Parting thoughts and thanks on product-defining speech and anti-knowledge

One of the advantages of having an opportunity to blog here at Volokh Conspiracy is that the experience spurred conversations and correspondence that help clarify what the more ambiguous and difficult free-speech issues are. I'll use this last post to say a couple more things about product-defining speech and anti-knowledge in light of reader reactions.

Product-defining speech

I argued that speech is used as an element to define when some products are regulated. Let's stick with the example of pharmaceuticals. My argument was that speech is a necessary element to identify when a "mislabeled drug" is introduced to treat an off-label medical problem. Some readers continue to believe I am confusing the use of speech as mere evidence for the use of speech as a required element.

Those readers are in good company-that's precisely the argument my colleague Chris Robertson has made. (By the way, if you ever want to clear a crowd, just invite Chris and me over and then get us talking about the Food and Drug Administration.) Under the speech-as-evidence theory, speech is used as mere evidence of "intent," but intent of what?

The off-label drug example differs from cases in which promotional speech is used to show intent to aid and abet criminal or civilly illegal conduct related to dual use products and services. In cases in which a gun manufacturer uses speech to encourage buyers to use their guns to intimidate or harm other humans, or a hotel owner encourages renters to use their rooms for prostitution, speech can be used as evidence of intent to commit (or to abet) the downstream crimes.

In the case of off-label drug promotion, there is no downstream illegal conduct. Doctors can prescribe drugs, pharmacies can supply drugs, and patients can take drugs for any purpose that the doctor deems fit. The FDA could attempt to apply its "mislabeling" laws any time a drug has expanded from its on-label market to unvetted off-label uses (although there would be significant political backlash to preserve the autonomy of the physician-patient relationship.) If it did regulate off-label uses of a drug, just as some other products can be legally used for some purposes but not others, then the FDA's use of promotional speech by drug manufacturers would be evidence of abetting and participating in an illegal market.

But this is not the case. Instead, "intent" to introduce an off-label drug is actually a mislabeled element. Drugs are introduced for off-label uses all the time. Only the encouragement of those off-label uses, and only encouragement by one set of speakers (those who have a direct financial interest in the market), is regulated. Thus, I think it is inevitable that courts will have to think through not only the drug promotion regulations but other areas, too, where commercial speech is used as the trigger for product regulation.

By the way, I do not necessarily think that a category of speech regulations that I'm calling "product-defining speech" is a terrible idea. There's something parsimonious about letting a manufacturer's own labels and speech determine whether a product has to comply with a set of regulations. But I am confident that it requires some deeper thinking about how it fits with First Amendment doctrine and goals. It is very weird that some products (a drug, a Google search of a person, even a Q-Tip) that everyone knows is routinely, perhaps even predominantly, used in a particular way (for off-label use, for employment purposes, inside the ear canal) are not regulated as such as long as the key participant in the marketplace never mentions it.

Liability for anti-knowledge

In another post, I suggested that courts should be more open than they are to liability for influential junk-science anti-knowledge that causes physical harm.

Some readers are highly skeptical that anti-knowledge shared in public discourse should ever lead to liability, particularly because the government (perhaps a court or even—ack!—a jury) would have to decide which claims were contested and which were beyond the pale. I agree that distinguishing between "anti-knowledge" and "contested knowledge" (which would be fully protected) is the most discomfiting part of my proposal. (This is one reason I made sure not to equate "accepted knowledge" with truth.) Even so, my proposal is less radical than it seems.

Comparing my proposal to a couple other well-established areas of speech regulation shows that evidentiary gaps and epistemological problems are not usually deal-breakers when there are important competing social interests. First, the censorship and fining that the FDA and Federal Trade Commission already do when health and science claims are made without sufficient support has significant effects on public discourse by removing one set of interested and knowledgeable speakers.

We have grown comfortable with a stark division between commercial and non-commercial speakers, but that strong division is starting to collapse as the Supreme Court takes more seriously the availability of diverse sources of information to listeners. If commercial speakers ought to be held responsible for harmful junk science, and I think they should, then I have to be willing to carry that logic to other speakers, too. The rules for commercial speakers and public speakers do not necessarily have to be the same, but a total immunity for non-commercial speakers unnecessarily deprives the public of a set of legal remedies that can reduce harm.

The second comparison is to defamation. Defamation is a subset of anti-knowledge—"false" factual statements that harm a victim's reputation. Defamation cases have to be decided using the best available evidence about whether the statement was true or false, and that evidence is fallible, of course. Moreover, defamation uses a mere preponderance-of-evidence standard for determining "falsity," meaning that a speaker can be held liable for making a statement that is 49 percent likely to be true. I actually think this is a serious flaw in need of correction within defamation law; I would love to see a heightened standard for the requirement of falsity—something equivalent to clear and convincing evidence or evidence beyond reasonable doubt.

But at the same time, if we are comfortable punishing false statements in public debate for reputational harms, I don't think that having judges/juries distinguish between anti-knowledge and (fully protected) contested knowledge would be aberrational or indefensible for physical harms.

That said, I have to admit this is the part of the project that I feel most uneasy about. It's one reason I took up the challenge to blog about it. But I should also confess that I carefully chose a hypothetical in which the victim of junk science (vaccine-related) was a third party rather than the listener. Listeners have a duty to take information in the public domain with a grain of salt, and this could be relevant to liability. It may make sense to use a liability rule that distinguishes between a listener who harms himself and a listener who harms a third party. (Liability rules for getting people drunk and letting them drive home are often designed with this distinction—the drunk driver can't sue the social host, but innocent third parties often can.) Alternatively, a legal restriction on anti-knowledge could require evidence that the speaker was likely to be particularly influential to listeners, maybe by brandishing credentials and experience.

Clearly I am still working through what I think the most responsible contours of anti-knowledge liability ought to be. But when I take a birds-eye view of speech regulation, I see a lot of path-dependent thinking about which parts of public discourse are untouchable and which aren't.

Thanks again for reading!