Rebecca Tushnet

Research Directions

Discussants:  Laura Heymann: What is the end of empirical
work? Which of our legal principles are based on assumptions or are not meant
to be connected to the Q of how consumers actually behave?  Lots of research about design remains to be
done: what aspects of design are performing the various functions we care
about, etc. 

Are there additional
costs from having trade dress protection that don’t already arise from design
protection?  V. versa.  Damages questions: if we are to have
apportionment, how shall it be done? What motivates consumers w/r/t specific
features versus entire collection of features/general reputation?

Aesthetic
functionality came up a lot.  Figuring
out how to parcel out the issues that are now grouped under that umbrella. 

Unfair competition:
relates to question of remedies. 

Graeme Dinwoodie:
Private practice/private ordering: uncovering what’s there. We might be in a
pretty good position to exploit that.

Empiricism is
fashionable, but it’s hard to engage w/that legally w/o a normative structure;
we often gloss over normative foundations b/c we think we know them but they’re
often quite contested, especially when it comes to remedies.  European remedies also deserve reexamination.  Often taught last in a rush even though they
may be the most important part of the case.

Lemley: A lot of
theoretical justifications for TM, and for defenses/limitation, are or ought to
be essentially resistant to evidence. We might want to know what people think
when they see a can of Coke appear in a movie. Almost all his students think
there’s a sponsorship relationship, but should that inform TM law?

Bone: Empirical
results are often a mess. We as scholars may take it to extremes: pick studies
that support us, or review a lot of studies but ignore their methodological
limitations.  Sounds a note of caution.  If you’re uncertain, do you default to
current rules? Reform nonetheless?

McKenna: think more
broadly about remedies and relation b/t TM and unfair competition, if remedies
are starting to divide more b/t technical TMs and others. There’s much history
to be mined.  Key structural difference
b/t technical TM and unfair competition was precisely remedies, so we should
reengage that project.

Dinwoodie: Echo Bone’s
skepticism re empirical work—additional knowledge can’t be harmful if we
understand what it does and doesn’t tell us. 
Also, empirical work is really good if you have a particular utilitarian
view of the world. Otherwise, you need other value systems.

Bone: not so much
utilitarian as consequentialist. 

Yen: we might be
horrified by what we found out about empirics.

Lemley: It doesn’t
follow that we don’t need the evidence, but we must think intelligently about
how to use that evidence. Even a nonconsequentialist would want to know the
impact of rules.

McGeveran: Empirics
doesn’t mean big data; Silbey’s qualitative work is empirical and interrogates
TM myths. Some may unravel tidy assumptions about consumer use of
info/autonomy. There’s also lots of myths that are used in normatively
undesirable ways. Take the bitter w/the sweet.

Dogan: we should see
who’s suing whom in product design: qualitative examination of cases; also the
registry.

Dinwoodie: CTM
regime may be clogged; companies are getting more sensitive to costs.

McKenna: more
evidence on clearance costs may be available. 
One woman he talked to advises clients not to change packages b/c
clearance costs are so extraordinary: need to clear TMs, designs across EC and
nationally. Interview studies about how people decide on packaging could be
very useful.

Lemley: A nominative
fair use doctrine for product configuration: communicating similarity but not
confusion.

McGeveran: would you
care if there were empirical evidence of confusion from certain nominative fair
use? I wouldn’t.

Bently: Basic
confusion level is pretty high—no matter what, some people will misattribute a
Visa ad to Mastercard.

Yen: Does that
confusion interfere w/the operation of the market?  It doesn’t seem to make people skeptical
about what they’re buying in stores.

Heymann: there are
always anecdotes of people who accidentally buy the house brand in the CVS.

[Discussion about
search attributes and the complexity of search; plus consumers may well just
have general impressions when they go to the store.]

Dogan: how would you
fix secondary meaning??

McKenna: there’s no
good standard even for word marks. We assume that, if people say, “I think they
make it b/c of that word,” there’s secondary meaning and move on.  W/design, consumers can assume that but also
not be using the design as an indicator of source.   Length
& manner of use, advertising etc. don’t answer the question, just tell you
people are familiar with the thing—not how they interpret it. Same is true for
word marks, but we’re more willing to assume that long exposure to word = treat
it as a TM.  If we actually try to do it
right, it will be so hard that we will rarely succeed/it will be very
expensive.  Litigation surveys
particularly.

Bently: accept that
there are many myths around TM.  People
think the largest word on a product must be the TM. 

McKenna: Tom Lee’s
study shows it doesn’t matter what the word is [unless it’s generic];
majorities think it’s the brand if it’s presented in a certain way on the
package.  This is to say that we really
say a term is descriptive when it would be too costly to prevent other people
from using it.

Session 3: Remedies

In recent years,
trademark scholarship has focused largely on questions of subject matter and
scope, including much critical assessment of standards for infringement and dilution.  In this brief discussion, we will turn our
attention to remedies, and will consider whether creative approaches to
remedies might offer alternative means for reining in trademark law’s perceived
excesses, and/or for balancing competing interests in trademark disputes.  To this end, we will consider some of the
following questions:  What impact has the
Supreme Court’s eBay opinion had (and what impact should it have) on the
availability of injunctive relief in trademark suits?  Is it true, as many courts have held, that
the risk of permanent harm from infringement makes trademarks different from
other forms of intellectual property? 
Does it make sense to allow courts to grant limited injunctive relief
when they perceive a strong risk of confusion from a defendant’s use of a
symbol that indicates source, but has no trademark protection because of
functionality, genericism, or some other exclusionary rule (cf. Blinded
Veterans, Thermos, Singer)?  How
do/should courts measure damages in trademark cases, and how does it compare to
the approach in different areas of IP litigation (cf. the approach for design
patents)?

Introduction:  Mark Lemley: All the circuits that have ruled
on the issue have said eBay applies
to TM, so that we are to apply the four part test for injunctions preliminary
and permanent.  I don’t disagree w/idea
that eBay should apply but have
substantial concerns w/applications, particularly in 9^th Cir.
Doctrinal error: they’re more guidelines to be balanced than elements to be
satisfied.  In the course of dismantling
presumption of irreparable injury, they then bar injunctions w/o such
proof.  That’s not what eBay thought it was doing. Also
problematic in TM sense; it’s hard to prove irreparable injury in a large array
of TM cases once you take away likely consumer confusion as evidence of
irreparable injury.  Evidence of actual
sales diversion = is it reparable? 

Factor 2 is
indistinguishable from 1: adequacy of remedy at law. But unlike patent/©, where
if you win you will get damages, in TM it’s not just possible but relatively
common not to get damages. Sales diversion is usually required. Profits still,
in most circuits, require willful infringement. May be changing for a kind of
dumb reason, which is the enactment of §43(c), which has now been read as
intending retroactively to change the rule in regular infringement, since
willfulness is specifically required for dilution profits.  But in many circuits you still have to show
willfulness. Combination means that you can win your TM case and be entitled
neither to an injunction nor to damages or profits b/c defendant wasn’t willful
infringer.  We could have a conversation
about whether that’s ok, but it strikes him as problematic.  TM ought to have a different spin on public
interest: consumer confusion is bad. We should count that.

Injunctions ought
not be automatic, but more common than in the wake of eBay; ought to be tightened up in favor of the TM owner.

Can you use eBay as policy lever to push back
against TM overreaching—Sandy Rierson & Peter Karol have argued for doing
that.  Prove TM infringement that
actually injures you—back door into materiality; way of distinguishing between
classic claims for confusion about source/product and confusion as to
affiliation or sponsorship.

Rebecca Tushnet

The harm stories
accepted by courts to expand TM are at long last being challenged at the remedy
stage, because showing irreparable harm requires plaintiffs to explain why
their harm is special and not sufficiently measurable to be compensable in
money damages.  This isn’t ideal, in part
because of its patchy nationwide implementation, but it’s still a positive
development and may encourage courts to have greater skepticism about expansive
harm stories in the first place.  Unlike
Lemley, I think eBay makes sense applied to TM. 
Where he sees a wrong in the failure to balance harm to the public
against the irreparability of the harm, I see a threshold that challenges the
court to ask whether the public is really being harmed in the first place.  The cases where eBay poses a serious barrier
to injunctive relief are almost never classic direct substitution cases.

Irreparable/not
compensable: easy case of distinction, where D lacks money to pay otherwise
reparable harm. Note too that if the D continues in a case declining to grant
an injunction, it likely becomes a willful infringer. I wouldn’t counsel
someone to continue under those circumstances.

Let’s assume a court
does find likely consumer confusion, in a case where a symbol is unprotectable
under TM law as such—unregistrable under §32. 
Should the court grant limited injunctive relief?  I have to admit, I’m a skeptic.  Such cases are usually at the core of classic
unfair competition—that is, they involve the same goods or services, even
though in the case of territoriality issues there may not be direct competition
or a real possibility of substitution in the ordinary sense.

Why am I a
skeptic?  A consumer rationale and a
producer rationale.  Consumers: we know
that disclaimers rarely work unless they are very carefully done.  Consumer has to be motivated to perceive the
difference.  Maybe this is an area where
disclaimers could work, but I would want empirical evidence that this is
so.  There’s one study that’s not very
promising, involving Bayer versus Bayer AG, the German entity that is
separately owned; the disclaimer there just didn’t work, and in fact was
indistinguishable from a so-called “claimer,” a statement that the two entities
were in fact related.  Essentially, as
many critics of disclaimers say, we’d be imposing the disclaimer requirement to
make ourselves feel better about the fact that consumers are going to be
confused; we can then put the blame on them for not being perceptive
enough.  Only if you view unfair
competition law as also being about commercial morality, which independently
justifies certain constraints, does the disclaimer requirement have an obvious
warrant.

Which leads me to
problem #2, the producer side, and that’s of course the strike suit.  I am skeptical that the game is worth the
candle if the only acceptable output is a disclaimer requirement.  I am skeptical that threat letters will admit
that the only thing a plaintiff in such cases can legitimately seek is a
disclaimer. Of course, in many cases whether there is no TM protection will
itself be hotly contested, so the additional encouragement to overreach
provided by an unfair competition rule is not as big of a problem as it might
otherwise be.  Further: how often do
producers make generic products, or functional products, with no other branding
but their shape or name?  The makers of
competing thermoses used their own branding, for example. 

Territoriality is
probably the situation where the risks of abuse by potential defendants are
highest—Belmora does not have attractive facts in terms of Belmora’s initial
conduct—but I think false advertising rather than more amorphous unfair
competition might well be able to address that. 
I do think that space probably should be opened up between TM and unfair
competition, but I am unconvinced that this space should be created by having
unfair competition get bigger rather than by having TM get smaller.  I know I’m kind of a one-trick pony here, but
the better-defined requirements of false advertising, especially materiality,
should at least be taken into account in considering the possible scope of an
unfair competition remedy.  This would
also allow us to think about how reasonable consumers should behave, since the
reasonable consumer standard tends to be more demanding in the false
advertising space.  Can’t always have
your cake and eat it too.

Lemley: Willfulness
when you continue conduct after determined to be illegal—patent courts have
said you’re not a willful infringer even though you know you’re infringing a
valid patent; the rationale is that we’ve baked the question of whether we
should punish you into the standard for granting relief. Maybe that’s different
b/t patent and TM because patent involves treble damages if willfulness is
shown, or maybe 2^nd Circuit “you don’t need willfulness” standard
becomes dominant.

RT: I do think that
there would be a patent/TM difference, but that’s fair.

Discussant: Bill McGeveran:
reenergizing the understanding of TM adjudication as an equitable job is a
benefit of applying eBay. Judges have
been ordered to be more flexible about their job in TM.  Maybe some intermediate remedies are
inadequate, but it’s worth learning more about them.

Willfulness/damages
is happening independently from different sources—courts are not consciously
reasoning that injunctions are harder to get so damages should be easier to
get, but maybe that’s opened ground for them to reconsider what the proper
balance should be.

Lemley: From TM
owner’s perspective, can trade things off, but as we move from injunctions to
damages we lose benefit to public. Not a fair tradeoff to say you can confuse
consumers by paying the P.

McGeveran: there are
both dials—some injunction, and some damages.

Kur: We have only
one ECJ decision about fines for disrespect of injunctions. Even if infringer
is clear that it will never happen again—infringer was victim of fraud—you have
to impose fines; no leeway for national course. 
Would also apply to injunction as such. 
We see an element of compulsory licensing here.  There is a difference b/t patents, designs,
and TM—no compulsory licenses in TM are allowed.  There is the public interest. 

Disclaimers: we need
research on how well that works.  There
must be cases where companies have the same name but are separate.  Budweiser: they don’t want to coexist but we
could think about making that easier.  It
shouldn’t just be continue infringing, but it should be some requirement or
there’s a risk of conflict w/international law.

Dogan: how does this
relate to yesterday’s conversation: is there something unique about product
design that might make it more suited to alternative remedies.

RT: compulsory
licensing: but that gets back to whether an interest has truly been
invaded.  TM harm stories (very different
in Europe) just aren’t that convincing.

European approach if
there is no mark, as w/genericity: can there be prohibitory injunction?

Kur: w/unfair
competition, there’s no common answer across EU. In Germany, it would be case
by case remedy; hopes it wouldn’t be pure prohibition on use of
generic/functional feature.  Wouldn’t run
up against bar on compulsory licenses b/c it is outside of TM.

Dinwoodie: normal
remedy for slavish copying is pure prohibition.

Kur: but might not
be the case if the feature were truly functional.

Lemley: similar
issue in patent/©: what if injunction reaches noninfringing conduct?  Fits into balance of hardships factor. If
injunction shuts down use of whole product including noninfringing pieces, that’s
a reason not to grant.

Dinwoodie: In
Denmark, Lego lacks a TM in the brick. Can they get relief against slavish
imitation?

Kur: speculative;
doesn’t think it has come up.

Dinwoodie: Slavish
imitation is about commercial ethics, not consumer understanding. The case law
under 43(a) in the US is the smidgen of consumer understanding—not enough to
bar us from saying it’s generic or functional. 
The relief under 43(a) is a lite version of consumer protection, but Continental
cases may be much more driven  by
commercial ethics. When we say unfair competition, these are two different
variants.

Grynberg:
Intermediate remedies may just not work. 
Not just a problem for generic/functional.  Also true for non-source confusion,
affiliation claims. Consider possible dynamic effect: if we just stopped
protecting trade dress, consumers might change; we should consider the possibility
of standardizing a disclaimer remedy, which might not work in the near term.

Peter Karol: Though
every circuit has applied eBay, there’s
a huge variance in what that means to them, amounting to a rejection: sometimes
they cite eBay but allow courts to
presume irreparable harm (11^th Cir.), and say that the ordinary
remedy is injunction (district courts). 
Normatively: a lot of whether eBay
should apply relates to what you think it was doing. Most courts do think they
have flexibility, but rubber hits the road if you think that presumptions are
suspect.  Good reasons to distinguish
patent and TM for presumption. 
Presumption of injunction for infringement could be overcome by laches,
delay, balance of harms—that was consistent w/equity, which didn’t require
chucking presumptions.

W/r/t difficulty of
getting damages, that developed b/c of Champion
Spark Plug & similar cases which said “injunction is ordinary, so
damages should be rare.”  We’re now so
distrustful of what TM is doing (or uncertain) that you might not get anything—that’s
a bad situation.

Bone: The equitable
factors are no mystery. The first two factors are threshold, and the balance of
harms is a balance.  I for one am glad to
see it introduced in more traditional way into IP.  Maybe should make eBay context-specific. For product design, I’m with Tushnet.  No presumptions, requires proof of
irreparable injury. False negative error costs are lower—denying an injunction
isn’t that bad b/c there are other marks to distinguish the item.  Willfulness: depends on reasons why you’re
denying injunction. If denied b/c no irreparable injury/no adequate remedy at
law, then continuing to do it could be characterized as willful. 

The problem
w/consumer harm is that you can’t pay them directly. But indirectly, you might
be able to affect that by requiring D to pay P in a way that might induce
change.

Dogan: especially
for affiliation confusion, if we believe that consumer harms are illusory or
minimal.  Materiality through the back
door.

Dinwoodie: In Europe
we don’t have a common view on unfair competition.  It is optional to apply the enforcement
directive to unfair competition. British debate at the moment: patent case on
injunctions indicated it was willing to go more eBay-like in private law cases, and patent judges have indicated
their attraction to that.  In continental
Europe, injunction comes automatically from violating property-law right; there
may not be harmonization, but proportionality and other rules in the enforcement
directive have enough give to create a range of remedial options.  Hollister case, TM context, has been an
effort to use proportionality to temper relief in TM cases, particularly
parallel import cases—repackaging where infringement was only due to technical
failure to give notice to a party.

Bently: Unused
registered mark—clear sense that it would be inappropriate to give injunction
where it was now being used by a large business. We will probably be flexible
unless ECJ tells us we can’t.  Enforcement
directive makes it a bit messy b/c says injunction shall be granted except in
special circumstances.

McKenna: There’s an
enormous amount of additional work to figure out the role of empiricism. I’m
sympathetic to the need to focus on whether harms are real. Law has decided not
to do that at wholesale level: whether sponsorship confusion is harmful, for example.
Retail level is second-best. In some ways, I’m with Tushnet to say that if we’re
raising the Q at the remedial stage, b/c we have doubts about fact of empirical
harm, it’s weird to impose remedies w/o empirical basis. Why go halfway on
empiricism?  OTOH, they’re operating at
different levels. We’re accepting a generic harm story.  Disclaimers might not work, but consumers do
adapt, so there may be other things courts can do.  But we have to identify the problem we are
trying to solve when we pick a remedy. A disclaimer doesn’t work if I’m trying
to disabuse consumers of certain notions; changes in packaging might work for
other purposes [e.g., Belmora maybe?]. 
We’re doing this w/presumptions built in a world of TM law we don’t live
in any more, from when confusion really meant passing off, now applied to
enormous range. But b/c we didn’t have a real conversation about harms of other
kinds of confusion, we have to do it in nonideal circumstances.  [“[Lawyers] make their own [law], but they do
not make it as they please; they do not make it under self-selected
circumstances, but under circumstances existing already, given and transmitted
from the past.”]

McGeveran: dynamism
is an important point.  Certain types of
disclaimers—consumers could learn.  [But
what is your reason for believing that?]

Presumptions are not
all created equal: rebuttable, bursting bubble, etc.  We have choices.

Remedial modesty has
different implications for different kinds of cases.  [This is maybe one more reason to identify
classes of cases, the way Pam Samuelson has for copyright, which affects both
the confusion tests/defenses but also the remedies.]  Public values exist on other side—first sale,
comparative advertising, consumer information interests, etc.  Peripheral confusion (endorsement etc.) +
clear values on the other side = good case for no injunction, as a matter of
standard civ pro thinking.

Lemley: why isn’t
the right response then “you’re not liable”?

McGeveran: that’s
the first best solution.  Second best
would be limiting doctrines. Third is remedies, but that’s something.  Tabari: lower court enjoined use of “Lexus”
in any circumstances, though Kozinski accepts some residual likelihood of
confusion on affiliation, and makes rules about how to tailor nominative fair
use injunctions.  Can get it out of
domain name in some circumstances, but not full ban.

McKenna: if you
flunk nominative fair use, the injunction can only order you back within its
bounds. 

RT: Bone’s
suggestion for eBay being stringently
applied to product design: seems plausible, but what about descriptive terms, where
there are also error costs issues—are we working back around to technical TMs?
Might be good w/that.  Grynberg: dynamic
consumer response is great, but the problem is that response to disclaimers decays over time, so the suggested
remedy has a mismatch with the ideal result. 
If the market responds, it probably won’t be b/c of the disclaimer but
b/c of independent consumer learning.  Package
redesign is not ordinary disclaimers so there is a lot to be learned about how
that might work.  Graphic design elements
might turn out to be really important.

Tabari is a really interesting example of empiricism: nominative fair use is
very explicitly about deciding that things are nonconfusing as a matter of law,
which is to say that it’s not empirical.

Kur: It would be
useful to have access to covenants that firms sign to prevent too much
collision in the market.  Our knowledge
about how one could do that as a remedy would be improved by such learning.

Heymann: consumer
research about how consumers are trained to look for particular
features/interact more with the product. 
That could help w/specificity and tailoring.  Though standardization also produces gains on
the consumer side (e.g., nutrition labels). 
Relatedly: what the goal is for these kinds of remedies.  We have a sense of what damages are supposed
to do.  But what do we want injunctions
to do? Where is the consumer in all this?

Bently: yesterday we
were expressing doubt that product design is relied on, but now we’re
suggesting it as a remedy?

RT: I understood
myself to suggest that color on packaging might be helpful, but I wasn’t
talking about the design of the product itself [though others think that might
also be worth investigating].

McKenna:
second-best; we might think product design is rarely relied on, but at the
remedy stage the court has rejected that. 
Also, even if there’s some source designation, it’s also doing other
things, and we might want those other things available while trying to do something
about the source designation that might not be a full fix but mitigates the
harm.

TM as info
transmission system; yet admitting that there are lots of things that have nothing
to do w/info and more to do w/equitable position of parties.

Bone: Effectiveness
of disclaimer has a lot to do about whether consumers care; if disclaimer doesn’t
work, it might be b/c consumers don’t care, and aren’t harmed.  Empirical work on disclaimers: our results
would have very high variance. Some rely and some don’t. Depends on
context.  Sometimes they reduce confusion
somewhat; leaves normative question of how much reduction is worth the
candle/we will require to accept disclaimer.

Heymann: Product
liability, product warnings—the law doesn’t care much about whether they
actually work.

McGeveran: this isn’t
the only area where there’s existential angst over subjects’ reaction. Privacy
policies that no one reads.  He’s happy
with rules that make clear that the most you can require a Lexus dealer to say
is “we’re not affiliated w/Lexus.”

Yen: Costs imposed
on consumers should be considered. Injunctive remedies on D impose search costs
on consumers; the ones who learned to distinguish the brands and are happy—if you
take one brand away and make it be different from existing mark, consumers
incur search costs to find it (or to avoid it!).  If we’re interested in disclaimers, the
analogy to tort has to be careful. 
Warnings in tort are used in the context of fault.  Even though products liability is strict, the
Q is whether it’s reasonable to sell product w/ or w/o warning. If you sell
product w/reasonable warning and consumer gets hurt anyway that’s not your
problem. But TM is supposedly  not fault
based—if consumers are confused, you’re liable.

Burrell: enthusiasm
for highly fact specific injunctions was a little surprising. If you can’t predict
what injunction will look like, you’re less likely to reach settlement, which
is to be welcomed.  Similar issues exist
w/nuisance, where injunctions are delicately shaped.

Dogan: Think holistically
about ways of avoiding confusion, presale and point of sale; even highly fact
specific cases can start establishing norms allowing commercial practices to
evolve, shrinking the scope of product design trade dress. Wouldn’t require
that level of precision in every case.

McKenna:
historically unfair competition had to show intent; for technical TM we
presumed it b/c same mark/same goods. What happened? We lost sight of that.
Presumption of intent in technical TM made sense in same mark/same goods, but
doesn’t as applied to other things. 
Also, when test asks for reasonably prudent consumers, you’re asking a
normative question. So we’re willing to live with unreasonable confusion. So
analogy to tort is actually good.

Thinks that old
unfair competition cases aren’t about: this thing is functional but we’ll give
you some remedy. Rather, the cases are about: this thing is functional, but D
is making the problem of confusion worse when it doesn’t have to. It’s fault
based.  Not necessarily going to fix the
problem. But your entitlement to do X doesn’t mean you have to do Y, Z. Fault
based.

Bone: he sees an
affirmative obligation to take measures to limit confusion.

Dogan: sees
both.  The language of the courts is
affirmative obligation, but the facts show D doing something else besides using
the functional feature.

Yen: doesn’t think
that modern TM law today admits of fault as a doctrinal requirement.  We can read it in, but courts don’t say it.
If we talk about fault, we need to talk about intentional tort and
negligence.  Unreasonably incurring risk
of bad result is something courts haven’t worked out in TM.