View PDF Version

No. 08-565

 

In the Supreme Court of the United States

MICHAEL B. MUKASEY, ATTORNEY GENERAL OF THE UNITED STATES, PETITIONER

v.

AMERICAN CIVIL LIBERTIES UNION, ET AL.

ON PETITION FOR A WRIT OF CERTIORARI
TO THE UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT

REPLY BRIEF FOR THE PETITIONER

GREGORY G. GARRE
Solicitor General
Counsel of Record
Department of Justice
Washington, D.C. 20530-0001
(202) 514-2217

In the Supreme Court of the United States

No. 08-565

MICHAEL B. MUKASEY, ATTORNEY GENERAL OF THE UNITED STATES, PETITIONER

v.

AMERICAN CIVIL LIBERTIES UNION, ET AL.

ON PETITION FOR A WRIT OF CERTIORARI
TO THE UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT

REPLY BRIEF FOR THE PETITIONER

The court of appeals has affirmed a nationwide injunc tion invalidating an important Act of Congress whose con stitutionality this Court has already seen fit to review. In deed, the court of appeals has permanently precluded the Act's enforcement "at any time for any conduct." Pet. App. 148a. That sweeping decision is not only erroneous but, if allowed to stand, will leave millions of children without the protection that Congress carefully crafted against the harmful effects of commercially marketed pornography on the World Wide Web.

Respondents contend that the court of appeals should have the last word on the constitutionality of the Child On line Protection Act (COPA), Pub. L. No. 105-277, Div. C, Tit. XIV, 112 Stat. 2681-736 (47 U.S.C. 231 & note), and insist that this Court has already settled the relevant legal principles. But the legal principles that the court of appeals treated as settled were primarily drawn from its own prior panel opinion, an opinion that this Court pointedly declined to endorse when this case was previously before it.

Far from resolving COPA's constitutionality, this Court examined only one question at the preliminary-injunction stage-whether the promotion of filter software might be a superior alternative to COPA-and it deemed the record inadequate to determine the answer. That question re mains a key issue, as respondents acknowledge. Indeed, they offer scant defense of the court of appeals' rulings on other issues. This Court invited the government to demon strate that promoting filters will not cover the necessary ground, and it has done so.

The court of appeals critically erred in considering that crucial issue. Nearly half of all parents do not use filters; that figure is undisputed. Those children receive no protec tion at all. Nor do children whose parents use filters but who gain access to the Internet on unfiltered computers. The court of appeals dismissed these serious difficulties by positing that parents who do not use filters "trust their chil dren." Pet. App. 44a. That supposition is no substitute for the analysis this Court called for on remand. This Court should again grant review to correct the court of appeals' misconceptions and should consider the First Amendment implications of this important statute with the benefit of a full, more current record about the evolving technology.

A. This Court Should Resolve The Question Of COPA's Con stitutionality, With Due Regard To The Evidence That Pro moting Filters Is Not An Equally Effective Alternative

1. This Court's last opinion, rendered at the prelimin ary-injunction stage and under an abuse-of-discretion stan dard, did not resolve how the relative effectiveness of COPA and filters should be analyzed once a full record was developed at trial. Indeed, it would have been impossible for this Court to resolve that question in advance, as re spondents suggest it did (Br. in Opp. 10), without an up-to- date record about "the deficiencies of filters," the degree to which parents install filters, and the relationship between the two. See Ashcroft v. ACLU, 542 U.S. 656, 668 (2004); Pet. 28. The government has now shown that filters are imperfect; that their defects keep them from being univer sally accepted; and that a substantial share of home com puters leave children unprotected.

Yet on remand, neither the district court nor the court of appeals gave sufficient attention to these key points, es pecially the undisputed facts that nearly half of all parents do not use filters and that filters' consistent tendency to block even legitimate sites is a contributing factor in par ents' decision to deactivate them. See Pet. App. 90a (dis trict court makes no findings about filter use and almost none about overblocking, opining only that "underblocking is the more important concern"); id. at 44a (court of appeals dismisses fact that 46% of parents do not use filters).1 Nor did the courts below attach significance to the finding that one increasingly common tool for children to browse the Web-the mobile phone-does not currently offer a full range of filter technology. Id. at 86a. Instead, the court of appeals noted only that filter use had increased 65% over the previous four years. Id. at 44a. But that increase still left nearly half of all computers in homes with children un protected. And the court of appeals did not address whether the rate of use can be expected to increase to the point of making filters an effective alternative. The evi dence, which the court overlooked, shows that it cannot. See Pet. 28-29.

2. Respondents contend that this Court settled the ap plicable law and left only the facts for development on re mand. But the points of law to which respondents point (Br. in Opp. 7-8) are First Amendment generalities that were not disputed the last time this Court granted review. What was disputed then, and remains disputed today, is how those principles apply to the concededly severe prob lem recognized by Congress: access by children to harmful pornography on the Web. That issue is now presented on a full record. The court of appeals' attempt to apply the relevant First Amendment principles to that record was deeply flawed and warrants this Court's review.

This Court noted in its previous opinion that Congress may take steps to promote filters. Ashcroft v. ACLU, 542 U.S. at 669. And the Court stated that the government had not yet proved that this less restrictive alternative would be less effective. Id. at 670. The government proceeded to shoulder that burden on remand, introducing evidence that approximately half of all parents do not use filters and that government promotion of filters would have only a limited effect. Pet. 28. The government showed, therefore, that even promoting filters would not address the significant problem COPA targets: millions of children with unfiltered access to harmful pornographic content on the Web. Re spondents' suggestion (at 10) that "[t]he government made no attempt to meet that burden" is simply incorrect. Rather, it was the courts below that gave little or no atten tion to the rate of filter use and the prospects for increasing that rate. The court of appeals simply hypothesized rea sons why more parents do not use filters, relying not on pertinent factual findings (there were none) but on evidence submitted by respondents and controverted by the govern ment. See Pet. App. 44a; Pet. 28-29. The invalidation of an important federal statute based on that erroneous analysis warrants this Court's review.

B. Respondents' Position Threatens Effectively To Foreclose Protection Of Children From Pornography On The Web, Despite This Court's Repeated Indications That Appropri ate Regulation Is Possible

1. Congress enacted COPA in response to Reno v. ACLU, 521 U.S. 844 (1997), in which this Court invalidated an Internet-regulation statute because, inter alia, Con gress should have chosen a less restrictive "possible alter native[]," like "regulating some portions of the Internet- such as commercial Web sites-differently." Id. at 879. That is what COPA does. See 47 U.S.C. 231(a)(1), (e)(1) and (e)(2). But respondents now effectively contend that there can be no permissible regulation of sexually explicit content on commercial Web sites that is harmful to chil dren. Respondents give two reasons; neither is persuasive.

First, respondents essentially argue that so long as mi nors can obtain access to pornographic Web sites that are maintained in other countries, COPA's efforts to shield mi nors from such content in this country will be for naught, and therefore "does nothing to advance the government's asserted interest." Br. in Opp. 13. Second, respondents contend (at 11-12) that COPA is per se ineffective because of the exclusions for noncommercial speech and for Internet protocols other than the Web- limitations that Congress wrote into COPA precisely to comply with this Court's decision in Reno. Pet. 5, 6.2

Respondents' bleak themes of surrender and futil ity-both their contention that the present inaction of other countries has rendered the United States helpless to ad dress its own serious problems, and their contention that Congress's good faith efforts to comply with this Court's decision in Reno are, of necessity, self-defeating-provide no basis for denying certiorari, much less invalidating Con gress's effort to address a critical national concern. The First Amendment does not require Congress simply to throw up its hands in the face of the still-present, still-com pelling governmental interest in affording children in the United States adequate protection from the concededly harmful effects of pornography on the Web. And even the courts below did not adopt these absolutist positions. See Pet. App. 23a-26a (declining to adopt argument that not including foreign sites would make COPA underinclusive); id. at 135a-136a (rejecting argument that covering only the Web is underinclusive). Moreover, this was so even though they misread this Court's 2004 opinion to establish conclu sively that COPA cannot be enforced against foreign Web sites at all. Pet. 29-30; see also pp. 10-11, infra (discussing the geographic scope of enforcement).

At the preliminary-injunction stage, this Court was well aware that COPA likely would not prevent access to at least some foreign content (though that content might be blocked by filters,3 where they are used) and that COPA does not apply to non-Web protocols. See Ashcroft v. ACLU, 542 U.S. at 667-668. Indeed, the district court and the court of appeals had endorsed a theory of underinclusiveness like the one respondents advance here. Pet. App. 188a. Had this Court agreed, it could have declared the case over. Cf., e.g., Legal Servs. Corp. v. Velazquez, 531 U.S. 533 (2001) (invalidating statute on First Amendment grounds at pre liminary-injunction stage). Instead, it declined to adopt any of the court of appeals' reasoning, noted that the eviden tiary record was silent on relevant aspects of the filter-soft ware alternative, and remanded for a trial principally on the relative efficacy of enforcing COPA and promoting fil ters. This Court presumably would not have remanded for a pointless exercise; the natural conclusions are that it un derstood that both filters and COPA are imperfect solu tions, and that it remanded to allow the government to es tablish that COPA satisfied constitutional standards.

2. Respondents, joined in this respect by the court of appeals (Pet. App. 44a), also contend that if a parent does not prevent her child from gaining access to harmful online content, the government has no further interest in protect ing that child. Br. in Opp. 13. Once again, respondents effectively contend that no regulation of pornography on the Web is possible and that the protection of children from commercial online pornography must depend on voluntary undertakings. And once again, the First Amendment does not leave Congress so helpless to protect children (and to assist parents in protecting their children) from concededly harmful material.

There is no question here of "overriding the decisions of parents," as respondents suggest. Br. in Opp. 13. Under COPA, parents can permit their children to have access to any site they wish-for instance, by allowing them to use one of the access credentials permitted by the statute, such as a credit card. Pet. 6 n.1; see Pet. 24-25 (discussing the operation of COPA's affirmative defense). Without COPA, however, many children would remain unprotected, even though their parents do not want them to view porno graphic Web sites, because their parents would fail (for whatever reason) to install filters on the family computer. And even though some parents do install filters, that does nothing to address the possibility that their children will access the Internet somewhere else, "hack" the filter soft ware, or simply bypass the filter by stealing their parents' passwords. Pet. 29. Thus, even if filters worked perfectly and did not cause parents to remove them out of frustration with overblocking, see pp. 3-4, supra,4 they would not be universally adopted, and they would not be foolproof. COPA works in these situations where filters do not, and in most such cases COPA will be effectuating, not overriding, parents' wishes to keep their children's Web browsing safe.

3. In each of these respects, respondents contend that no matter how well or how poorly filters work, and no mat ter how widely or how narrowly filters are actually used, filters are better than COPA because they are voluntary, catch foreign material, and regulate some non-Web proto cols. That reasoning, if accepted, would mean that despite Congress's concededly compelling interest in protecting children and the clear import of Reno that there are valid means to achieve that interest, Congress cannot regulate the Internet beyond encouraging voluntary measures by parents. But if our Constitution compels such a meager and defeatist approach to this critical problem, such a hold ing should come from this Court, not from a single court of appeals whose nationwide injunction resolves the matter forever.

C. The Court Of Appeals' Misconstruction Of COPA Warrants Review

The court of appeals' holding that COPA is ineffective, overinclusive, overbroad, and vague rested in large part on the court's mistaken interpretation of several of COPA's key elements. Pet. 21-26. Those interpretations came pri marily from the preliminary-injunction panel's opinion, which this Court declined to endorse. Respondents suggest (at 19) that those misinterpretations are "irrelevant," be cause at the earlier stage they "did not prevent this Court from affirming the preliminary injunction" based on uncer tainty about the relative effectiveness of COPA and filter promotion. To the contrary, properly understanding COPA's scope is essential to understanding its likely effec tiveness.

1. Most significantly, COPA's affirmative defense re buts respondents' argument that COPA "criminalizes valu able speech" and that "adults will be deprived of constitu tionally protected speech." Br. in Opp. 1, 19. That argu ment depends on the court of appeals' conclusion that the affirmative defense is "effectively unavailable," Pet. App. 29a (quoting id. at 127a). If the defense is available, then respondents may display as much sexually explicit (but non-obscene) material on the Web as they wish; they need only use one of various "reasonable measures" to limit chil dren's access to material covered by COPA. 47 U.S.C. 231(c)(1).

Respondents offer no defense at all of the court of ap peals' reasoning, which wrongly concluded that because these "reasonable measures" are not perfect their good- faith use would not be honored under COPA. See Pet. 24- 25. Respondents say only that the district court "made over 50 specific findings of fact" on this issue, Br. in Opp. 19, but that has nothing to do with the lower courts' legal interpre tation of Section 231(c)(1), which the government has con sistently challenged and which is reviewed de novo.

2. Respondents also give only a cursory defense of the court of appeals' conclusion that, as a matter of law, this Court's previous opinion establishes that COPA has no ap plication to foreign sites. See Br. in Opp. 11 n.2. As already explained, that conclusion was erroneous. Pet. 29-30. In stead, respondents principally argue that the court of ap peals' error does not matter because "practical consider ations" would ensure that COPA is not enforced against foreign sites. Br. in Opp. 11 n.2. That contention is inade quate to defend the court of appeals' incorrect reasoning.

In fact, as the government showed at trial, a significant portion of content that appears "foreign" from its Internet domain name is actually hosted on servers in the United States, and even truly foreign sites operating for commer cial purposes contract with American credit cards for pay ment. Pet. 30; C.A. App. 444-445. These contacts give the United States regulatory leverage that respondents ignore.

And in any event, the government's evidence shows that even reaching only overtly domestic Web sites-which by themselves amount to half the world's online pornogra phy-would have a disproportionately salutary effect. An overwhelming majority of "feeder" adult Web sites (which offer explicit content to induce visitors to pay for more ac cess) are hosted in the United States. C.A. App. 1264.

3. Respondents similarly offer no defense of the court of appeals' holdings on narrow tailoring, vagueness, and overbreadth. They say only that the embedded statutory interpretations (of the terms "commercial purposes," "seri ous * * * value for minors," and "as a whole") were cor rect and, in any event, irrelevant. Br. in Opp. 16, 17. But because, as set forth above, the court of appeals' discussion of filters was flawed, it is necessary to reexamine the court's application of these other First Amendment princi ples as well. The court of appeals deliberately left in place its previous holdings that this important federal statute is unconstitutional. This Court granted certiorari on those issues before, and should do so again. Congress has sought to address an important national problem concerning the welfare of our children in a manner consistent with this Court's precedents. This Court should review the decision invalidating that effort on the full evidentiary record devel oped below.

* * * * *

For the foregoing reasons and those stated in the peti tion for a writ of certiorari, the petition should be granted.

Respectfully submitted.

GREGORY G. GARRE
Solicitor General

 

 

DECEMBER 2008

1 The district court did enter a single finding quibbling with some of the overblocking rates found by the government's expert in his study of filter technology. Pet. App. 95a-96a. Respondents contend (at 12 n.3) that this finding made the government's overblocking evidence "unre liable." But in fact the district court quarreled with only the expert's findings about the America Online (AOL) filter. The expert's findings about 12 other filter products and configurations remain unrebutted, and he found overblocking rates as high as 19.6% even on settings designed to filter only pornography. C.A. App. 773, 796. The AOL fi lter has no such pornography-only setting, see Pet. App. 96a, and accor dingly parents would have the same problem that the expert did.

2 Yet respondents contend at the same time that COPA's limitation to sites operated for "commercial purposes" narrows COPA only "slightly." Br. in Opp. 16.

3 Because filters often rely on key words and contextual clues, see Pet. App. 80a-82a, the software would often have trouble blocking sites whose text is in a foreign language.

4 Respondents suggest that parents do not use filters because they trust their children, but they cite a survey of parents who have never used filters. See Br. in Opp. 13. Parents who have removed filters-a highly relevant subsample-were surveyed separately, and report that the principal reason for removing filters is that they are too restrictive. C.A. App. 1568.