Thursday, May 19, 2016

1201 roundtable, evidentiary and procedural issues in rulemaking process

SESSION 2 Rulemaking Process – Evidentiary and Procedural Issues
This session will explore the general operation of the triennial rulemaking process under section
1201, including the evidentiary showing required for an exemption, and the procedural aspects of the rulemaking.
Andrew Moore, Kevin Amer, Regan Smith, Jason Sloan
40,000 written comments.  Participants noted significant effort/burdens.  How is it working?
Rebecca Tushnet, Organization for Transformative Works: 500-600 hours; burdensome.  Cyberlaw Clinic eloquently runs through the evidence about legislative intent and appropriate interpretation of the requirements, won’t rehearse here.
In terms of what a showing of adverse impact on noninfringing uses should require: The Copyright Office should consider actual knowledge and behavior among potential users of the exception when interpreting whether there are alternatives to circumvention—that is, it should consider whether the alternatives are in fact known and used.  CO has not even addressed this argument, which we have made across three rounds of rulemaking, much less given a reason to reject it.  That is frustrating as well as contrary to due process principles and notice and comment rulemaking.
Bruce H. Turnbull DVD Copy Control Association and Advanced Access Licensing Administrator, LLC: In general, it works well. Basic approach is correct: those who seek the exemption have the necessary info.  Some issues about number of rounds.  Respondents: we had one opportunity but proponents had multiple opportunities.
Q: was there something you had left unsaid?
Turnbull: Other than through the hearings, imperfect way of doing it, there were things in the responding round from the proponents that we would have liked to have an opportunity to respond to.  CO could reject new evidence, or give another round, but there were definitely times we would have responded.
Raza Panjwani, Public Knowledge: we filed general comments at the outset: identification of classes of works; standard for identifying noninfringing uses; adverse effects.  Cyberlaw clinic’s comments on adverse effects were very appropriate.  Distinguishing tablets and phones; distinguishing K-12 students and college students and teachers; distinguishing Blu-Ray and DVD—not clear where these distinctions come from, and increases lots of burdens.  Expectation that proponents must ID affirmative precedent showing noninfringement, as opposed to showing that there’s no precedent showing infringement; given 1201, there’s a lack of opportunity for case law development in that case. Effectively denies courts jurisdiction to define what copyright law is in certain areas.
Q: statute requires finding of likely noninfringement.
Krista L. Cox, Association of Research Libraries: Extraordinarily time-consuming. Lucky to have clinic assistance; otherwise it’s not doable in time in resources.  Write a brief on why your use is noninfringing.  Takes a year or more.  Public interest resources/time are extremely limited.  Putting together evidence is difficult b/c courts have lacked the opportunity to develop precedent in the area. Turnbull says that proponents have info necessary, but there’s an asymmetry of interests, where the corporations opposing the exemptions v. public interest groups w/o resources have to walk through this complicated and burdensome and formalistic system.
Chris Mohr, Software & Information Industry Association: Distinguish problems w/statute and problems w/rulemaking. The statute is working quite well overall, encouraging new business models. This was designed to be a fail-safe.  Specifically, statute has to be a source of causation of harm; rather than inconvenience—it has to be the statute causing the adverse effect.  To the extent the rulemaking has gone off the rails, might be useful to examine how closely the library has hewn to that standard.
Q: Elaborate?
Mohr: Library changed its position on whether an exemption can issue to a class of user.  Allowing users makes it a lot fuzzier.  Other statutes may prohibit a course of conduct, and if both statutes are valid and proscribe the same same conduct, 1201 can’t be identified as a cause; has nothing to do with ©.  Tie goes to no exemption.  That’s what I mean by statutory causation.
Matt Williams, Association of American Publishers, Motion Picture Association of America, and Recording Industry Association of America: Comments contain criticism, but you also deserve good bit of praise for taking tricky statute and making it overall work quite well. It’s not broken, so be careful trying to fix it. Functioning as intended. Frequently unhappy with some outcomes, and we have suggestions for improvement. We’d love to see an advanced draft of the language in order to comment on drafting choices.  One filing v. three, we also felt that there were some things left unsaid, in part b/c at the hearings proponents were given a lot of opportunity to explain their cases, so we ran out of time.  PK raised the issue of distinguishing b/t different types of users; grew out of the alteration of the approach to consider a class of users, though my clients have come to find that change helpful, though we do have some concerns.  Band raised the issue of how it will keep growing and growing until the whole building works on it, but that’s a result of the CO trying to accommodate proponents and make their case and alter the definition of a class; you have to take a little bad with a lot of good, which is why the proceeding has grown.  1201 inhibiting the development of normal © case law: never understood that argument. You have to counsel a client to address both. If there are the number of threats that exist, there are lots of ways to pursue declaratory relief w/o connection to the 1201 threat and still pursue their argument; other exceptions-related cases that have nothing to do w/1201.
Q: how to define a class of works: subsequently have defined w/ reference to set of users.  If this were to define wider pool, would there be fewer exemptions?
Peter Decherney, University of Pennsylvania: Not a question of larger or smaller, but about bringing the exemptions in line with fair use law; makes the logic of exemptions much cleaner.  The obsolete video games exemption is really about archivists.  Should rethink de novo review etc.  How evidence is preserved and used and reused for the future.
Q: reform for renewals might take some pressure off.
Panjwani: Clinic help is not sustainable.  Consumer appliances aren’t covered w/any exemptions; that will be an additional class. Comment periods: in my limited experience, the party w/burden of proof gets opening and reply; surreply is typically not a matter of right. If the burden is on proponents, you can have your exemption denied w/o any opponents as happened w/recommendation exemption for blind in 2010.  Cyberlaw clinic delineates 9 factors CO apparently considers, but you can’t ID them from any one place, you have to parse the NOI and the NPRM.
Office has to use classes but grant exemptions for noninfringing uses, which creates a problem. There’s a middle ground. Motion pictures for classes of fair use, instead of types of formats, where that gets too sliced thin.  Urge a movement away from current level of specificity.
RT: Agree.  Too difficult.  Witnesses testify about good pedagogy, not about media studies and film studies or about kindergarteners or 5th-graders or 12th-graders.  That justifies an exemption going to teaching, not to K-12 or college students.  That also tracks the statute and fair use cases, which talk about education not about particular grades.
Turnbull: Focus on K-12 and different types of higher education has enabled exemptions; a broader category resulted in no exemption.  Broader category = much more likely to find that there isn’t a real problem.
Williams: Agree w/Turnbull.  If you only altered that approach, it would be narrowed. You have to decide that the use is noninfringing, not just likely noninfringing; but even w/likely noninfringing, you have to engage in the linedrawing being criticized, like uses of short portions. That helps you get to the conclusion that it’s a lawful use.  Also need conclusion that there are no reasonable alternatives, which requires you to consider different formats. DVD quality footage needs aren’t Blu-Ray quality footage needs.  Allows compliance w/statute (though he disagrees) as well as giving proponents some of what they want.
Burden of proof was debated in 2000 and 2003; lots of disagreements b/t NTIA and CO but both agreed that burden was on proponents, in legislative history.  Fed. Reg. 64588: initial reasoning about burden on proponents.
Cox: Use-based definitions make sense.  K-12 and disciplinary distinctions don’t make sense b/c we’re talking about educating students. Drawing distinctions makes the text of the exemptions less usable.  Most recent process: 2006 exemption for educational uses was 44 words long; now 1000 words long.  Average teacher has a much easier time understanding the former than the latter.  Would make the process less long and drawn out if you focused on uses and not on slicing users apart.
Robyn Greene New America’s Open Technology Institute: CO should be uniformly applying a standard of “likely noninfringing”—Cyberlaw clinic’s comments.  Narrow interpretation of adverse effects also allows for better implementation. 
Decherney: in 2006, we were criticized for limiting to media professors; we had support from many different fields, AAUP support.  DVD/Blu-Ray distinctions are not about noninfringing uses; level of need/quality/useful for education—what is the standard?  Not clear to us, and possibly not to you.
Q: Burden of proof: CO has said that the burden of proof is on the proponent.  Substantive and procedural: reforms that give less opportunity to submit evidence?
Mohr: administrative notice could streamline the proceeding.  Someone who’s proven an exemption should be able to provide abbreviated info that they’ve used it, provided that all that’s being sought is the exact same exemption unless someone comes forward and shows the exemption is being abused or that it should be revisited.  As admin agency, have more flexibility in how you handle that going forward.  [Actually, a court could exercise “de novo” review without revisiting any facts.]
Q: is there agreement than preponderance is the proper baseline framework?  Or do we have latitude?
Panjwani: we don’t take issue with the preponderance as such but what we have to show.
RT: No, don’t change the process to give us less opportunity to make our case, but burden of proof/preponderance on what?  The statute’s requirements: noninfringing use, adverse effects, not the 9 things added.
In terms of interference w/ development of case law: (1) my own experience, counseling remixers who wanted to counternotice—how was this footage made; mostly it was through DVD ripping because that’s the best way to get high quality source—before the exemption they couldn’t counternotice no matter how committed they were to fair use because they’d lose regardless; now that we have the exemption they can go ahead and be ready to make the fair use case; no counternotice I have worked with has ever proceeded to a case; (2) dicta in the Corley case opining on what was not before it—quality—explicitly disavowed by subsequent 2d Cir precedent but still haunting us 16 years later.
Williams: confusion in the comments about what the standard means: more likely than not, which isn’t very difficult to meet.  Comments seemed to think it meant something different. You could clarify that in NOIs. Manager’s report: Don’t come to us with hypos, philosophical issues, only real world issues. Burden has been handled properly.
Q: re totality of circumstances.
Greene: Manager’s report is not appropriate legislative history. Text of the statute doesn’t say any of that. It says likely to be adversely effected by virtue of the TPM. The appropriate standard is whether or not the use is likely noninfringing, which would be in keeping with preponderance of evidence.  Seek to determine adverse effect as a result of TPM, not heightened standard as to level of adverse effect.  Not the report from a single member of Congress—no substantial adverse impact contemplated.
Panjwani: Breaking Down Barriers to Innovation Act offers amendments by explicitly placing into statutory factors a lot of the things we’ve talked about here. Still our position that statute can be interpreted to allow for many of the things if not all.  DVD space-shifting, which we apply for every 3 years. CO disagrees w/us in analysis of case law, but NTIA looking at same case law comes to the same conclusion as us; same happened with NTIA and narrative film clips for fictional filmmakers. The appropriate authority to settle that dispute is a federal court, and the only way we can do that is to get it to a federal court w/an exemption.
Q: how could we fix that?  Would you tip the scales in the case of disagreement?
Panjwani: appropriate application: barring affirmative caselaw saying it’s infringing, the tie should go to noninfringing as that element. If CO is wrong, rightsholders can close that exemption off by bringing suit.
Q: is that disagreement w/outcome or process?
Panjwani: both: if done right, fact of agency disagreement + no affirmative precedent = should have come out differently.
Mohr: Statute says nothing about affect by TPMs; now says affected by the prohibition. That change is the manager’s report issue—that textual change wasn’t mentioned in the Harvard Cyberlaw submission, and that’s fairly significant.  W/r/t manager’s report itself, Sutherland on Statutory Construction—nice summary of case law.  Statements by managing committee’s member—general rule, properly applied by CO, is contained there.
Turnbull: Puzzled by statutory background; changes made later on, but the need for evidence was emphasized by the committee that created the process, though the process was changed later on and moved from Commerce Dep’t to CO.  Commerce became advisor.

Cox: exemption for the blind/ereaders.  Came with number of books they couldn’t access, and Joint Reply from AAP etc. said that the submissions didn’t give any indication that the exemption had been used, and no evidence that exemptions helped.  Difficulty around this is confusion about what you need to show: that the past exemption has made things better? Difficult to get that for the blind. Should be enough to show that there are a significant # of inaccessible literary works out there. Touches both about evidentiary standards and difficulty of the process.
Q: Would statutory change help?
Cox: renewal w/o showing things have changed would help.
Williams: Commerce Committee and Manager’s report are both completely valid legislative history, but either way there are other reports taking the same position.  Likely noninfringing: likely to be adversely effected in their ability to make noninfringing uses—a judge doesn’t tell you that it’s likely that this is what the law is.  The CO has taken a different approach, but is already favoring proponents by saying “likely” noninfringing. To take it further and require adverse precedent against proponents upends © b/c default is that you’re making a copy you’re committing an act of infringement; you then have a burden
RT: compare © preemption in 301: the statute was changed after the report came out, and courts have agreed again and again that the legislative history is not helpful. Same here.  Commerce became advisor and not decisionmaker: that isn’t a trivial change, it’s a complete revision, changing both the entire branch of gov’t in which the decision sits and the competence/focus of the decisionmaker, not to mention the other changes made to the process. “Upends ©” in previous comment: making a copy is not the point of 1201.  The people I represent, for example, have legitimate access; they want to make remix.
Decherney: some of these issues are matters of degree. We might all agree that a remix is noninfringing, but what is the level of quality is needed? This is not really a measurable issue, though we can have anecdotes.
Q: should the period be longer or shorter?
Panjwani: both too short and too long. For dealing with advances in tech, too long.  For coming back again and again, too short.
Williams: no perfect solution.  3 years has worked pretty well.
Mohr: fine from our perspective. Resist the premise that a change requires tinkering w/the statute.  [Not what the content folks said at 512 hearings, though in fairness that occurs on our side too; as Mohr pointed out to me in a followup, SIIA in particular did not ask for changes to 512.]
Cox: Panjwani is right.  To resolve this, could have a shorter period for a new exemption, keeping up with advances in tech, but permanent or streamlined process for existing exemptions.

Decherney: 3 years can be a long time for an educator.  Recognize that law school clinics, which do a lot of the representation, have an academic calendar—that would help a lot. Post-hearing questions complicate things.
Q: If we stay in this sandbox, what can we do to make this better? 
Williams: Hearings should take place when students can attend and do oral advocacy.  Currently hearings come just after students leave for the year.  We’d love to see drafting approaches at the end of the process.  Time is a concern, but given the period for post-hearing letters built into the process, if drafting issues were just presented to people who appeared at the hearing, we could give you some feedback, and that would be helpful.  Hearings are often very helpful, but if we prepare a fact witness that relates to every way that a studio uses access controls, it’s relevant across all different proposals.  Last cycle, it wasn’t clear whether the CO could use the testimony from one proposal in considering another, and that harms both proponents and opponents.  [This is one consequences of too much splitting.] More clarity on what evidence can be presented for the first time at a hearing.  Would be helpful to clarify. 
RT: how to make this better?  Cyberlaw clinic’s suggestions.  Picking up on Adler, who talked about how access was special: access/rights controls have now been merged by actors making strategic use of 1201, destroying balance Congress intended. The CO has referred to this problem several times in its rulings.  Remixers have lawful, paid-for access; what they need is the ability to make creative, fair uses. This is the problem of the merged access/rights control.  Could recognize that as “other factor” justifying exemption, both w/in traditional © categories and outside them.  Congress may have been contemplating Celestial Jukebox, but definitely not celestial fridge or tractor!
Greene: making rulemaking more accessible by allowing proponents to submit confidential versions, particularly in the context of security research but also in some other circumstances. Sometimes we withhold info for fear of liability, confidential business info, or info that could allow others to exploit existing vulnerabilities.  SEC etc. allow.
Q: just for CO?  Public version & private version? 
Greene: yes, there should be public version and full public discussion about parameters. 
Turnbull: no problem with confidential submissions, but we’d like a protective order allowing some witnesses/opponents access to the information w/o exposing it to the public.  There are a number of precedents for that in other admin proceedings.   On access v. rights: the access granted in a DVD is to that content in a particular context in a particular form/format and under the rules of the system. In the clear on the TV, but not in a usable way in the system.  They own it, but they have access only visually & if they want to point a camcorder at the screen go ahead [Can I quote that?] [That is exactly the problem I’m ID’ing about the merger of access and rights controls: you configure it as an access control so that you can control the rights.] Clear that Congress contemplated this rulemaking to deal w/the uses of the content in the context of the access control. That’s why it’s 1201(a) rulemaking and not 1201(b).  Mischaracterize the nature of this proceeding, which is about uses in the context of access control.
Williams: Agree w/Turnbull.  Wrong to think that Congress didn’t anticipate that some access controls would also be copy controls; that doesn’t mean that in every case an exemption should result.  If it is both access & use control and inhibits a lawful use, you go through the factors and see if there are alternatives.  Red herring to say merger has created unanticipated problems.  [Tony Reese has a good article about this.]  No problem w/confidentiality; we might also like to file confidentially, and counsel should be allowed to see them.
Q: will certainly look at confidentiality. But 3-year rulemaking is a much quicker pace than what other federal agencies do, so we wouldn’t want a year of fighting over a protective order.
Q: statutory language about “such other factors.”  Role of non-copyright issues.  Several commenters argued that CO shouldn’t consider non-copyright issues.  Is it within our authority to reach out to other agencies or are we limited to consulting with NTIA.

Mohr: as long as you disclose your sources, you can amass them—though there should be public comment on them.  Get information from experts on particular subject matter and weigh their views even if you didn’t agree with them.
Panjwani: “Other factors” are directed at the fact that the proponents must prove noninfringement, allowing CO to consider other harms; interagency process isn’t necessary to figure out whether exemption is warranted. The q is whether © liability will attach—TPMs weren’t supposed to cover anything under the sun, from product safety to medical devices. Conflation of “there’s software in here and that’s expression” with “that’s our issue”—medical devices have existed for a long time, and law and policy have developed to deal with that, e.g., whether you can modify your car to defeat emissions; if you’re not VW, we have annual inspections to deal with that. This isn’t the appropriate venue.
Q: petition about internet of things.  Examples of hacking subway system, nuclear power grid. Should we entirely ignore public safety?
Panjwani: Hackers are going to hack regardless of whether there’s a 1201 prohibition.
Williams: reiterate: be careful about what constitutes a core © concern. That would have to be carefully defined to have any real benefit. There is a distinction b/t access and copy controls, and Congress clearly intended to protect unauthorized access to works available for subscription/on demand [but not fridges and tractors; movies on demand].  A work protected by both access and copy controls: you could hack for access and watch the movie, and without circumventing the copy control there’d be no nexus to infringement; Congress clearly intended to bar that.
Decherney: All these considerations often cut both ways—found out about hearings that VW was actually the problem, not hackers.  Same is true w/fair use: that is the foundation of many productive businesses and economic successes.
Turnbull: other factors include the integrity of our licensing system, which doesn’t interact w/fair use—whether the system can continue to exist. More generically, as you address “not core copyright” you need to be really careful not to eliminate considerations that matter in the context of expressive works.
Panjwani: role of legislative history: Mohr’s correct about approaches, but in the specific case of the House Manager’s report, there are specific criticisms.  “Substantive diminution” language in the report, specifically criticized by Nimmer in his article on DMCA legislative history.

No comments: