Picker MobBlog

Closing this Mob

The experiment was to group, blog and disband, and so it is time to disband.

As to the experiment, I am very interested to hear views on that, either as comments on this post or via email. What worked? What didn’t and why? I personally have been very pleased with the discussion, within the mob through posts and comments and across the Web, through links and blogs and the mainstream media.

I should say that I am dazzled by SCOTUSblog’s success—92,000 hits on Monday. Wow. SCOTUSblog has helped to turn Supreme Court opinion releases into movie premieres, minus the cameras and red carpet. That is really terrific, and I hope/expect that SCOTUSblog will make this a regular part of what they do. I also expect other blogs to do this next year, and obviously the mainstream media will do this as well. Dahlia Lithwick and Slate have been doing this for a number of years already. And SCOTUSblog continues to be a key source for information about the Court; see today’s post on the end of the term statistics.

To thanks. I want to thank my co-mobbers for posting and discussing: Doug Lichtman; Jessica Litman; Jim Speta; Larry Solum; Lior Strahilevitz; Phil Weiser; and Tim Wu. Larry, of course, will continue to post on his Legal Theory blog, and be sure to check his post from yesterday on Grokster.

I also want to thank people who commented and who just clicked and read. Until a week ago, clicking and reading was my only experience with blogs, and it obviously still will remain the main way I interact with blogs.

Main but not exclusive. So in setting up this blog, I indicated that mobs would be the heart of the blog and that will be the case.

But between mobs, Doug Lichtman, Tim Wu and I will continue to do some posts. As I have mentioned before, Tim and I co-taught an innovation seminar in the Spring; classes are over, but we want to continue to talk. Doug and I have co-authored before; co-taught a number of times (including the locally-infamous Technology, Innovation and Society class (once)); and are co-residents on the sixth floor of the University of Chicago Law School. Doug, Tim and I had emailed a couple of months ago about trying a blog, so I think that we will try this one.

As to the next mob, Doug, Tim and I are emailing, but again, if you have thoughts, please post a comment or email me. Thanks.

Randy Picker

The Court and Copyright

Bill Patry sent me an email expressing his concern about how the Court is dealing—or perhaps more precisely not dealing—with important copyright issues:

I have been attending Supreme Court oral arguments in IP cases since 1983. Until somewhat recently, one could always predict the outcome by a vote or two. In the 1980s and very early 1990s, it was hit or miss how well the Court grasped the matter, though. What has changed for me is that I am always impressed by the Court's grasp of issues at oral argument, but am disappointed in the quality of the resulting decision. Grokster is the most disappointing of all. Like Lotus v. Borland, where the Court split 4-4 after a week (lazy!), Grokster raises to me serious issues about the ability of the Court to deal with hard copyright technnology issues. I had predicted after Grokster there were not 5 votes for either side and that proved true. But I didn't expect a phony 9-0 unanimous opinion, phony in the sense of tossing something out as if the Court had really done something. In my view they didn't. I'm happy they slapped down the Ninth Circuit's misreading of Sony. And I agree! that the staple article of commerce doctrine shouldn't by itself insulate someone who actively induces infringement (although query whether the betamax would itself meet that test), but I find disturbing the real result of the case a 3-3 decision with three on the sidelines. The Court is only deciding 75 cases a year. It is shameful to me that 5 votes couldn't be found and that the Court instead took the easy way out.

Network Neutrality Redux

Despite Randy's call for Tim and Jim, let me add my two cents.  First, I think Jim is right to note that the law of unintended consequences is something to be wary of.  There were very good reasons for and very good consumer benefits that came from the anti-tying rule that promoted competition for equipment that connected to the telecommunications network.  In the case of a broadband network neutrality rule, however, its advocates have not met (at least by my lights) the challenge of explaining--based on the complex economics I referred to previously--why a categorical rule providing an attachment right is appropriate.

Second, even without the case for a categorical rule, I agree with Tim that a standard presuming an attachment right (or an interoperability right, as it were) is appropriate.  The benefits of a more contingent right, which could be evaluated on a case-by-case basis with an eye to whether a legitimate business justification explained the restriction, is that it allows for dynamism and consumer benefits that might arise only from vertical integration.  To date, however, the FCC has not shown the institutional fortitude for such approaches so whether it will in the future raises a challenging institutional question.

And since Randy and Jim mentioned Trinko, it is fair to state that its thumb on the scale against antitrust courts and in favor of regulatory agencies will look increasingly questionable if the FCC (and other agencies) do not develop the competence (or have the legal authority) to address antitrust-type matters effectively.  I noted before that Posner's concerns about antitrust courts intervening in high technology markets is on the mark; even in light of these concerns, however, I remain concerned about policies that call for deference to regulatory agencies that may be asleep at the switch (like the FCC was during the Bell System's parade of horribles) or to technologists who are developing products with plausible (as opposed to actual) efficiencies.  With that, I will sign off and get back to my work-in-progress on this theme before Randy amends his caution about "you can write or you can read, but not do both" to "you can write or you can blog . . . (except if you are Tim Wu)."

Picker: Implementing Network Neutrality

So a couple of questions for Jim and Tim on network neutrality.

Jim, in your last post, you seem to favor creating neutrality through antitrust, meaning I assume, the court system. You do mention Trinko, as you must, but I would see Trinko as a strong statement against creating mandatory dealing obligations through antitrust, at least in the presence of a substantial statutory scheme that addresses mandatory dealing obligations, namely the 1996 Telcom Act. I should look again at Trinko before I post this, but what do you see Trinko as saying about nondiscrimination? The heart of the discrimination in a mandatory dealing case in a network industry is that the incumbent is favoring itself and excluding outsiders with regard to access to a key input

Jim and Tim: how many providers would you need before you would be willing to walk away from neutrality rules? 3? 4? More? We have cable and DSL; Intel is making me lust after WIMAX; and there will be other wireless approaches.

Speta: Back to Network Neutrality

Enough frothing about administrative law and constitutionalism, and back to telecoms.

The interesting questions are:  what is a network neutrality rule and should we do it?  Tim says that it would forbid the blocking of network attachments and access to legal applications and content.  But Tim also says it would be triggered by a finding of consumer harm.  This last is where the action is, for some have proposed mandatory, or at least default, access rules when describing network neutrality.  I don't think a mandatory access rule is a good idea, because of harms to dynamic competition and innovation, and I don't think such a mandatory rule has yet been proved necessary.  But, if it the obligation/punishment is conditional on a finding of consumer harm, then that would require a finding that markets aren't providing the optimal level of access, opportunities to use attachments, etc.  And, I would be in favor of such a system.  Though I think that more familiar rules like "unfair competition" or even "antitrust" (assuming we can get past the institutional problems that Trinko identifies by using agencies instead of courts) do exactly the same thing and are superior to inventing new rules.

Speta: Do statutes matter ... XYZ

Tim’s right that Brand X ignores what "telecommunications services" is. The majority has two things: DNS and caching. But telephone systems provide signalling and addressing, and caching is entirely optional and separate like voicemail.

But I come at the rest of his post from a different starting point. The fundamental rule of administrative law – and a fundamental of the rule of law – is that agencies can only exercise law-making power if it is given to them by Congress. I do not believe in a nondelegation doctrine. But, if one believes in the Constitution or representative democracy, the power to make law just has got to begin with Congress. Tom Merrill has made this argument with his customary power and clarity. Academics can tell the FCC what to do with the power it has; but (I hope) we can't confer that power in the first instance.  Maybe the FCC should have power to "do good" for the Internet (except where, in Broadcast Flag, Congress has been really, really specific).  I don't think it should, but I can imagine such a world.

To my mind, the whole point of the Brown & Williamson case a couple terms ago was that, even though the broadest terms of the statute might include an authority for the FDA to regulate tobacco, that result is so significant that there ought to have been a clearer expression by Congress that it intended the agency to have that power.  (Not "ought" in the sense that Congress must satisfy a "clear statement" rule, but "ought" in the sense that, given how significant the claimed regulatory authority would be, one can't imagine that Congress would have given it away in such an off-hand manner.) (Step -1 has always existed.) Maybe section 1 of the 1934 Act, which says that Congress is creating an FCC to “regulat[e] interstate and foreign commerce in communication by wire and radio” includes the power to regulate the Internet. But that is implausible (in my view), because the Act then says specifically what lawmaking authority the FCC has – over common carrier communications services (Title II) and over spectrum licensing (Title III).  More.

If I do the thought-experiment Tim proposes, I get a different result.  My guess is that the Court did not accurately “predict[]” that Congress wants the FCC to have essentially uncabined power to regulate the entire Internet, because Congress intended the FCC not to have regulatory authority over lots of forms of communications (like private networks and services) and the 1996 Act specifically pointed in the direction of the FCC’s regulatory demise (sections 160 & 161).  Of course, it’s really the case that no one can make any such prediction, and the rule of law requires the matter to be kicked back to Congress.

I happen to think that statutes do matter to agencies, especially when the D.C. Circuit is watching. Not all the time, of course, for all of these are human institutions and knowledge and skill are imperfect. But regulators talk like statutes matter. Courts talk like statutes matter. And market participants talk like statutes matter. I don’t think this is just talk. The rule of law and freedom to innovate both require it.

Lichtman: Reading Wu, Reading Grokster

Tim sees all sorts of limitations and constraints announced in the Grokster opinion. He sees a distinction between protocols and business models; he perceives a safe harbor; and so on. That leads him to reject the worries raised by Fred, Rebecca, me and others. But I think Tim misses the mark.

Start with the text of the unanimous opinion. I take it that there are only two definitive statement in there that will be hard for litigants to later reinterpret. The first is the one that opens the opinion: “We hold that one who distributes a device with the object of promoting its use to infringe copyright, as shown by clear expression or other affirmative steps taken to foster infringement, is liable for the resulting acts of infringement by third parties." The second is the much-discussed footnote 12. Beyond that, the rest looks like a free-wheeling conversation of factors that mattered in this case but might or might not be dispositive in later cases.

Broaden, now, to the concerns that Tim seems to doubt. First, lots of us think that the Grokster decision will chill legitimate innovation. The worry, put simply, is that honest descriptions of business models and protocols (ie “Rip. Mix. Burn.”) will open the door to plausible litigation. Yes, the makers of a new bittorrent-like protocol might ultimately emerge victorious on an inducement theory; but the music industry will have a credible threat to sue, and an ability to keep the thing in discovery for a while. That will change the dynamics of innovation and negotiation. (Remember when ReplayTV caved in the face of a threat of litigation?)

Second, lots of us simultaneously think that Grokster is too easy for bad actors to evade. The point there is that I can dishonestly describe my new protocol, and in that way fully duck inducement-style liability. So I built the new protocol and advertise it as a great way to move Shakespeare – ignoring that it can also trade music files – and inducement is no worry. Then, as Lior points out, other folks will spread the word for me that the Shakespeare network can also carry Madonna’s prose.

Third, Grokster leaves fully open the question of vicarious liability. It also renders more precarious the Sony rule, which lots of people had wishfully interpreted as a bright-line immunity for dual-use goods. Here again legal uncertainty for everyone.

Overall, then, I think Tim might be too optimistic. The Court left intact most of the uncertainty that was already in place, and (at best) added to the risk associated with inducement (by elevating its prominence) and that associated with Sony (by showing that there is not overwhelming support for the bright-line version of its rule.)

Ok -- Tim, where am I all wet?

Wu: Brand XYZ

So to my mind, the basic upshot of Brand X is this:  FCC, go for it, carpe diem, whatever.  Do whatever you think needs to be done to get us out of this embarassing broadband slump.  How else to explain ignoring the plain language of the statute? 

Can we be honest about this, fellow techies and telecom aficionados?  Justice Thomas's argument that there's no basic transport service offered by cable operators is a joke -- so far-fetched it makes Anakin's conversion to the dark side look plausible by comparison.  For one thing, as everyone knows, the cable company does sometimes sell you a nearly naked transport (layer 2) service, and lets someone else sell you the information (layer 3) service.  In fact, here's the real joke: that's what Brand X was all about.   Cable operators in Brand X were fighting for the right to offer a bundled service. The Supreme Court gave them that right by assuming that they always do offer a bundled service.  (Evil laugh).

So the Court is dead wrong about the statute, but then, who cares?  To elaborate on what Phil said, I think the Court knew something when deciding Brand X.  While Congress obstensibly wrote the 1996 Act, there's probably not an elected person in Congress who has a clear memory of the definition of "telecommunications service" nor now has a discernable intent as to what that was supposed to mean for broadband.  And given that, the Court feels free to do whatever it feels like, which in this case means giving the FCC the green light to try and make some noise.  Go FCC!

Stated more legally: the Mead / Chevron standard, as of now, has step 0 (Skidmore or Chevron) step 1 (plain language?) step 2 (reasonable interpretation)?  But in truth, there's a step (minus one), -1.  And that's called, does Congress care?  If not, ignore Chevron / Mead, and do whatever you think is right, and that's the Ad. law holding of Brand X.  To restate Holmes, we might say that ad. law deference is a prediction of what Congress will do.

Picker: iTunes Again

So a number of posts and comments need responses; I think that I will do that in sort of briefish chunks so that people can read what they want to read.

Let me start with the continuing dispute Tim and I are having about iTunes—I feel as if we are back in our joint Spring innovation seminar where the disagreements were frequent and lively (how nice!)—and that will get us to some useful issues about business and product definition.

So Tim suggests that iTunes might “be accused of facilitating illegal sharing as between its customers.” I have to know more about the product (I listen to music the new old-fashioned way: I buy CDs at Amazon and rip them to my laptop with Windows Media player).

Wu: iGrokster

An already mainstream view, held by many people I respect, including Doug, Fred VL, Cory Doctorow, Rebecca Tushnet, and others, goes like this: Grokster's intentional inducement is too vague a standard and will chill innovation. How can an innovator know whether their typewriter, pen, VCR or filesharing machine will be perceived as illegal? Will Rip, Mix, Burn put you in jail? Who can know?

The opposite view, held at times by Lior, and perhaps others, is that Grokster makes it too easy to evade contributory liability. All you need to do is pretend to write Kazaa without the ads and away you go. Grokster is a joke.

With respect, I think both of these perspectives are mistaken.

First, I think its important to appreciate the superficiality of the new Grokster rule. It is a decision that regulates business practices rather than technologies -- more technically, it is a rule for clients, not protocols.  Nothing in the decision condemns SMTP, HTML, Bittorent, ICQ, or any of the other great inventions that have made it easier to share information.  Protocols, aka "standard articles of commerce," are still covered by the Sony rule.

To be clear about this, we might say that Grokster is the law of clients, and Sony is the law of protocols. For those who like layered models, we might say that Sony covers everything but layer 7 (and also transport protocols in layer 7). And while the 3 votes against Sony are regretable, they aren't the holding.  Nor is Aimster.  Sony is still Sony.

So what is Grokster like as client-regulation? The criticisms of Grokster described above lack confidence in intent-based standards. Both criticisms assume that intent is either dangerously easy (Fred VL) or dangerously hard to prove, respectively.  But I believe intent-based tests are fundamental to law, and in practice do not necessarily suffer from the problems described.

As I see the Grokster test, it is akin to a species of unfair competition law. It damns a business-practice -- it says you cannot run a business whose revenue model is premised on encouraging illegal acts. Doctorow calls that a thought-crime, but it is common for businesses to be so limited. For example, most obviously, business may not create price-fixing agreements, which you might also call a "thought-crime." The point is that intent based-standards are common in the regulation of business, and it is neither trivally easy or trivally not easy to prove intent.

Second, while many seem to disagree, I see the Grokster intent model as subject to various safe-harbours designed for operations like iTunes, and also subject to FN12. I wrote about that earlier, and won't repeat it here.

The results of Grokster, then, might be summarized as follows:

Legal, and safe-harbored
  (1) File-sharing protocols of all kinds (SMTP, FTP, HTML, Bittorrent, Gnutella)
  (2) Clients that are neutral and designed for legal uses (Browsers, email, ICQ)
  (3) Clients that take safe-harbor measures (iTunes)

Illegal
  (1) Clients that can control illegal use and don't stop it    (Napster)
  (2) Protocols that are designed only for illegal uses (Sony).
  (3) Clients that encourage illegal uses (Grokster).

Let's take Bittorrent as an example, as Ed Felten already has...