Saturday, September 10, 2016

Stanford Sociology and Psychology of IP

Session 5: Fairness, efficiency, and distributive justice

Discussion leaders: Stephanie Bair: Concepts of ownership/justice in ownership may be universals—studies of children across cultures.  Purpose of IP as understood to be to prevent plagiarism—asking for permission is sufficient, though, in many of these lay understandings. Organizational context studies: Also a distributional component to fairness. Contributors to project: rewards distributed unequally = offended sense of justice.   Creators perceive work environment as fair if there’s respect and trust. A procedural aspect: more likely to think distributional outcomes are fair if the procedures used to determine outcomes are non-arbitrary or otherwise perceived as fair.

Psych literature can link efficiency and fairness. Fromer proposed: moral rights are incentives for creators b/c creators really care about them.  Silbey’s interviews w/creators found much the same. Organizational behavior literature backs this up.  In companies, people choose voluntary creative behaviors more often when they perceive the work environment as more fair; they experience more intrinsic motivation.  Leads to objectively more creative outcomes.  Might be important to pay attention to fairness to continue intrinsic motivation.

We think more innovation is good; if asked why, we might say it contributes to economic growth/social welfare. If you had distributional concerns, you might think we could use tax to deal w/distribution. But there are people who take issue w/the idea that economic growth is necessarily a good thing or the best way to measure welfare. Maybe we should be using measures of happiness or subjective well-being (Buccafusco), or a capabilities approach as Sunder has argued.  How/to what extent should IP be promoting these alternative visions of social welfare?

Rob Merges: W/studies you can say that policy implications are unclear; if you reveal an instinct, that doesn’t mean that instinct should be driving policy, since some of our instincts need to be controlled. But it’s a good case study: we tend to interpret empirical work in a way consistent w/our normative starting points.  Consumer of literature on natural instincts, so needs to be careful.  Still thinks there’s something powerful in the idea: when empirical work lines up with philosophical theories about the right or need for ownership, that’s two different methodologies pointing in the same direction.

Strongest finding across methodologies: sense that we don’t give enough respect for attribution in the US.  US is deficient in reflecting intuition that attribution is important. That’s low-hanging fruit—people will take attribution rights in place of extra compensation. Almost all the studies—experimental, interview, theory—show that it’s something we really should care about. Policy insights would say: we probably got Dastar wrong, and we need some kind of more formal attribution right in US IP.

When we look at how important design is in the modern economy—our theories about TM are running into empirical evidence.  There’s this sense that design is important and we want to encourage it but our TM theory is stuck in consumer protection mode and we’re having trouble adapting it.  If theory tells us that all this solicitude for designers isn’t what TM is about, one response is to kick it out; but if we want to respect design work, maybe we should look at the existence of the design profession exploring new ways of being creative and adjust our TM theory rather than adjusting that evidence.  There are also other areas where theory is out of step w/empirics; need to be more flexible.

Case about how we interpret the studies we used. In the MRI studies, associations around brands light up same part of brain that responds to religious icons.  It’s easy to criticize that; materialist might say religions are very old brands, or might say that brand designers have tapped into something very deep, which takes a lot of talent, and that means it’s worthy of protection [the protection we grant to religions to bar other people from using their icons?  Herein of “if value, then right.”]

Innovation w/o IP versus IP without IP.  IP as exception rather than rule (older version); IP in context (becoming more modest about role of formal IP rights; willing to substitute with tax policy, credits, etc. if it can do the job better). Related to negative space, but broader policy canvas: contextualizing IP as a modest part of a big set of social norms and a big set of policies. Requires some humility in our field.  Private initiatives like the movie title registry MPAA runs; agreements not to sue (smaller scale)—that’s transactional efficiency in the presence of IP. Strong form of IP w/o IP is where it doesn’t exist or has been replaced by informal norms. Entertainment industry is intermediate b/t complete stranger to stranger interactions and very close-knit organizations like roller derby.  Spectrum from anonymity to somewhat shared interest to close-knit; IP plays into norms differently in those situations b/c social distance is a fundamental idea in sociology.

We have to be careful about areas where IP seems to have been displaced: don’t extrapolate close-knit groups to other areas like intermediate or anonymous. Norms only work where there is social proximity and the prospect of some sort of sanction.  Book on licensing samples by Peter DiCola: ethnography of that little part of the music industry, but we need a lot more like that to sort out the role IP plays.

Property instinct literature: I see that IP w/o IP reflects intuition about credit/attribution that need not be formal IP; the fact that people reinvent something like IP when IP doesn’t work indicates it’s in sync w/people’s basic instincts.  When we see IP intruding into these spaces, we have to be careful—won’t always displace norms; formal IP may be useful in final period problem when people are leaving the community/won’t be repeat players. Case law in those contexts won’t be representative and won’t displace social norms, but it’s a special situation.

Distributive justice: seems to me that lots of what we’ve been talking about takes the form of worrying about old tradeoff b/t consumers and rightsholders.  Need for fairness leads him to talk about proportionality; details of IP rights are important b/c they’re where we carry out the job of recognizing the property instinct and the need to limit it.  Property instinct = not libertarian strong form; people’s intuitions are that property should exist but also that it should be limited.

Meta-points: care in mining the literature.  Mindful about what’s trendy. Brain studies = materialist moment. Be careful about making big jumps from study to policy conclusions.  Studies aren’t designed to tell us anything about IP policy: interpretation of those studies/parts of brain lighting up has to be careful.  We have our own group dynamic and we’re not immune from the phenomena we study.

Commentators: Mark Lemley: If fairness matters to production, that needs to be part of our incentive theory. But you can overstate the degree of what’s learned v. social. Fairness is socially constructed and defined. Many things we view as fair today would be viewed as unfair 100 years ago and vice versa.  People 100 years ago would probably take 100 years’ ago’s IP laws as largely fair, with perhaps some outliers; same today. 

People care less about money and more about credit than the law does. That’s a lever point for change in multiple different directions.  One solution: stronger attribution right, but probably not by reversing Dastar. Also suggests low-hanging fruit to reduce social costs by reducing economic control that we give to reduce deadweight loss w/o reducing incentives.

Also consider fairness to consumers. Easier to think about fairness to creators, or fairness as between early and late creators, but consumers have to think they’re getting a fair deal too. Instinctive moral reaction to 6x price increase in Epipen that kills kids; same thing seen in internet piracy cases, where personal justifications for filesharing are about price & greedy people, and some are about ability to have access to information.

Fairness maps differently to patent rules than to the rest of IP, b/c patent rules are so far removed from our instinctive reaction. 95-98% of patent lawsuits outside pharma are against independent innovators, not even alleged copiers; this doesn’t align fairness instincts. We tend to align instincts by telling implausible stories about “theft.”  Either we need to think about why rules are different and realign them, or come up with a story why this should be treated as fair.  There’s a perfectly good economic story in which it’s hard to administer an independent invention defense.

Distributive justice: I resist this even more than fairness as an economist, not b/c it’s not important but b/c economist’s instinct is to first think about expanding the pie and then separately thinking about expanding it. Literature indicates that people generally aren’t willing to think that way and are upset about expanding pie if it’s sufficiently unfair (e.g., dictator game and rejection of beneficial deal).  That said, enlarging the pie is generally a much better idea in the long run—modern tech v. equally distributed 18th century tech.  Average person here is better off than richest people in the world in 18th c.

IP mostly hurts rather than helps distributive justice. Increases costs of things to poor people, decreases access. Also it is very strongly oriented towards winner take all system: music, movies, tech.  Maybe we compensate for that by having outside-IP systems, but badly enforced IP might also be the best of all worlds: people who can afford to pay for music pay for music, but poor people who otherwise might not have access get access for free through “piracy.” May not be a sustainable equilibrium, but consider when infringement promotes distributive justice.

Contracting around IP/IP-free zones.  Important to consider what we should put into the legal rules and what we can rely on the non-positive law to do. People take strong IP regimes and contract down; people take weak IP regimes and contract up. It doesn’t follow that we should have strong or weak rights. It may be that baseline legal entitlement matters somewhat less if people are free to move regimes to a happier place.

Greg Mandel: Policy justifications for IP are richly contested.  Somewhat of a cultural debate; very hard to resolve, see Lemley’s discussion of “Faith-Based IP.”  Attributions sounds good to people of various backgrounds: natural rights, fairness, efficiency: a cheap right to provide in terms of social costs while providing incentives. We should think about these overdetermined policies that can mean something different to different people and may be easier to agree on.  Very hard to move the needle in people’s beliefs. Patent attys tend to believe overwhelmingly in the incentive theory, but we’re never going to give everybody that experience. His guess: many policymakers, judges lean more towards the lay side of the camp and have more disparate views about justifications than experts. 

We still have to worry about lots of distributional concerns and other norms of fairness; there’s agreement among people from different perspectives that IP at a minimum shouldn’t inhibit creativity and innovation, and should support it.

RT: case against attribution rights.  Try and write it!  I did, and I ended up concluding that I couldn’t.  Lemley’s point: Some fairness rules can’t be written into law, at least not law as the US system knows it (maybe civil law, but I don’t think they’ve obviously done a bang-up job either, and even if you did think that the differences in systems are so great that I don’t think we could project success in US).  This debate was had in family law (divorce, ok reasons for) and 1A (defamation, often too bad so sad) and we generally decided it was a better idea not to try to match fairness norms exactly because of the systematic biases in who’s making decisions and information/administration costs borne by the system. Family law’s embrace of no-fault divorce: Makes the system look bad to be constantly judging fairness w/in private relations.  I urge everyone who thinks this is worth exploring to sit down and explore it: draft what you think should be done.

McKenna: What we make of the attribution desire: is that a claim that we aren’t currently getting enough output? Is it a claim that we should give credit regardless b/c it’s right?  The practical difficulties are enormous. To what would it attach? Anything w/snippet/substantial similarity?  It guarantees repeated conflicts b/t TM and ©; attribution of things other than physical entities; every time TM law has encountered that, it’s struggled to find a meaningful answer.  It’s not clear we’re all talking about the same thing—authorship, ownership, © concept, TM concept?

Buccafusco: There’s nothing about IP that makes us welfarists, Kantians, etc.; nothing intrinsic to IP will decide that.  We still need to distinguish preferences and welfare.  You could think satisfying preferences is always conducive to welfare, but I don’t think so; we need to talk about why that might be. Anxious about idea that concerns about fairness are valuable corrective to economic accounts of IP; that people value stuff is not evidence that it is good for them in the absence of other evidence. These intuitions are likely to be weak proxies, esp. where strength of preferences can be manipulated by framing of question.  Hard to say anything useful about welfare effects—my guess is that music is good for welfare, but all 18-25-year-olds are likely to like the music they’re given; no matter what’s popular, it will be popular.  Lemley says: think instead about enlarging the pie.  But he’s not confident we can do that either.  At what point do we have “enough” movies?  Not clear as a matter of © law, or on the consumer side, even though we might think that we can judge producer side welfare.  Sometimes creating is good for people and they get pleasure for it, and that’s a value that counts. But it might be easier to say things about consumer side benefits of patent law.  Length of their lives matters.

MacCoun: working on a computational account of distributive justice. There’s a remarkable consensus about what procedural justice is. Distributive justice is quite different.  You need to define your version.  Allocation by need—there are different ways of interpreting that; same as to allocation proportional to contribution; Pareto-type rules of making no one worse off; equality (minimizing difference, maximizing joint gain), and so on.  You recognize these all. 

Sprigman: Sprigman, Buccafusco & Burns on what’s a name worth: there’s a puzzle at the heart illustrating difficulties w/attribution. We ran a photo contest w/real money. They could choose $ or less $ plus attribution. We showed they valued attribution.  Second experiment, which varied the baseline: we told them once that they had the right but website might want to buy your attribution right; once that they didn’t have the right but could buy it. 4 to 1 difference in valuation: when people had the right they wanted 4x as much as they were willing to pay to get it. Default rule really constructed the valuation. What do you take from that?  Previous experimental work showed valuation gap in transactions—authors demanded more than buyers were WTP.  Default rule of no attribution right will tend to push the parties closer together; European rule will drive them further apart.  Seller will value the right higher, reducing the number of transactions, all else being equal.  Upshot to us: well, it’s hard to know, but there will be a cost associated w/attribution: you might satisfy preferences endogenous on the rule themselves but drive down a market that will clear at a lower rate of output.  If you’re Candide, everything happens for the best, but he thinks that a preference so heavily endogenous on the legal rule should be treated differently than a less manipulable preference.

Silbey: attribution is a place where norms play effective and strong role; communities police them very well. It’s a place where shaming works well. Although I see the anxiety a lot, and the feeling of violation, private ordering has grown up around those misalignments w/law that have been very effective.

Discomfort w/fuzziness of fairness and distributive justice—but efficiency is a social construct too. It’s just dominant (among us).  That doesn’t mean we should align whatever the dominant fairness mechanism is today with our IP system.  When there is profound misalignment, for example in the piracy context—we have a rule of law problem that’s profound. We shouldn’t just accept that we have “pirates.”  Illegal aliensàundocumented immigrantsànew policies. You can imagine a reframing around infringement, a new word that might start changing the fairness construct to make it look no longer like an infringement.  Misalignment can be a canary in the coalmine about disruption of categories that are necessary to the rule of law. ,”We can’t have a system that depends on the existence of rule-breakers.  [It’s called “how many municipalities fund their systems,” which is not to say I disagree with the normative conclusion!]

Pedraza-Farina: Huge emphasis on attribution with oncofertility coalition (boundary crossing organization). How do you deal with patents?  Informant said that she wanted to be on patent to get credit, but she also didn’t want patients to think she was conflicted w/them, profiting from their illness—credit had good and bad features. Attribution is related to patents, but there’s a dark side.

Lemley: there’s attribution litigation even when the economic rights are assigned; people have even tried to litigate about the order of names on the patent, and a few have tried to get their names off patents.

Burk: If attribution is fairness, maybe we don’t try to draft amendment to copyright law, but maybe we have to try to create conditions where people can bargain into the rules that make sense for them—labor law as the solution, or employment law, or some other form of law.  Maybe © and TM and patent are all different situations. Germany has a required royalty sharing agreement with inventors.  Other forms of law.

Ben Depoorter: fabled double distortion argument: if you use law for distribution you get inadvertent effects on allocative decisions w/r/t that activity. W/IP the assessment is even more difficult b/c we understand so little about incentives.  Maybe it doesn’t matter at all, according to our research, but that’s a hurdle.  Ex ante, if you have a notion about inadvertent effects, that’s why you get so much pushback about attempts to change distribution in IP.  If we don’t know what promotes creativity, we also don’t know what distorts it.

Greg Mandel: Attribution as a concrete potential suggestion.  Push back against critiques: I hardly think it’s a panacea, but we’re starting to see potential.  Re: challenge of drafting, might be true but that’s b/c your standards are higher than those for ordinary statutory drafting.  The fact that would be highly imperfect isn’t necessarily a reason not to do it. [I think it’s a reason an attribution right codified into law is likely to turn out to be more trouble than it’s worth.]  Sprigman & Buccafusco seem to describe an endowment effect. All IP rights will do that. If our goal is to make rights transfer easer, any reduction in rights would accomplish that goal.

MacCoun: it’s 2x as large as the standard endowment effect.

Buccafusco: when people own stuff, they may value it in ways conducive to welfare/consistent w/preferences, or they may not; different sorts of people differ in their reactions.  If I as mug owner am differently reactive to endowment effect, it might be less good as a proxy for my welfare than it would be in other hands. We’re anxious that the people given these attribution rights are subject to higher levels of bias.  When you’re trying to use WTP/WTA as measure of welfare and they differ, you need a story about which one to use. Sometimes there are pretty good stories.  Our other studies: Valuation anomaly in creativity studies comes mostly from overoptimism; that’s just a bias leaving money on the table. If the higher valuation is just “this is special to me” we are more likely to say that’s not really a bias. 

Mandel: we could provide a very weak remedy to deal w/this.

Fagundes: what do we do about knowing people’s intuitions about IP?  Normative and descriptive must be sorted.  But you can still say something: (1) be aware of preferences/intuitions as a constraint in the same way you’re aware of gravity; if you aren’t paying attention it will screw up what you want to do, without having normative force. (2) proxy for empirical evidence of actual preferences, but it’s still a descriptive project that constrains normative arguments you might want to make.  My paper was called IP norms in roller derby, but it was really about property norms, not patent or TM.  We might identify different norms in each realm; what constitutes welfare in one domain might not translate into another, for example technical innovation v. artistic creation.  Keep profound differences in mind.

Peter Lee: Universities in patent licensing often reserve rights for nonprofit research, and for use in the developing world: private actor working to increase distributive justice. Do we want access to IP, or to fruits thereof? Do we want communities to have their own IP?

Mark Suchman: Robin Hood loophole: idea that rich will pay for the rule and the poor, judgment-proof, get the benefit of breaking it—redistributive role for the judgment-proof. But there is this other story about how the rules affect the legal and illegal, which is much more symbolic. The rule that creates those who can pay as legal and those who can’t pay as illegal has broader effects.  Later in life, the judgment-proof college student may be stigmatized by past history.  On one hand, I like the flexibility that gives you in crafting policy, but there is a symbolic dark side.  [Consider who gets prosecuted for other “lifestyle crimes” like walking in the street instead of on the sidewalk.] 

Lemley: if system is only working b/c we turn a blind eye to all these acts = need to change the law.

Suchman: consider disparate effects of Prohibition on certain ethnic groups.

Silbey: Tim Wu’s argument about tolerated uses—traceable nature of digital use means there’s less opportunity for obscurity, which is another reason to be concerned.

Sunder: written about composer of “The Lion Sleeps Tonight,” Solomon Linda, whose identity was erased as the music circulated.  His style was considered traditional and “born” to the public domain, even though it was unique/unusual. Implicit biases, global inequalities should affect our consideration of attribution.  See it also w/jazz and race issues around patent/who’s an inventor.

RT: Robin Hood: also social effects on rights holders of constructing it as a right: we see them conscripting intermediaries, including our universities, and arguing that technology needs to be deployed to do the enforcement work that law can’t do.  Imagined plaintiffs play a huge role in these debates: the big owners who come to Congress say they’d never sue really sympathetic plaintiffs, and largely but not entirely that’s right, but they’re not the only ones who’d use the rights they’re asking for.  We need stories to match their stories, because anecdotes end up making policy.  And we need our anecdotes to reflect reality.

McKenna: People who create are often erased in favor of corporate statements of authorship, which no one has highlighted—in many contexts we’re totally willing to have the humans erased. Intangible = special.  Barton Beebe would say: Why is it that people who put together this computer have no ability to be reflected in any way?  Maybe the TM answer is that people know their work will be merged into the corporation, but why not the same thing with ©.

Buccafusco: Hammer and nail: as IP lawyers, when we see distributional inequalities we look for IP to fix it. But there are other policies to deal w/distributional issues.

Silbey: interesting feature of contemporary culture that IP is more a feature of cultural conversation today than 20 years ago. Maybe we can’t talk about tax politically.

Burk: Labor law is decidedly less part of public discourse.

Silbey: challenge is to frame debate in way that has political purchase. Everyone can be an author is a very powerful frame.

Suchman: The other thing to remember is that we’re not just talking about economic redistribution. Dignity, free speech, identity formation—there may be other bodies of law that go after that too, but it’s harder to say there are clearer policy levers. Some of these things are distributive in some sense but very close to IP.

Silbey: you see this a lot in equality debates: does giving someone more equality give someone less? Not normally, but in IP if it’s framed as a money transfer then you get that characteristic.

Pedraza-Farina: we stop short of explaining what we mean by “access.”  Meaningful ability to engage w/it and use it in subsequent acts of creativity/productive recombination. The ability to do that requires some things outside of IP: labor law, education, knowledge infrastructures (Frischmann).

Fromer: Fair use can have this sort of effect if it’s structured right; lack of clarity can be a problem, but it is often trying to implement redistributive justice, e.g., for education. Productive access can come by putting things in the right place in the law.


Suchman: Mandatory licensing can also have some of these effects: owner has no right to exclude, but right to money.  Beyond that it’s fair use; then beyond that it’s public domain; but beyond that it’s public domain with publicly supported dissemination structures.  Policy levers are inside & outside IP.

No comments: