I often hear arguments that the Supreme Court has held that school boards are limited by the Free Speech Clause in their ability to remove books. A commenter on the first thread in this chain offers an example:
The distinction is that the school library already had the copy and then removed it. While many will debate whether that distinction makes a difference, it does under Supreme Court jurisprudence.
In fact, the U.S. Courts Web site says the same. [UPDATE: I e-mailed the Administrative Office of the U.S. Courts about this, and they removed the erroneous description.]
But this turns out not to be an accurate statement of what the Supreme Court has actually held. This issue was indeed before the court in Board of Ed. v. Pico, and four Justices did take the view that library removal decisions were generally unconstitutional if they were motivated by disapproval of the ideas that the book expresses (though would be permissible if they were motivated by other, supposedly more neutral, factors, such as the book’s vulgarity or age-inappropriateness or inaccuracy). And the four Justices were joined by one Justice in affirming the lower court’s decision, which refused to grant summary judgment in favor of the school board.
But there were precisely the same number of votes — four — for the view that even viewpoint-based removal decisions were generally constitutional (except in the narrow situation where the disagreement was based on pure partisanship, for instance if a Democrat-run board removed books because they were written by Republicans or because they praised Republicans) as there were for the view that such decisions were generally unconstitutional. The swing vote, Justice White, deliberately did not opine on the question; Pico thus left the issue 4-4.
Why did Justice White agree as to the bottom line action — affirming the court of appeals decision — with those Justices who thought viewpoint-based removals were unconstitutional? Simply because he did not want the issue resolved at that point, and procedurally the way to avoid that was to affirm. Here’s what happened, in Justice White’s own words (emphasis added):
The District Court found that the books were removed from the school library because the school board believed them “to be, in essence, vulgar.” Both Court of Appeals judges in the majority concluded, however, that there was a material issue of fact that precluded summary judgment sought by petitioners. The unresolved factual issue, as I understand it, is the reason or reasons underlying the school board’s removal of the books. I am not inclined to disagree with the Court of Appeals on such a fact-bound issue and hence concur in the judgment of affirmance. Presumably this will result in a trial and the making of a full record and findings on the critical issues.
The plurality seems compelled to go further and issue a dissertation on the extent to which the First Amendment limits the discretion of the school board to remove books from the school library. I see no necessity for doing so at this point. When findings of fact and conclusions of law are made by the District Court, that may end the case. If, for example, the District Court concludes after a trial that the books were removed for their vulgarity, there may be no appeal. In any event, if there is an appeal, if there is dissatisfaction with the subsequent Court of Appeals’ judgment, and if certiorari is sought and granted, there will be time enough to address the First Amendment issues that may then be presented.
I thus prefer the course taken by the Court in Kennedy v. Silas Mason Co., 334 U.S. 249 (1948), a suit involving overtime compensation under the Fair Labor Standards Act. Summary judgment had been granted by the District Court and affirmed by the Court of Appeals. This Court reversed, holding that summary judgment was improvidently granted, and remanded for trial so that a proper record could be made. The Court expressly abjured issuing its advice on the legal issues involved. Writing for the Court, Justice Jackson stated:
“We consider it the part of good judicial administration to withhold decision of the ultimate questions involved in this case until this or another record shall present a more solid basis of findings based on litigation or on a comprehensive statement of agreed facts. While we might be able, on the present record, to reach a conclusion that would decide the case, it might well be found later to be lacking in the thoroughness that should precede judgment of this importance and which it is the purpose of the judicial process to provide.
“Without intimating any conclusion on the merits, we vacate the judgments below and remand the case to the District Court for reconsideration and amplification of the record in the light of this opinion and of present contentions.”
We took a similar course in a unanimous per curiam opinion in Dombrowski v. Eastland, 387 U.S. 82 (1967). There we overturned a summary judgment since it was necessary to resolve a factual dispute about collaboration between one of the respondents and a state legislative committee. We remanded, saying: “In the absence of the factual refinement which can occur only as a result of trial, we need not and, indeed, could not express judgment as to the legal consequences of such collaboration, if it occurred.”
The Silas Mason case turned on issues of statutory construction. It is even more important that we take a similar course in cases like Dombrowski, which involved Speech or Debate Clause immunity, and in this one, which poses difficult First Amendment issues in a largely uncharted field. We should not decide constitutional questions until it is necessary to do so, or at least until there is better reason to address them than are evident here. I therefore concur in the judgment of affirmance.
So that’s why Pico has no precedential value on this question. The Court’s fractured decision in United States v. American Library Ass’n — where there was also no majority opinion — doesn’t resolve the issue, either. And no other decisions outside the library context dictate, or in my view strongly suggest, a result. If the library were treated as a “designated public forum” that’s generally open for a nearly limitless variety of speech, then the library wouldn’t be able to set up viewpoint-based restrictions on such speech. But library shelving decisions have never been treated as such a forum, because the choice of what books to select in the first place inherently involves some content-based and often some viewpoint-based judgment. There’s no caselaw that squarely tells us whether there are nonetheless constitutional constraints on such judgment, or whether removal decisions are constitutionally different from selection decisions.
In some ways, NEA v. Finley is somewhat on point, because it too involves government action that aims to impose some quality judgments, and that is inherently content-based. And Finley does suggest that “invidious viewpoint discrimination” in such judgments may be unconstitutional, though presumably non-invidious viewpoint discrimination would be permissible.
But Finley‘s point is just a suggestion — the Court specifically stresses that “we have no occasion here to address an as-applied challenge in a situation where the denial of a grant may be shown to be the product of invidious viewpoint discrimination,” and while there’s also wording there that suggests such invidious viewpoint discrimination would be impermissible, there’s no square holding. There is also no definition of when viewpoint discrimination becomes “invidious,” and it’s not clear to what extent the Finley case, involving arts grants, would carry over to the public library context (and especially public library books aimed at children).
(Note that Pico was a school library case, and one could certainly argue that decisions as to other public libraries, and especially the adult-aimed collections of those libraries, are constitutionally different from decisions as to school libraries or as to the children’s collections of other public libraries. But while this isn’t an implausible argument, it is again not one that is firmly supported by existing precedent.)
So all this should make clear, I think, that there’s no answer from the Supreme Court on the subject; nor to my knowledge is there a broad and firm consensus of lower courts. My sense is that, when it comes to shelving and removal decisions, then-Justice Rehnquist’s argument in Pico is the more persuasive one, at least as to the decisions in the inherently content-based field and often viewpoint-based field of what is to be on public school library shelves (though not necessarily as to Internet access decisions, which could be content-neutral and especially viewpoint-neutral). (I also think it’s quite proper for people to fault certain kinds of book removal decisions on the grounds that those decisions show narrow-mindedness, or deny library patrons — including children — valuable information, and are thus improper even though they aren’t unconstitutional or even more broadly rights-violating. Of course, as with many ethical judgments, such a judgment will turn considerably on the details of each case.)