Friday, February 16, 2018

WIPIP Concurrent 2

Nancy Kim, The License v. Sale Puzzle after Impressions v. Lexmark
Sales exhaust the patentee’s rights to any item regardless of what the contract says. But restrictions on licensees are different because a license doesn’t implicate the same concerns about restraints on alienation as a sale.  Even if the contract is enforceable, that doesn’t mean that the parties can’t contract about everything. They can’t recharacterize the world—you can call a contract w/a worker an independent contract, but if the facts establish an employment relationship you have an employee.  Same with license/sale.  “Authorized” sale is also important.

Contract can give rise for breach and infringement claims, but not for the same provision; not every breach exceeds the scope of the license and constitutes infringement. Principles: if the first transaction by the patentee is a sale, all subsequent transactions are sales. License agreement may or many not be followed by sale. If license allows licensee to sell patent items w/o restrictions, then the subsequent transactions to customers are sale.  If license imposes restrictive conditions on licensees’ ability to sell, they’re sales only if the licensee required with those conditions—e.g., limited to sales in North America, and then sales in France are unauthorized.  If licensee has obligation to impose restrictions on customers, then subsequent transactions are sales.  If mfgr didn’t have customer sign contract restricting use, mfgr might be in breach of contract, but it’s still a sale. 

With a standard form contract/adhesion contract if the patentee/© owner retains rights to reposess or control item, it should be considered a license, but otherwise a sale.  If negotiated or individual contract, distinction should depend on whether contract limits one of the patentee/© owner’s exclusive rights (that is, does it affect the scope of the license?).

Whether the purchaser/licensee infringed depends on whether the contractual provision was a covenant or condition, which means did it limit the scope of the license/implicate one of the exclusive rights of patentee/© owner, but the ©/patent cases differ.  Sale of © item may be treated as still having a license.  Example: a nonexclusive, nontransferable license to make 2 copies. This is not conditional language; transferability is not one of licensor’s exclusive rights as a © owner.  Also: Licensee agrees not to transfer, assign or sell—that’s still drafted as a covenant not a condition.  “provided that licensee may not transfer, assign, or sell … license is expressly conditioned upon compliance w/this provision”: if result would be forfeiture, courts in contract law generally are disinclined to find a condition.  How about: if licensee attempts to transfer, assign or sell this software, the license immediately terminates and is void: clearly a condition, and that’s what it should take.

Lemley: How confident are we that 9th Circuit’s condition/covenant distinction survives Lexmark?  Lexmark suggests © and patent are kin. 

A: she’s concerned about that.  Ct doesn’t directly address license v. sale b/c they say it’s a sale. Also different exclusive rights: with ©, you can sell the product and still not exhaust all the rights (e.g. public performance).

David Olson, Why Increasing IP Exhaustion Leads to Decreasing Ownership of Goods
Information costs for servitudes on chattel property are high; recording burden is high; may restrict resale/reuse of goods forcing inefficient manufacture of new goods instead.  With information goods/internet age cost of checking restrictions can be very low relative to value of item.  IOT goods that phone home/alert you as to any restrictions.  Cost of knowing them is just lower [depending on what you mean by “knowledge” given human practices & functional capacities].  Digital copy of book can also be created more cheaply; not much waste in idling/deleting my copy.  Allows for price discrimination, as does monitoring via IOT/collection of $ for downstream uses.

Still, most goods are sold free/unconstrained.  It’s often more profitable to simply make a first sale w/o monitoring or transaction costs.  E.g., cars.  Other times it can help buyers who can’t afford the total cost to give them only a partial right & have owner retain residual value—rental cars, tuxedos, etc. 

For the IOT, your refrigerator can be shut down if you violate a term.  HP is also licensing/renting printers on this basis—rent charged based on amount of use, and they send you the toner as you need it.  The relationship no longer looks like a sale; it is ongoing and involves patches, etc.  If your credit card expires, the printer stops working.  Pretty effective lockup through contract, and enforcement w/o any privity of contract necessary.  [Pretty good opportunities for fraud?]

Is this good/bad and how if at all should the law respond?  B/c price discrimination can be quite valuable to society, especially for cheap goods, the next move to preserve ownership would have to be inroads into contract law to stop these characterizations/prevent these restrictions being enforced via self-help.
Reasons to be cautious about making that move: now that exhaustion is so strong, we’ll see the same attempts through contract law; we’d be better off to allow experimentation and see how the market reacts. We’ve never seen mfgr limit all use/control all aspects; it’s not profitable and there’s bad PR from people not liking price discrim. Instead we’ll see monitoring and price discrim w/customization for users’ needs.  Consumer will not be able to see she’s being charged differently b/c the good will not be exactly the same for each one.  Like cars with different packages [this is a really disturbing example—we know that the poor/women/minorities pay more].  Can charge people willing to pay more, more, and charge low prices to those who can’t afford more b/c arbitrage will be impossible.  [When has this ever happened, compared to poor paying more?]

Better to use antitrust law than sale/license law to deal w/abuses such as market foreclosure.

RT: questions above.  Tell me why you’re hopeful.

A: discrimination law would still apply, but he was talking about standard packages; discrimination comes in at the overall price of the car.  People could still create data [unless contractually prohibited]; they might find that women are willing to pay more, which could be a problem.  Women sometimes are never offered a low enough price and they walk away; the transaction fails.  AI, which has the potential to lessen discrimination [do we know that?], means that if your goal is profit maximization you will try to figure out what people will pay and get past proxies like age and race.  [That does not seem to me to describe how the current algorithms work—they absolutely use your categories to be predictive about your characteristics, including WTP. Plus your other non-race and gender characteristics are tightly correlated w/them, so “non proxies” (e.g., where you live, whether you’ve gone to jail, how much money you make) are also proxies.  And then your reaction to the offers you get—which people in other groups won’t get—determines what happens next, so the discrimination just gets baked further in.]

Rosenblatt: rich could be allowed to make remix, poor not.  If what people want to do is on-label, then your logic makes sense. But if they want to do off-label stuff like remix, reverse engineering, follow-on innovation, doesn’t this bar that?

A: if we restrict producers to one price, the Q is do poorer consumers get to buy anything, or nothing?

Michael Burstein, Indiscrete Property
Information is often incompatible with the notion of thingness in ways that go beyond the mere costs of defining the thing: it is indiscrete and continuous, as opposed to discrete and separable as most tangible goods are. While a stock tip is relatively discrete, more often info conveys different value to different actors; info is more of a platform on which transactions occur.  Thus value depends less on definition than the structure of the underlying institution.  Should focus on rules that allow people to structure info flows for best use.

Case for info as discrete property, from two parallel movements suggesting strong IP to economize on use of info.  First, incentive/reward theory is unsatisfying; market failure resulting in underproduction doesn’t dictate IP as solution and other solutions could produce incentive w/less cost.  As empirical matter, it’s well understood that in many situations the IP incentive, or any incentive, may not be necessary for production of intellectual goods.  Next move: commercialization theory, where IP is about the subsequent use of creation/invention by giving incentive to solve market failure structurally similar to the earlier market failure theory, or by serving as beacon/signal/contracting device for people to come together and exchange info. New private law is aimed at reclaiming centrality of tort/contract/property in US law. Scholars offer info cost theory about economizing the cost of info specifying use rights.  Thingness of property matters b/c by defining a thing and marking it w/rough exclusion signals we create relatively clear signals about what you can and can’t do with it. Use rights in a car; you need no info about underlying owner of car to know not to steal it.  You just need to know it’s not your car.

Nonrivalrous info: inputs to info use are rivalrous, so we want to manage them. New public law theorists like Henry Smith argue here’s a benefit from reducing complexity of coordination, if the thing can be defined easily and the uses cannot.  In such cases, exclusion strategy is more economically justified than governance. Smith argues that patent is more thing-y than ©. This is not uncontroversial or unproblematic. Clarisa Long has made the exact opposite argument.

Problem: relies on thin concept of info such that all you need to worry about in defining a good is the cost of delineating the good. But other characteristics make it almost incoherent to talk about definition. Info is additive, communicative, and layered.  Additive: info bits build on one another so value is different depending on type or info.  “Data information knowledge wisdom” framework: data is info asset that’s product of observation, relatively discrete and objective; info is relevant or usable or significant—info uses data to answer a Q; knowledge is usually knowhow, turning info into instructions/decisionmaking aids, synthesizes across types of info.  Literature usually ignores wisdom and so will I.  As you move up, increasing correlation w/social meaning and value.  Info has value, but it exists in different forms w/different sources of meaning and value.

Communicative: JP Barlow: information is a verb, not a noun.  To have an effect in the world it must be transmitted & create meaning for recipient.  Does this not through ethereal transmission, but is contained in artifacts—physical manifestations that are vessels for the info. Some artifacts are more self-revealing of the info than others.  Sometimes info can be sticky/tacit (Eric von Hippel) and needs more than writing down. Info can be modulated self-consciously, as w/software modularity so you can hide certain elements of code & reveal others.

Layered: Heterogeneity—small molecule pharma—most direct way to communicate info is to show its structure, but you could also talk about the molecule/generate info about that.  You could talk about chemical characteristics, biological effects, efficacy & toxicity, all w/o revealing its structure.

Info exists in continuous indiscrete state in many cases.

In land, things are rough signals of value. But if info is continuous then thingness may send wrong signals.  Lose out on ability to communicate across users.  May be over and underinclusive.  Answer to Smith’s question about how you specify rights & duties up front may be “you don’t have to.”  Spectrum of discreteness may also help rationalize our intuitions, e.g. about patentable subject matter. If abstract idea is relatively indiscrete, and open platform approach is most efficient, that may help explain why IP doesn’t protect them. Likewise for doctrines about timing of patentability such as utility/reduction to practice—when the info becomes more discrete.

Rebecca Kai-Fang Lin, Artist’s Resale Right: Past, Present and Future
What makes art so special?  Work of art can be one of a kind; reproduction doesn’t create the same thing.  Low price when first sold.  Resale right was exception to exhaustion introduced in France.  Visual artists usually earn less than average income; this allows recognition and fair treatment and also helps them stay connected to the works.  Also can support cost of storage, conservation and restoration.  No evidence that resale right harms the art market.

Berne Convention introduced resale right in 1971 as reciprocal right based on country of artists.  Rightholder is author/authorized person.  Covers original works of art/original manuscripts of writers and composers.  Nature: inalienable right; applies in any resale.  Collecting societies should be able to collect the royalty and should have a right to gain information.  US, China, France and UK are the main fine art markets now, representing about ½ of global sales.

EU has harmonized: rightholder/authorized person though member states can limit who those can be; liable person is generally the seller but member states can diverge; the covered works are original works of graphic or plastic art; must be physical but copies can be included.  Nonassignable and inalienable right; applies in resales unless b/t private people or acting for nonprofit museum.  Sliding rate by sales price. Its share of the art market grew over the past years. 

France: liable person is acquirer & law officials jointly; applicable sales are those via a public auction or dealer.  Uniform 3% rate, and artists must publish in Official Journal if they want the right.  UK: seller and relevant person jointly & severally liable; copy of work is generally excluded; collecting society and right to information exists; sliding rate based on sale price. 55% of dealers not affected; average processing time of 10 min. 81% of artists use the income to pay for living expensives.

In US, only California—royalties of 5% of sale price, held to violate dormant commerce clause.  Japan lacks resale right.  Potential artist benefit would be big on market of EUR 2.19 billion; 280,000 visual artists could be covered; manga should also be included.  China has no resale right—proposal to cover limited to auction sales and limited to increase in value.  Int’l treaty may be a good idea.

RT: Collecting societies have a bad reputation for not distributing money fairly especially where the uses are varying and the market is thin (i.e. especially outside music).  If you require registration, what’s the point of a collecting society?

A: Tracking use is a big problem with other collecting societies but discrete sales of physical objects are more identifiable, and collecting societies can handle the administrative costs more easily than individual artists.

Q: Why would we want to benefit successful artists and not unsuccessful ones, when resale rights are only valuable to a very few?

A: mitigating inequality.

Chiang: What is the problem you’re trying to solve? It’s not free riding.

A: musicians can make lots of money from sales of copies, but visual artists can’t, especially for works sold early in their lifetimes.  Sellers benefit from the subsequent creations of the artists so artists should also benefit.  Unfair.

No comments: