rests what strikes me as the heart of the substantive First Amendment analysis -- and it's a nugget that merits more discussion than it's likely to get.

First, some background (which I guess means I'm burying the nugget, too). The the opinion involves a federal statute that lets the government issue special subpoenas (called "National Security Letters") while prohibiting the recipient -- usually a phone company or an Internet Service Provider -- from disclosing the existence of the NSL.

This is one of the issues I discussed in my Crime-Facilitating Speech, 57 Stan. L. Rev. 1095 (2005). Publicizing the existence of this sort of government surveillance can seriously interfere with the investigation, for instance by informing criminals and terrorists that they're under suspicion and that certain phones, bank accounts, or tactics are no longer safe to use.

But such publicity can also substantially inform public debate about government action and possible government abuse. Concrete and timely examples of alleged abuse -- whether or not the alleged abuse rises to the level of illegality or unconstitutionality -- may be necessary to persuade the public or opinion leaders to press for changes in government policies: A general complaint that some unspecified abuse is happening somewhere will naturally leave most listeners skeptical. And even if the revelation of the surveillance is only delayed for some months or years, rather than being prohibited forever, such a delay may make it much harder to get timely political action, especially since people tend to be much less interested in alleged abuses years ago than in alleged abuses that are happening right now.

So should the speech be protected because it's valuable to public debate, despite the potential harm to law enforcement or even national security? Or should law enforcement and national security prevail despite the potential harm to informed public debate on the merits of the surveillance tactics? That's the tough question that the Crime-Facilitating Speech article tries to deal with, as to this question and as to others, and that the Second Circuit was asked to confront.

Now at first the Second Circuit's decision seems like a victory for free speech maximalists, because the court holds both reads the statute narrowly and concludes that even in its narrowed version it is procedurally flawed. As our first post on the subject said, the court held that,

[We] rule that [the relevant provisions] are unconstitutional to the extent that [1] they impose a nondisclosure requirement without placing on the Government the burden of initiating judicial review of that requirement, and [2] rule that subsections 3511(b)(2) and (b)(3) are unconstitutional to the extent that, upon such review, a governmental official's certification that disclosure may endanger the national security of the United States or interfere with diplomatic relations is treated as conclusive.

So there have to be extra procedures: A means for the recipient to demand that the Government quickly ask a court to review the NSL, and independent review by the court of the Government's arguments. This early story likewise focuses on the procedure, and understandably so, since that's what the bulk of the court's opinion discusses.

But what about the substance? May the government indeed order you not to, for instance, go to the newspapers to complain -- with details -- that you're being dragooned into turning over information that you believe the government shouldn't be able to demand? Here's the court's answer, on pp. 46-47 of the 55-page opinion:

A demonstration of a reasonable likelihood of potential harm, related to international terrorism or clandestine intelligence activities, will virtually always outweigh the First Amendment interest in speaking about such a limited and particularized occurrence as the receipt of an NSL and will suffice to maintain the secrecy of the fact of such receipt.

What's more, this strikes me as the extent of the substantive First Amendment analysis on this point (except for a brief mention on p. 37 that "no governmental interest is more compelling than the security of the Nation"). The passage I just quoted has no footnotes and no citations attached to it.

To be sure, many people might conclude that such a passage needs no citations, and that of course the First Amendment is virtually always trumped by "reasonable likelihood of potential harm, related to international terrorism or clandestine intelligence activities." But that's not entirely clear, it seems to me. Even in wartime, the First Amendment protects some speech that may indeed jeopardize national security (see, for instance, the wartime election hypothetical). Likewise, whether the government may punish speakers who publish classified information that is leaked to them (as opposed to people who breached secrecy agreements that they voluntarily entered into as a condition of government employment) is at least a difficult question, though a recent district court case (U.S. v. Rosen) answered the question "yes," at least in some circumstances.

Nor is it clear that the free speech interest is much less because the statement is about "limited and particularized occurrence as the receipt of an NSL" -- general arguments about government power and alleged government abuses often rest on specific "limited and particularized" facts, such as the fact that the government has tried to coerce a particular entity to reveal a particular item of information. Even the opinion acknowledges, on p. 36, that the factual information "is relevant to intended criticism of a governmental activity," and thus has substantial First Amendment value. There's thus some reason to think that the circuit's "will virtually always outweigh the First Amendment interest" statement would apply to other constitutionally valuable speech as well.

So I think there's a good deal more that needs to be discussed here, and that I hope that future courts will indeed discuss it, whether while applying the procedures that the Second Circuit mandates (especially if those procedures can be applied informatively in a published opinion), while considering the matter in other circuits, or while reviewing the Second Circuit's decision (if the Supreme Court agrees to hear the case). And in the meantime, we should be conscious about the Second Circuit's substantive conclusion -- one that authorizes a pretty significant speech restriction -- and not just its procedural conclusion.