Mark McKenna: what does it mean to compete in IP? Does Pepsi compete with Coke? Seems obvious that they do (except for with me, since as Mark points out I accept no substitutes) given how they’re sold and their relative pricing. What competition means: They’re sufficiently close in the mkt under some sufficiently meaningful measure. So we need some way to measure closeness (utilitarian—price; Louis Vuitton and Gap handbags are not close on price) and when the closeness is sufficient.
Examples of problems: functionality in TM. Aesthetic functionality and even some mechanical functionality cases require courts to determine whether exclusive use of a feature would put competitors at a significant non-reputation related competitive disadvantage. The more broadly you define the market, the less necessary a feature will be. Example: Dippin’ Dots flash-frozen spherical ice cream, claims against another maker of flash-frozen spherical ice cream. Response: can’t compete in the flash-frozen ice cream business if denied access to shape and general colors. Dippin’ Dots says Frosty Bites is free to make its ice cream like anyone else does. Court rejects that: Frosty Bites want to compete in the flash-frozen ice cream market. This seems intuitively right, but how do you reach the determination that it’s a discrete market?
With patents lost profits cases: may need to know how patented good interacts with the other unpatented components. Has developed some degree of sophistication in identifying alternatives. Reasonable royalties: market definition is important because you have to identify alternatives that are close enough to serve as functional substitutes—the law here is much less developed/coherent than lost profits methodologies. Overall courts have no consistent methodology, which makes cases unpredictable and hard to argue. There probably is a best way to define markets, which we should do. Also, lack of definition allows significant and strategic inconsistency.
Courts/scholars are dismissive of ideas that IP rights create market power. Markets are broad and robust when it comes to the cost of IP rights, but narrow when it comes to awarding damages (courts assume there was no equally attractive substitute to the right you infringed).
Mark Lemley: When you start to apply antitrust tools in IP cases, really strange things start happening. Paradigm of last couple of decades: to define a market, ask what would happen if I had a small but significant nontransitory increase in price. Would people switch or stay? If they wouldn’t switch, then the two things aren’t really competing. However: even if you’re a monopolist, something constrains your price. A monopolist prices up to the point at which people would stop buying goods. Okay, so we need a reference price: could I raise my price 5% over marginal cost. If you start asking that in the context of IP, an astonishing array of IP rights have market power. Maybe Coke and Pepsi have some price constraints; brand loyalty may prevent switching—Rebecca Tushnet is not unique in being willing to pay $0.05 more for Coke.
Clorox sells at a significant margin over generic bleach, chemically identical. We know the answer to the question of whether Clorox could sell at a higher price over marginal cost, we know the answer is yes because it does. Likewise, people will not generally switch from J.K. Rowling to Stephen King in most cases, certainly not within anything close to marginal cost or fixed cost.
Even smaller brands suffer from this phenomenon. 2 bottles of organic green iced tea, Tazo and Honest Tea: 30% price differential, side by side on the shelf. Drugs: Advil has 70% price premium over gov’t-guaranteed identical ibuprofen sitting next to it on the shelf with a “compare to” sign. Under traditional analysis, then, these products do not compete with one another because there’s not sufficient price competition to constrain the price of the branded good.
For decades, we’ve talked about IP rights as not monopolies but really ordinary property rights creating boundaries. True that IP rights create boundaries and prevent substitution of the kind you’d ordinarily expect in a real antitrust case. If you raise your price, usually, other people may come in and compete with you. But people can’t come in and make Tazo iced tea: IP creates a barrier to entry insulating parties from price competition.
A lot of monopolists in the world? Monopolists are subject to significant legal constraints. Exclusive dealing/licensing arrangements turn out to be problematic if the company is a monopolist. Rule of reason analysis for exclusive licensing of publishing to one publisher: antitrust would presumptively condemn that.
Antitrust has dismantled limits on vertical restraints (exclusive territories). You can’t agree with competitors on prices, but if these things aren’t in the same market, then there’s no problem with companies agreeing on what they’ll do.
We have a digital rule (in or out) for an analog rule. Even if we abandoned marginal cost for fixed cost, we’d still find many cases of IP conferring price power significantly in excess of fixed cost.
We used to know there was a significant conflict between antitrust and IP, but then we’ve been saying that they both serve long-term dynamic competition. We should question that for antitrust and look at implications for IP: conferring rights often means conferring market power.
McKenna: The point of an IP right is to force people to use imperfect substitutes or pay your extra price. Some IP rights are better than others at this. One takeaway: this is just a reminder that something we’ve been trained to dismiss is true: IP rights have costs. They aren’t just speech costs or access costs, though those are real; they’re competitive costs. How to deal with that: a bigger question, bound up with how persuaded one is for IP justifications in the first place. If we want better substitutions, for example, we could cut back on the scope of substantial similarity in copyright and even on the derivative works rights—might be better substitutes than the next available substitute even though they are often also complements. Patents: might want to limit scope of claims. TM: pay more attention to mark similarity—get closer to draw on some of that power. Basis to criticize sponsorship/affiliation and dilution, where what’s driving decisions is brand meaning or value. Across all IP, strongest implication: robust first sale doctrine. Second-hand products may be better substitutes than the originals of another product.
Lemley: Broader sweep of history: both antitrust and IP, alone and in combination, have been cyclical throughout history—fear of dangerous monopolies gives us strong antitrust and weak IP, then we get desire for innovation and idea that we don’t need to worry much about monopolies. We’re in a strong IP/weak antitrust point, maybe starting to move in the opposite direction. In the 1930s and 40s it was common to speak of patents and TMs conferring market power. Question: can we smooth out the cycles?
Deven Desai: This is a call to limit IP before it blows up. Build better boundaries between what IP covers and what’s beyond its reach. Market definition reminds us that abstractions and line-drawing play major roles in figuring out what law will do. Open our minds to the unstated forces driving IP to see if that’s really working.
Focusing on markets may leave too much on the table. Proposed solutions: enhance substitutes—do they really do that? Enhancing comparative advertising might do so. But divergence between analysis and solutions suggests the real issue is reducing the scope of IP. How much market power is sufficient? Even if we limit rights, we’ll still have problems defining substitutes.
[apologies that I had to leave for class at this point]