition, that here seems a reasonable likelihood that it could be upheld as a contextual Congressional specification of obscenity under Miller that is consistent with the spirit of Miller, consistent with the spirit of Free Speech Coalition, and fully cognizant of the need to modify Miller in its non-essential (in free speech terms) to take account of the special dimensions of the modern production and use of material portraying explicit sexual acts by children. Adding elements of pandering would make the argument even stronger.

By contrast, H.R.4623's proposed modification of §2256(8)(D) would almost certainly fail to survive a constitutional challenge after Free Speech Coalition. The addition of the word "indistinguishable" does narrow the class of covered materials substantially, but does not change the nature of the government's interest. Even if no person at all could tell the difference between materials using real children and materials using computer-generated images, the absence of real children in the latter case is exactly why the Supreme Court in Free Speech Coalition refused to find Ferber applicable, and no degree of indistinguishability in he image can create a real child where none existed before. §2256(8)(D) can be rehabilitated only by moving its category closer to the category of Miller-defined legal obscenity, and the insertion of indistinguishability, which is on a different dimension entirely, thus makes H.R.4623's approach no more likely to be upheld than was its predecessor.

3. The Affirmative Defense. A recurring prosecutorial problem in child pornography cases is the difficult of proving, given modern computer technology, that an actual child used in the materials is in fact an actual child. The Child Pornography Protection Act of 1996 attempted to address this problem by not requiring the prosecution to prove the presence of an actual child, but by giving the defendant an affirmative defense if the defendant could show that no children were used in the production.

The Supreme Court expressed some skepticism about whether First Amendment values could be carried by an affirmative defense, but in fact this is not an uncommon feature of First Amendment law. The First Amendment-motivated Noerr-Pennington defense against an antitrust prosecution, for example, is an affirmative defense, and the prosecution or plaintiff need not initially prove that the defendants' collaboration is not for the purpose of petitioning the government. Similarly, in a civil action for invasion of privacy or copyright the First Amendment concerns of newsworthiness and political commentary are permissibly embodied in an affirmative defense. Where a law of general application imposes an incidental burden on First Amendment activities, it is up to the defendant to make the case that the First Amendment is implicated. Arcara v. Cloud Books, 478 U.S. 697 (1986). Where a dismissed public employee claims that the dismissal violates the First Amendment, the employee must make that showing, and it is not the "burden" of the dismissing employer to show that the First Amendment was not violated. Connick v. Myers, 461 U.S, 138 (1983). And insofar as there is a First Amndment privilege against compelled disclosure of journalistic sources, the privilege is something that must be raised by the journalist and not initially negated by the subpoenaing authority. Branzburg v. Hayes, 408 U.S. 665 (1972)(Powell, J., concurring).

Although affirmative defenses have traditionally been allowed to carry First Amendment values, the Supreme Court in Free Speech Coalition was concerned that this particular affirmative defense was defective because it omitted possession offenses and because it did not allow the defense in cases where actual adult actors who appeared to be children were used. S.2520 cures both of these defects, and thus respects the authority of Free Speech Coalition while at the same time going a long way towards addressing what has become an important procedural problem in the effective prosecution of genuine - actual child - child pornography cases. S.2520 explicitly includes mere possession cases, and explicitly allows the affirmative defense where only adults - but real adults - are used, and thus appears to satisfy the primary concerns expressed in Free Speech Coalition. It is possible that the Court's skepticism about the wisdom of allocating the First Amendment concerns to an affirmative defense will make even S.2520's response insufficient, but because of the frequency with which such an allocation pervades all of First Amendment doctrine, and because of S.2520's specific response to the Court's detailed concerns about the content of the affirmative defense, there seems a substantial likelihood that this provision will survive further constitutional challenge.

4. Other Provisions. S.2520 contains a number of other provisions that are largely constitutionally non-controversial. Some of these provisions, such as those regarding additional staffing for child pornography prosecution and the inclusion of a separate offense for using child pornography in actual child molestation, will be useful in the fight against child pornography. Others, such as the constraints on unlimited venue, will be important in protecting against overzealous and ultimately counter-productive prosecution, even though broader venue provisions would likely be constitutional.

It is common to talk about "the" problem of child pornography, but in fact there are several problems. One is the abuse and exploitation of children in the actual production of child pornography. Another is the use of already existing child pornography in the seduction and exploitation of children. Another is the way in which child pornography provides reinforcement for pre-existing proclivities sometimes created but much more often simply exacerbated by the existence of child pornography. And another is the way in which the very existence of child pornography, even apart from their consequences to real children, debases the society and the environment in which we all live. Yet although all of these concerns are real, pursuit of all of them is not equally constitutionally permissible. If modified as I have suggested, S.2520 is likely to provide a constitutionally safe way of pursuing some but not all of these objectives. Legislation that seeks to pursue all of them simultaneously, however, or that seeks to pursue them without regard for existing constitutional limitations, will likely wind up serving none of these objectives. Tailoring legislation to existing First Amendment constraints represents neither approval of the content that the First Amendment protects nor agreement with the Supreme Court decisions protecting them. It does, however, represent the most effective and the quickest way to deal now with the gaps the exist in the present legislative portfolio, and thus represents the best and fastest ay to increase the effectiveness of the fight against child pornography.