On December 15^th, the Ninth Circuit heard en banc oral arguments in a contentious copyright case:Garcia v. Google. The core questions in the case are interesting enough: whether performer Cindy Lee Garcia can claim copyright protection for her five-second performance in the film Innocence of Muslims, and whether Google should consequently be enjoined from showing the movie. The case raises important questions about online speech, about creativity, and about the recourse available for people who allege harm stemming from online content. A number of professors, myself included, wrote and joined amicus briefs on Google’s behalf. For a great write-up of the en banc hearing, check out Cathy Gellis’s summary over on techdirt.

But the surprise issue of the day was raised (unsurprisingly) by Judge Alex Kozinski. Judge Kozinski wrote the original panel opinion holding that Garcia should receive copyright protection, and he spent the en banc oral argument vigorously defending his stance. One avenue of defense Judge Kozinski chose was to reference the language of a recently negotiated treaty: the Beijing Treaty on Audiovisual Performances. Judge Kozinski referenced both the treaty itself, and a Fact Sheet on the treaty created by the PTO. (See the oral argument for J. Kozinski’s reference to the AVP Treaty at 23:30, and reference to the USPTO factsheet on the Treaty at 24:25.)

The Beijing AVP Treaty was negotiated at the World Intellectual Property Organization (WIPO) by the PTO, the State Department, and the Copyright Office. The treaty appears to speak clearly to one core Garcia issue: Article 6 gives performers the exclusive right to authorize both “the fixation of their unfixed performances,” and the communication of those performances to the public. As Jacob Victor has pointed out, this isn’t really copyright. But it’s hard to think offhand of another U.S. law that would indicate that the U.S. is in compliance with this requirement. And if the Ninth Circuit (in my view, correctly) finds that copyright law does not cover Garcia’s 5-second performance because she did not fix the performance herself, then U.S. law as it stands is probably not in compliance with the Beijing treaty’s requirements. Google’s attorney attempted to argue that the U.S. approach to joint authorship would keep us in compliance with the treaty, but as Kozinski responded, this is a stretch.

As a question of interpreting international law in domestic courts, Google’s attorney correctly pointed out that the Beijing Treaty shouldn’t matter. The United States signed the Beijing Treaty in 2012. But under the treaty’s own terms, it is not yet in effect because 30 parties have not yet deposited instruments of ratification (Art. 26). Even if the treaty eventually does enter into force, the Senate has not yet voted to ratify it by Article II’s required two-thirds majority, and there’s serious discussion in the legal academy over how much weight signed but nonratified treaties should be given in domestic courts (see, e.g., Rebecca Crootof’s great piece on the Charming Betsy doctrine, at n.62, noting this debate and observing that “in practice, courts are using nonratified treaties codifying customary international law in conjunction with the Charming Betsy canon when interpreting ambiguous statutes”). If the Beijing treaty articulated a norm of customary international law, it might be referred to in U.S. courts even without ratification, but the recent nature of the performance right and its specificity counsel against this impact. (Thanks to my colleague John Quigley for this point.)

Once the treaty is ratified, concern over its content would be more justified. Jane Ginsburg has explained that “[i]f the United States ratifies the Beijing Treaty, it will be necessary either to interpret the present Copyright Act to extend to the broadly defined class of ‘performers’ the relevant exclusive rights (subject to transfer to the producer upon authorized fixation), or to enact equivalent copyright or neighboring rights protections.”

But the important question here is how much, if at all, U.S. courts should be restricted in their interpretation of U.S. copyright law by the actions of the domestic agencies negotiating our international obligations. In Fact Sheet on the Beijing Treaty, the PTO explained that “[u]nder U.S. law, actors and musicians are considered to be authors of their performances providing them with copyright rights.” Judge Kozinski latched on to this language, asking Google’s attorney to explain why it does not indicate that Garcia owns a copyright in her performance under U.S. law. Judge Kozinski implied that the PTO Fact Sheet on the treaty, and its interpretation of U.S. law as being in compliance with the treaty, should in some way constrict the Ninth Circuit’s interpretation of a U.S. statute. Should the Ninth Circuit give deference to the interpretation of U.S. law held by an agency that negotiated the treaty? If so, what kind of deference?

What should first strike most people is that the PTO generally administers patent and trademark law, but Garcia is a copyright case. Chevron deference applies to an “agency’s construction of the statute which it administers.” Mead has been read to suggest a “Step Zero” to Chevron, asking first whether Congress gave “express authorization” to the agency to “engage in the rulemaking or adjudication process that produces the regulations or rulings for which deference is claimed.” The Supreme Court has held that Chevron can still apply in limited circumstances to an agency’s interpretation that doesn’t go through congressionally authorized rulemaking or adjudication. But usually, without this kind of evidence that Congress has authorized the agency to “speak with the force of law,” the less deferential Skidmore framework applies. (Thanks to my colleagues Peter Shane and Chris Walker for a very clear walk-through of the unclear mess of interpretive approaches).

The PTO—tasked primarily with handling patents and trademarks—has not been given express authorization to create rules interpreting U.S. copyright law. Nor is there any other statutory indication that has been implicitly delegated with implicit authority to speak with the force of law on the meaning of copyright. The PTO has, to be sure, been given authorization to advise the president on international IP policy issues (35 U.S.C. §2(b)(8), (9)). It would be a considerable stretch to say that this authorization to be involved with international IP negotiations gives the PTO the ability to create deference-worthy interpretations of U.S. copyright law in the process of participating in international negotiations.

Furthermore, the statutory authorization is abundantly clear on the point that authorizing the PTO to be involved in international IP commitments, including copyright commitments, does not take power away from the Copyright Office: “Nothing in subsection (b) shall derogate from the duties and functions of the Register of Copyrights or otherwise alter current authorities relating to copyright matters.”

And on the issues raised in Garcia, as Google’s attorney pointed out, the Copyright Office has directly spoken: it declined to register the work as copyrightable. When you get further into the details, determining the exact level of judicial deference to the PTO and the Copyright Office gets incredibly messy. (See, e.g.: Patry on deference to Copyright Office rejections; the Third Circuit’s refusal to reach deference to the Copyright Office, n. 5; recent thoughts on deference or lack of deference to the PTO.) But the idea that the Copyright Office might be better equipped as a matter of expertise to determine copyrightability than the PTO does not feel like a stretch.

As a policy matter, there is a strong argument that courts should not give deference to the PTO’s interpretation of current U.S. copyright law as outlined in the Fact Sheet. Granting deference to the PTO’s interpretations in the Fact Sheet would allow an agency to exploit ambiguities in statutory law and nudge court interpretations towards treaty language before the Senate or Congress as a whole get a chance to weigh in on whether U.S. law comports with treaty requirements. The default rule should remain that U.S. law remains unaffected by a treaty not dealing with customary international law, at least until the Senate and President ratify it. Otherwise, deference to the negotiating agency gives power to that agency to bypass both the Senate’s and the House’s—and even the President’s—meaningful involvement in determining the relationship between international and domestic law.

International IP lawmaking, I and others have observed, is prone to capture by private parties. The easier it is for international IP lawmaking to have direct effects on U.S. law, the more incentives powerful private actors will have to take their domestic legal battles abroad—or to the domestic agencies that negotiate the international agreements. Agencies are easier to capture than Congress (or two-thirds of the Senate), and thus placing this power in an agency would create an alternative and easier route for narrowing U.S. domestic law. Less coordinated, smaller actors in conflict with those aware of the international system may then find that statutory ambiguities they relied on have vanished—and not in domestic courts.

Policy laundering—the idea that policy makers can use international law to make legal changes domestically—is familiar to IP attorneys. But such laundering has occurred in the past when Congress used new international obligations as justification for altering domestic law. Judge Kozinski’s thinking on the PTO Fact Sheet would create a new, sleeker, opportunity for policy laundering, one in which Congress is almost entirely uninvolved.