The Child Online Protection Act of 1998 would Suppress Free Speech among Adults

E-commerce links for hardcopy of book containing this chapter (DADT 2002).  

In June 1997, the Supreme Court overturned many provisions of the Communications Decency Act, those measures in the law that would have effectively prohibited “indecent” (though not “obscene”) postings on the Internet in any area publicly accessible to minors. The CDA had even made certain subject matter (as having to do with promoting abortion) indecent for the purposes of the law. The Court held that much of the Act was overbroad and that it unreasonably and unnecessarily burdened legal communication between adults as protected by the First Amendment. Although this is not generally known, some provisions of the CDA actually promoted free speech and were left intact. For example, it contains a provision to protect Internet service providers from liability if their customer websites post illegal content, although today there may be some exceptions (such as potential ISP downstream liability associated with the Patriot Act as well as a few suits against ISP’s that allegedly may have knowingly hosted particularly violent content).

In October 1998, Congress tried to remedy this constitutional defect with the Child Online Protection Act (“COPA” or “CDA II”). The law is limited in application to commercial sites (those with anything at all to sell) and it replaces the notion of indecency with the nebulous “harmful to minors” standard. “Harmful to minors” essentially means “obscene with respect to minors.” A “commercial” posting available to minors (of any age under 17) is unlawful if an “average person” would find the web page patently offensive or prurient with respect to minors and without redeeming value with respect to minors (that is, presumably all minors). Apparently, non-pornographic adult material or subject matter could be found harmful to minors. An affirmative defense is secured by requiring credit card access or adult verification for any online file having such materials. The law did not deal specifically with violent material, and it did (somewhat clumsily) try to exempt Internet service providers from liability for the crimes of their customers.

The law defined “harmful to minors” with a list of three prongs, which essentially enumerated as follows: (1) appealing to the prurient interest of minors; (2) sexually explicit, either with graphics or detailed written description or possibly even by reference and inference (the law did mention the post-pubescent female breast); (3) not having serious literary, social, political or scientific value for minors. Content apparently would meet the legal definition of “harmful to minors” only if it satisfied all three prongs. The law used the notion of community standards borrowed from the well-known custom with obscenity. 

In November 1998, a coalition of seventeen plaintiffs filed suit in federal court to have the law overturned. As a member of Electronic Frontier Foundation, I am named as an “indirect plaintiff.” I had authored my ambitious book, Do Ask, Do Tell, about libertarianism and gay rights, which I maintain with supporting materials at my site,, and some of my material is moderately adult in subject matter. For me, requiring credit-card or other adult access to see my materials would have been totally impractical and would have shut down or discouraged much traffic to my site, even if I could afford to implement a credit-card facility.

The law appears to have been motivated by pre-election posturing, particularly to impress voters that Congress would protect children.  In fact the law does little to shield children from inappropriate materials. Most hard-core pornography on the web already requires credit card or adult access, although some operators “tease” consumers with free “previews” (the Justice Department has maintained that stopping this was the “real” intention behind COPA, however clumsily and laboriously worded).

 More significant is that (even given technology as of the time of passage in 1998), parents already have several effective ways to keep adult materials from their kids, all of these established voluntarily by private Internet service companies for their customers. They can choose ISP's that offer kids' accounts with limited access to a predefined list of “G-rated” sites approved by their ISP's. More recently, some ISP's such as AOL have offered several ages ranges for minors with appropriate predefined content for each range (in a manner analogous to movie ratings). With only a little computer knowledge, parents may set their browsers to accept only sites rated to specific levels as to adult content (according to the Platform for Internet Content Selection). Or they may install one of several screening programs that inspect all sites for objectionable material.

There have been cases where children and young teens have discovered grossly indecent materials, but in practically every such case the parents had failed to install controls. Furthermore, all reputable service providers have codes of conduct (such as AOL's “Terms of Service”) which are much more understandable and reasonable than laws like CDA or COPA. Then, one should ask Congress, why is COPA necessary?

Arguably, commercial operators may have some responsibility to protect the public from their products regardless of what parents do. Yet the “affirmative defenses” proposed by Congress would (without much affecting commercial pornography) effectively hamper access to free content and circulation of lively discussion about important issues, both personal (such as safer sex) and public.

The web has become an effective way to present difficult materials with more depth than used to be possible. In the past, political issues about sensitive moral questions tended to be aired in a polarized fashion and to be settled by taking sides, raising money and counting votes while airing the simplest (and most emotional) rhetoric possible. Some problems, such as public health, gays in the military, and recently the conduct of our president, require delving into adult concepts before they can be understood in depth. Individuals and small businesses with intellectual property to “sell” can, by presenting these issues, often counter (and perhaps undermine) well-funded political organizations, certainly with more effectiveness than individuals simply venting personal opinions with letters to editors and or venting sentiments with pre-written broadcast faxes to public officials.

COPA, while perhaps intended by some lawmakers merely to stop commercial pornographers from “advertising” indecent exposures in public, may be taken in some communities as a warrant to stop all adult discussions on the Internet within possible sight of children. If certain things are right or wrong, then, according to some people, they just don't need to be discussed in public. Leave that to the pastors and professionals! Of course, this attitude does run counter to the First Amendment.

COPA indeed brings up the “cultural war” hedge issue: How much freedom should adults sacrifice to protect children from inadvertent harm? It also invokes the question, whether people should expect simple answers to difficult cultural questions (as through religion) or should work out tough issues on their own. (And, yes, I wonder whether the Minnesota Profiles in Learning would ultimately stunt the ability of people to look at issues on their own.)

On February 1, 1999, Judge Lowell Reed of the Eastern District of Pennsylvania issued a Preliminary Injunction prohibiting the Justice Department from enforcing COPA. In the early passages in his Opinion (at my site he writes that the Internet has taken much of the power to influence public opinion out of the hands of the corporate owners of the press and placed it in the hands of individuals with moderate means but definitely valuable messages to publish. Indeed, the technological changes of the past fifteen years have emphasized economies of scale in some industries but individualism and entrepreneurialism in others. The opportunity to publish valuable, provocative materials without bureaucratic supervision and without the approval of others is surely controversial to some people who make see information as a consumer good that should be regulated.

COPA is not the only threat to this. Some will call for holding authors, publishers or Internet service providers responsible when others use their writings or facilities to commit crimes (whether securities fraud or murder), and thereby keep the right to influence public opinion or achieve in the creative arenas under the control of politicians or other vested interests. For example, there is a lawsuit (settled out of court in 2001) against the author and publisher of an “assassination manual” over a crime committed by someone who apparently followed its recipe closely.

On April 2, 1999 the Justice Department appealed this injunction to the Third Circuit Court of Appeals in Philadelphia.  The Appeals Court upheld the injunction in June 2000, with an argument that focused particularly upon the topology of the Internet. The Court held essentially that the application of community standards would make the law unconstitutional, because it would place the most conservative communities in the nation in a position to heckle materials off of the open Internet for mainstream and liberal communities.

In June 2001 the Supreme Court accepted an appeal from the Justice Department. On the surface, it appeared that the Court might review all of the different levels of concern over the implications of a law like this for the First Amendment. Soon, however, it was clear that the Court’s greatest concern was the particular rationale offered by the Third Circuit. If COPA could be invalidated because of the failure of community standards, then a whole body of obscenity law might also be at risk.

The Justice Department wrote energetic briefs that downplayed the potential harm done by use of community standards. It invoked several important cases, such as Miller,[1] Hamling,[2]  and Sable.[3]  The “serious value” or third prong, it argued, had always been understood, even in obscenity law, to refer to a national standard. Furthermore it conceded that the serious value clause could be invoked if material has value “for a legitimate minority of older minors.”  In informal terms, this means that material that was valuable for  mature teens like the fictional characters Clark Kent (Smallville), Peter Parker (Spider Man) or maybe even Harry Potter would be acceptable.

As for the other two prongs, the definitions were clear enough, in its view, so that only pornography (possibly of the soft-core Playboy variety) could be affected. The separation of quasi-pornography by an “electronic brown wrapper” would be comparable to keeping such material in separate adults-only sections of convenience stores or gay book stores. Since most of the plaintiffs (myself included) offered adult but putatively non-pornographic materials, we (as plaintiffs) likely did not have standing to sue, according to this theory.

Oral arguments took place on November 28, 2001, about eleven weeks after 9-11 and, I will add, on the fortieth anniversary of my undignified expulsion from the College of William and Mary. The Court in posing ad hoc questions did show appropriate concern that a conservative community could essentially mediate the Internet for everyone. But it (particularly in the comments by Justice Breyer) also proposed the idea of a national community standard.  It further gave credibility to the idea that a jury could be instructed to apply a national standard (for all of the prongs) rather than the particular standards of the specific geographic communities that the jurors may have come from.

On the other hand, the ACLU pointed out that the use of a national standard still could deny adults ready access to constitutionally protected materials, a situation different from that with obscenity law. One controversy that developed during the hearings was how many of the plaintiffs really could be at risk of prosecution. The government seemed to think that there were only three such plaintiffs.  The ACLU pointed out that the threshold for application of “harmful to minors” definition could change at government whim, and that therefore there would exist a “chilling effect” on speech.

The Supreme Court ruled on May 13, 2002 (ironically while I had traveled back to Washington). It vacated the Third Circuit’s opinion but left the injunction in place. It instructed the Third Circuit to rule on whether COPA, given the troubling questions about the mechanics of how it could actually work, would pass the least intrusive means tests required by First Amendment strict scrutiny. However, it did hold that the application of community standards would not by itself render the statute unconstitutional.

The Court’s opinion did show considerable respect for the content of a number of the sites, some of which (like mine) had proportionally only a small volume of material that might be viewed as harmful to minors but for whom this material was a critical component to what the publishers or authors had to say. It also paid some attention to the questions over what is a commercial provider, particularly if that is an individual self-publishing political materials through a home-based business and not intending to sell pornography. 

But Justice Thomas, in a statement that was not concurred by other justices, suggested (based on Hamling and Sable) that publishers of harmful to minors materials should merely use publishing mechanisms (that is, mail or other hardcopy delivery) that permit them to physically sieve different materials for different destinations (that is, that the Internet was not a legitimate means to publish harmful to minors literature commercially at all). To quote: “If a publisher chooses to send its material into a particular community, this Courts jurisprudence teaches that it is the publisher’s responsibility to abide by that community’s standards. The publisher’s burden does not change simply because it decides to distribute its material to every community in the Nation.” Justice Thomas’s comment would sound like an invitation for a state or locality to define “harmful to minors” very broadly to heckle out Internet traffic, but the Court concurred that case law establishes a national standard for at least the “serious value” prong.

Here, I want to back up and reiterate my basic concern, as expressed in my affidavit, of how COPA might have affected a “new man in the block” of creative or controversial Internet writing and self-publishing. The sum of my argument depends upon presenting some material having particularly to do with male homosexuality and the way a segment of society views homosexuality as “cheating the system,” that is frankly mature in nature, probably suitable (even important) for high school teens but not for young minors, who might view such material as sensational or prurient.  Since I am self-publishing with my own money and limited resources, a prosecutor and jury might view me as promoting myself in front of children. An important component of my strategy is the offering of most of the material for free on the Internet to attract readers (“passive advertising” as opposed to “guerilla marketing”), as well as to offer hardcopy for sale through the Internet and retail outlets. It would be possible to maintain several arguments.

 First, should not material available to minors be segregated in a way similar to movie ratings? (My material, if filmed, would probably correspond to the R rating.) Second, if I want commercial gain, should I not be expected to have the scale that would protect children even from moderately adult materials? In late 2000 I discussed all of this with a reported at the Minneapolis Star Tribune, and she asked me why I should not be expected to make a (mutually exclusive) choice between print only, or totally non-commercial free Internet content without books to sell. Why, she asked, would this not satisfy First Amendment concerns?

One possible answer is that commercial speech does have considerable protection in prior case law, and a commercial component often enhances the reach, political effectiveness, and public credibility of an individual’s speech. Another possibility is that some communities might view some kinds of subject matter, especially gay-related materials, as prurient with respect to minors by definition. The ACLU raised this possibility in briefs, with mention of the military “don’t ask don’t tell” policy in one place (possibly because of prodding by me—and I’ll add here that I’m pretty sure that the legal clerks in the Supreme Court read all of the legal theories about DADT in my writings, as if I had succeeded in getting the DADT question before the Court in a case that it would actually take).  And end result could be, in theory, that it could be illegal to advertise any gay-related product for sale on the Internet.

Nevertheless, it is possible to read the text of the harmful to minors definitions as being particularly focused upon sexual acts or sexual parts themselves (particularly with the Second Prong) and if so, categorical material like general discussions of homosexuality would not fall into concern (even ignoring the community standards problem) until the speaker became specific with respect to sexual activity itself (as with conveying safer sex information).  The Court shows some concern over the possibility that this would not be the case, as when it mentions discussions of gay parents. Another important concern is that, as a practical matter, the government could use civil lawsuits rather than criminal prosecution to harass speakers when the government believed that only a lower standard of proof, often based on the sympathy of a jury, was achievable.

But this would lead to what the Third Circuit must take up now, the least restrictive means analysis, and the effectiveness of various schemes to separate young minors from adult content. The Supreme Court, in the oral arguments, asked if a self-rating system attached to browsers was feasible. The ACLU responded that this would not be possible for small Internet content providers like myself, but that assertion is not correct.

 The Internet Content Rating Association, for example, is developing a scheme to label web pages with meta-tags, and then parents will be able to configure browsers and ISP services like AOL to block content that has been self-rated in various ways (and the ICRA acts as a third party to corroborate the self-ratings and give them credibility). This is inherently inexpensive for individuals and small businesses, although it may require more software development in browser developers such as Microsoft.

 A more advanced method might be to place adult-like material into XML pages (rather than native HTML), maybe dumped off a relational knowledge database, and then use parsers, schemas and style sheets (associated with the java or similar object-oriented programming language or with other facilities offered, for example, by Microsoft .NET) to filter out content elements according to the profile of the viewer (again, parents would have some responsibility to set this up) and, moreover, the geographical location of the viewer (considering some kind of artificial intelligence profile for the community standards for that location).[4] Software companies would be able to develop and refine such a mechanism if they perceived an economic payoff according to the final legal disposition of this issue. 

As for credit cards and adult-ID services, the ACLU rather than the government is certainly right. Credit cards are only suitable for sites with large transaction volumes, not for content providers who offload their merchant processing to places like Amazon or Barnes and Noble.  And so far adult-ID companies have been offering services only for viewing outright pornography, not for educational adult content. But there is no reason why venture capitalists could not come up with a new style for adult-id services, if the market for them was there.

There are several other pieces of legislation that are sometimes confused with COPA. For example, a Children’s Internet Protection Act (CIPA) effective 2001, would require schools and libraries getting federal funds to install filters for pornography and adult material. This is less draconian than CDA or COPA but still involves government-sponsored censorship, using technology that cannot reliably separate pornography from adult non-pornographic (and First Amendment protected) material. Accordingly, the Third Circuit struck down this law in May 2002 in an opinion that acknowledged an imperfect world and that, given the limited technological reliability of filters with non-pornographic adult content, it would be impermissible to interfere with the communication of some intellectually legitimate ideas to minors because of the imperfections of technology. Unclear is whether filtering only some computers in a children’s room might be legally acceptable, whatever common sense tells us.

There is also a Children’s Online Privacy Protection Act (COPPA, 1998) associated with COPA, and this merely prohibits web sites from collection information it knows to be children under thirteen without consent of parents.  This provision has not been challenged.

Also in 1995, Congress considered making it a crime (at least in pictures) to suggest that a minor is engaging in sexual activity even if an adult is used or the image is generated by computer. Remember the film The Tin Drum, which a district attorney in Oklahoma tried to ban on the theory that a particular bedroom scene constitutes child pornography.  Congress passed the Child Pornography Prevention Act (CPPA) in 1996. On February 12, 2001, ABC “Nightline” covered this problem thoroughly. The Act provides that material  (on the Internet, in books, or in film) which “looks like” child pornography (that is, virtual child pornography) is prosecutable even if no child is used to produce it.  Specifically, in the language of the statute, “`child pornography' means any visual depiction, including any photograph, film, video, picture, drawing, or computer or computer-generated image or picture, whether made or produced by electronic, mechanical, or other means, of sexually explicit conduct, where¾
(A) the production of such visual depiction involves the use of a minor engaging in sexually explicit conduct;
(B) such visual depiction is, or appears to be, of a minor engaging in sexually explicit conduct; or
(C) such visual depiction is advertised, promoted, presented, described, or distributed in such a manner that conveys the impression that the material is or contains a visual depiction of a minor engaging in sexually explicit conduct.” [5]

The book Lolita, which describes a love affair with a twelve-year-old girl, might be illegal.  What about a novel that describes a sex act with a person underage in some states even if (in the fictionalized setting) the person looks like an adult or the character committing the crime in the plot of the novel does not know that the person is underage (an interesting pretext for a novel)?  Is all of this “thoughtcrime”?  What about sex education materials?  Is there a principled distinction between cartoons and “virtual reality” depictions of sex acts?  Of course, many will argue that computer generated images of children in sex acts encourages (by “mental effects”) actual sex acts against children by unstable people. This law has some of the problems of COPA. (Below is a good reference on this Act:  It would appear from this analysis that an affirmative defense is provided if the speaker or distributor does not advertise the material as portraying sex by minors and if real adult actors are used.  However, Jerry Hall of Tate & Bywater raises troubling questions about state jurisdiction at  The Supreme Court heard oral arguments on this, Ashcroft v. Free Speech Coalition, 00-795, on Oct. 29, 2001.  See  The movie Traffic was mentioned, and I know of others that could conceivably be affected by the law (The Deep End).

On April 16, 2002, The Supreme Court overturned this law, voting 6-3. Justice Anthony Kennedy, writing for the majority, said, “The mere tendency of speech to encourage unlawful acts is not a sufficient reason for banning it. The right to think is the beginning of freedom and speech must be protected from government because speech is the beginning of thought.”

It is well to note here that the federal government has recently become very aggressive in pursuing child pornography customers, to the point of setting up stings (such as “Candyman” in early 2002) and then obtaining search warrants to examine the hard drives of customers for cached or deleted images. Possession of child pornography is a strict liability offense much like drug possession, so even clicking upon an image that one knows to be child pornography is considered a crime. Persons caught in these dragnets generally have no criminal records and have been “guilty” only of idle or inappropriate curiosity.[6] It is easy to imagine how such enforcement efforts could someday ensnare persons who may have illegal images on their hard drives without their awareness.

The most important issue for me, in looking at COPA and similar legislation intended to protect children, goes deeper than just hindering access of adults to constitutionally protected materials. It raises the question as to whether certain conventional ideas (particularly those grounded in religious faith), because they are more acceptable to an “average person,” will achieve more attention in public debate than will more daring and challenging notions. The possibility of individuals and small business or political interests to widely disseminate their ideas, at some possible risk to younger children who may be inadvertently exposed to material over their heads, does tend to level political debate in the long run.







Copyright 1999, by Bill Boushka and the Libertarian Party of Minnesota. All rights reserved, subject to fair use. For permissions, please email Bill Boushka ( or Charles Test. Reprinted from the Minnesota Libertarian, July 1998, page 6.

[1] Miller v. California (1973) sets forth a similar three-prong test for obscenity and relies upon community standards. “There is no constitutional barrier under Miller to prohibiting communications that are obscene in some communities under local standards even though they are not obscene in others.”

[2] Hamling v. United States (1974). “This Court, however, rejected Justice Brennans argument that the federal mail statute unconstitutionally compelled speakers choosing to distribute materials on a national basis to tailor their messages to the least tolerant community: The fact that distributors of allegedly obscene materials may be subjected to varying community standards in the various federal judicial districts into which they transmit the materials does not render a federal statute unconstitutional.”

[3] Sable Communications of California v. FCC (1989). Sable addressed the constitutionality of 47 U.S.C. 223(b) (1982 ed., Supp. V), a statutory provision prohibiting the use of telephones to make obscene or indecent communications for commercial purposes. The petitioner in that case, a dial-a-porn operator, challenged, in part, that portion of the statute banning obscene phone messages. Like respondents here, the dial-a-porn operator argued that reliance on community standards to identify obscene material impermissibly compelled message senders to tailor all their messages to the least tolerant community. Relying on Hamling, however, this Court once again rebuffed this attack on the use of community standards in a federal statute of national scope: There is no constitutional barrier under Miller to prohibiting communications that are obscene in some communities under local standards even though they are not obscene in others. If Sables audience is comprised of different communities with different local standards, Sable ultimately bears the burden of complying with the prohibition on obscene messages.

[4] Tracey Baker, “The Dark Side of the ‘Net,” Smart Computing, August 2002, p. 40. Baker mentions at one site, , is  already offering software to determine the location of a client computer viewing a site. It is not yet clear how affordable or practical this is for a small business. Perhaps geographical sensitivity could be implemented in newer object-oriented languages as interfaces


[5] There are many sources of this text, such as


[6] Steve Silberman, “The United States of America v. Adam Vaughn,” Wired, October 2002, p. 126.