Thursday, January 19, 2017

National Constitution Center Podcast on Tam

Deborah Gerhardt, Ilya Shapiro,  and I discuss Tam.

The podcast can be found at the National Constitution Center's website, as well as on iTunes and other platforms.

Seen around town: transformative works of the day, inauguration edition.

Seen on my way back from the Tam oral argument:
Activist Comics

9th Circuit rejects Slep-Tone's anti-copying claim couched as TM claim

Slep-Tone Entertainment Corp. v. Wired for Sound Karaoke & DJ Servs., LLC, No. 14-17229 (9th Cir. Jan. 18, 2017) (per curiam)

Slep-Tone makes karaoke music tracks under the trademark “Sound Choice.” They’re released on CDs in a format known as “CD-G,” which accompanies karaoke music with graphics, lyrics, and singing cues when played on compatible equipment.  Karaoke operators may prefer ripped files for convenience; Slep-Tone now allows one digital copy per source CD-G subject to various conditions.  It alleged that defendants performed shows using unauthorized media-shifted files, and that this infringed its trademark rights.  The district court concluded that “[t]his attempt to stuff copyright claims into a trademark container fails.” 

The court of appeals affirmed: “[w]hether we call the violation infringement, unfair competition or false designation of origin, the test is identical—is there a ‘likelihood of confusion?’” And there was not.

Slep-Tone argued that, because the ripped files showed its marks and trade dress when performed, consumers would be confused into thinking that the tracks originated with Slep-Tone, rather than with defendants.  But that theory didn’t identify “consumer confusion about the source of an appropriate ‘good,’” only possible confusion over the source of content. Dastar says “origin” has to mean tangible goods offered for sale.  And, though Dastar was about §43(a), the same standard applies to §32.  Slep-Tone identified the new digital files created by ripping as the relevant tangible goods.  But those “new goods” weren’t relevant goods.  The “good” at issue must be the “ ‘tangible product sold in the marketplace’ rather than the creative content of that product.” Phx. Entm’t Partners v. Rumsey, 829 F.3d 817, 828 (7th Cir. 2016) (quoting Dastar, 539 U.S. at 31).

There wouldn’t be confusion about the source of the tangible good sold in the marketplace, because “[c]onsumers are not aware of the new, media-shifted digital files about which Plaintiff asserts confusion.”  Any confusion isn’t about the source of the goods.  There’s no use of the Sound Choice marks “in connection with the sale, offering for sale, distribution, or advertising” of the files under 15 U.S.C. § 1114(1)(a) or “in connection with” the files under § 1125(a)(1).

[Query whether this result is consistent with the Second Circuit’s recent pronouncement that confusion over sponsorship or affiliation must also be considered before rejecting a confusion theory.]

This is really a case about unauthorized copying, and Dastar cautions against creating a “species of mutant copyright law” under the Lanham Act.  Slep-Tone was really complaining about allegedly unauthorized use of the content of the karaoke tracks, and Dastar bars such claims.  “The words of the Lanham Act should not be stretched to cover matters that are typically of no consequence to purchasers.” Hey, it’s materiality!  Hi, materiality! 

Lee v. Tam post-argument

Lee v. Tam

My thoughts on observing the argument: I was surprised how much attention the Justices paid to the copyright comparison.  The SG centered his response around the idea that copyright is historically a mechanism to stimulate expression, while trademark lacks the same speech-promoting role (and indeed serves a speech-suppressing function).  I might have spent more time saying that, under unconstitutional conditions doctrine, an exclusion from copyrightability for disparaging works would almost by definition be trying to affect speech outside the government program at issue—that is to say, it would be trying to leverage the government’s power to grant protection to expression to deter people from speech that includes disparaging ideas; therefore the unconstitutional conditions analysis would be sharply different if Congress decided to disallow copyright in disparaging works.

I have no prediction at all.

Post-argument panel at AU:

John Connell, Archer Law (Counsel for Respondent)
Denial of registration is always b/c gov’t finds mark disparaging, offensive to 3d party.  Gov’t disfavors that. That’s inherently viewpoint discrim b/c positive/neutral views are permitted to register.  Lanham Act is widely available to broad range of mark holders.  Selectively denying mark b/c of negative views expressed by mark is viewpoint discrimination.

Ilya Shapiro, Cato Institute (representing Cato Institute on amicus brief): added examples of offensive marks like Raging Bitch beer.  Beyond black letter law on viewpoint discrimination or unconstitutional conditions. Not thrown into jail, but denied a gov’t benefit for exercising his first amendment rights.  TM registration is an important statutory right—easier to defend infringements. Record label won’t sign w/o registration.

Mike Carroll: how does Wash. Football team dispute bear on this?

Jesse Witten, Drinker Biddle (representing Amanda Blackhorse, et al. on amicus brief)
No question that our case hangs over Tam, especially locally.  Overshadowing our case is the n word.  If you open the door to the Slants, you have to issue registrations for everything. Maybe that’s good policy, maybe not, but many judges will be leery of official registrations for incredibly racist marks.  We have examples of horrific marks registered before the statute was changed in the 1940s.  Gov’t felt it couldn’t be quite as brazen.  Intended to complement gov’t’s brief.  Simple argument: there’s no burden on the speaker.  Can still make claims under §43(a).  TM empowers you to silence somebody else. It’s not a 1A issue, it’s an exception to the 1A.  Refusal of registration reduces marginal ability to silence, but that’s not a 1A right.  J. Sotomayor asked some relevant questions. 

Q: unregistered owner can bring claim?

Prof. Rebecca Tushnet, Georgetown University (representing Law Professors on amicus brief)
TM is a right to suppress speech.  If Tam gets trademark protection for the expressive aspects of his mark, he is able to suppress other speakers—including a racist band using the term to express its racism.  This characteristic of TM and TM registration means that 1A analysis targeted at traditional punishments is inappropriate—there is no neutral baseline.  It also highlights that the gov’t is regulating only the registration aspects, not the expressive aspects of the mark.

Not denied a gov’t benefit outside the contours of the gov’t program itself. Unconstitutional conditions is a doctrine that prevents attempts to leverage a gov’t benefit to change constitutionally protected behavior outside the program.  Gov’t can decline to pay for abortion, but it can’t decline to provide other health care to someone who has exercised her const’l right to abortion.  Simon Tam can register TMs in logos, names, other features; if he were denied any TM registration until he changed the name, Shapiro would be right.

Why is viewpoint discrimination bad?  Because it targets a particular group for suppression.  Disparagement bar doesn’t do that.  Hating gays is a viewpoint; hating is not.  Materiality: do consumers care about the feature at issue?  If a blanket ban on disparagement is viewpoint based, then a ban on material falsehoods and not immaterial falsehoods is also viewpoint based because it depends on whether consumers care about the feature you’re discussing or not.

Shapiro: Agree that this is about unconstitutional conditions.  Also issues of scope of IP more broadly. If you’re going to have a registration system, it’s the gov’t administering and not speaking.  It can’t discriminate in this way.

Joel MacMull, Archer Law (Counsel for Respondent)
Notion that Simon Tam has not been burdened.  Keep in mind: this is a facial challenge to a constsitutional provision.  In all instances, §2(a) is discriminatory. When you have a facial challenge, you don’t have to demonstrate individual burden.  He can only continue using the mark if he’s not second in time in a given territory.  He’s based in Portland Oregon—someone could beat him if he didn’t have the benefit of the nationwide priority from the registration, which would allow him to prevent other people from using it nationwide.  43(a) isn’t a good enough fallback for those reasons.

Jesse Witten: the burden is that it’s harder for him to silence people outside of Portland?

MacMull: he may not be first in these other territories.

Witten: SCt has identified TM as an exception to free speech, not an instance of free speech.  Gay Olympics case: passes muster as a legit. regulation of commercial speech, satisfying intermediate scrutiny.

[RT: I hate that case but think it’s distinguishable.]

Shapiro: none of that has to do with registrability. Bracketing suppression, we take the law as given, and it’s an equal protection question.  Can your registration be denied b/c the gov’t doesn’t like it.

RT: on 43(a) and distant users: why wouldn’t they have 1A rights to do so? Denying them rights based on registration’s nationwide priority can’t be about protecting consumers in Michigan who don’t know about the Portland Slants and can’t be confused.  It is about another objective: helping businesses manage relations.

Carroll: what happened at argument?

Connell: the Ct wanted to rule unanimously against both sides. [h/t Mark Lemley]

Shapiro: Ct was struggling w/how to write the opinion. Justices weren’t satisfied w/gov’t or petitioners in terms of bright line for what they’re going to do.  He thinks 5 votes for pet’r; Breyer on fence; Sotomayor leaning towards gov’t; Thomas was silent but was deciding vote on Confederate flag case.  That’s distinguishable b/c nobody thinks every TM is endorsed by the gov’t.

Witten: no established doctrinal bucket.  I felt like a surprisingly large number had an open mind.  There are good arguments & weaknesses on both sides.

Connell: false association/no names w/o consent is a right of publicity issue, a completely different animal.

Shapiro: surprised Court wasn’t that interested in vagueness issues.  Dykes on Bikes: name approved but logo rejected b/c had name in it.  Denied to Have you heard Satan is a Republican but not The Devil is a Democrat.

Carroll: SG said no one rejected non-disparaging marks that were rejected.

Shapiro: Fag-dog has been registered 2x and refused 2x.

MacMull: it’s the gov’t’s burden.

Witten: open Qs: can you do a facial challenge on vagueness grounds?  SCt has said yes and no.  There may be some cases that are really hard and others, like N-word, are not hard.  Should not facially invalidate in such cases.  Marsh v. Chambers: SCt articulated a test that prayer was fine as long as it didn’t proselytize or disparage. If that’s the word employed to interpret Establishment Clause, hard to find that “may disparage” is vague.

MacMull: threw in constitutional stuff as a side issue; initial issue was an evidentiary one.  Vagueness, I long thought, was one of the strongest arguments, but it was never the one the Court or the Fed Cir en banc put in front of the parties. 

Q: reappropriation as potential distinction?

MacMull: PTO shouldn’t care about applicant’s intent, for many reasons not the least of which is that intent can change.  Tam’s intent could be different tomorrow.  TMs are freely assignable [w/goodwill]. 

Shapiro: vile marks are registrable, but that’s not a guarantee of success in the commercial marketplace. College football teams changed names b/c they found it untenable. It wasn’t the gov’t telling them to do so. Private solutions matter.  If The Slants’ name hurts ticket sales, that will take care of the problem.

Tuesday, January 17, 2017

Monday, January 16, 2017

Transformative work of the day, Lin-Manuel Miranda edition

What's Next?  Since I'm working on 512 roundtable reply comments right now, it seems particularly appropriate to remind everyone that clips can be fair use.

Friday, January 13, 2017

It's a Tam, Tam, Tam world

Along with a podcast for the American Constitution Center and an NPR interview running the morning of the Tam argument (next Wednesday), I'll also be on a post-argument panel at AU--the panel before features none other than Tam himself.

American University Washington College of Law 
Program on Information Justice and Intellectual Property 
Supreme Court Series:
Lee v. Tam
Post Argument Discussion
Wednesday, January 18, 2017
3:15pm Registration
3:30pm Policy Panel
4:15 Legal Panel
Reception to follow

NT01 Ceremonial Classroom in Warren Building
American University, Washington College of Law
4300 Nebraska Ave., NW Washington DC 20016 

Registration & CLE information:Lee v. Tam
In PIJIP’s ongoing Supreme Court Series, a panel of counsel for amici and parties will discuss the case on the afternoon following oral argument before the Court.

Issue in Lee: Whether the disparagement provision of the Lanham Act, 15 U.S.C. 1052(a), which provides that no trademark shall be refused registration on account of its nature unless, inter alia, it “[c]onsists of . . . matter which may disparage . . . persons, living or dead, institutions, beliefs, or national symbols, or bring them into contempt, or disrepute” is facially invalid under the Free Speech Clause of the First Amendment.

Prof. Michael Carroll, American University Washington College of Law (Moderator)

3:30pm - First Panel - Policy Discussion
  • Simon Tam, The Slants (Respondent)
  • Prof. Victoria Phillips, American University Washington College of Law
4:15pm - Second Panel - Legal Discussion
  • John Connell, Archer Law (Counsel for Respondent)
  • Joel MacMull, Archer Law (Counsel for Respondent)
  • Jesse Witten, Drinker Biddle (representing Amanda Blackhorse, et al. amicus brief)
  • Prof. Rebecca Tushnet, Georgetown University (representing Law Professors on amicus brief)
  • Ilya Shapiro, CATO Institute (representing CATO Institute on amicus brief)

Trump's trademarks

Selling, first and foremost, would not eliminate possibilities of conflicts of interest. In fact, it would exacerbate them. The Trump brand is key to the value of the Trump Organization’s assets. If President-elect Trump sold his brand, he would be entitled to royalties for the use of it, and this would result in the trust retaining an interest in the brand without the ability to assure that it does not exploit the office of the presidency.

First, there is absolutely no requirement that any transaction divesting Trump of the Trump trademarks must have royalties attached.  Lump sums—that is, outright sales—are perfectly acceptable methods of carrying out such transactions.  Fixed license fees with no relationship to sales are also possible.  The lawyer’s claim is a misstatement of the law, and one that should not have been made by anyone with even minimal trademark transaction experience.  I take it that in essence this statement is meant to be coupled with the later, and contestable in its own right, claim that President-Elect Trump shouldn’t have to (comply with the Constitution if that would mean he’d) suffer financially from being President, plus the implicit assertion that lump sums or fixed fees would be less financially attractive than royalties.  But “we think he couldn’t make as much money from a true sale” is not the same as “a sale would entitle him to royalties” and thus continue his financial interest in the brand no matter what. 

Second, to the extent that Trump is in fact (if you believe he is) isolated from information about and management of the brands, he has a bigger trademark problem: that would mean abandoning the mark.  If he is to be the owner, he needs to control the quality of the products and services sold via that mark.  If all he promises to do is step back so that he’s not involved in any decisions but just reaps the rewards, he’s just engaging in naked licensing, which in the US results in the abandonment of the mark.  So this “explanation” is flat wrong at best, backwards at worst.  [I wonder what would happen if someone brought a cancellation petition after he takes office.  I suppose one could claim that the non-blind trust is the legal owner, and is making management decisions, which just happen to inure to his benefit.  At last look, he was personally the owner of record of a bunch of the Trump registrations according to TSDR; I wonder if/when that will change.  But the real point is: without true divestment, no trust arrangement can keep him from having a financial interest in the Trump marks and businesses and thus from being vulnerable to influence in his official duties.]

Now, that doesn’t mean that these misstatements will ever have any consequences for the President-Elect, any more than other, more significant ones have done.  But let’s make a record anyway.

Question for those who know European privacy law

Is this ad campaign for Spotify a problem?
"Dear person in the Theater District who listened to the Hamilton soundtrack 5,376 times this year, can you get us tickets?"
"Dear person who played 'Sorry' 42 times on Valentine's Day, What did you do?"
Bonus: if these aren't actual Spotify customers, is this false advertising?  (By my calculations, you couldn't listen to the full Hamilton soundtrack 5,376 times in one year.  Maybe if you listened to mostly Act I, to avoid too much gross sobbing.)

Thursday, January 12, 2017

Harvard IP fellowship: applications sought


The Project on the Foundations of Private Law is an interdisciplinary research program at Harvard Law School dedicated to scholarly research in private law. Applicants should be aspiring academics with a primary interest in intellectual property (especially, patent, copyright, trademark and trade secret) and its connection to one or more of property, contracts, torts, commercial law, unjust enrichment, restitution, equity, and remedies. The Project welcomes applicants with a serious interest in legal structures and institutions, and welcomes a variety of perspectives, including economics, history, philosophy, and comparative law. The Qualcomm Postdoctoral Fellowship in Private Law and Intellectual Property is a specifically designed to identify, cultivate, and promote promising IP scholars early in their careers. Fellows are selected from among recent graduates, young academics, and mid-career practitioners who are committed to spending two years at the Project pursuing publishable research that is likely to make a significant contribution to the IP and private law, broadly conceived. More information on the Center can be found at:

PROGRAM: The Qualcomm Postdoctoral Fellowship in Private Law and Intellectual Property  is  a  full-time,  two-year  residential  appointment,  starting  in  the  Fall  of 2017. Like other postdoctoral fellows, IP Fellows devote their full time to scholarly activities in furtherance of their individual research agendas in intellectual property and private law. The Project does not impose teaching obligations on fellows, although fellows may teach a seminar on the subject of their research in the Spring of their second year. In addition to pursuing their research and writing, fellows are expected to attend and participate in research workshops on private law, and other events designated by the Project. Fellows are also expected to help plan and execute a small number of events during their fellowship, and to present their research in at least one of a variety of forums, including academic seminars, speaker panels, or conferences. Through organizing events with outside speakers, helping to run programs, and attending seminars, fellows interact with a broad range of leading scholars in intellectual property and private law. The Project also relies on fellows  to provide opportunities for interested students to consult with them about their areas of research, and to directly mentor its Student Fellows. Finally, fellows will be expected to blog periodically (about twice per month) on our collaborative blog, New Private Law (

STIPEND AND BENEFITS: Fellows have access to a wide range of resources offered by Harvard University. The Center provides each fellow with office space, library access, and a standard package of benefits for employee postdoctoral fellows at the Law School.  The annual stipend will be $55,000 per year.

ELIGIBILITY: By the start of the fellowship term, applicants must hold a J.D. or  other graduate law degree. The Center particularly encourages applications from those  who  intend  to  pursue  careers  as  tenure-­track  law  professors  in  intellectual property and private law, but will consider any applicant who demonstrates an interest and ability to produce outstanding scholarship in the area. Applicants will be evaluated by the quality and probable significance of their research proposals, and by their record of academic and professional achievement.

APPLICATION: Completed applications must be received at by 9:00 a.m. on March 1, 2017. Please note that ALL application materials must be submitted electronically, and should include:

1.         Curriculum Vitae

2.         PDFs of transcripts from all post-secondary schools attended.

3.         A Research Proposal of no more than 2,000 words describing the applicant’s area of research and writing plans. Research proposals should demonstrate that the applicant has an interesting and original idea about a research topic that seems sufficiently promising to develop further.

4.         A writing sample that demonstrates the applicant’s writing and analytical abilities and ability to generate interesting, original ideas. This can be a draft rather than a publication. Applicants who already have publications may also submit PDF copies of up to two additional published writings.

5.         Three letters of recommendation, emailed directly from the recommender. Letter writers should be asked to comment not only on the applicant’s writing and analytical ability, but on his or her ability to generate new ideas and his or her commitment to pursue an intellectual enterprise in this area. To the extent feasible, letter writers should provide not just qualitative assessments but also ordinal rankings. For example, rather than just saying a candidate is “great,” it would be useful to have a statement about whether the candidate is (the best, in the top three, among the top 10%, etc.) among some defined set of persons (students they have taught, people they have worked with, etc.).

All application materials with the exception of letters of recommendation should be emailed  by the applicant  to  Letters  of  Recommendation should be emailed directly from the recommender to the same address.

For questions or additional information, contact: Bradford Conner, Coordinator,

Federal Circuit Symposium at AU on Patent & TM law, Jan. 27

Jan. 27, 2017 at American University College of Law


Monday, January 09, 2017

10th Annual Evil Twin Debate,"Free Speech or Free to Reject?: Trademark Registration & the First Amendment”

Sponsored by the University of Richmond School of Law and the AALS IP Section and hosted at Golden Gate University School of Law.   Mark Lemley & I debated, and though I usually try to point out that my summaries may not be complete, here my summary of Lemley's statements is very disjointed, because in this case I was trying to prepare my rebuttal while taking occasional notes, so for a fairer explication of his arguments you should read his amicus brief on behalf of Dykes on Bikes, linked below.  I also hope a recording of the event will be made available.

My presentation: Four big things the parties and amici disagree on, and some thoughts about vagueness and inconsistency in 2(a) refusals:

First: Whether 43(a)(1)(A) incorporates the same policy based limits as registration under section 2.  If 43(a) protection is still available when use of an unregistered symbol causes confusion then the impact of denying registration will be much less.  The problem for those arguing that 43(a) is still available is that the text of the provision leaves us with no real way to distinguish the bars in 2 that we definitely want to apply to 43(a)—like the bar on protecting generic terms as such—from bars that we don’t.  So, you get cases like Belmora in the 4th Circuit, which essentially removed territoriality as a limit on 43(a).  By contrast, the Renna case from New Jersey said that a section 2 registration-barred mark couldn’t be protected by 43(a) and explained why it thought that such a result was sensible in light of the underlying policy behind the Lanham Act and questions of administrability.  Regardless of which way one comes out on Belmora and Renna, the point is that this question has implications far beyond disparagement, and I think that it’s a bad idea to try to answer that broader question merely as a waystation to the First Amendment question about disparagement.

Second: TM’s effects on non-owners or non-registrants. In what way, if at all, should it matter that granting Tam a registration enhances his capacity to suppress the speech of others? Most of those on Tam’s side, Lemley’s brief excepted, don’t say much about this.  Relatedly, and the source of troubling inconsistencies in those on Tam’s side, is a trademark commercial speech? Tam says it isn’t always commercial speech and some amici say it never is because there’s always non-trademark meaning in a mark; if that’s so we’re subjecting infringement cases to far less scrutiny than necessary. Implicit in Tam’s argument seems to be that trademark owners have messages but infringers don’t – except that Tam also says one of his objectives is to prevent a racist band from calling itself the Slants.

From Lemley’s amicus brief: Dykes on Bikes wants to control commercialization of the mark.  Which is to say it wants to prevent other people from using the mark in expressive ways such as on T-shirts and mugs.  Mark Lemley and Stacey Dogan, among others, have argued that this shouldn’t be understood as trademark use, but at the very least it has both commercial and message-delivering aspects.  The difference between the infringer’s desire to speak and Tam’s situation is, supposedly, the harm done by confusion—but as we know, there’s usually no such evidence of harm, and no legislature has made factual findings that confusion itself does harm.  Lemley’s brief gives another example of the benefits of registration by pointing to his client’s ability to use the Sunrise registration process to control a domain  name registration: like it or not, though, that website would have expressed a message, and not one inherently based on confusion with the existing organization.

Third: What framework to analyze 2(a) under: is it government speech, commercial speech under Central Hudson, viewpoint discrimination, unconstitutional conditions, or something else?  One of the problems is our ornate First Amendment doctrine, which grows new protrusions every year.  But another is the problem of applying the First Amendment to government programs that change the structure of the marketplace; trademark registration, not to mention trademark infringement doctrine, does that, and so does the existence of public schools.  When the government is deeply embedded into the structure of an institution, how does that affect our understanding of the government’s freedom to make choices about what to support, as long as it’s not fining people for speech or sending them to jail? I conclude that the best framework is unconstitutional Conditions: the government can’t leverage its program to affect speech outside the program.  And on that framework, it’s clear that there is no effect on Tam’s ability to communicate his message outside of the trademark context. He doesn’t have to give up the name to get trademark protection for the content that includes the name, the way that a prohibition on copyright for disparaging works would require the author to give up on disparagement in order to get protection.

Fourth: Viewpoint versus content discrimination.  The line between the two can be unclear, and I think especially in combination with the previous problem it is very important to understand exactly what the government’s trying to do.  Is the government trying to suppress discussion of racial issues?  Stigmatize abortion?  Or something else?  For better or worse, the government in 2(a) disparagement is trying to limit registrations to those not perceived as disparaging to a targeted group.  To me, that is content discrimination, not viewpoint discrimination.  The government doesn’t care which person or group you target. 

The government also doesn’t care why you chose the symbol at issue—maybe you want to reclaim the number fourteen for non-Nazis, or maybe you want to show your Nazi bona fides; either way, it’s the reaction of the targeted group that matters and not your malign or benign viewpoint.  Of course your choice will express a viewpoint; that’s not the issue.  The issue is whether the overall system, not just the outcome in your case, discriminates against the holders of any coherent class of viewpoints.  One can say that true threats express a viewpoint, but that just highlights the point that it’s possible to reframe almost any content regulation as a viewpoint regulation.  In my opinion, the better path is to ask: why do we care about viewpoint regulation more than content regulation?  As the Supreme Court has previously said, viewpoint discrimination is when the state is “attempting to give one group an advantage over another in the marketplace of ideas,” and that’s just not what the disparagement bar is.

Relatedly: The applicant’s viewpoint about the term or the targeted group matters not at all. The understanding of the targeted group is the key. And many on Tam’s side says that consulting the views of the targeted group constitutes viewpoint discrimination, even though heretofore we’ve used the idea of viewpoint discrimination to identify discrimination based on the speaker’s viewpoint.  You can, I suppose, create a new category of viewpoint discrimination “in the air,” to borrow a phrase from Learned Hand.   But if relying on the reaction of the audience to determine registrability makes disparagement viewpoint based then so are descriptiveness and deceptiveness, also dependent on audience understanding.

Final issue: vagueness.  I think this is frankly the most serious issue for the government, and here I do have to appeal to the nature of registration as a government program in its defense. We’re up to half a million applications a year.  If we are going to have registration as an individualized determination at that scale, we are going to have inconsistencies.  My analogy is to public schools: if we’re going to have them, then some history teachers are going to give better grades to the same essay than other history teachers (which some plagiarists have surely learned to their sorrow).  We cannot have the scale of the government participation in the marketplace without inconsistency; if we want the program to continue to exist—which not everyone does for public schools, but Tam’s supporters definitely want registration to continue to exist—then we have to deal with inconsistency in procedural and not substantive ways. That is, we have to provide mechanisms for making records and challenging adverse decisions.  And we have this, though there’s a reasonable argument to be made that the TTAB should take more account of precedent.

If we decide that disparagement is too vague a standard, then it will be difficult to distinguish almost anything in trademark that we discuss with our students—the line between suggestive and descriptive, which has significant consequences for registrability; so too with the line between descriptiveness and genericity; the standard for confusion refusals under 2(d) as well as the infringement standard under 32 and 43(a); the definition of false advertising; and so on.  Someone like Justice Breyer, who likes to balance, could readily conclude that the game isn’t worth the candle for disparagement but is for these other exclusions, but most of the Justices on the Court are uncomfortable with that kind of ad hoc balancing.  I fear that if they accept the vagueness argument they will just encourage other courts to lie about how vague the other parts of trademark law are, and that makes our law less law-like.

Vagueness & the contradictory Dykes on Bikes rulings, discussed in Lemley’s brief, where the word mark was allowed but a later stylized word mark was refused even though it was just a narrower version of the word mark: Great example because there’s a recent exact match in the genericity space: Churrascos, which was allowed as a standard word mark for restaurants, then rejected for being generic when the registrant tried to register a stylized mark.  Is this dumb?  Yes.  Should we change the rule so that there’s more issue preclusion at the PTO?  Maybe, though I’m concerned given the installed base of decisions, and given the lack of true adversity in many of these cases in which the applicant is the sole source of information about the meaning of the applied-for matter.  But more to the point, I think we should be asking administrative procedure questions about dealing with inconsistencies—if we constitutionalize this process, then we’re inviting the process to grind to a halt, or convert into a filing-based rather than truly examination-based regime.

Mark Lemley: Viewpoint/content doesn’t matter for the outcome because Tam wins either way.  Under any kind of reasonable way, this has to be viewpoint based. Redskins, if it goes down in TM office, goes down b/c it’s offensive to Native Americans. But “Native Americans” is ok.  They want to discourage bad messages and not discourage good ones.  Sorrell: Regulation b/c of disagreement w/message it conveys is subject to heightened scrutiny.  NAACP is registrable but not Nat’l Ass’n for Colored Persons—that’s classic viewpoint discrimination.

At the very least, this is content based restriction on speech. They are applying this new and special law b/c of the topic discussed. That makes it content based.  It is true that the justification is that people are upset by disparagement.  Heckler’s veto.

Our odyssey: the only evidence was that lesbians were perfectly happy w/our use of the term dyke.  What we did see was that the second examiner looked it up and found bad terms including dykes. Counsel said “I’m a dyke,” and you could hear an audible gasp—this examiner was personally shocked.  Even if we say we will apply it neutrally, we use gov’t as listener, gov’t as decisionmaker, and that’s why you get crazy inconsistent decisions.

Can’t pass strict scrutiny under content based. Gov’t has no interest justifying rule other than suppression of speech itself. Less restrictive alternatives. Gov’t can disassociate itself from the speech. Could say on each registration.  Could even probably get away with only not endorsing some of them.  But there’s no way in which they can survive strict scrutiny.

Commercial speech/not doesn’t matter. TM can promote and repress speech. But that fact doesn’t change the fact that what we’re doing when we regulate disparaging speech is regulating things w/real expressive impact.

We’ve all assumed that 43 protection was available w/refusal of registration, but there aren’t any cases saying that. Most of the other aspects of 2 have been incorporated into 43, so it wouldn’t be surprising to have this one either.  If you’re restricting the right to have TM protection at all, that’s well beyond unconstitutional conditions.  This isn’t just a gov’t benefit; it has real significant consequences in enforcement.  It was impossible for a small nonprofit to prevent blatant knockoffs w/o registration: eBay, Etsy, CafePress all want registrations to pull merchandise. If you can’t show them TM registration, they say go away. That’s perfectly reasonable; they don’t know who has common law rights or who’s the first user.  At least TM registration makes them comfortable. Practical effect isn’t just additional remedies.  It’s the difference b/t effective enforcement and lack of effective enforcement.  Even if that weren’t true, unconstitutional conditions cases don’t necessarily apply—just b/c I can impose a harm on you, doesn’t mean I can do it for reasons or purposes themselves directed at the repression of speech.  Even w/ no right to benefit, some reasons on which gov’t may not rely.

Real problem: even if we try to fit this in unconstitutional conditions doctrine, doesn’t avoid scrutiny of purpose.  US v. Playboy: distinction between burden and ban is just a matter of degree; burdens must satisfy the same rigorous scrutiny as bans.

If this is unconstitutional, what else goes down with it?  This is a slippery slope argument. Refuse to do the right thing today for fear of being forced to do the right thing tomorrow.  If we applied rigorous First Amendment, most results would be salutary. We might lose scandalousness.   One way might be to decide this on viewpoint discrimination and then say the rest are just content-based.

RT: It’s not viewpoint discrimination.  Other groups are protected to the same extent as Native Americans. That’s not viewpoint based b/c it doesn’t promote any group over any other.  Content based sure, as is the rest of TM law.

Note all the quotes in Lemley’s presentation about prohibition, and ban, and suppression, and punishment—these quotes make sense when you are talking about sending people to jail or fining them for their speech. But the government deciding not to include specific words in their program is different.  If Tam couldn’t get a registration until he gave up using Slants, everything Lemley said would be correct.  But he can, so it isn’t.

Sorrell: disagreement w/the message = heightened scrutiny.  Consider this statement: The Holocaust happened. Can public schools teach this and give students bad grades if they don’t acknowledge it on tests?  Don’t want to impose Sorrell’s standard when the gov’t is creating the program, and that’s why we have unconstitutional conditions.  Can fund abortion and not contraception or vice versa.  Democracy promotion but not authoritarianism (or perhaps, upcoming, vice versa). 

All of TM law is content based.  That’s the problem. The rest of TM registration can’t survive strict scrutiny either; the less restrictive alternative is always fraud, or even the common law.

Gov’t as decisionmaker: he says that vagueness of disparagement is why you get crazy inconsistent decisions.  But it’s not—Churrascos is the same. The size of the program is why you get crazy inconsistent decisions.

Gov’t interest is nonparticipation in disparagement by gov’t.  Shelly v. Kraemer type rationale.

He says: “what we’re doing when we regulate disparaging speech is regulating things w/real expressive impact.” But that’s the point: because the alleged infringers will have an expressive impact too, there is no neutral position from which to evaluate a rights-granting program.

If we applied rigorous First Amendment, he says most results would be salutary. I like that with infringement, sure, but not so much ITUs and nationwide protection and extension of protection under Paris Convention.   And TM plaintiffs get to yell property rights and not have the same rules apply to them.  Note the majority in Tam said that protecting investment was unquestionably one of the okay purposes of TM law, so I fear that only some parts would get real scrutiny.

Jake Linford: why doesn’t SCt pretend that TM are just like © and find some traditional contour—there’s enough 1A protection in the system already?

Lemley: That would be wrong. There are substantial differences explaining why the court hasn’t said a 1A free ride. You’ve built in more speech protections to © than TM. Fair use might or might not be sufficient, but it’s an effort to protect expressive uses.

The Court is influenced by existence of Patent/© clause in Constitution; couldn’t say that with a straight face re: Commerce Clause.

Linford: recent UHS of Delaware case: pleading standards for 32 and 43 increase the requirements for success under 43: does that shift your perspective on harm to TM owners who can’t get registrations?

RT: No, b/c I think it’s ok not to give 43 protection to disparaging marks.

Jennifer Rothman: Whether it matters who’s filing w/r/t disparagement.  It does seem that perspective is based on who’s filing.  Dykes on Bikes. Even if it’s just content based, not convinced that gov’t interest is sufficient, b/c it’s not gov’t speech.

RT: I don’t think the record bears that out. Democrats and Republican examples, not to mention all the Bitch and anti-woman marks, suggest differently.  They go all different ways.

Gov’t doesn’t have to be speaking to have an interest in not endorsing.  Shelly v. Kraemer is not about public perception but about what positions the gov’t can accept without implicating itself in discrimination, e.g., it's acceptable to call women the c-word.

Irene Calboli: A child wearing Dykes on Bikes would be sent home from public school. That’s suppression of speech. How do you reconcile that?

Lemley: True in elementary school; untrue in college; shouldn’t be true in high school.  If they throw me out of college, I’ve got a slam-dunk First Amendment case.  We pretend that kids aren’t actual people.

Ann Bartow: My sense became that it mattered if someone complained; that can help make sense of inconsistent outcomes.

Lemley: that certainly matters.  In Dykes on Bikes there was one guy who thought lesbianism was bad; lost for lack of standing.  My sense is same as Rothman’s—this is not neutrally applied, but gendered, even as it is supposedly protecting minority groups.  Systematically reinforces some standards.

Bill McG: prediction?

Lemley: affirm.  Viewpoint; or content discrimination with bad justification; may not say anything about the rest of the system. As a practical matter, there’s not much reason to be concerned.  Beyond scandalous marks and maybe emblem or insignia, where there is a plausible confusion based or search cost based justification, courts will just continue as is.

Tyler Ochoa: Does not registering suppress speech?  Granting TM gives Tam a right to suppress others’ speech: on balance registration suppresses more speech than it encourages.

Lemley: always true that there are speech interests on both sides. It’s burdening Tam’s speech to shunt him towards certain types of speech rather than others. That’s problematic under 1A even if the result is more infringement.

RT: The registration will also shunt potential defendants towards certain types of speech rather than others.  Lemley is right that it’s impossible to add this up and that we should probably look to reasons instead.  Then we disagree about whether the reasons are ok.

Q: Institutional counterfactual about vagueness: The PTO is weak; it can’t write substantive rules. In an alternative universe where it were like the EPA or SEC, would a code have helped?

Lemley: it would have helped if you think the problem is vagueness, not an affirmative burden on speech.

Friday, January 06, 2017

Reading list: anti-surcharge rules and behavioral economics

I found this quite interesting, though I am much less skeptical of behavioral economics than the authors.  Most significantly, I am very sympathetic to the Occam’s Razor argument that the no-surcharge rule serves obvious consumer protection functions across a range of plausible circumstances, and I’m intrigued by the authors’ claim that countries that allow surcharges are plagued with complaints about excessive charges and insufficient disclosure.


During the past decade, academics—predominantly scholars of behavioral law and economics—have increasingly turned to the claimed insights of behavioral economics in order to craft novel policy proposals in many fields, most significantly consumer credit regulation. Over the same period, these ideas have also gained traction with policymakers, resulting in a variety of legislative efforts, such as the creation of the Consumer Financial Protection Bureau. Most recently, the efforts of behavioral law and economics scholars have been directed toward challenging a number of state laws that regulate retailers’ use of surcharge fees for consumer credit card payments. In part as a result of these efforts, the issue has come before multiple courts, with varying outcomes.
In 2016 the issue reached the Supreme Court, which granted certiorari in Expressions Hair Design v. New York for the October 2016 term. The case, which centers on a decades-old New York state law that prohibits merchants from imposing surcharge fees for credit card purchases, represents the first major effort to ground constitutional law (here, First Amendment law) in the claims of behavioral economics.
In this article we examine the merits of that effort. Claims about the real-world application of behavioral economic theories should not be uncritically accepted—especially when advanced to challenge a state’s commercial regulation on constitutional grounds. And courts should be especially careful before relying on such claims where the available evidence fails to support them, where the underlying theories are so poorly developed that they have actually been employed elsewhere to support precisely opposite arguments, and where alternative theories grounded in more traditional economic reasoning are consistent with both the history of the challenged laws and the evidence of actual consumer behavior.
The Petitioners in the case (five New York businesses) and their amici (scholars of both behavioral law and economics and First Amendment law) argue that New York’s ban on surcharge fees but not discounts for cash payments violates the free speech clause of the First Amendment. The argument relies on a claim derived from behavioral economics: namely, that a surcharge and a discount are mathematically equivalent, but that, because of behavioral biases, a price adjustment framed as a surcharge is more effective than one framed as a discount in inducing customers to pay with cash in lieu of credit. Because, Petitioners and amici claim, the only difference between the two is how they are labeled, the prohibition on surcharging is an impermissible restriction on commercial speech (and not a permissible regulation of conduct).
Assessing the merits of the underlying economic arguments (but not the ultimate First Amendment claim), we conclude that, in this case, neither the behavioral economic theory, nor the evidence adduced to support it, justifies the Petitioners’ claims. The indeterminacy of the behavioral economics underlying the claims makes for a behavioral law and economics “just-so story”—an unsupported hypothesis about the relative effect of surcharges and discounts on consumer behavior adduced to achieve a desired legal result, but that happens to lack any empirical support. And not only does the evidence not support the contention that consumer welfare is increased by permitting card surcharge fees, it strongly suggests that, in fact, consumer welfare would be harmed by such fees, as they expose consumers to potential opportunistic holdup and rent extraction.
As far as we know, this is the first time the Supreme Court has been expressly asked to consider arguments rooted in behavioral law and economics in reaching its decision. It should decline the offer.

Wednesday, January 04, 2017

Reading list: 2016 TM year in review

 Mark P. McKenna & Shelby Niemann, 2016 Trademark Year in Review:
This brief essay reviews some of the most significant developments in trademark law during the past year. In most cases we have interpreted “year” fairly liberally, particularly to highlight some longer-term trends. We focus on six areas: (1) the constitutionality of § 2(a) of the Lanham Act; (2) the Fourth Circuit’s Belmora decision and the availability of § 43(a) claims when the plaintiff has not used a mark in the United States; (3) the effect of B&B Hardware; (4) injunctive relief and the presumption of irreparable harm; (5) nominative fair use; and (6) initial interest confusion.

Tuesday, January 03, 2017

Lexmark harm requirement reopens gap between registered and unregistered TM infringement

UHS of Delaware, Inc. v. United Health Services, Inc., 2016 WL 7474801, No. 12-CV-485 (M.D. Pa. Dec. 29, 2016)

If other courts follow this logic, this could be big.  Plaintiff UHS Delaware sued United Health Services, Inc., and related defendants for trademark infringement under §32 and unfair competition under §43(a), and related state claims.  The parties cross-moved for summary judgment, and the court denied most of the motions, with a significant exception.  And there’s a bit about how (not) to admit evidence of consumer confusion through hearsay exceptions.

UHS Delaware is the healthcare management company for Universal Health Services, Inc., a for-profit company whose subsidiaries run more than 235 acute care and behavior health facilities and surgery centers in a number of states, though not New York. UHS Delaware doesn’t itself provide healthcare services, but rather manages an affiliated network of healthcare service providers. UHS Delaware owns incontestable federal registrations for two trademarks: a UHS word mark and a UHS stylized mark, for use in connection with hospital services and hospital management services.
UHS Delaware stylized mark
United Health Services is eight nonprofit corporations which together comprise an integrated healthcare system headquartered in the southern tier of New York.  Defendants began using “UHS” around 1982.  In 1997, it adopted a new system-wide brand:

In 2010, it rebranded again with a new logo, and adopted a policy that, for all its entities, “UHS” would be followed by a specific location or services identifier, such as “UHS Delaware Valley Hospital.” UHS Delaware sued in 2012.
New United Health Services logo
“Claims for trademark infringement and unfair competition share the same essential elements under both the Lanham Act and Pennsylvania law.” Except: Lexmark applies to all §43(a)(1) claims, given the clear language of the opinion and the statute.  To maintain a cause of action under § 1125(a), “a plaintiff must plead (and ultimately prove) an injury to a commercial interest in sales or business reputation proximately caused by the defendant’s misrepresentations.”

UHS Delaware submitted no evidence of injury to sales or reputation.  It withdrew its claim for actual damages, and argued that damage to its business and goodwill wouldn’t be calculable without undue burden and expense.  “This deficit of proof necessarily precludes any assessment of proximate cause.”  Thus, defendants won summary judgment on the §43(a) false association claim and state common law unfair competition claims.

This conclusion, while it could shake up quite a lot of §43(a) cases, wasn’t as important here because of UHS Delaware’s federal registration.  Turning to §32: Likely confusion is fact-intensive, and the court found this case was not one-sided enough to justify summary judgment for UHS Delaware.  The marks were highly similar.  The Third Circuit has said that when “products are directly competing, and the marks are clearly very similar, a district judge should feel free to consider only the similarity of the marks themselves.” However, there were genuine disputes about the overlap, if any, between plaintiff’s and defendants’ respective target markets, so the court didn’t stop there.

UHS Delaware maintained that its mark was inherently distinctive (though not because of incontestability, which doesn’t bear on strength); defendants didn’t disagree, presumably because of their own interests, because an initialism like UHS is really just as descriptive as the words for which they stand.  However, defendants argued “that any conceptual strength the marks may have in isolation is diluted by widespread third-party use of the ‘UHS’ logo in both medical services markets and beyond,” and the court found they’d submitted a fair amount of evidence to that effect. Similarly, there wasn’t strong evidence of commercial strength, and general ad expenses don’t equate to promotion directed specifically at the UHS marks. Nor did any consumer surveys or focus groups establish actual market recognition.  Without such evidence, “[t]he dilutive effect of third-party usage clearly undermines the marks’ overall strength. Consequently, this factor is neutral.”

Though the consumers included ordinary patients, the sophistication of that group could still be expected to use “a heightened standard of care” in selecting medical services, favoring defendants.

Length of use/actual confusion: isolated and idiosyncratic examples of confusion don’t count.  The court discounted a pre-rebranding email in which one of defendant’s employees observed to a colleague that “there are a few UHSs out there” and “we sometimes get e-mails and inquiries meant for them.”

Other evidence was: a CNBC broadcast in November 2010 in which defendants’ logo was displayed on screen while plaintiff’s chief executive officer was speaking about Universal; testimony of three of defendants’ officers that both UHS Delaware and United Health Services employees received telephone calls and other communications meant for the other; testimony about a subordinate’s 2013 report to her that a news report had run in Binghamton, New York, referring to a United Health Services facility as a Universal facility in both online and on-air stories; a 2015 tweet including United Health Services’ Twitter handle when referencing Universal; testimony of a Universal employee about a misdirected 2015 email received from Healthgrades about a potential award to defendants’ Wilson hospital; and a 2016 billing dispute letter addressed to “United Health Services” but mailed to both United Health Services and to Universal.

Defendants argued that this evidence was both inadmissible and irrelevant. Kos Pharmaceuticals, Inc. v. Andrx Corp., 369 F.3d 700 (3d Cir. 2000), held that the first level of hearsay—employees recounting customer statements evincing confusion—aren’t excludable hearsay because they aren’t submitted for their truth, but rather for their error.  “It is the second level of hearsay—statements by employees to the deponent describing telephone calls—that must satisfy an exception to the hearsay rule.”  UHS Delaware failed to identify an exception justifying the admission of this second-level hearsay. “Officers may permissibly testify to the fact that they received reports of confusion from their employees, but the hearsay rule prohibits any substantive recount of underlying conversations.”  Anyway, the excluded evidence wouldn’t change the summary judgment ruling, because none of the conversations seemed to establish actual confusion based on the marks, and “[a]necdotal evidence of misdirected telephone calls or letters is usually insufficient to prove actual confusion, particularly when the communications are relayed by interested employees and do not concern the purchase of products or services.” 

The court also pointed out, as to the CNBC interview associating defendants’ logo with Universal’s CEO, “that error may just as likely have derived from the incredible irony that both companies’ CEOs shared the name ‘Alan Miller.’” The other instances of alleged confusion were “few and far between,” “negligible” compared to more than six years’ use of similar marks.

Defendant’s intent: this requires intent to confuse, not just intent to copy.  “[M]ere knowledge” of a competing mark is insufficient to prove intent to confuse.  All UHS Delaware showed was “a potential awareness of Universal’s presence in the general healthcare marketplace,” and defendants’ stated objective was, at least in part, “to distinguish itself from competitors such as United Healthcare and Universal Health System.”

The existence and degree of overlap among the parties’ trade channels and target markets: the parties both advertised hospital and healthcare services through print, online, and television media.  But there was a factual dispute about things like how much UHS Delaware advertising NY consumers would be exposed to; UHS Delaware runs ads in national publications, including the NYT and the WSJ, but doesn’t target New Yorkers.  Some New Yorkers receive services from Universal entities, but not many—only 1% of clients were New Yorkers, except for one facility with a 4% New Yorker population.  “The court cannot ascertain from the record whether the cited examples of overlap derive from UHS Delaware’s marketing efforts and consumer brand recognition, or the sheer happenstance of New York residents requiring hospital services on vacation.”

Relationship of the goods in the minds of consumers: the services are very similar, favoring UHS Delaware.

Likely expansion: This isn’t actually a great shorthand, because it’s supposed to encompass “other facts” “from which a consumer might expect that the trademark owner, through expansion or otherwise, is affiliated with the alleged infringer.” This was neutral, because there was no evidence about consumers’ expectations about expansion; the fact that New York law bars for-profit corporations like UHS Delaware and Universal from operating healthcare facilities in the state wouldn’t necessarily be known to consumers.

The court also denied United Health Services’ motion for summary judgment on its §33(b) senior user defense, given questions about whether its rebranding in 1997 constituted an abandonment of “UHS.”  Most of the uses after that seemed residual, which isn’t enough to avoid an abandonment finding.  Third-party uses of an abandoned mark in news reports wouldn’t be enough, either.