« Bifurcating Habeas | Main | Alvarez and the First Amendment's Epistemological Crisis »

Tuesday, October 18, 2011

Guest Post by TJ Chiang: The Stolen Valor Act as Trademark Protection

The following post is from GMU's TJ Chiang, an occasional guest with Prawfs:

On Monday, the Supreme Court granted certiorari in United States v. Alvarez, on whether the Stolen Valor Act, which criminalizes falsely representing oneself as having been awarded a military medal, violates the First Amendment.  Most of the debate so far has focused on whether the First Amendment contains a categorical exclusion for false speech, and the primary analogy the Solicitor-General has tried to draw is to defamation.  In my view, this is the wrong way to view the problem.  The far better analogy is trademark law.

If one thinks about it, labels like “Medal of Honor” and “Purple Heart” would not have much intrinsic value but for the fact that Congress has chosen these names as the names of military decorations.  The only reason that these labels have value is that Congress has built a brand around them, associating the brand only with people having desirable characteristics such as courage and valor.  This is much the same as the fact that the label “Prada” has built-up value, because the company has associated the brand only with products that have desirable characteristics.

What Congress is trying to counter with the Stolen Valor Act is basically what a manufacturer is trying to counter when it sues counterfeiters for trademark infringement: If other people who do not have the desirable characteristics come to be associated with the brand, then the brand will lose its value.  And this means three things.  First, the closest common law analogy to the Stolen Valor Act is not defamation, but the old tort of “passing off”—the product being passed off is the defendant himself, who wishes to pass himself off as having desirable characteristics such as courage and valor, but it is still passing off.  Second, if one regards federal or state trademark law as constitutional, then there is almost no argument that the Stolen Valor Act is not.  Third, the problem with stealing valor is not its falsity as such, but also the fact that the falsity diminishes Congress’s investment in the brand, which the presentation of the problem as a falsity issue does not capture.


Posted by Administrators on October 18, 2011 at 09:45 AM in Constitutional thoughts, First Amendment, Intellectual Property | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Guest Post by TJ Chiang: The Stolen Valor Act as Trademark Protection:


I agree that there's a trademark analogy to be drawn in Alvarez, but I don't know that it's especially helpful to the government. It's important to distinguish between different categories of trademark liability here--even within infringement doctrine. If you'll forgive the shameless self-promotion, I address this issue in a forthcoming article in the Minnesota Law Review (http://ssrn.com/abstract=1798867).

The problem with the counterfeiting analogy, to pick up on James's point, is that the harm flowing from the false speech expressed through a trademark can be difficult to identify and is not consistent across all types of trademark claims. In some cases--post-sale confusion cases in particular--it boils down to nothing more than government-subsidized social exclusion. In the Minnesota piece I argue that liability for trademark counterfeiting--at least in the limited but economically crucial case of luxury brands--is problematic for reasons quite similar to those discussed by Judge Kozinski in his en banc concurrence in Alvarez.

As for dilution, I think there is at least a theory of harm there beyond just "falsity=bad," though it's empirically dubious for a number of reasons that others--particularly Rebecca Tushnet--have identified.

Posted by: Jeremy Sheff | Oct 18, 2011 2:58:09 PM

Joe and James each anticipated my two responses to Joe's point. The first is that I'm skeptical that consumer (or listener) confusion is required for trademark law to survive First Amendment scrutiny, because it would invalidate all the anti-dilution provisions. The second is that, even if confusion is required, the question is whether the law is categorically unconstitutional because it fails to require proof of this element. In First Amendment jurisprudence, this falls under the over-breadth doctrine. The question then becomes whether cases of listener non-confusion, i.e. the defendant misrepresents himself as an awardee but everybody knows he is not, are substantial. There is at least a plausible case that such cases of non-confusion will be rare, especially if we take a narrow view of what it constitutes to "falsely represent" oneself as an awardee and exclude cases of obvious parody and satire (which the U.S. attempts to do in its petition for certiorari).

Posted by: TJ | Oct 18, 2011 12:25:43 PM

I think TJ's point is well-taken. I think of the Stolen Valor Act, trademark laws, and other similar laws as involving "falsity-plus": the speech must both be false and cause some further harm for government to be able to prohibit it. The harm must be of a sort that government has a legitimate interest in preventing (this is what distinguishes Stolen Valor from seditious libel, for example) -- but Congress is allowed to presume harm at the categorical level rather than requiring an individuated showing. Falsity is in the background the whole time: the discussion of categories of speech causing harms starts by asking what made the statement false.

Posted by: James Grimmelmann | Oct 18, 2011 11:48:28 AM

I think TJ is quite right to point to trademark law as the more fitting comparison in evaluating the First Amendment challenge to the Stolen Valor Act. But I doubt it leads to the conclusion TJ suggests.
The key to proving a valid trademark infringement claim - of which the common law passing-off claim is an example - is proving harm to consumers. That harm goes by the name "likelihood of consumer confusion as to source," and it refers to the likelihood that people who encounter the accused item will mistake its source.
I'm not a First Amendment scholar by any stretch. At the same time, I find it at least plausible to suggest that the First Amendment price for prohibiting misleading speech such as a confusingly similar trademark is proof that it is, in fact, likely to mislead.
The Stolen Valor Act's defect, on this theory, is that it does not require any showing that the accused's use of / wearing of / claiming entitlement to wear the medal is likely to confuse someone about the accused's status as a legitimate medal winner. And it would be a simple enough fix for Congress: add likelihood of audience confusion to the elements of the offense.
(Finally, I'm putting aside a separate trademark law idea that I know TJ is aware of - trademark antidilution protection. Flipping TJ's argument, if the Stolen Valor Act is unconstitutional for the reason I suggest above, then both federal and state andilution statutes are also seriously vulnerable to First Amendment attack. But this is a separate matter from the more conventional trademark infringement idea TJ raises here.)

Posted by: Joe Miller | Oct 18, 2011 10:33:52 AM

Because the "TM owner" is the government, the speech considerations are different than the considerations present when private parties litigate TM disputes. Think of the myriad cases where TM owners have silenced others using TM law as a vehicle to do so. I doubt we'd want to allow the U.S. government to start censoring speech based on confusion, dilution, and affiliation claims. Do the movies Independence Day, the Bourne Identity, etc. imply U.S. Gov. affiliation with or sponsorship of the movies? If not, how are they different from similar uses of TMs in movies and TV shows that courts have found to be infringing?

Posted by: anon | Oct 18, 2011 10:24:42 AM

The comments to this entry are closed.