By Mike Dorf

Monday's decision in Holder v. Humanitarian Law Project (HLP) is, of course, most notable for upholding the federal "material support" statute against a First Amendment challenge, but in the course of getting to that result, the majority opinion by CJ Roberts says some things about free speech doctrine more generally that will likely have the effect of broadening protection for free speech.

A threshold point of contention was whether the law should be judged as content-based (and thus subject to strict scrutiny) or content-neutral (and thus subject to what is nominally intermediate scrutiny but under the SCOTUS precedents amounts to a test that is fairly easy to meet).  The government argued that the law should be deemed content-neutral because it is a regulation of conduct--here, material support--that is often non-expressive (e.g., providing money, goods, or services to a terrorist organization) and only happens to be expressive in this particular case.  Not so, said CJ Roberts.  He noted that strict scrutiny applied in Cohen v. California, even though Cohen was charged under a breach-of-the-peace law that applied to conduct regardless of whether it was expressive in the particular case.  (Cohen wore a jacket that said "Fuck the Draft" on it.)  According to the Court in HLP:  "The law here may be described as directed at conduct, as the law in Cohen was directed at breaches of the peace, but as applied to plaintiffs the conduct triggering coverage under the statute consists of communicating a message."

As Eugene Volokh notes on his blog, this analysis appears to be inconsistent with prior doctrine defining a content-based law by reference to the law's terms, not the particular conduct that is alleged to infringe it.  I agree and I would note that there is a genuinely difficult puzzle here about how to go about defining what counts as content-based.

The HLP formulation strikes me as too broad, as a hypothetical example should illustrate.  Suppose a statute that makes someone guilty of murder if he "intentionally causes the death of another human being without justification or excuse."  Now suppose that a mob boss is accused of murdering a rival by instructing one of his henchman as follows: "Make sure that Don Tattaglia sleeps with the fishes."  Following the communication of this message from boss to henchman, henchman kills Tattaglia and dumps his body in the river.  We could say, using the HLP formulation that the murder law may be described as directed at the conduct of murder, but as applied to the mob boss, the conduct triggering coverage under the statute consists of communicating a message.  It seems plainly crazy to subject the murder law to strict scrutiny; yet that seems to follow from the opinion of CJ Roberts in HLP.

There are two ways in which the foregoing hypothetical could be tamed.  First, we might note that the murder law will undoubtedly survive strict scrutiny, so there's ultimately no harm in subjecting it to strict scrutiny.  But this answer, I think, misses the point.  Suppose that the mob boss ordered his henchman to commit a less serious offense, such drug trafficking.  It's not obvious that our drug laws could survive strict scrutiny because they arguably do more harm than good (in creating a black market, fostering violence, etc.).

That leads to the second possible escape hatch: Perhaps any crime committed by a speech-act would fall within the unprotected category of speech identified in Giboney v. Empire Storage. As most recently elaborated in the Stevens case, that category consists of "speech integral to criminal conduct."  Okay, then, suppose that the mob boss gives instructions to his henchman to commit tortious but non-criminal acts.  Now does strict scrutiny apply?

To be clear, I am not saying that the Court was wrong to apply strict scrutiny to the material support law.  I am saying that the Court's definition of content-based laws seems quite broad.

Meanwhile, there are difficulties in the other direction as well, which may explain why the Court errs in the direction it does (and note that there is no dissent on the point I'm discussing).  Suppose a state university adopts a rule forbidding students, staff, and faculty members from storing explicitly sexual computer image files on university-owned servers, even if the images are not legally obscene.  Now suppose that the rule is challenged and so the university changes its policy: It adopts a broader policy forbidding "misuse of university property," and disciplines students, staff, and faculty members for storing pornography on the university servers under the broader policy.  I can see upholding the porn policy on its own as a legitimate outcome, but it's very difficult to justify subjecting the porn-only policy to a higher level of judicial scrutiny than one uses to evaluate the porn-as-misuse approach.

What all of this may show is that the distinction between content-based and content-neutral regulations of speech is simply too blunt to capture the real stakes in many free speech cases--a point that my colleague Steve Shiffrin has made for some time.