Third Breakout Session
Copyright and Competition
Copyright’s Technological Interdependencies
Technological patrons depend on copyright for their motivations. Copyright is an interdependent part of broader creative systems. Dichotomy between patronage and copyright as alternative systems is wrong. They work together to bring about creation and innovation.
Examples: software and hardware development kits and related platforms. (Apps on Amazon, Apple.) Providing these products facilitates production of/access to more creative works. Those creative works then trigger additional innovative activity. Content creation/hosting (YouTube). Netflix/Amazon original series—produced in order to lure people to the technological products they want people to use, so they subsidize the creation of content/strike deals with content owners. Free & open source software.
Legal examples: digital fair use (Google Books)—innovation to take advantage of works/provide additional value; digital first sale (e-lending, shared accounts); software copyrightability (Oracle v. Google). Growing interdependence of tech innovation and creative activity.
Doesn’t mean to suggest that all is well. Tech patrons can have so much power in an area tha they can cause problems. Troubles: Amazon-Hachette fight (intermediary undermining creative production?); YouTube v. indie labels, YT exercising market power; antitrust law is a necessary part of the broader landscape.
IP clause treats inventors as important to creativity—“Science and the Useful Arts”—so recognizing interdependence is important.
Copyright Authorship and Interpretation
The Authorship Rights of Performers
Context for emerging conflicts: digital tech allows DIY recordings; creators overlook WFH b/c they’re amateurs or very very inexperienced professionals. Impromptu on the street videos of strangers. Recording artists sending termination notices to record companies. Athletes. Subjects who pose for photos, like celebrities, fashion models, etc.
Verbal formulations of authorship standard assume sole authorship: default is one person. Many formulations focus on control, which also lead to idea of sole authorship. Standards thus disfavor claims by performers; intent to share authorship status is required though they can’t really articulate what that intent is. Usually director/producer has no intention of sharing authorship with performers. Also requirement that each contribution is separately copyrightable can disfavor performers because who contributed what can be hard to determine after the fact—did the director instruct the actor? Presumption tends to be that actors don’t do much. Subtleties of facial expression—are those contributions of copyrightable expression?
WFH: CCNV v. Reid cited the Restatement of Agency, but didn’t track source accurately. NLRB cases are much more like the Restatement. Compares copyright and NLRA factors used—almost half the factors on which CCNV relied were not in the Restatement (right to assign additional projects, when and how long to work, hiring and paying assistants, employee benefits, tax treatment). NLRA cases: nonexhaustive list: some used in CCNV, but also include whether parties believe it’s employment, whether individual shares in entrepreneurial risk/reward, individual engaged in distinct occupation or business (shared factors: right to control manner and means; skill required; who supplies tools; who supplies place of work; duration of employment; payment by time v. payment by project; relation to hiring party’s regular business; whether employer is in the business of doing X). NLRA has now gone in direction of entrepreneurial risk/reward.
Where did the non-Restatement factors in CCNV come from? Right to assign additional projects came from copyright case, but the others came from tax, ERISA, and labor law cases—a mongrel. The copyright case also mentioned whether the artist worked for several parties at one time as evidence of independent contractor status, and whether the artist was hired through channels the employer customarily uses for hiring new employees—SCt chose to ignore those factors.
Conclusion: awfully malleable test; can pick any one you want. Not sure instructive for considering who really is an employee for © purposes. Someone could be employee for labor/tax purposes but not ©. E.g., musician represented by AF&M in his/her employment—is that © employee status? Could be divergent outcomes in different circuits. Labor contexts routinely disregard labor loanout arrangements.
Garcia v. Google: doesn’t resolve these issues satisfactorily. No agreement on employee status; not enough in the record. Author of performance even though not author of film—how can her performance be a copyrightable thing except insofar as fixed in a tangible medium which is the film? Dissent says she’s author of nothing. Better answer: joint work, but neither side argued that.
Inadvertent performers: people who don’t intend to perform, or don’t intend to be recorded. If they do consent to fixation, recognizing authorship could allow the person to control/suppress distribution. That is a problem, but joint authorship might help solve it.
Penalty defaults: affect music negotiations. Applying this to collective rights organizations, not just statutory licenses as in previous work. CROs can have similar problems of being one size fits all, and can function as penalty defaults. Parties who look to cut around CROs have some advantages—they can get a higher rate—but drawbacks include adverse selection b/c big powerful companies are the only ones that can decide to go on their own, leaving behind a weaker CRO to serve weaker entities. Potential for anticompetitive conduct, “bordering on collusion” (Pandora v. ASCAP).
Pandora is her case study. SDNY is the rate court. Major music publishers (there are only 3 left) decided for the first time ever to withdraw digital rights from ASCAP. They didn’t say outright that they needed ASCAP to help with physical world enforcement but not digital. They said they didn’t like that record companies were getting more money than they would.
A vote in favor of private ordering—got up to double the rate that ASCAP got. The ultimate ruling by Judge Cote: Nope. Can’t withdraw just digital rights. Paper argues that §106 explicitly says that copyright owners can split up rights. But sympathetic to judge—they’re colluding. Still, shouldn’t contravene federal law.
System isn’t working. The DOJ allowed us to get to the point where there are only 3 major publishers left; the rate court contravened federal law; adverse selection causes problems for musicians and songwriters who were promised higher rates, when companies actually negotiated big advances and lower rates. Time to reconsider competition policy in copyright. Should not assume robust competition in the marketplace. Statutory licenses are circumventable/opt-out from CROs both possible based on the assumption of robust competition. If that’s wrong, we need to change. Inverse competition model would assume monopolistic power—licenses could be mandatory, like §111 and §119, with a license to petition for ability to circumvent. Telecom Act has similar provisions for incumbents to petition for exception to interconnection rules if they show robust competition exists. While technically songwriters can withhold rights, in reality there’s no practical way to do so; CROs may in essence be a statutory license.
CRO could compete with withdrawing member in her proposal—that would be a statutory license proper. These mechanisms would reduce anticompetitive harm while allowing private ordering under conditions of competition. Use existing regulatory authorities like CRB. Requires minimal amendment to Copyright Act.
RT: but isn’t the rate court in existence because of antitrust law? Why does preventing a violation of the Sherman Act contravene §106? Compare: violation of defamation law, or even right of publicity under current doctrine. May prevent you from splitting rights as you’d prefer (cf. Wendt v. Host which precludes licensing creation of derivative works) but current doctrine says that’s not a problem, and if it is a problem it’s a conflict preemption problem. Preclusion would be the analogous issue here. (If it’s possible to abuse a copyright and be subject to an antitrust remedy (cf. Microsoft), I don’t think that divisibility of copyright should have separate force in the antitrust analysis. Interfering with divisibility seems to me the same thing as interfering with an exclusive right or a right to license for these purposes, unless there is something special about divisibility.)
A: thinks there’s something different about precluding divisibility.
Carroll: the remedy for a violation of antitrust laws is this consent decree, and the judge is determining the scope of the remedy—this isn’t a copyright judgment, but giving effect to an antitrust remedy to which you’re subject.
Asay: but the judge said you’re in or you’re out of the consent decree.
Carroll: that’s an antitrust ruling: you can’t use a little bit of market power in one area and not in another. If you’re going to exercise market power by being part of ASCAP, you have to be part of the consent decree.
Q: precedent for simultaneous licensing by CRO and copyright owner for broadcast networks—ASCAP can make license available, but they’re required to let you negotiate program licenses with the copyright owners too. ASCAP can’t be the only place you can go = precedent. Music companies would argue that they can withdraw, but it’s a precedent for avoiding supercompetitive rates.
A: that’s her theory
Information Product Redesign as Commercial Expression: Antitrust Treatment of Speech and Innovation
Each speaker has expressed some malaise with respect to competition policy/antitrust. (NB: LaFrance was a different session I ducked into b/c of my interest in Google v. Garcia; the remaining presentation in this session was about DRM and ebooks.)
1st Circuit, reversed in Sorrel v. IMS: just because their product is information instead of beef jerky, plaintiffs contended that any regulation was a regulation of speech. How seriously will we take this? What shall we do with ever-increasing speech defenses that your effort to impose antitrust law on me violates my free speech rights? Wrapped up with core challenge in antitrust: how to analyze innovation-based defense arguing that innovation is pro-competitive offsetting anti-competitive effects. Current treatment is all-or-nothing and that’s not a sound approach.
Google’s search dominance as an example (Nielsen is another in the paper). 70% of general search engine market; changes in algorithm have sometimes resulted in antitrust investigations or litigation for disadvantaging vertical competitors. Volokh & Falk white paper: First Amendment protects search engine results against antitrust law.
Core value in antitrust: managing tradeoffs. But antitrust is better with apples to apples tradeoffs. Speech in this context is a noneconomic right that antitrust law doesn’t handle very well. Innovation is something antitrust law really supports, not just price and quantity. Notwithstanding, it still has a really hard time operationalizing that value and making tradeoffs: you see a binary treatment for both speech and innovation, all or nothing.
Political speech/speech petitioning gov’t is immunized, no matter how anticompetitive; but pricefixing as communication gets no protection. Both of these outcomes are good. But it breaks down in the middle. E.g., to protect NAACP boycott in Claiborne Hardware, Court had to ignore facts in record (some boycott participants had economic interests in conflict with boycottees); even though the result was right, the analysis was wrong and an example of distortion caused by all or nothing approach. In commercial speech and defamation, courts have alternatives.
Similarly with innovation: courts sometimes claim to balance innovation against anticompetitive effects, but they never actually do—don’t admit that product redesigns can have both pro and anticompetitive effects; they allow small innovations to overwhelm larger anticompetitive effects.
We need to admit this is hard but start picking off the easy suspects (innovation small, anticompetitive effect large). Similarly, immunize political speech but not pricefixing, and then give low level of solicitude to speech inherent in Nielsen ratings and Google results—low but not nonexistent. A plea for a middle ground.
Q: the Google argument is that Google is engaging in some form of fraud. On the other side, Google might be afraid to say that a page owner’s willingness to pay to raise results is good evidence that the algorithm got it wrong, so that’s not a problem.
A: First Amendment route is to say that the algorithm is an opinion. The other route is to say “I’ve made technical decisions; don’t second-guess me.” Nielsen: says it makes decisions about how to sample. Real terror of having to second-guess the market. You see that also with Google. Assumptions built in to what kind of harm to innovation will occur if court gets into second-guessing. That is indeed worrisome but it’s unclear how much of a morass that will be. Thus her idea is to start in the corner where there’s trivial innovation but clear anticompetitive harm.
Q: framing similar to net neutrality debate. First Amendment interests of third parties are also relevant—interest in access to search results. Competing First Amendment interests also need balancing.