Cites & Insights: Crawford at Large
ISSN 1534-0937
Libraries · Policy · Technology · Media

Selection from Cites & Insights 4, Number 10: August 2004

The Censorware Chronicles

The major development is still a work in progress: COPA was sent down by the Supreme Court for the second time. Beyond that, a California congressman introduced a new bill to turn parents (or almost anyone else) into prosecutors—and Mary Minow offered a thoughtful commentary on how libraries should deal with CIPA.


The Child Online Protection Act (COPA) passed Congress in 1998. That’s right: Clinton signed it. With censorship as with strong copyright, party lines don’t matter. Briefly, COPA imposes a $50,000 fine and six months in prison for the knowing posting, for commercial purposes, of web content that is “harmful to minors”—unless the company or person doing the posting can demonstrate that they restrict access to such materials to people 17 and older.

As I’ve been reminded, CIPA was not a direct response to COPA—but COPA was a fairly direct response to the Supreme Court overturning the Communications Decency Act of 1996. The ACLU and others filed suit for a preliminary injunction against COPA enforcement. The District Court granted that preliminary injunction and the Third Circuit affirmed the injunction based on COPA’s use of “community standards” as a basis for determining the status of material. When the government appealed to the Supreme Court, it remanded the case back to the Third Circuit, saying that the community-standards issue alone did not make the law unconstitutional. The Third Circuit affirmed the injunction a second time on broader grounds, including the assertion that COPA was not the least restrictive means available for the Government to serve its interests. Note that all of these proceedings have to do with enjoining the government from prosecuting COPA cases—not with overturning the law itself. That requires a trial, which has been in abeyance through the two rounds of appeals.

Arguments before the Supreme Court

These arguments took place in early March 2004. Solicitor General Theodore Olsen, arguing to overturn the injunction, used a web search (probably Google) to illustrate the extremity of “online smut.” Type in the words “free porn” and you get a list of 6,230,000 websites, he said: “I didn’t have time to go all the way through those sites.” According to an AP story, Olson called internet porn “persistent and unavoidable” and said the government has a strong interest in shielding teenagers and younger children from it. “There is a compelling government interest in protecting minors from the effects of material that is not obscene by adult standards but that is nonetheless harmful to minors. The Web poses a serious threat to that compelling interest.”

Here’s where things get interesting and confusing, if you’ve followed arguments on CIPA. ACLU lawyers argued that other tools are less intrusive—such as filtering software (censorware). So ACLU is arguing in favor of censorware in one case and against it in another? Yes—and that’s not necessarily a contradiction. More on that later.

Immediate Commentary

Seth Finkelstein offered two commentaries at his Infothought blog shortly after the Supreme Court arguments. The first one deconstructed Olson’s “6,230,000” claim; the second addressed the real difficulty in the anti-COPA argument.

In the first, Finkelstein did “something often unrewarded in this world—think.What search did [Olson] do exactly?” Finkelstein assumed that it was the two words “free” and “porn” (without quotes) in Google. He got “about 6,320,000” results the next day, which (as he notes) is “close enough; the total number returned often varies a bit.”

But what does that search mean? Not that 6.3 million web pages offer free pornography; not even that 6.3 million contain the phrase “free porn.” It means that this many pages apparently have the word “free” somewhere on the page or in links to the page, and also the word “porn” somewhere on the page or in links to the page. The AP story from the previous day is one such page (it even has the phrase), as are all the other reports of Olson’s testimony—none of which could remotely be considered pornographic. So, to be sure, is Finkelstein’s page.

That’s not all. Finkelstein started digging into the results “to see if I could find some non-sex-site mentions before the Google 1000 results display limit.” (Olson could not have gone all the way through those sites if he had wanted to: Google doesn’t allow it.) What happened? He got to the 876th result and got the Google “omitted entries” message. You know the one:

In order to show you the most relevant results, we have omitted some entries very similar to the 876 already displayed.

If you like, you can repeat the search with the omitted results included.

Trying a couple more times, he got different numbers—but always less than 900. I’m going to quote two paragraphs (separated by other paragraphs) with the note that it wouldn’t take much to remove the first word of each paragraph and treat them as arguments deserving as much weight as Olson’s “6,230,000” claim:

Joke: Hear ye! Hear ye! Instead of “6,230,000 sites available,” there’s really uniquely less than 900! At least, according to Google.

Humor: If the evidence from a Google search was good enough to be used to justify censorship when it said “6.2 million,” why isn’t it good enough to justify no censorship if on further investigation it says less than 900? That is, if you thought it was valid before, with a big number, why isn’t it valid now, with a small number?

Finkelstein attaches “(garbage in, garbage out)” at the end of that second paragraph—but I think he’s made a good case. How much “porn” is actually freely available on the web? Nobody knows—and estimates tend to be made by people with heavy stakes in the results. Never mind that many of the 876, or 6.3 million, or some number in between, aren’t porn sites at all. Never mind that the definition of “harmful to minors” is either nebulous or, as in COPA and CIPA, extremely difficult to prove.

Finkelstein’s second post deals with the problematic nature of ACLU’s argument that censorware is better than COPA. The problem is that this argument requires these advocates to speak favorably of censorware—which means they tend to discredit those who say that censorware just plain doesn’t work. Which apparently accounts for much of Finkelstein’s difficulties before he gave up on censorware investigation. There’s strong documentary evidence that Mike Godwin and others did precisely that: Discourage people from demonstrating how defective censorware is because they saw it as the “intelligent” alternative to COPA and its ilk.

How can I reconcile the ACLU’s argument in COPA and ALA’s argument against CIPA? One problem with CIPA is that it’s a federal mandate for censorware—and for a kind of censorware that may or may not exist (censorware that only blocks images and that blocks images in the narrow categories of child pornography, obscenity, and the “harmful to minors” equivalent). Saying that optional parental use of censorware at home, using choices that suit the parents, is preferable to heavy-handed legal actions against smut peddlers, isn’t inherently at odds with saying that enforced library use of censorware on all computers used by all ages is a bad idea. (I’ve said more than once that I disagree with ALA’s absolutist policy on age neutrality. I have no real problem with sufficiently tailored censorware being used in children’s areas, or even [and perhaps preferably] children’s computers restricted to whitelists of library-approved sites. But when I say “children,” I mean a group that certainly ends at an age lower than 17!)

The Decision

The Supreme Court issued its decision on June 29. You should be able to find it easily enough (it’s No. 03-218 of the October term, 2003, if that helps). The Supreme Court held that “the third circuit was correct to affirm the District Court’s ruling that enforcement of COPA should be enjoined because the statute likely violates the First Amendment.”

This ruling does not, in and of itself, strike down COPA, although one might hope the Government wouldn’t go to the expense of insisting on an actual trial that’s so likely to result in striking down the law. The ruling keeps the anti-enforcement injunction in place. Meanwhile, technically, the six-year-old law is also in place but unenforceable.

The majority ruling does rely on censorware to support its finding that there are less restrictive ways to meet Congress’ goal of discouraging minors from gaining access to online “harmful to minors” material. For that matter, censorware could do a better job: By estimates cited in the decision, at least 40% of the “free porn” originates outside the U.S. and is immune to COPA—and, presumably, intelligent pornsters would move their servers offshore if COPA was upheld, increasing that percentage.

I’ll only cite a few comments from the 15-page opinion signed by Justice Kennedy. “Content-based prohibitions, enforced by severe criminal penalties, have the constant potential to be a repressive force in the lives and thoughts of a free people.”

How well would COPA protect minors? Quite apart from the foreign-smut problem, a site can avoid prosecution by requiring use of a credit card, debit account, digital certificate verifying age, or any of a number of other certifications of adulthood. In other words, any kid who wants to look at the nasty stuff need only copy down one of their parents’ credit card numbers—and I’d guess that a lot of people under 17 have their own credit or debit cards.

Kennedy points out that content-based restrictions on speech must meet a high standard. Even assuming that certain protected speech may be regulated, the court “asks what is the least restrictive alternative that can be used to achieve that goal.”

The purpose of the test is not to consider whether the challenged restriction has some effect in achieving Congress’ goal, regardless of the restriction it imposes. The purpose of the test is to ensure that speech is restricted no further than necessary to achieve the goal, for it is important to assure that legitimate speech is not chilled or punished. For that reason, the test does not begin with the status quo of existing regulations, then ask whether the challenged restriction has some additional ability to achieve Congress’ legitimate interest. Any restriction on speech could be justified under that analysis. Instead, the court should ask whether the challenged regulation is the least restrictive means among available, effective alternatives.

The First Amendment doesn’t actually say “The Congress shall only make laws regulating freedom of speech when such laws forward other legitimate government interests.” It says something about Congress making no laws—but, you know, the Supreme Court doesn’t want to slap Congress around just because it ignores the Constitution. Getting back to the decision, Kennedy notes that the Commission on Child Online Protection, a “blue-ribbon commission” created in COPA itself, evaluated different means of restricting minors’ ability to gain access to harmful materials on the internet. “It unambiguously found that filters are more effective than age-verification requirements.”

Kennedy admits censorware isn’t a perfect solution as it both overblocks and underblocks—but the government offered no evidence that COPA would be more effective than filters.

Stevens (Ginsburg joining) adds a concurring opinion noting that COPA’s reliance on community standards should have been enough to strike it down (he dissented in the earlier case). Stevens goes on to consider just how restrictive COPA actually is, particularly since the burden of proof would be on the defendants: They would have to prove that they had age restrictions in place. Stevens also notes the fuzzy nature of obscene material and the untested new category of “harmful to minors.” “Attaching criminal sanctions to a mistaken judgment about the contours of the novel and nebulous category of ‘harmful to minors’ speech clearly imposes a heavy burden on the exercise of First Amendment freedoms.”

Justice Scalia dissents in a brief opinion that mostly cites his other dissents. He doesn’t believe that smut deserves constitutional protection: “Nothing in the First Amendment entitles the type of material covered by COPA to [strict scrutiny].” To Scalia, apparently, smut simply isn’t speech, so the First Amendment doesn’t apply. Indeed, he states flatly that porn “could, consistent with the First Amendment, be banned entirely.” Whew.

Breyer (Chief Justice and Justice O’Connor joining) also dissented, in a statement that’s longer than the majority opinion. He asserts that the “harmful to children” category covers “very little more” than legally obscene material, considers the burden imposed by COPA to be “no more than modest,” and goes on in some detail about the modified test. As I read Breyer’s discussion, one might conclude that there is no category—which means that COPA is pointless. That is: Young children normally don’t have prurient interests. Material that appeals to the prurient interests of adolescents probably also appeals to such interests for some adults. The third test? “One cannot easily imagine material that has serious literary, artistic, political, or scientific value for a significant group of adults, but lacks such value for any significant group of minors.” If that’s true, then COPA should be struck down as being redundant: Obscenity is already illegal. If there’s no new category, what’s the point?

Breyer also objects to the decision because censorware already existed when Congress passed COPA. Thus, it was part of “the status quo, i.e., the backdrop against which Congress enacted the present statute.” He argues against censorware on several other bases—it costs money, it depends on parents’ willingness to control their children, and it lacks precision.

Then, late in the dissent, he slips. After consistently suggesting that COPA would only add some tiny amount of “borderline-obscene” material to that which can already be regulated, he says something quite different: COPA “significantly helps to achieve a compelling congressional goal, protecting children from exposure to commercial pornography.” Most of which, according to most of the court, is protected by the First Amendment.

A Few Post-decision Commentaries

Andrew Mutch posted a brief report on Web4Lib the day of the decision. He found it interesting that “the majority opinion looked to filtering software as a less-restrictive alternative to achieve the goals of COPA”—and pointed out that the quandary faced by ACLU and ALA affects both sides of the case:

Having convinced the Court in the ALA case [CIPA] that filters were effective controls on pornographic material, the government suddenly found that argument turned against its arguments in COPA that filters weren’t an effective tool to stop children from accessing pornographic web sites. On the flip side, the free speech coalition found itself in the position of arguing that the use of filters by parents and others was a preferred alternative to the restrictions that COPA would have imposed.

Seth Finkelstein cited those comments and provided a little more well-documented history:

This is actually an extremely old horns-of-dilemma, going back to the original Internet censorship debates in the mid-90’s concerning the Communications Decency Act. The politics is often counter-intuitive, quite different from what would naively derive from simple models of censorhip positions.

In fact, when I first decrypted censorware in 1995, I was specifically asked not to publish the results, because the civil-liberties strategy was to argue that censorware works, as part of a legal and social campaign…

Tony Mauro offered an analysis at the First Amendment Center Online, noting that COPA may return to the Supreme Court once more—“And by the time it looks at the law again, the Supreme Court’s composition may have changed.” Mauro notes that the language of the majority decision “was heartening to First Amendment advocates.” He also quotes Justice’s Mark Corallo who was predictably “disappointed” that the Court let the First Amendment get in the way of prosecution:

Congress has repeatedly attempted to address this serious need and the Court yet again opposed these common-sense measures to protect America’s children. The Department will continue to work to defend children from the dangerous predators who lurk in the dark shadows of the World Wide Web.

I like Mauro’s reading of Breyer’s dissent:

A frustrated Justice Stephen Breyer, writing in dissent, reviewed the history of the law and asked, “What else was Congress supposed to do?” He said Congress had read the Supreme Court’s Reno decision [striking down CDA] “with care” and shaped its second legislative effort to respond to its criticisms.

A Supreme Court overwhelmingly sensitive to the Bill of Rights would have a simple answer to Breyer’s question: “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press...” In other words, give it up.

Finally, Marjorie Heins published “The right decision; the wrong reason” at the Free Expression Policy Project. She says it’s the wrong reason because “the Court endorsed a technology with the potential for far greater censorship than COPA or similar laws—Internet filters.”

As to Kennedy’s argument that filters are less restrictive than COPA, she agrees in part:

Filters are less restrictive, in part, because nobody goes to jail for failing to use them. But their use is not always voluntary. [As in CIPA]… And, equally important, filters are notorious for mindless overblocking….

Equally troubling as its intoxication with filtering technology, the Court majority said not a word about the vagueness and subjectivity of such standards within COPA as “patently offensive” or “serious value.”

She discusses that issue in more detail, then notes that Justice Breyer’s dissent was the only one to address the question of presumed harm from sexual expression, “although he did so unintentionally.

It’s a good brief discussion; you’ll find it at

Parent’s Empowerment Act

H.R. 4239, introduced April 28 by California Congressman Duncan Hunter (R), would “provide a civil action for a minor injured by exposure to an entertainment product containing material that is harmful to minors, and for other purposes.”

That’s what the summary says, but that’s not at all what the bill says—which should come as no surprise to anyone who’s looked at legislation related to either copyright or censorship. You can find the bill on Thomas (start at What the bill really does is to allow a parent—or any other “person acting on behalf of the minor”—to prosecute a civil action in federal district court to obtain relief

against any person who knowingly sells or distributes in interstate or foreign commerce an entertainment product containing material that is harmful to minors, if—

(1) a reasonable person would expect a substantial number of minors to be exposed to the material; and

(2) the minor as a result of exposure to that material is likely to suffer personal or emotional injury or injury to mental or moral welfare.

“Relief” means “compensatory damages” to the minor of “not less than $10,000 for each instance of any such material in any such product to which such minor was exposed,” as well as punitive damages, a “reasonable attorney’s fee (including expert fees),” and “any other appropriate relief.”

If the kid’s exposed to smut because the parent or guardian has it lying around, there’s no case.

The definitions are also interesting—and, I suggest, loose enough that H.R. 4239 would be on the fast track to judicial oblivion if it was ever passed. “Entertainment product” means any medium, from photo to video game to pamphlet to sound recording. “Material that is harmful to minors”:

The term “material that is harmful to minors” means any pornographic communication, picture, image, graphic image file, article, recording, writing, or other pornographic matter of any kind that is obscene or that—

(A) the average person, applying the contemporary standards of the adult community in which the minor resides with respect to what is suitable for minors, would find, taking the material as a whole and with respect to minors of the ages that the person reasonably would expect to be exposed to the material—

(i) is designed to appeal to, or is designed to pander to, the prurient interest in nudity, sex, or excretion, with respect to minors; and

(ii) depicts, describes, or represents, in a manner patently offensive with respect to minors, an actual or simulated sexual act or sexual contact, an actual or simulated normal or perverted sexual act, or a lewd exhibition of the genitals or post-pubescent female breast; and

(B) a reasonable person would find, taken as a whole, lacks serious literary, artistic, political, and scientific value for minors sufficient to overcome the pernicious effect of that material. [Emphasis added]

This bill also defines “minor” as under 18, not 17.

Note the use of community standards—and, even more seriously, the phrase (italicized here) that weakens the “merit” provision of the usual test for obscenity. In essence, this law would be an open invitation for the “Family” organizations or any other group to file massive numbers of civil lawsuits in the most favorable jurisdictions, taking to court Internet service providers, librarians, bookstores, what have you. Unhappy with a juvenile book at the library that includes any vaguely sexual language? Can’t get the library to remove it? Take them to court—as long as you have a handy juvenile around who finds their “moral welfare” likely to suffer from that nastiness.

Hunter’s blatant about it: He wants the bill to “turn parents into prosecuting attorneys fighting a wave of obscenity,” according to a post at the Comic Book Legal Defense Fund. But of course it’s not obscenity Hunter’s on about—it’s a much broader area. I’d hope this bill will be laughed out of committee, but I know better than to overestimate the concern of Congress for the First Amendment.

Lawfully Surfing the Net

Minow, Mary, “Lawfully surfing the net: Disabling public library internet filters to avoid more lawsuits in the United States,” First Monday 9:4 (April 2004).

Yes, it’s 34 pages long, but that includes ten pages of notes—and Mary Minow is always readable. As a librarian and lawyer, she also knows what she’s talking about. This article was extremely timely, coming as it did three months before the deadline for CIPA compliance. If you’re in a public library and still not sure whether your approach to CIPA is correct (assuming you didn’t turn down the e-rates), read this article.

That’s especially good advice if you’ve implemented censorware that blocks text as well as images, that doesn’t have narrowly-defined categories relating directly to CIPA requirements, or—and this is important—that doesn’t have provisions for instant disabling on proof of age.

Minow introduces two libraries with different CIPA implementations:

Cautious Public Library installs filters and tries to follow the law as written: its policy is to unblock a site when an adult patron asks a librarian to unblock the site for bona fide research. Cautious Public Library will disable the entire filter rarely, if ever.

Quick Public Library also installs filters, choosing a vendor that offers minimum blocking. Its policy allows patrons to sit at a public terminal, and select FILTERED or FILTER DISABLED access, after clicking that they are at least 17 years of age. No librarian intervention is required.

Now, if you’ve read my “CIPA special” and think that Minow interprets the Supreme Court’s actual decision in the same extreme way I did, you know the outcome: Quick Public Library is doing it right, while Cautious Public Library is setting itself up for an “as applied” lawsuit. As this article makes clear, Minow—who, unlike me, is a lawyer—does read the decision similarly. I find this heartening. If you’re in a library that would just as soon keep the filters operating full time on all the computers, you’d better have a library policy that provides a legal basis for that decision—and you shouldn’t be surprised when the local ACLU chapter sues.

I’m not going to summarize or annotate a 24-page article. There’s a lot here, including a convenient recap of CIPA’s key provisions.

One further note, though. “A quick disabling policy is not only truer to the professional ideals of intellectual freedom, it’s also legally safer than the cautious disabling policy.” Minow believes that CPL is far more likely to be sued than is QPL, “which in fact bears very little risk of litigation under federal law.” As she notes later, six of the nine justices interpreted CIPA to mean that an adult patron “need only request unblocking, and the library will do so,” without further discussion or investigation of motive. That’s not how the law originally read, but “the Supreme Court’s interpretation trumps the plain text of a statute.” The question of “bona fide research” is also moot: “or other lawful purposes” means that a patron need only say “I feel like lawfully surfing the Net, and I don’t want it filtered.”

Cites & Insights: Crawford at Large, Volume 4, Number 10, Whole Issue 53, ISSN 1534-0937, is written and produced by Walt Crawford, a senior analyst at RLG. Opinions herein do not reflect those of RLG. Comments should be sent to Cites & Insights: Crawford at Large is copyright © 2004 by Walt Crawford: Some rights reserved.

All original material in this work is licensed under the Creative Commons Attribution-Non­Commercial License. To view a copy of this license, visit or send a letter to Creative Commons, 559 Nathan Abbott Way, Stanford, California 94305, USA.