Tag: Internet

Whitehouse.gov Switches to Drupal

There was some buzz earlier this year when the White House used the free, open-source Drupal content management platform for Recovery.gov. Now the administration’s marquee Web site Whitehouse.gov will be using it.

The AP story linked just above does a good job of recounting the benefits of open source in this application: chiefly, low cost and high security.

Arnold Kling wrote recently on the Library of Economics and Liberty blog relating the work Elinor Ostrom did to win the Nobel prize in economics to how the Internet enables private provision of public goods—no regulation, little to no centralized authority at all.

Open source is nothing if not an example of that, and it’s good to see this use of open source joining many others across the big, beautiful Internet.

From the Oxymoron File: The Neutral Subsidy

Peter Van Doren points me to some revealing passages in a new article in the Journal of Economic Perspectives. In “Subsidizing Creativity through Network Design: Zero-Pricing and Net Neutrality,” Robin S. Lee and Tim Wu caution against tiered pricing for Internet access services, writing:

[U]nless sufficient bandwidth and quality of service can be guaranteed for the “free” Internet, there is a risk that … tiering will serve to sidestep de facto prohibition on termination fees… . [A] priced-priority system could simply become a de facto fee charged for all content providers if the “free” Internet was of sufficiently poor quality and consumers shifted their usage behavior accordingly… . [T]his might dampen the introduction of new content and services and eliminate the subsidy for content innovation currently provided by net neutrality.

Locking in net neutrality by regulation would lock in a subsidy to content providers. Lee and Wu prefer it, and many of us may like the results, but it’s hard to call a subsidy regime “neutral.”

Is This Intervention Necessary?

So asks the Washington Post in a cogent editorial about FCC Chairman Jules Genachowski’s speech proposing to regulate the terms on which broadband service is provided. (More from TLJ, Julian Sanchez, and me.) The WaPo piece nicely dismantles the few incidents and arguments that underlie Genachowski’s call for regulation.

As the debate about “ ‘net neutrality” regulation continues, I imagine it will move from principled arguments, such as whether the government should control communications infrastructure, to practical ones: Will limitations on ISPs’ ability to manage their networks cause Internet brown-outs and failures? (This is what Comcast was trying to avoid when it ham-handedly degraded the use of the BitTorrent protocol on its network.) Will regulation bar ISPs from shifting costs to heavy users, cause individual consumers to pay more, and hasten a move from all-you-can-eat to metered Internet service? We’ll have much to discuss.

Preemptive Regulation of the Internet

Julian Sanchez has already done a fine job of assessing FCC Chairman Julius Genachowski’s speech announcing his plan for federal regulation of the Internet. There was nothing really new in it. No substantial problems justifying regulation have emerged, and—Genachowski’s claims to modest aims aside—any ‘net neutrality regulation is likely to be a substantive morass. Says Julian:

[I]t absolutely reeks of the sort of ad hoc ‘I know it when I see it’ standard that leaves telecoms wondering whether some innovative practice will bring down the Wrath of Comms only after resources have been sunk into rolling it out.”

If the FCC goes ahead with regulating the Internet, the public will get a good look at what closed systems are really like. The FCC’s retrograde “Electronic Comment Filing System” doesn’t even allow full-text searches of submissions. This is but one failing the Internet’s engineers all over the country—and not just in big telcos—will run into dealing with the FCC.  It’s laughable that this outdated telecommunications bureaucracy is trying to take over the Internet.

A complex array of network protocols and business processes make up “the Internet.” The Internet’s end-to-end architecture is good engineering because it is naturally open, flexible, and conducive to communications freedom. The Internet empowers consumers to fend for themselves, such as in their dealings with Internet Service Providers. When Comcast degraded the Bitorrent protocol, it took just weeks for consumer pushback to end the practice. The FCC opened an inquiry long after the matter was settled.

But some politicians and the FCC’s lawyers think their slow-moving, technologically unsophisticated bureaucracy knows better than consumers and technologists how to run the Internet. The FCC’s “net neutrality” plans are nothing more than public utility regulation for broadband. With federal regulation, your online experience will be a little more like dealing with the water company or the electric company and a little less like … well, the Internet!

As Julian said, Tim Lee’s is the definitive paper. The Internet is far more durable than regulators and advocates imagine. And regulators are far less capable of neutrally arbitrating what’s in the public interests than they imagine either.

Eye of Neutrality, Toe of Frog

FCC Chairman Julius GenachowskiI won’t go on at too much length about FCC Chairman Julius Genachowski’s speech at Brookings announcing his intention to codify the principle of “net neutrality” in agency rules—not because I don’t have thoughts, but because I expect it would be hard to improve on my colleague Tim Lee’s definitive paper, and because there’s actually not a whole lot of novel substance in the speech.

The digest version is that the open Internet is awesome (true!) and so the FCC is going to impose a “nondiscrimination” obligation on telecom providers—though Genachowski makes sure to stress this won’t be an obstacle to letting the copyright cops sniff through your packets for potentially “unauthorized” music, or otherwise interfere with “reasonable” network management practices.

And what exactly does that mean?

Well, they’ll do their best to flesh out the definition of “reasonable,” but in general they’ll “evaluate alleged violations…on a case-by-case basis.” Insofar as any more rigid rule would probably be obsolete before the ink dried, I guess that’s somewhat reassuring, but it absolutely reeks of the sort of ad hoc “I know it when I see it” standard that leaves telecoms wondering whether some innovative practice will bring down the Wrath of Comms only after resources have been sunk into rolling it out. Apropos of which, this is the line from the talk that really jumped out at me:

This is not about protecting the Internet against imaginary dangers. We’re seeing the breaks and cracks emerge, and they threaten to change the Internet’s fundamental architecture of openness. [….] This is about preserving and maintaining something profoundly successful and ensuring that it’s not distorted or undermined. If we wait too long to preserve a free and open Internet, it will be too late.

To which I respond: Whaaaa? What we’ve actually seen are some scattered and mostly misguided  attempts by certain ISPs to choke off certain kinds of traffic, thus far largely nipped in the bud by a combination of consumer backlash and FCC brandishing of existing powers. To the extent that packet “discrimination” involves digging into the content of user communications, it may well run up against existing privacy regulations that require explicit, affirmative user consent for such monitoring. In any event, I’m prepared to believe the situation could worsen. But pace Genachowski, it’s really pretty mysterious to me why you couldn’t start talking about the wisdom—and precise character—of some further regulatory response if and when it began to look like a free and open Internet were in serious danger.

If anything, it seems to me that the reverse is true: If you foreclose in advance the possibility of cross-subsidies between content and network providers, you probably never get to see the innovations you’ve prevented, while discriminatory routing can generally be detected, and if necessary addressed, if and when it occurs.  And the worst possible time to start throwing up barriers to a range of business models, it seems to me, is exactly when we’re finally seeing the roll-out of the next-generation wireless networks that might undermine the broadband duopoly that underpins the rationale for net neutrality in the first place. In a really competitive broadband market, after all, we can expect deviations from neutrality that benefit consumers to be adopted while those that don’t are punished by the market. I’d much rather see the FCC looking at ways to increase competition than adopt regulations that amount to resigning themselves to a broadband duopoly.

Instead of giving wireline incumbents a new regulatory stick to whack new entrants with, the FCC could focus on facilitating exploitation of “white spaces” in the broadcast spectrum or experimenting with spectral commons to enable user-owned mesh networks. The most perverse consequence I can imagine here is that you end up pushing spectrum owners to cordon off bandwidth for application-specific private networks—think data and cable TV flowing over the same wires—instead of allocating capacity to the public Internet, where they can’t prioritize their own content streams.  It just seems crazy to be taking this up now rather than waiting to see how these burgeoning markets shake out.


Online Privacy and Regulation by Default

My colleague Jim Harper and I have been having a friendly internal argument about Internet privacy regulation that strikes me as having potential implications for other contexts, so I thought I might as well pick it up here in case it’s of interest to anyone else. Unsurprisingly, neither of us are particularly sanguine about elaborate regulatory schemes—and I’m sympathetic to the general tenor of his recent post on the topic. But unlike Jim, as I recently wrote here, I can think of two rules that might be appropriate: A notice requirement that says third-party trackers must provide a link to an ordinary-language explanation of what information is being collected, and for what purpose, combined with a clear rule making those stated privacy policies enforceable in court. Jim regards this as paternalistic meddling with online markets; I regard it as establishing the conditions for the smooth functioning of a market. What do those differences come down to?

First, a question of expectations. Jim thinks it’s unreasonable for people to expect any privacy in information they “release” publicly—and when he’s talking about messages posted to public fora or Facebook pages, that’s certainly right. But it’s not always right, and as we navigate the Internet our computers can be coaxed into “releasing” information in ways that are far from transparent to the ordinary user. Consider this analogy. You go to the mall to buy some jeans; you’re out in public and clearly in plain view of many other people—most of whom, in this day and age, are probably carrying cameras built into their cell phones. You can hardly complain about being observed, and possibly caught on camera, as you make your way to the store. But what about when you make your way to the changing room at The Gap to try on those jeans? If the management has placed an unobtrusive camera behind a mirror to catch shoplifters, can the law require that the store post a sign informing you that you’re being taped in a location and context where—even though it’s someone else’s property—most people would expect privacy? Current U.S. law does, and really it’s just one special case of the law laying down default rules to stabilize expectations.  I think Jim sees the reasonable expectation in the online context as “everything is potentially monitored and archived all the time, unless you’ve explicitly been warned otherwise.” Empirically, this is not what most people expect—though they might begin to as a result of a notice requirement.

Now, as Jim well knows, there are many cases in which the law sets defaults to stabilize expectations. Under the common law doctrine of implied warranty, when you go out and buy a toaster, you do not explicitly write out a contract in which it’s stipulated that the thing will turn on when you get home and plug it in, that it will toast bread without bursting into flames, and so on. Markets would not function terribly well if you did have to do this constantly. Rather, it’s understood that there are some minimal expectations built into the transaction—toasters toast bread!—unless the seller provides explicit notice that this is an “as is” sale. This brings us to a second point of divergence: Like Jim, I think the evolutionary mechanism of the common law is generally the best way to establish these market-structuring defaults. Unlike Jim, I think sometimes it’s appropriate to resort to statute instead. This story from Techdirt should suggest why:

It’s still not entirely clear what online agreements are actually enforceable and which aren’t. We’ve seen cases go both ways, with a recent ruling even noting that terms that are a hyperlink away, rather than on the agreement page itself, may be enforceable. But the latest case, involving online retailer Overstock went in the other direction. A court found that Overstock’s arbitration requirement was unenforceable, because, as “browserwrap,” the user was not adequately notified. Eventually, it seems that someone’s going to have to make it clear what sorts of online terms are actually enforceable (if any). Until then, we’re going to see a lot more lawsuits like this one.

Evolutionary mechanisms are great, but they’re also slow, incremental, and in the case of the common law typically parasitic on the parallel evolution of broader social norms and expectations. That makes it an uneasy fit with novel and rapidly changing technological platforms for interaction. The tradeoff is that, while it’s slow, the discovery process tends to settle on efficient rules. But sometimes having a clear rule is actually more important—maybe significantly more important—than getting the rule just right. These features seem to me to weigh in favor of allowing Congress, not to say what standards of privacy must look like, but to step in and lay down public default rules that provide a stable basis for informed consumers and sellers to reach their own mutually beneficial agreements.

Finally, there’s the question of whether it’s constitutionally appropriate for federal legislators, rather than courts, to make that kind of decision. I scruple to say how “the Founders intended” the Constitution to apply to e-commerce, but even on a very narrow reading of the Commerce Clause, this seems to fall safely within the purview of a power to “make regular” commerce between the several states by establishing uniform rules for transactions across a network that pays no heed to state boundaries. A patchwork of divergent standards imposed by judges and state legislators does not strike me as an especially market-friendly response to people’s online privacy concerns, but that appears to be the alternative. If there’s a way to address those concerns that’s both constitutionally appropriate and works by enabling informed choice and contract rather than nannying consumers or micromanaging business practices, then it seems to me that it makes sense for supporters of limited government to point that solution out.