4. Summary In sum, we reject a public library’s interest in preventing unlawful or otherwise inappropriate patron conduct as a basis for restricting patrons’ access to speech on the Internet. The proper method for a library to deter unlawful or inappropriate patron conduct, such as harassment or assault of other patrons, is to impose sanctions on such conduct, such as either removing the patron from the library, revoking the patron’s library privileges, or, in the appropriate case, calling the police. We believe, however, that the state interests in preventing the dissemination of obscenity, child pornography, or in the case of minors, material harmful to minors, and in protecting library patrons from being unwillingly exposed to offensive, sexually explicit material, could all justify, for First Amendment purposes, a public library’s use of Internet filters, provided that use of such filters is narrowly tailored to further those interests, and that no less restrictive means of promoting those interests exist. Accordingly, we turn to the narrow tailoring question. 2. Narrow Tailoring
Having identified the relevant state interests that could justify content-based restrictions on public libraries’ provision of Internet access, we must determine whether a public library’s use of software filters is narrowly tailored to further those interests. “It is not enough to show that the Government’s ends are compelling; the means must be carefully tailored to achieve those ends.” Sable Communications of Cal., Inc. v. FCC, 492 U.S. 115, 126 (1989). “[M]anifest imprecision of [a] ban . . . reveals that its proscription is not sufficiently tailored to the harms it seeks to prevent to justify . . . substantial interference with . . . speech.” FCC v. League of Women Voters of Cal., 468 U.S. 364, 392 (1984). The commercially available filters on which evidence was presented at trial all block many thousands of Web pages that are clearly not harmful to minors, and many thousands more pages that, while possibly harmful to minors, are neither obscene nor child pornography. See supra, Subsection ii.E.7. Even the defendants’ own expert, after analyzing filtering products’ performance in public libraries, concluded that of the blocked Web pages to which library patrons sought access, between 6% and 15% contained no content that meets even the filtering products’ own