TMSR Session 2: Administrative Agencies and Specialized Courts

Introduction: Robert Burrell

IP Australia exerts extraordinary control over legislative agenda; if you can persuade IP Au. that something needs to change, then it can change quite quickly. But the flipside is that it’s seen in light of what the Office needs to worry about. There have been problems with defenses; the Office doesn’t have to worry about infringement, so nothing changes. But to deal with different holidays in different states, emergency legislation was passed within weeks so the Office would know how to count deadlines.

Volume of decisions by hearing officers: registrability decisions have first mover advantage. They ought to be able to exert significant control over case law: courts should be slow to depart from existing practice.

Limited oversight: fed court worried about undermining confidence in specialized tribunals.

Office can also react if not happy w/court decisions. What matters is what the TMEP says the court said, which isn’t always the same as what the court said; sometimes says “the court can’t have meant that.”

IP Australia also regulates patent & TM attorneys through professional standards board domination.

Academic research: you have to demonstrate you’re bringing in grants; a great way to increase the chance of that is an industry partner; the best funded, most active one of those in TM is IP Australia, so that controls the focus of research. Recently appointed first in-house economist; has now metastasized. IP Austl. doesn’t know what to do with all of its money (it’s self-funding), so we have the Wallace & Grommit World IP Exhibition, with Grommit chasing off copycats.

Most important countervailing factor: competence. Leading case on distinctiveness, where Australian High Court lowered the distinctiveness bar—until it reached the high court, it was being litigated by a construction partner. In fairness to the high court, none of the evidence you’d have expected to have in about the importance of “Italian” when you are selling coffee was before the court, so it’s not as surprising they did something done. So in litigation, no one says “the Office has done this 20 times before.”

Mark Lemley

Agencies in TM law: PTO is sort of a black box b/c ignored TM so long on the policy side. It would never occur to anyone to put a TM person at the head of the PTO; it’s a tiny percentage of their policy work. The good news is that there is an opportunity to do policy work not tainted by existing capture. If and when that changes, INTA will show up. A broader/systemic problem is the problem of scope: we see this in patent side when Office tries to intervene. The business of the PTO is granting patents/TM. They have a lot of experience w/distinctiveness and functionality, as well as failure to function. But they have virtually no experience, except in TTAB, with other side of the equation, remedies, 1A limitations, etc. When they bump up against those in cases like Tam they do a lot of running in place w/o useful guidance. Even if you could navigate around capture, the scope one is harder b/c it’s systemic; you can’t give them deep knowledge on the infringement side. One of the risks of looking at only one side is that you can’t actually grant a “narrow” registration.

TTAB is better suited, b/c at least get microcosm of LOC cases b/c of competing TM claimants. Though they aren’t in the same position to do agency rulemaking and also suffer from Fed Circuit gatekeeping, which has its own rule for LOC—twenty other factors that no other circuit pays attention to; the Fed Cir refuses to follow SCOTUS on functionality while other circuits comply. It’s a tough process to get SCOTUS to fix.

And of course, agency deference is going away this year. Benefit of being able to point to those rules seems much less.

Two other decisionmakers in practice making lots more decisions: The UDRP panels in cybersquatting. There were a lot of benefits: quick, cheap, but also out of control. Could learn to facilitate simple cases more cheaply; most people who have legit defenses can’t afford to assert them in court. A system that gave an opportunity to do that outside the normal court system might have some benefits, but UDRP offers cautionary tale about structuring that. Who gets to decide where cases go is important.

Fromer/McKenna paper on Amazon’s TM system: that’s not a positive story; they identify a number of problems. But consider potential positives: Fromer/McKenna show that, b/c you need a registered mark to participate in the system, there are a lot more people registering marks than they used to be, especially small businesses. This has risks/problems of clutter and depletion, but it doesn’t have to be a problem b/c you’ve got to have the same mark—except that we don’t look at images, which is a huge problem—but Amazon only cares about identical marks. Thus, registering nonsense words. They function for the Amazon system but not in the minds of consumers. That can have problems, but a whole bunch of marks that are extraordinarily narrow/cast no shadow and are only going to be applied against identical copying doesn’t necessarily create a terrible system, especially under conditions of congestion. Replacing human mnemonics with a system that uniquely identifies a good & where it’s from & can connect it to rankings might be good! But caveat that it’s not clear that Amazon has that system. We might see a separation of TM function of TMs and the brand loyalty function of handbags. We’ve been struggling w/the growth of the brand in TM context, distorting source identification function; Amazon could help us break that linkage.

Jeremy Sheff

2 years of expungement: what’s been happening? Director’s initiated 33 expungements; 8 proceeded to final disposition of expungement (1 partial, 7 full). 771 petitions for expungements; only 122 have gotten to institution stage which involves an Office Action. 74 have had an outcome: 32 partial cancellation, 42 full cancellation. In 80 cases Office decides not to go forward for insufficient evidence of expungement, though they then proceed with their own action including evidence dug up by Examiner. So PTO may be cleaning up after petitioners.

Filing bases for director-initiated: 15 44(e), 14 1(a), 4 66(a). Non-director: 381 1(a), 233 66(a), 149 44(e).

16b reexamination of use: 1134 director-initiated, 954 third-party initiated. 304 director-initiated proceeded to full cancellation, 4 partial. (Question to explore: is the large number of these connected to post-registration audit?) Non-director initiated, 92 were not instituted, 140 instituted; 19 partial cancellation and 57 full cancellation.

With three million active registrations, we might be trying to empty the ocean with a teaspoon.

Post-registration audit percentages have very different numbers; they choose about 5000 registrations/year and about half end up deleting goods/services. So a random pull has a 50% chance of finding lack of use on something. For 44(e) and 66(a), about 2/3 of those audited delete goods/services, whereas it’s 45% for 1(a). That’s of registrations w/a response. Total cancellations (3883 of 29022 audited) may be for nonresponse.

What to make of this? Shoveling sand against the tide, b/c we have ¾ million-1 million new applications/year. It’s not just the ocean, but the iceberg melting into the ocean, which means that bad actors flooding the system w/new applications to gain advantage in private systems, we’re dealing with that process with a bespoke, artisanal examination process.

If we’re running out of TMs, that’s a fixed stock that could go away. The examiner cohort is a flow resource, not a stock; they can get congested. We could increase the price! Raise registration and renewal fees; make it more difficult to get a registration for Amazon, which might lower the burden on the system.

In response to question: we don’t charge by number of goods listed in a class, only by class, but listing more goods does increase the chance of auditing.

Nobody who’s been audited would have selected goods/services on basis of concern over auditing (in response to Burrell’s question about whether that increases the incentives to use broader generic terms to describe goods/services).

Fighting a forest fire with a water pistol won’t work.

Dinwoodie: could be useful if actually internalized—if it happens to repeat filers. Could also add penalties. [There are costs for deleting goods, but they’re small fees.] Could make a penalty fee.

Burrell: the problem is that it runs against the story “you’ll hurt SMEs!” Self-filer makes mistakes, poor self-filer.

Dinwoodie: could you do it more granularly?

Burrell: not sure that’s possible to avoid the innocent SME.

RT: Carol Rose was right all along. Property registries are themselves a common that can be overused. And her insights in that article might suggest something about what comes next, after the TMA’s tiny tweak that clearly won’t be enough.

Expertise issue—FTC just issued a rule on impersonation; only heard from INTA and the like, and apparently doesn’t know anything about TM doctrines or how far INTA thinks “impersonation” extends. Calling scholars to submit comments as FTC considers supplemental notice of proposed rulemaking: whether to impose “means and instrumentalities” liability on third parties like OpenAI and whether to extend its new rules to pretending to be an individual, whether a real person or a completely made up one—that is, the FTC might make it an unfair trade practice to use an alias.

URS as extension of UDRP: faster, a little cheaper, more out of control.

Fromer: definitely true that TM is swamped by patent at PTO—hard to get attention on TM issues, and internally it’s created a very sleepy culture in the TM office. You might get a dynamo like the people who self-started the audit program, and more of that would be great, but it doesn’t emerge very often. Circular: the reason patent swamps TM is patent brings in more money, so there’s a very rational explanation. But it doesn’t have to be that way. It would be great for the PTO to charge Coca-Cola more than a few hundred dollars to renew. Money brings attention.

FTC is a place that could do more TM and might be interested [they already have opinions on ©] and need education on non-patent IP. TM examiners used to be specialized b/c it didn’t seem fair that some units were clearing applications faster b/c of the different demands of the different areas; that would be hard to undo, but technical devices really are different from beverages. Think about reinstituting specialization.

Amazon [like the Trademark Clearinghouse for new gTLDs] does not distinguish in the kinds of goods for which a mark is registered, so you can use your registration far more expansively than the PTO designed it.

Ramsey: one suggestion is to require a specimen of use for each listed good, not just one good in each class. Also, PTO should bring audit results to Congress’s attention for further action.

Treat classes commonly used for expressive uses differently—clothing, mugs versus technical equipment.

Lemley: suggests maybe the existence of the opposition process changes things. Someone is immediately looking at your work with a critical eye—that never happens with a patent. But doesn’t know what examiners’ involvement with that is.

Dinwoodie: intensive review for productivity on the TM side.

Farley: past practice was collaborative, knowledge sharing. Remote work has changed that.

Silbey: the incentives are bad: they’re supposed to issue TMs.

Lemley: system is set up to push you in a particular direction.

Silbey: what if the expectation was that some number of applicants should not get TMs?

Sheff: there’s other stuff we haven’t talked about that are interventions: Domestic residence requirement has done a lot of work. 150 cases in which Office identified people who filed 100 applications/week and issued order to show cause and ultimately sanctioned by forbidding them from filing more; resulted in 19,000 applications from those applicants. TMA seems to have led to a decrease of 100,000/year.

Burrell: If we were serious about this, we would have to think about automation/scale. Austl. doesn’t even revoke TMs when a company is deregistered—why not do more information sharing? So that if the mark were to be transferred as an asset, some work would have to be done.

Sheff: linking systems costs a lot!

Burrell: That’s what increased fees are for.

Silbey: expertise can be about narrowing a claim rather than not issuing a TM at all.

Dinwoodie: that’s already a distinctive feature of the US system, which does it more than anyone else.

Burrell: US examiners are better than anyone else in the world at narrowing nonsense, even if that’s a low bar.

Silbey: you need expertise to do it really well, though. And it’s harder if you have less time per application.

Study showed that if there’s a 6 month decision rule, decisions issued in the last 5 days of that period are more likely to be overturned on appeal.

McGeveran: narrowing someone, or rejecting them, isn’t as fatal here as it is in other jurisdictions.

Mid-Point Discussants: Jake Linford

If registration doesn’t matter to TM pathologies, then fixing examination won’t matter. There’s some evidence of heterogeneity among examiners. Hydraulic pressure: Tam & Brunetti put pressure on TM office; they don’t want to grant disparaging/scandalous—failure to function steps in. Not clear whether that was top down or bottom up response, but examiners can find ways to deal w/at least some pressure, and TTAB is willing to support this exercise of discretion. Brunetti doesn’t get FUCK registered on f2f grounds. Student note criticizing f2f as standardless/needs concretization—make it more like aesthetic functionality and double down on competition. But do examiners think through what we would call speech-related concerns? You could say f2f has or includes speech concerns. Though you do see PTO explicitly invoking speech concerns in defense of its rule in Elster.

You could increase the cost of renewal over time, as Posner suggested for ©, but we might not want that for policy reasons. You could increase fees, but that might exacerbate the problem of seeing applicants as the customers.

Fraud on the PTO as an option? The Fed Cir cracked down on that; we could go back to that.

INTA was concerned that TTAB wasn’t publishing enough of its opinions; would that give better guidance? Only 6% of TTAB opinions are published; 40% of signed opinions of courts of appeals.  [It would only increase the scatter plot.]

We complain about costs of TM litigation but don’t like the Copyright Small Claims board; INTA represents Ps and Ds but only very sophisticated ones, united in their sophistication; could we think about a small dispute procedure? [Isn’t that one view of registration?]

Stacey Dogan

Where to go? Structural competence versus personnel-based competence. Appellate courts are supposed to have a broader eye to the interests at stake than just the parties; we want them to be equipped to define legal rules that achieve balance in the law. This aspiration for them requires competence, expertise, capacity to engage in a flexible normative project that can respond to new questions and situations. When that competence is absent, either b/c of structural constraints or institutional failure, that’s disorienting, b/c where do you look for correction?

SCOTUS failures don’t seem structural; not unrelated to the larger legal ecosystem. The PTO is inherently limited. The courts will have to play a significant role in defining the shape of TM rights. The nuggets of constraint the Court gives us seem plucked from the ether, without context or explanation.

We’re not going to solve the personnel issues in the near term. Strategies in the short term: we’re left with what we have, and find ways to work w/in doctrine like RT suggested. In months since Jack Daniel’s, the caselaw is not giving lots of hope. The precise uses that the majority takes for granted—titles of expressive works—are being treated as source-indicating by the lower courts. There’s an opportunity to work on lower courts, both trial and appeals, on giving more texture and principle and normative content to the LOC factors. We’ve seen courts do this in some contexts. Use in the internet, we saw courts willing to limit factors in particular cases; could do that with expressive works as well in a way that’s less “doctrinal” or test-based. Allows for some level of instruction and consistency in cases of certain types, like use in expressive ways.

Movement lawyering? In communication with affected constituencies, working at defining more clearly the problems we’re trying to solve. Identify the greatest threats and capture a theme. We have to accept the lack of traction we get describing TM harms and make it more saleable. Look to antimonopoly movement in antitrust space. It’s a long term project. Using rhetoric + empirical evidence about the costs of concentration. Speak to the public. Partner with law & political economy folks; make it into a larger project about the economy. Project: what is public interest TM law?

Leah Grinvald

CBP: engaged in development of and enforcement of TM law w/o any judicial oversight. Others have noted a shift of the costs of private enforcement to public/government. TM holders can record with CBP, and is used when scanning shipments that come in through ports. Invites TM owners to provide training materials for their officers; even have a powerpoint format for them. Invites TM holders to come to offices, bring samples to train the officers (swag).

They ask TM owners to ID counterfeits. Their main focus is on detaining and seizing counterfeit goods; they say they’re protecting the public from counterfeits, whether handbags or otherwise, as a safety issue. They’ll send the TM owner photos unless they pay a bond for samples. Regardless of how TM owner verifies it’s counterfeit, they take the TM owner’s word for it.

Importer has 7 days to respond if they have a valid reason for importing; if not, the products are seized and mostly destroyed, but there are provisions for donations (despite safety concerns) if they need the product—by auction, to nonprofits, or if gov’t needs them. TM has to be removable from product and product must not pose harm, though how they know that is a good question if the reason for seizure is to avoid harm. Anecdata: mostly the products are destroyed. [Some may go home w/the agents.]

They can also seize goods with copying or simulating marks, not just counterfeits. The regs don’t give guidance on what that is; the importer has 30 days under this provision to show either that they have good reason or that they can remove the TM. There’s nothing in the regulation that restricts officers to detain only products within the registered class, despite the fact that marks are supposed to be registered per class. Backdoor dilution claim w/o need to prove fame or any kind of association.

All of these decisions are done at the level of the officer working hand in hand w/the TM owner; officers aren’t trained in law; there isn’t anything like an internal appeals process—importers rarely go to court b/c they are mostly small businesses. No consequences for mistakes to the TM owner or the officer; just release the goods!

Dinwoodie: US complained to WTO about China not destroying goods; shouldn’t we be doing better?

Grinvald: not the first time we’ve been hypocrites; the regulation allowing it is still on the books.

Burrell: this system is being exported through the world because of the USTO, and it’s now in Australia. Timeframes are so tight; even if SME wants to challenge, must find a law firm that is willing to work around the clock, which is massively expensive.

McKenna: they won’t even tell you what they’ve seized. They have an internal appeal process that’s a joke. The TM owner gives them an image; they call the TM owner and ask (w/o a picture) if they should seize the products and the TM owner says yes. You have to sue the gov’t to get them back and there is no remedy against the gov’t. With LKQ, a whole bunch of the seized parts were covered by a design patent license, but they said “too bad, it’s TM.”

Dogan: you can get substantive law right and still get wrong outcomes if you get the incentives wrong, and these stories might be helpful in getting the harms out to the public. Are there stories about seizures of useful spare parts/works that the public would be outraged about? Is there a way to make that come more alive?

Should we have conflicts rules that limit what TM owners can give agents?

Lemley: confront the rhetorical power of the anti free riding impulse.

Why does this system work at all? How do they know what’s mapped to relevant goods/services? Has anyone figured out whether it’s a random walk, mostly circumstances where they were alerted to a problem, what percentage of goods subject to seizure end up being seized, etc.?

Another form of seizure popped up ten years ago: internet domain name seizures. CBP is using the analogy to seizure of goods to seize domain names. They’re quite clearly at the behest of particular TM/© owners.

Grinvald: definitely an audit basis: they randomly open packages and check them against a list. Latest stats are from 2021; Hong Kong and China are top countries inspected. They also absolutely encourage anyone to submit tips.

Silbey: seems like a story similar to Mass. Anticounterfeiting law where industry was policing in front of stadium on behalf of NFL and just arresting T-shirt sellers. It’s an MO, not only at the CPB—officers and industry leaders work together. What are the narratives one might mobilize? Abuse of enforcement systems, collusion. Patent troll narrative was effective.

Lemley: but TM bullying stories don’t seem to move the needle in the same way.

Silbey: Eldred turned out to be a mistake; could not convince the Court there was a problem. Worry about SCOTUS and TM in the same way: the story we have to tell is not one they are willing to hear. Empirics, costs, and efficiency don’t matter.

Lemley: yes, the Justices clearly think people wouldn’t be confused, but VIP must nonetheless lose.

Silbey: might as well have said “traditional contours of TM” in VIP; so what can we learn from the patent trolling story?

RT: Better story: Emojis? Schedule A defendants? [Thinking about Lawrence Block’s Batman’s Helpers. Police corruption]

Just a reminder of my FOIA lawsuit against CBP. Happy to share the underlying responses with anyone doing research in the area.

Linford: bonding mechanisms? Schedule A story that got traction was the country music fan selling Luke Combs stuff. This person’s side hustle was sympathetic. But is an anti-merchandising right view popular in the US? That’s an empirical question. How many errors are enough before people stop saying “errors happen”? The procedural problem is less persuasive the lower you think the error rate is.

Goldman: we don’t know the error rate b/c it’s not researchable, and it’s not researchable b/c they’re not following the procedures.

Burrell: where’s the evidence that we need reliable data to persuade people? We need anecdata that is powerful: the government took my stuff!

Dinwoodie: When the RIAA sued a grandmother and a teenager, they looked really bad.

Linford: but the two cases that made the court of appeals, the RIAA won.

RT: but what does it mean to win? They aren’t doing it any more and they didn’t collect the money.

McGeveran: people see themselves as branders now. They identify with the “brand” idea.

Lemley: here’s what TM owners think you can’t do on Tik Tok: making fun of a brand; wearing branded clothes w/o permission—that might be a way to flip the narrative. As the world separates TM from brand, we should think about what it means to separate those doctrinally—would help us have narrower scope for TM law but also free us to rethink what a “branding right” would look like independent of confusion over sales of products. Not sure it would be based on confusion! A lot of bogus confusion survey stuff would not be important.

Originally dilution was an alternative cause of action in the states; you couldn’t bring both claims.

Burrell: IKEA case in Europe: “IKEA plan” used to describe anti-immigrant plan; court is considering whether there is “due cause.” Most likely will cut & paste from their copyright decision saying that satire & parody are ok and send it back to the Belgian court. In working out whether a parody implicated free speech rights, Belgium is entitled to take into account its equality guarantee, so being political speech wasn’t super protective.