Wednesday, December 12, 2007

Internet Freedom and FUD

Adam Thierer has just written another thought-provoking post on TLF/PFF about the well-trod net neutrality debate. He is riffing on a ZDNet article by long-time net neutrality critic Larry Downes. The heart of his (and Larry's) argument is that the Internet should remain free from meddlesome regulation. I must say that I wholeheartedly agree. Where I take issue is whether most net neutrality proposals are necessarily meddlesome. This is a critical distinction too fine for Thierer or Downes.

I agree with Thierer on a great number of issues. He is spot-on when it comes to the perils of presumptive content regulation in the name of child protection. The risks of caving to DOPA-style regulation or other ill-conceived technological "solutions" to the issues that youth face online cannot be underestimated. Closing off vast swaths of the internet to the next generation of online artists, innovators, and entrepreneurs is self-evidently foolish. What Thierer fails to recognize is that the risks he envisions under the banner of "child protection" are paralleled in the world of net neutrality -- just not in the way that he assumes.

A libertarian of his flavor fears government at the expense of overlooking risks from industry. Thierer makes much of Bruce Owen's article from earlier this year, examining the legacy of railroad common carriage. I analyzed Owen's selective memory in an earlier blog post, and have yet to read a persuasive argument to the contrary. Weak rhetoric about "the value of keeping the government out of Internet content" is no more than FUD. Take the following statement, written by Downes and repeated by Thierer, for example -- "So why do the same civil-liberties groups that recognize the value of keeping the government out of Internet content want to open a loophole large enough to drive several Mack trucks through?"

Sounds like a bad idea to me.

But, let's take a look at this suspiciously hyperbolic language. In addition to the highly questionable railroad analogy, Downes rallies airline deregulation and Sarbanes-Oxley in favor of his argument. The role of the Civil Aviation Board (CAB) seems bewilderingly distant from the shape of any net neutrality legislation. No neutrality proposals include anything resembling routing requirements (very different from source-based prioritization), subsidization, or rate regulation (other than the eminently simple "you can't charge certain sources more than others"). Indeed, it's disingenuous to the extent that all net neutrality advocates embrace client-side speed price tiering. Sarbanes-Oxley might as well be a stab in the dark. Please explain, Downes, in what way the corporate costs of net neutrality compliance (adding no additional discriminatory technology) resemble SOX obligations?

Are we to fear government intervention in all of its instantiations? The Thierer rhetoric would certainly indicate so. If Downes believes that "the Internet has thrived in large part because it has managed to sidestep a barrage of efforts to regulate it..." he may wish to brush up on which title of the Communications Act governed last-mile internet access until 2005. The trouble is that the black-or-white government/liberty position undermines Thierer's position. To be sure, the PFF would like to see enhanced antitrust regulation of our communications infrastructure. Antitrust and ex ante rulemakings are both forms of government management. If Thierer/PFF is right, there must be an argument or normative reason why antitrust is a more promising check on the power of infrastructure owners than something like neutrality legislation.

If so, let's do away with the regulation-vs-liberty rhetoric and engage the alternatives on the merits. Why is open-ended ex post antitrust review more straightforward than clearly defined up-front rules? Why do the last two years of broadband consolidation argue against the prior ten years of common carriage that fostered the internet explosion? Why should the proven technological structure of the IP "hourglass" be upset in favor of uninhibited carrier discretion?

Tuesday, November 13, 2007

Copyright and Campus Networks

I'd like to share an email I sent to the staff of Vern Ehlers, congressional representative for my home district in Michigan and attendee of my alma mater, Calvin College. This bill goes for markup tomorrow at 9AM 1:30PM, so it will be interesting to see what happens to the clauses in question. [February 7, 2008: The House passes it with the clauses intact. August 4, 2008: The Senate adopts a version of these clauses.] [August 14, 2008: Signed into law.]

Thanks for taking the time to listen to my concerns about a couple of intellectual property-related clauses in the current draft of the Higher Education Act reauthorization scheduled for markup on Wednesday. At the moment, I'm actually writing from Cambridge, MA where I'm completing my Masters degree at MIT focusing on telecommunications policy and the public interest. However, my interest in the effect of communications regulation on higher education began at Calvin, where I earned a dual-degree in Computer Science and Philosophy. Among other things, while at Calvin I hosted a campus-wide debate on web filtering technologies.

The language in question is on pages 411-412 in the current draft: "Sec. 494. Campus-Based Digital Theft Prevention". The intention here is good -- to curb copyright infringement that takes place over campus networks. The problem is that the mechanism is 1.) mandating questionably effective technical measures to solve this problem and 2.) inequitably punishing all students at any school that is deemed to have not met this poorly defined standard. Specifically, all students would lose their federal financial aid. With respect to the technical measures, the draft calls for "alternatives to illegal downloading" and "technology-based deterrents". The "alternatives" are likely to be a handful of overly restrictive services already available over the internet on college campuses. The "technology-based deterrents" are particularly troubling because they threaten substantial collateral damage to legitimate academic activities (see for example the letter from MIT that I have attached). Technology cannot exercise fine-grained discrimination of data based on the nuanced details of the law (not to mention important carve-outs targeted at "fair use" specifically for academic purposes).

On balance, these clauses seek to defer copyright enforcement to universities rather than the copyright holders themselves. They do so by ham-handedly pushing imperfect technological solutions on our institutes of higher education -- incurring substantial academic and financial costs. I hope that the committee will recognize that this is a poor solution to a real problem that deserves much more exploration.

(Attachment: MIT-P2P-Letter.pdf)

Here is some other coverage of the bill:

[Update 11/15: The bill has passed in committee without amendment to the offending language, so the battle now goes to the House floor. Ehlers' staff was nice enough to get back to me, and I sent them this response:]

Thanks for your timely and helpful reply. I was disappointed to see the bill pass in committee this morning without proposed amendment of the provisions or debate on their merits. I did see the one-page document published by the committee, purporting to address unfounded "myths" raised in opposition by "supporters of intellectual property theft." (attached)

As the bill goes to the House floor, I hope that Rep. Ehlers will be perceptive enough to see beyond this combination of straw men and ad hominem. Whereas the document characterizes the opposition argument as, "H.R. 4137 would take financial aid away from colleges and students who engage in illegal file sharing," this is a misdirection. Rather, it has been clearly argued that the bill would take financial aid away from students at colleges that do not follow the vague mandate to plan alternatives and technological deterrents to illegal file sharing -- which this bill unquestionably does. The document also states that the bill would not "force" colleges to use alternative file-sharing programs, which is technically true but elides the "plan to explore" language that clearly requires the first step in such a process. Finally, the document claims that the bill "does not ask colleges to enforce copyright laws." This is absolutely true. We would not want colleges to be tasked with enforcing the law, and we would certainly not want to task them with enforcing unproven "technology-based deterrents" that are a poor approximation of copyright law and generate considerable collateral damage.

The provisions attempt a "have your cake and eat it too" approach in which the bill does not technically force copyright enforcement on universities, but still mandates action that has that effect. Universities already have ample incentive to deter students from such illegal activity, and (as the MIT letter I forwarded earlier shows) they are exploring the nascent technical approaches to this problem. A grant program to further fund these efforts might be helpful. A vague, unfunded mandate to take a combined technology/alternatives approach is likely to generate only non-productive costs and to have a chilling effect on aspects of academic study.

(Attachment: 20071114COAAFileSharing.pdf)

See also:
September-October 2008 Update:

The Department of Education holds a series of public meetings to determine what specific rules will be used, with input from representatives from higher ed giving testimony. EDUCAUSE explains:

Although the provisions are now law, it is still necessary for the Department of Education to hold a “negotiated rule making” process in order to define exactly how the law will be enforced on colleges and universities. The final rules that spring from that process may not take effect until July 2010. It is essential, however, that institutions make a “good faith effort” to comply with the law in the meantime.

During the negotiated rule making process, colleges and universities will have an excellent opportunity to present their views on how this law should be applied. This is a very serious issue - one that, if done incorrectly, could result in significant costs in money and time, as well as the disruption of networks. Consequently, it is critical that the Department of Education hear from those of us who are directly affected, so they are aware of how important the P2P issue is for our institutions.

The general message in testimony from universities is "make the rules as flexible as possible, allowing each university to decide its own approach."

On December 31, 2008, the Department of Education publishes its intent to develop five negotiated rulemaking committees. The copyright provisions fall under "Team V."

Monday, November 12, 2007

Why Manage?

Bruce Owen recently published an article in Cato's Regulation, entitled "Antecedents to Net Neutrality". Owen's central point is that net neutrality is simply a recapitulation of the failed telecommunications policy of common carriage. According to him, history shows that regulatory structures are subject to capture and that they do a worse job at enhancing social welfare than market forces—even when those forces create at best a Schumpeterian monopoly. Owen's article reads as though this is a new argument, but in reality it's just the latest salvo in the broadband access back and forth that he and Lessig have already had in the publication. (Incidentally, Owen's article is essentially a copy-paste from this earlier AEI-Brookings/Stanford working paper/Free State Foundation cross-post that has footnotes not present in the article version).

The fundamental question is whether history indeed shows that regulation of communications infrastructure fails and that alternatives provide greater benefit for the public. In short, why use regulatory rulemaking to manage communications when property or antitrust can manage it better?

The title of this blog comes from LBJ's remarks upon signing the Public Broadcasting Act of 1967. He explained that, "today our problem is not making miracles—but managing miracles. We might well ponder a different question: What hath man wrought—and how will man use his inventions?" He was speaking in the context of mass media dominated by the broadcast model, and governed by now-abandoned principles like the Fairness Doctrine. At the same time, he spoke of "a great network for knowledge—not just a broadcast system, but one that employs every means of sending and storing information that the individual can use." As predictions of the future go, this is pretty good. But, what about this notion of managing our communications miracles?

To begin with, we need to recognize that property-like rights assignment or antitrust oversight amounts to management just as much as regulatory rulemaking. The rhetoric of restrictive government regulation vs. free property markets is at best a false dichotomy and at worst a justification for arbitrarily picking winners. When it comes to media, and especially broadband access, we are not anywhere near the point at which scarcity has disappeared or at which competition has generated widespread consumer choice. In the interim, we are left with the difficult but necessary task of managing this remarkable invention—one way or another.

Owen posits that common carriage has been an historical failure, and that antitrust should intervene where markets fail. To be sure, net neutrality derives some of its power from the legacy common carriage. But is he right that common carriage in communications law has empirically failed?

Owens' retelling of history is remarkably selective. He begins with railroad regulation, the statutory source these principles. After concluding that over a century of railroad common carriage regulation resulted in "a series of highly discriminatory and dysfunctional regional transport cartels," he begins to discuss common carriage in the context of the U.S. v. AT&T antitrust proceedings in the early 1980's. It is not surprising that he locates his analysis in these proceedings, given his personal connection with the Department of Justice during that period, but such an angle hardly offers a representative view of the doctrine. All we truly learn from his account is that the DoJ analysis of AT&T was difficult, and that the antitrust-based remedies were imperfect.

One might seek instead to start with Title II of the Communications Act, the regulatory home of telecommunications common carriage, and then to focus on the most relevant examples. Most notable are the Computer Inquiry proceedings, which dealt with the interaction of long-held common carriage requirements and newly developing information services. In particular, the Computer II rulemaking sought to retain common carriage at the "basic" services layer (ie. placing a phone call) while allowing less controlled competition at the "enhanced" services layer (ie. ISPs and other dial-up services). Whereas most scholars recognize that Computer II allowed the discrimination-free operation of dial-up ISPs that facilitated the explosion of the internet, Owen inexplicably concludes that the rulemakings ended in "morasses of complex, unworkable, and ineffective or self-defeating regulations."

Perhaps Owen is thinking of other elements of the Computer Inquiries, or perhaps his assessment of the enhanced/basic distinction differs radically from prevailing opinion. He doesn't explain. Nevertheless, he assures us that in any event market competition will give consumers better services at lower costs. After all, telephone companies and cable companies already compete in the broadband market. On his telling, vertical integration and exclusive control over the "pipes" only increase providers' incentives to invest and improve market efficiency. Whether or not his claims about broadband competition hold water is a topic for another day, but either way this argument does little to bolster his weak historical account.

Of course, Computer II is only one example of the negotiation between common carriage and competition. A more nuanced analysis might seek to separate the particular discrimination-oriented policies from the monopoly-contingent clauses. Although net neutrality, in all its variations, derives much from common carriage, it is not synonymous with it. In fact, proponents themselves have criticized the technologically siloed regulatory system in which common carriage operates. The historical success and durability of Computer II contributes to that critique. Owen's accusation that advocates have "apparent ignorance of more than a century of economic and regulatory history" might well be turned back on him.

Why manage? In reality there is no choice. The question instead becomes "How to manage?" Certainly one approach is to defer judgment at all network layers to a market that is deemed to be "well functioning." Antitrust might serve as a backstop in this scenario. In order to decide whether this is the best approach, the policymaker must have a robust understanding of history as well as a nuanced conception of present circumstances. In this post I've addressed the particular historical claim that Owen makes. A richer reading of history would not only tell the full story of common carriage, but would probe the boundaries between the sometimes-competing and sometimes-complementary approaches. It would reference landmark cases like Brand X and Trinko. It would recognize that the debate involves even the device layer.

In future posts, I'll examine some of these cases and spend more time with the contemporary details. My simple motivation is the fact that communications policy affects how we communicate. As such, it is concerned with more than just the efficient exchange of goods and services. It is susceptible to ideological wrangling. It both sparks rhetoric and shapes the platform by which rhetoric is exchanged. The fact that communications is central to our nature as human beings is not only why we manage, but also why informed management matters.

[Update 12/29: I noticed via TLF that in this 2006 NYT Op-Ed Tim Lee made arguments similar to those made by Owens. I respect Tim's work and read most of his stuff, but on this issue his vision is just as myopic as Owens'.]