The top line: No federal preemption, and FCC must reconsider the impacts of reclassification on public safety, lifeline.
Earlier this month, the D.C. Circuit published its opinion in Mozilla v. FCC. This was the case challenging the Commission’s reclassification of broadband internet access service (BIAS) from a “telecommunications service” to an “information service” and the accompanying repeal of the net neutrality protections the FCC put in place in 2015. CDT and others filed this lawsuit in 2018 because we believe the internet should remain as open (where your ability to access any and all lawful content is the same) and flat (with equally low barriers to entry for everyone) as possible. But preserving an open internet against an uncompetitive landscape in which a few major companies control access requires regulatory oversight. For this reason, we also challenged the FCC’s legal argument for the reclassification, which attempted to permanently abolish the FCC’s own ability to regulate internet access providers.
On reclassification, the court upheld the FCC’s decision, but directed the Commission to rethink the potential impacts its policy reversal might have on public safety, the Lifeline program, and providers’ access to utility poles. Although the Commission downplayed these issues, both in the rulemaking and in its reaction to the court’s opinion, they are substantial. Agencies must consider how their policy changes might impact public safety, and that consideration should amount to more than handwaving. The same goes for the Lifeline program, which helps ensure that not only the wealthy are able to use a phone and get online. Finally, the FCC’s rules that guarantee equal access to utility poles and pipes only apply to Title II services. This means that the major broadband providers (who also have Title II services like landline telephone) can still get access to shared infrastructure, but broadband-only competitors cannot, making it even more difficult to introduce real competition into the market.
The court’s remand means that the FCC must go back and formally re-think these issues. Time will tell how quickly the FCC is able to go back and do it’s homework, but this is not the only issue standing between this administration and total de-regulation.
The FCC’s gambit to rid itself of authority to regulate backfired in a critical way: Because the FCC claimed it had no authority to regulate broadband, the court said that it also lacked the authority to prevent the states from enacting net neutrality laws. So while the Commission won a battle, it may now lose the war. The loss on the issue of state preemption is a major setback for ISPs who would prefer to remain unregulated. (The court did leave open the possibility that the FCC might be able to preempt states on a case-by-case basis, but it is unclear how this might play out.)
Bottom line: State actions will pressure ISPs and Congress toward federal legislation.
The court handed a big win to the 30+ states with net neutrality rules currently in place or in progress. The prospect of a looming patchwork of state regulation should be a strong incentive for ISPs to seek a federal legislative solution. Passing the Save the Internet Act, which has already passed the House (and the Senate in a previous congress) would be the quickest way to achieve this, effectively reinstating the 2015 Open Internet Rules. Other options will certainly take longer, even though the vast majority of Americans across the political spectrum support the 2015 rules. Regardless, while the litigation will continue, the policy focus likely will shift towards the legislative branch.
Between the lines: Court’s hands are tied by SCOTUS (and its own) precedent. It is time for SCOTUS and/or Congress to step in.
A few things stand out in this latest decision. First, the DC Circuit determined its hands were tied by the relevant Supreme Court precedent (NCTA v. Brand X) and by the statutory language of the Communications Act. Second, the court’s understanding of the technologies in question, although improved, still fails to account for the rapid technological advances that are imminent in the U.S. Finally, the fate of the regulatory approach for broadband seems to depend less on statutory interpretation than it does on a couple of relatively obscure technological features.
In both the unanimous opinion of the court and Judge Millett’s concurrence, the court found that it must defer to the agency’s judgments about how to map dense technological and policy considerations onto an equally complex set of statutory provisions. This is nothing new― the Chevron Doctrine has always been at the heart of the net neutrality litigation― but the court’s deference in this case stands out because of the court’s skepticism toward the agency’s reasoning.
The unanimous opinion barely even acknowledges the FCC’s primary argument for reclassification: that because ISPs offer a pathway to the internet, they offer the capabilities of retrieving, storing, processing, etc information. Instead, it focuses on the Commission’s back-up argument: that because ISPs operate DNS resolvers and also cache data locally on their networks, their broadband “offering” is a mix of telecommunications services and information services, potentially subject to classification as either one. We will discuss these in more detail below, but of note here is that DNS and caching were the same “services” considered by the Supreme Court in the Brand X case. Judge Millett’s concurrence expresses frustration with the limiting effects of this precedent:
“[Brand X] compels us to affirm as a reasonable option the agency’s reclassification of broadband as an information service based on its provision of Domain Name System (“DNS”) and caching. But I am deeply concerned that the result is unhinged from the realities of modern broadband service.”
“The Supreme Court, however, is not so constrained. It is freer than we are to conclude that the “factual particulars of how Internet technology works,” Brand X, 545 U.S. at 991, have changed so materially as to undermine the reasonableness of the agency’s judgments and in particular its “determinative” reliance on DNS and caching, 2018 Order ¶ 33 n.99. Or Congress could bring its own judgment to bear by updating the statute’s governance of telecommunications and information services to match the rapid and sweeping developments in those areas. Either intervention would avoid trapping Internet regulation in technological anachronism.”
The per curiam opinion fails to reflect an understanding of the technological underpinnings of internet access. For example, when the court discusses the FCC’s determination that mobile broadband is not an “interconnected service,” it buys the argument “that—even though users need to acquire equipment and software separately for mobile voice—the function of interconnection is provided by the purchased mobile service itself.” The court continues, “With VoIP, by contrast, the add-on application—and not the broadband service—supplies the interconnection functionality.”
In addition to the double standard here (for carriers’ voice service, it isn’t about the software, but for non-carrier VoIP it is about the software?), the court overlooks the fact that mobile voice is VoIP (or the equivalent, VoLTE) and that both are a subset of the functions supported by a broadband connection. The only difference for consumers is that your carrier’s mobile voice application (the one with the phone icon) is pre-loaded on your device, while third-party VoIP applications may require you to download them first.
The court then goes on to an even more flawed analysis: “The gap in Petitioners’ theory is shown most clearly in the obvious inability of a would-be caller from a NANP (North American Numbering Plan) number who seeks to reach a person with mobile broadband but no form of VoIP (or mobile voice service).” Sure, I cannot necessarily make a call from my landline phone to an (unknown?) person who managed to acquire a data-only mobile plan, but I also cannot call someone with a landline phone number who has not purchased and connected a physical telephone. Nor can I (as a person) communicate with a fax number.
What the Commission and the court get wrong here is seeing mobile apps as fundamentally different than physical devices (or customer premises equipment- CPE) like landline telephones and fax machines that enable different services. While older networks require us to use specialized, single-purpose equipment to connect and communicate across networks, modern mobile networks allow us to use a variety of specialty apps on a single multi-purpose computing device. So the VoIP apps, whether pre-loaded or downloaded after a device purchase, are the equivalent of landline handsets—one only needs to “plug them in” to a broadband network.
Despite disagreeing with the FCC’s interpretation (this time), CDT still believes that agencies need to be able to interpret statutory language, and that their expertise should get deference from courts. However, while the Chevron doctrine of deference (to an expert agency’s reasonable interpretations of ambiguous statutory language) is an important aspect of how our legislative, administrative, and judicial branches interact, it is also important to make policy based on an accurate understanding and characterization of the technology at issue. This system relies on expert agencies being expert, but when experts misrepresent a technology for regulatory or deregulatory purposes, they undermine the trust that deference is built on. Although it is perhaps unreasonable to expect judges to have sufficient expertise to question the validity of an agency’s technical analysis, their willingness and ability to do so seems critical to ensuring that deference is not abused.
Which brings us to the two aspects of network operation on which the FCC and the Supreme Court have focused on in the classification of broadband: DNS and caching. In its secondary argument, the Commission portrayed these “services” as an inextricable part of the “offering” made by ISPs. This reasoning tracks the Supreme Court’s opinion in Brand X (decided in 2005, based on facts from 2002), which said that it was acceptable for the FCC to classify cable modem service (early broadband) as an information service because cable companies also performed DNS lookups and caching. And that the average consumer perceived DNS and caching as part of the “offering.”
The overwhelming majority of Americans, Democrats and Republicans alike, support the protections laid down in the 2015 Open Internet Order; let’s get it done!
This was a weak argument, even back in 2005. Now, in the words of Judge Millett, this argument “blinks reality” for at least two reasons. One is the rise in popularity of alternative DNS providers―DNS is no longer an invisible ingredient in the broadband offering but a separate service available from multiple providers with varying degrees of security and privacy. Yes, ISPs still operate their own DNS resolvers, but more and more people are turning elsewhere for this task. There are even encrypted DNS options being built into web browsers and turned into apps (to the dismay of ISPs). And because of encrypted web content, the caching argument has gotten weaker as well. Yes, ISPs still cache some content, locally, but their ability to do so has steadily declined as more and more of the web converts to HTTPS. So why did the court let the FCC get by with this weak, outdated reasoning?
According to the D.C. Circuit, Brand X dictates that the FCC’s determination that DNS and caching turn broadband into an information service (even though this amounts to a tiny tail wagging a huge dog) is a reasonable one. The court goes on to point out that Brand X says nothing about the relative predominance of telecommunications or information services, so if ISPs interweave two golden threads into an ordinary sweater, it is still reasonable to call it a “golden garment.” It’s hard to say how far the court would let this go; could the FCC still call broadband an information service even if no one ever uses the ISP’s DNS? Is it enough to simply “offer” a service like caching that no one asks for or has the choice of using or not using? As long as Brand X is the controlling case, the answer may be yes.
The surest way to fix this unfortunate loophole is by making new law. Whether through the Save the Internet Act or another legislative measure to cement strong net neutrality protections and preserve FCC authority, it is time for Congress to represent the will of the people in law. The overwhelming majority of Americans, Democrats and Republicans alike, support the protections laid down in the 2015 Open Internet Order; let’s get it done!