Friday, August 12, 2016

IPSC Breakout Session IV

Empirical Copyright
Copyright Misuse: A Taxonomy
Ann Bartow
Occasionally called copyright misuse; non-US countries call it something else, but trying to taxonomize a certain type of misbehavior.  Some of this is not currently called misuse: using © to try to cover facts and ideas—books that value baseball cards, etc.  Judges use a “reverse merger” doctrine: you have copyrightable expression in facts if the facts look creative to the judge.  That’s suspiciously like sweat of the brow, but work + value = claims for expression.  Trying to cover things instead of works: Digital handshake cases, where “expression” is used to control a device.   Quality King: grey market goods.  Chinese are worried about Kirtsaeng: publishers may stop making $20 textbooks.  Gutting of first sale via contract.  Copyright trolling: buying up (or not!) copyrights just to sue.  Pornographers make it an art: special shaming techniques to make people pay—“barely legal,” racist, gay porn.

Lisa Ramsey: what happens if we call these all copyright misuse?

A: more coherence.  © owner loses more often.

Eric Goldman: I didn’t hear antitrust in this list—do you want to expand on the relation b/t © misuse and antitrust?  Or is misuse a broader principle?  My struggle: the whole point of © is to shut down competition—that’s its intrinsic nature. So rules of engagement say “this kind of shutting down competition is ok,” and then other types of competition you can’t shut down.  If you can make the good/bad distinction coherent that would be helpful.

A: that’s what I’d like, w/r/t things like “reverse merger.”  If judges could fit patterns into a larger scheme it might help deal w/pushing the boundaries of copyright. 

Sharon Sandeen: consider earlier cases. A lot of the problem w/abusive litigation is C&D letters, choosing to sue in improper venue; Ds may capitulate before it’s litigated. Additional category of misuse?

Ramsey: different remedy, like threats action?

Ariel Katz: Posner says misuse is just antitrust and thus unneeded, then changes his mind later about © misuse—antitrust articulates a principle that should be followed; he sees misuse as more abuse of process.  In many cases you can say that you just don’t have copyright over the matter claimed.

Ramsey: interacts w/remedy: suppose P is trying to cover facts; is entire copyright invalid?

Q: is this normative or descriptive? Licensing away fair use is permitted by case law. 

A: My intent is to describe and then make normative claims at the end.  [Perhaps if you fit together the different branches it will be clearer whether particular versions stand out.  You can’t contract away right to use facts, apparently; but you can contract away fair use—why?]

Q: Video games: using © licenses to prevent use of bots in games.  Is that ok?  Also, consider Brownmark v. South Park—letting fair use be considered on motion to dismiss; first published case to use “copyright troll.”

IP Theory
More Property-Like than Property: The Prevalence of Property Rules in IP Remedial Schemes
B.J. Ard

Property rules allow holdouts/irrationality; liability rules have pricing difficulties.  Dispute over these rules is often linked to whether IP should be thought of as property or not property.  Property = intentional trespass. But property today defies any easy simplification.  We have a number of liability rules for unwitting trespass, and for various holdout problems—most jurisdictions have negligence standards for unintentional trespass. When it’s permanent intrusion and nonnegligent, most courts will balance the equities and sometimes force a sale. Nuisance is another classic case.  We can imagine a strong property regime for nuisance, as in early English common law.  But a rule like that would have made industrialization practically impossible; the current American approach often awards damages.  And finally, direct state takings uses a liability rule.

Contrast w/IP: © has mandatory statutory damages [for timely registered works].  [See also: TM counterfeiting.]  Punitive damages bear little relation to harm caused; there’s often no solicitude for the inadvertent infringement.  Fair use can cut through a lot of potential transaction costs, but like the rest of ©, fair use doesn’t care about willfulness/infringement.  Its balancing test is about harm to owner and broader social benefit, but not benefits to the would-be fair user.  Damages may be out of sync w/harms.  Statutory licenses: set prices for certain copying/transmission.  Public choice theory shows limits on the effectiveness of this practice.  Won’t address problems faced primarily by consumers, or for startups.

Patent: looks possibly like liability rule, given eBay.  Reasonable royalty is the standard. However, courts still enjoin in the vast majority of cases in which there is actual competition.  Fed. Cir. rules have made royalties supercompensatory through various rules.

Seems backwards for three reasons. (1) Greater notice failures exist in IP.  Land records are better than for IP.  (2) Nonrivalrousness of intangible goods.  (3) Cumulative/overlapping nature of intellectual production. Smith argues that modularity is an advantage of property rules: nonowners know to keep out.  IP doesn’t fit as easily in the same model; what I do in producing my own film or machine may come from bundling together a lot of rights.  In tangible property, we often see liability rules deployed to facilitate this type of bundling.

Why not a negligence standard for patent? Inventor who diligently searches is no better off than one who does no search at all, and might be worse off if open to finding of willfulness. Negligence = patentees have clearer incentive to provide notice.  Short of that, we could at least have harder caps on non-negligent infringers’ damages.

Don’t squeeze IP into Procrustean bed of real property.

Q: literature on inadvertent infringement in patent—Monsanto—coming to the nuisance type.  Don’t buy the assertion that real property uses liability rules “more” than IP.  More than we expect?  Is this quantifiable?  What are your limiting principles for reform?  Maybe everything should be liability, but that also seems extreme.  Where do you draw the line?

A: mostly interested in notice failures/unwitting infringers. That would be a major improvement.  Another key principle: in service of designing patent and © to promote the forms of innovation we want to promote—another empirical Q. Are we deterring by punishing innocent infringers? Are we undermining incentives to innovate?  We’d still have a workable/productive system, but w/fewer problems.

Q: re: search.  If I know that there’s something out there, I’m worse off if I don’t look than if I do.

Q: political economy perspective: real property analogy is used to claim that cutbacks on IP are “takings.” You may want to point out that’s not what you’re trying to do.

A: not trying to take a position on whether IP “is” property—pointing out that “property” doesn’t work the way that certain people claim it does when making arguments about IP.

RT: timely © registration: w/o it, no statutory damages and you’re in liability-land subject to eBay; note move to add them in © Office’s small claims proposal.  Consider TM, at the property/tort interface and struggling w/remedies.

Indiscrete Property
Michael Burstein

Once you define a res, the question is how to manage it. Smith etc. argue that recognizing thingness of such assets allows benefits.  Info is often not subject to thingness in a way that goes beyond the costs of delineating the res. Info is often indiscrete and continuous; the logic of mixing ownership and governance strategies for it then becomes incoherent.

Commercialization may require coordination; if coordination costs are minimized, that can make commercialization easier and have social benefits. Thus the need for exclusion depends on the cost of delineating the thing and signalling its existence as a thing. If you can easily say “keep out” it might be easier to have an exclusion strategy. If you can more easily identify a use, then governance strategies may be more effective/efficient.

Smith is quite subtle about governance/exclusion strategies in IP, but others have taken position that if exclusion is a relatively low cost way to coordinate downstream use, then we should try to push more subject matter into exclusion strategies.

Our view is circumscribed by focus on delineation costs of defining/identifying boundaries.  This isn’t accurate depiction of info.  Info science: “data, information, knowledge, wisdom” hierarchy, usually depicted as a pyramid.   These correlate w/meaning and value—data is less valuable/meaningful than info, which is less than knowledge.  [Knowledge is knowing that Frankenstein isn’t the monster.  Wisdom is knowing that Frankenstein is the monster.]  At each stage there are transformations to move from one level to the next. This can turn philosophical: “information is a verb, not a noun”—info is something you need to do something with.  Economic literature: knowledge is something that can be codified and exchanged, which requires transformation, e.g., codifying tacit knowledge.  Economics talks about costs of codifying tacit knowledge.  Strandburg: Self-revealing v. non-self-revealing info; von Hippel: sticky v. nonsticky info. Some info is easier to transfer than other info.  Design theory: new private law has drawn on this for its concept of modularity, but in design theory, modules can be designed. Choice can be made consciously about what’s in and what’s out. Persons can design info flows/exchange info selectively.  Heterogeneity of info: how we develop info about info for purpose of exchanging/transforming it. Pharma: core info is structure of molecule, but you can develop info about the molecule that doesn’t reveal its structure but reveals enough to facilitate exchange.

Discrete info, in his view, is separate or distinct. Continuous info is inseparable.  Not all info is indiscrete. Especially lower on the hierarchy. But indiscrete info is different from real property. Bargained for exchange in real property is possible when different people put different values on private goods. But indiscrete info communicates value to different people in different ways.

When we propertize, the choice b/t exclusion and governance is more complicated than many think. Exclusion strategies can be underinclusive of social goal of promoting innovation; may also be overinclusive by preventing communication of valuable information.

What result?  Explain intuitions about content of IP—patentable subject matter.  Focus on rules and institutions that enable people to structure info flows as they choose.

Ted Sichelman: isn’t this also true of real property? A boundary around a piece of land: an entrant may not interfere w/uses of owner; uses outside boundary may affect uses of owner. So we just have to figure out whether over and underinclusiveness are worth the benefits we get from creating the boundaries.  [Information environmentalism redux!]  Most areas of IP don’t protect info directly, but uses w/r/t that info.  Making, using, selling, offering to sell is what patent covers, not “a molecule” as such.

A: under and overinclusivity operate differently, b/c it’s much more difficult to anticipate relevant uses of asset, b/c intellectual assets convey value continuously as opposed to land/bottle of water.  There are multiple uses of land/water, but the way in which uses are communicated to people depends on characteristics of the thing that are much easier to communicate/consistent than w/information.

Empirical Copyright
Pretty Please: Software Piracy Rates and Charismatic Appeal
Andrew Moshirnia

In combating illegal videogame downloads, is it more effective to play nice or dirty? Not much research on vg piracy, though it’s an enormous economic drain.  VG market is about $100 billion, about 6x recording revenue.  VG piracy is an ongoing resource drain not a single lost sale, if a game requires updates/skins—pirated game will continue to consume bandwidth; go on to help chats to get the game to work; people w/cracked versions can often cheat in online games, ruining it for everyone.  Major vector for malware, unlike music/movies which aren’t executable files. Constant fear of crackdown in modding community, which overlaps w/pirate community.  Artistic concern: move to freemium, server-side games—constraints driven by something other than the marketplace.  [Interesting definition of marketplace, as if it existed w/o law.]

The scene: private newsgroups/torrents; picked up and repackaged, often w/malware, into public torrents.  Justifications: cost; quality/sampling; DRM backlash; anti-corporate ideology.

Countermeasures: DRM: endogenous DRM, where game detects it’s pirated and messes w/player rather than locking them out.  Serious Sam: if you’re playing a pirated version, a giant pink scorpion starts shooting you about a minute in.  Open pricing, which partially addresses cost concerns.  Humble Bundle: dedicates chosen percentage to charity.  Charismatic appeal: forswearing DRM, making indie/personal appeal by developers.  They distinguish between themselves and EA (big bad). 

His belief: DRM will encourage piracy where DRM breaks the game, as when Sim City’s authentication servers were down for a month.  As long as DRM can last 21 days, it works.  You make your sales then; you just want to prevent zero-day piracy.  Open pricing may also work to limit privacy.  Emotional avenues are likely to be ineffective, but might have interesting effects. Game quality will encourage both piracy and legit sales. Attitude of publisher may weakly encourage piracy.  Appeals won’t have impact but may have interaction w/open pricing.

Study design: data gathering now. Examine downloaders/seeders of cracked games, identify factors influencing piracy of individual titles, evaluate factors against claimed philosophy and behavioral model. Looking at data from 2008 on; that was the explosion of torrenting w/Spore.  Looking at torrents, publishers; conduct regression and modeling to get a more tailored approach and avoid a draconian response.

Torrent trackers: can look at number of downloads, but that may be inflated b/c of multiple torrents; total number of seeders/leechers; rank in downloads; torrenting of software is very top heavy—top 10 will account for great majority of downloads at any given time.

Game data: DRM used, opening price, open price ever offered, critic score/was it a sequel, publisher’s market share/employees/attitudinal survey, charismatic appeal, number of legit copies sold.

Analysis will depend on data quality.  Difficulties: torrents aren’t only source; false torrents; multiple downloads; poor records b/c it’s an occulted activity. If I can only get ordinal rank, no parametric testing possible.

Matt Sag: is your plan to look at only games that have been downloaded, or broader population of games?  Don’t select on the dependent variable.

A: ID number of games and go through population.

Kerry Abrams: affordable alternatives as reducing piracy in video, music—what about VGs?

A: fewer subscription services for new games in this realm. There is the idea that we’re not concerned w/non-new games. Subscription services tend to be not new games. But in terms of ease of access, there are developments in digital distribution, mostly through Steam.  Online only access is also a move, but that will limit the types of games that can be made and will limit modding.

Katz: do you distinguish b/t new games v. old/noncommercially available? What about merchandising?

A: Most of the time, things out of the top 25 don’t have data collected. Civ III fourteen years after the fact is not going to be big; if it does show up I can control by year of release.  There are definitely economic advantages to tangible goods: Master Chief T-shirt can’t be downloaded.  I’m not trying to calculate lost sales and even pirates might buy the shirt.  I’m more interested in change away from moddability and what might be done about that.

Undetected Conflict of Laws Problems in Cross-Border Online Copyright Infringement Cases
Marketa Trimble

WIPO recommended training and soft law improvements to deal with crossborder cases; Trimble thinks more is required. WIPO report used only two US cases, Zippo (largely overruled in most circuits) and magistrate judge decision in Nevada, in its study of 56 cases.  Trimble sampled infringement cases filed in 2013, 364 cases, under 10% of those filed that year.  WIPO report is underinclusive, but also overinclusive in looking for cases involving conflict of laws.  Trimble’s sample is only copyright cases, which has different inclusivity problems.  WIPO looked globally; US is very specific/different.

Over 80% of cases were online infringement cases. 90% of those involved online digital copies; 63% of online cases involved bittorrent; 74.7% of online cases were filed against John Doe defendants.  But in the entire sample, there was only one case involving some conflict of law issue. Few defendants are foreign domiciled.  But you won’t see the conflict problem in many filed cases b/c the problems are so big for litigants that they don’t even file cases against foreign defendants b/c they know how difficult and costly it would be.


Need more coordination of rules, improvements in judicial cooperation, and streamlining of judicial proceedings in cross border cases. Maybe small claims proposal could offer a way forward.

No comments: