A ‘Third Way,’ Or a Bridge to Nowhere?

The Obama administration plainly is unafraid to press forward with policy reforms in the face of substantial obstacles — witness the passage of its health care legislation after several serious political setbacks. But not all setbacks are created equal, and the Obama administration’s goal of imposing “net neutrality” regulations on the Internet faces daunting challenges after a federal court in Washington last April struck down the most aggressive attempt by the Federal Communications Commission (FCC) to regulate the Internet, in the case Comcast v. FCC.

Advocates of net neutrality regulations continue to argue that they are urgently needed, and recent developments — including a new joint proposal from corporate behemoths Google (which has long supported net neutrality) and Verizon (which has long opposed it) — have kept the policy debate alive. But the Comcast decision creates an enormous legal obstacle for the FCC: the agency must establish that its net neutrality proposals are authorized by laws passed by Congress, and it must shoulder the weighty burdens that the law places on an agency wishing to overturn its own past decisions. To see why this will be so difficult for the FCC, we must first step back and understand the net neutrality debate, how that debate has evolved, and how the agency’s regulatory powers are defined and constrained by law and precedent.

The net neutrality debate cuts to the first principles of control of the Internet infrastructure: will broadband Internet service providers retain their longstanding control over the manner in which they provide services to various customers? The issue of net neutrality raises complicated technical and regulatory questions, but University of Pennsylvania law professor Christopher Yoo encapsulated it well in the Cato Institute magazine Regulation:

Pinning down a precise definition of network neutrality is difficult. Roughly speaking, it requires network providers to route traffic without regard to the source or content of the packets of data that move across the Internet, the application with which those packets are associated, or the sender’s willingness to pay. In the words of leading network neutrality proponent Lawrence Lessig, “Net neutrality means simply that all like Internet content must be treated alike and move at the same speed over the network.”

Net neutrality is a classic infrastructure regulation, intended to cure the ills of market power: the networks’ perceived ability to exert monopoly power, to choose favored or disfavored customers, and to price services accordingly. More concretely, net neutrality would prohibit networks from selling faster, more reliable service to preferred websites or applications while concomitantly degrading the service for disfavored sites and applications — such as peer-to-peer services for swapping bootleg music and video files.

Originating as a largely academic debate among thinkers like Tim Wu, Lawrence Lessig, Mark Lemley, and Christopher Yoo, the idea of network neutrality took on substantial political salience in the early 2000s, as Web content providers like Google and Amazon began to publicly press for federal measures prohibiting network providers from discriminating among content. Led by Bush-appointed chairmen Michael Powell and Kevin Martin, the FCC promulgated a set of principles, ultimately formalized in a 2005 policy statement, “to encourage broadband deployment and preserve and promote the open and interconnected nature of the public Internet.” But the FCC did not go so far as to mandate network neutrality, and various congressional efforts to codify Internet nondiscrimination bore no fruit.

As a presidential candidate, Senator Barack Obama leapt into the debate wholeheartedly. In an October 2007 appearance on MTV, he promised to make net neutrality a priority in his first year in office, describing himself as “a strong supporter” and offering his audience a dystopian scenario in which the Fox News website would receive preferential treatment from network providers. He assured viewers that as president he would appoint pro-net-neutrality partisans to the FCC. He delivered on that promise in March 2009, when he nominated Julius Genachowski — his law school classmate and the chairman of his presidential campaign’s technology policy working group — to be chairman of the FCC.

Once confirmed to the post, Genachowski quickly acted to promote net neutrality. Only three months after taking office, he proposed to turn the principles in the 2005 policy statement into formal regulations and to add net neutrality requirements. The full FCC embraced Genachowski’s proposal, formally promulgating his net neutrality regulations shortly thereafter. But the agency’s forward-leaning approach was premised upon a particular interpretation of federal telecommunications statutes — one that soon would be severely undermined by a major federal court.

A federal agency “literally has no power to act … unless and until Congress confers power upon it,” according to the Supreme Court. The FCC’s authority comes largely from the Communications Act of 1934 and its subsequent amendments, including the landmark Telecommunications Act of 1996. By that legislation, Congress established several categories of communications and erected a general regulatory framework for each.

Of these categories, the law heavily regulates “telecommunications services,” which provide “telecommunications for a fee directly to the public … regardless of the facilities used.” Providers of telecommunications services are regulated as “common carriers,” a legal term with deep roots in Anglo-American history. Telecommunications service providers, like all common carriers, are subject to severe regulation of their rates and terms of service.

By contrast, the law very lightly regulates “information services,” defined as those offering “a capability for generating, acquiring, storing, transforming, processing, retrieving, utilizing, or making available information via telecommunications.” As the federal government explained to the Supreme Court in 2005, “the Communications Act generally does not impose regulatory obligations on providers of information services,” and even to the extent that the Act does empower the FCC to impose minimal regulations, the FCC “has traditionally refrained from exercising its … authority to regulate information services.”

Finally, the FCC also has “ancillary authority,” rooted in Section 4(i) of the Communications Act, to take further measures reasonably connected to its performance of its statutorily-mandated responsibilities. (This is comparable to the Constitution’s “Necessary and Proper Clause,” which empowers Congress to take action reasonably necessary to exercise its enumerated powers.) But this supplemental FCC authority is not unlimited — a lesson the agency learned the hard way in Comcast v. FCC, as explained below.

The Communication Act’s categories are mutually exclusive: if a service involves both a “telecommunications” component and an “information services” component, then the service collectively falls into the information services category. And so in 1998, the FCC reviewed traditional Internet service providers (ISPs) and concluded that, because ISPs typically provide an integrated service that includes not just data transmission but also Web browsing and e-mail, they should be categorized as information services. Four years later, the FCC specifically reviewed cable ISPs and concluded that they too are information services, not telecommunications services. The agency again rejected calls to separately regulate cable Internet’s data-transmission component, finding that the “telecommunications component is not, however, separable from the data-processing capabilities of the service…. As provided to the end user the telecommunications is part and parcel of cable modem service and is integral to its other capabilities…. We are not aware of any cable modem service provider that has made a stand-alone [that is, without e-mail and other services] offering of transmission for a fee directly to the public, or to such classes of users as to be effectively available directly to the public.”

Critics of the FCC’s decision appealed to the federal courts, but the Supreme Court rejected their arguments in National Cable & Telecommunications Association v. Brand X Internet Services (2005). The Court declared that the FCC had reasonably concluded that cable ISPs qualify as “information services.”

After Brand X was decided, the FCC extended its cable Internet analysis to cover other kinds of Internet service providers, including DSL (phone-based) broadband, wireless broadband, and broadband-over-powerline services. (As a local-telephone-network-based service, DSL previously had been deemed a “telecommunications service” by the FCC.) That same year, the FCC also issued the aforementioned policy statement setting forth the agency’s general principles for light-handed regulation of broadband Internet services.

The FCC’s regulatory framework left broadband service providers with great discretion to control their networks. And only two years after the Supreme Court approved the framework, one broadband provider made headlines with its exercise of that discretion. In 2007, the Associated Press reported that Comcast was discriminating among its users, “actively interfer[ing] with attempts by some of its high-speed Internet subscribers to share files online.” Users of popular file-swapping services such as BitTorrent found their activities “stifled” by Comcast. Comcast defended its actions as necessary for the management of its scarce network resources.

Two advocacy groups, Free Press and Public Knowledge, quickly filed complaints with the FCC, arguing that Comcast’s actions violated the FCC’s 2005 policy statement’s principle that “consumers are entitled to access the lawful Internet content of their choice” and “to run applications and use services of their choice.” In a 2008 decision described by reporter Declan McCullagh as “the first time that any U.S. broadband provider has ever been found to violate Net neutrality rules,” the FCC ruled in favor of Free Press and Public Knowledge. The agency concluded that Comcast’s actions unnecessarily and “significantly impeded consumers’ ability to access the content and use the applications of their choice.”

But for all of the FCC’s technical analysis of the dispute, the truly explosive aspect of its order was a pure issue of law: the agency was attempting to impose substantial regulations on broadband Internet despite its own judgment that broadband Internet was an “information service.” Appealing the FCC’s decision to the D.C. Circuit, Comcast argued that the agency’s orders lacked any basis in law: the 2005 policy statement was a purely advisory document with no legally binding effect, and no other provision in the Communications Act authorized the FCC to impose its authority on the challenged Internet services.

In response, as in its Comcast order itself, the FCC did not argue that broadband Internet services were now heavily-regulated “telecommunications services.” Instead, the FCC invoked the doctrine of “ancillary authority” to take actions reasonably connected to its effective performance of its statutorily-mandated responsibilities. The FCC argued that its regulatory action against Comcast was ancillary to its administration of two sections of the Communications Act: Section 1, which announces Congress’s policy in favor of “rapid” and “efficient” communications services; and also Section 230(b), which announces Congress’s policy in favor of the development of the Internet and the encouragement of “technologies which maximize user control” of the Internet.

On April 6, 2010, the D.C. Circuit rejected the FCC’s argument and overturned the agency’s decision. It declared that the FCC’s ancillary authority can arise from a statutory statement of congressional policy only when the policy statement is tied to “an express delegation of authority to the Commission” — such as Title II’s authorization for regulation of telecommunications services. The FCC’s Comcast orders, by contrast, were not tied to any grant of regulatory authority; instead, the agency had attempted to rely on the broad, generic policy statements of Sections 1 and 230(b). That argument was wholly unsatisfactory to the D.C. Circuit, because “if accepted it would virtually free the Commission from its congressional tether,” giving the FCC limitless authority to regulate the Internet. (The FCC also invoked other specific Communications Act provisions authorizing it to take certain regulatory actions, but the D.C. Circuit noted that those provisions either did not actually authorize the type of regulatory action at issue in the FCC’s Comcast order, or that those provisions could not be considered on appeal when the FCC had not originally invoked them in the Comcast order.)

We can now return to the larger net neutrality debate. At long last, opponents and supporters could agree on something: the D.C. Circuit’s Comcast decision effectively killed the FCC’s October 2009 net neutrality proposal, by rejecting the selfsame “ancillary authority” arguments that the FCC had identified as the basis for its authority to impose net neutrality rules. In fact, the FCC’s net neutrality proposal had even explicitly invoked the agency’s Comcast order. By voiding the Comcast order, the D.C. Circuit had gutted the FCC’s net neutrality plans. In other words, unless Congress stepped in to give the FCC new authority to impose net neutrality regulations, the agency would have to erect an entirely new framework justifying regulations under the existing Communications Act.

And so FCC chairman Genachowski did precisely that, one month later. He decided not to attempt to rely on the now-discredited “ancillary authority” argument. And he chose also not to reclassify broadband Internet services as “telecommunications services.” Instead, he proposed a “third way”: redefine “the transmission component of broadband access service — and only this component — as a telecommunications service,” and apply some (but not all) of the Communications Act requirements applicable to telecommunications services, including the prohibition against unreasonable denials of service and other unjust or unreasonable practices.

As FCC General Counsel Austin Schlick detailed in a memorandum issued the same day, Genachowski’s approach would simply reverse the FCC’s previous conclusion, approved by the Supreme Court in Brand X in 2005, that broadband Internet is an information service, its component parts inseparable. One month later, a bare majority of the FCC’s five commissioners voted to put the “third way” proposal out for public comment.

Now, Genachowski and Schlick are correct that the Supreme Court’s administrative law precedents do allow agencies to change their minds, so long as their new conclusions are not unreasonable. The Supreme Court’s Brand X decision specifically noted that the FCC may change course in the future. Schlick also correctly concedes that any FCC reversal of its own prior conclusions would require a full factual review of the technical question of whether “there is a divisible telecommunications service within broadband Internet access.”

But one must not understate the burdens that the law places on a federal agency that desires to reverse its own precedents. Federal administrative law, as established by the Administrative Procedure Act and expounded by the Supreme Court and D.C. Circuit, strikes a balance between two crucially important interests. On one hand, it seeks to give agencies enough flexibility to allow new administrations to revisit old issues, lest ossified agency decisions prevent regulatory progress. On the other hand, it tries to restrain agencies from arbitrarily and capriciously reversing precedents, which would profoundly degrade the rule of law.

Thus, it is not enough for the FCC simply to claim, as Schlick suggests, that its new conclusion is a “reasoned interpretation” of the law and facts. The agency must also satisfy the standards that govern agency reversals of past precedents, which the Supreme Court most recently summarized in another FCC matter, FCC v. Fox Television Stations, Inc. (2009): When an agency’s “new policy rests upon factual findings that contradict those which underlay its prior policy,” the agency must “provide a more detailed justification than what would suffice for a new policy created on a blank slate.” Consistent with that standard, the D.C. Circuit requires that an agency abandoning past conclusions or policy must explain how “the relevant facts have changed or the [agency] has reasonably made a different policy judgment.”

That means that the FCC would need to confront the existing factual record, including the agency’s own prior conclusions. As we have seen, the FCC concluded in 1998 and 2002 that broadband Internet service was a fully integrated “information service,” based on the fact that broadband Internet users use high-speed transmissions alongside information-processing capabilities — browsing the Web, accessing Domain Name System (DNS) databases to locate Web servers, accessing e-mail, transferring files, and other services. If the FCC is to change its classification, it will have to explain either why the facts have changed or why the agency is abandoning its prior rubric and creating a new standard to apply to the same facts.

Can the FCC satisfy those standards? Perhaps. But it has not yet indicated how it would or could do so. General Counsel Schlick’s memorandum notes that the FCC would “take a fresh look at” the question of “how broadband access providers market their services, how consumers perceive those services, and whether component features of broadband Internet access such as e-mail and security functions are today inextricably intertwined with the transmission component.” That showing will be difficult enough: today the trend seems to be toward more, not less, one-stop-shopping among buyers of broadband Internet services. But even setting that problem aside, Schlick wholly ignores some of the other integrated components identified by the FCC and the Supreme Court last time: namely, users’ reliance on DNS to match website names to Internet protocol addresses, and users’ direct download of files from servers. How many people actually purchase broadband Internet service but do not use (in the Supreme Court’s words) their network provider’s “comprehensive capability for manipulating information using the Internet via high-speed telecommunications”?

Finally, the Supreme Court noted in its Fox decision last year that an agency carries a heavier burden when its new position displaces a policy that “has engendered serious reliance interests that must be taken into account.” That is precisely what broadband service providers will argue when they challenge future FCC net neutrality rules: that their investments in the Internet’s infrastructure over the past decade were premised upon assurances that they could control the terms of service and prevent users from devouring bandwidth that could be reallocated more equitably or in a manner that provides a revenue base necessary to maintain or expand the network.

In its current form, Chairman Genachowski’s and General Counsel Schlick’s proposal appears to be an example of what we might call “Hope and Change”: change the supposed facts and hope for the best. Administrative law allows the FCC to overhaul past agency precedents and erect a new net neutrality framework, but not without first carrying heavy burdens imposed by the rule of law. Short of that, the only way to impose net neutrality is through an act of Congress.