Federal Communications Commission	FCC 24-52


DISSENTING STATEMENT OF
COMMISSIONER NATHAN SIMINGTON

Re: 	Safeguarding and Securing the Open Internet and Restoring Internet Freedom, WC Docket Nos. 23-320, 17-108, Declaratory Ruling, Order, Report and Order, and Order on Reconsideration (April 25, 2024).

Net neutrality is one of those catchphrases that tricks you into thinking that there is no other side of the argument. Who, after all, would support net non-neutrality?  Who would support net discrimination? That sounds almost like opposing milk for babies.

But while political activists can hide behind empty catchphrases and slogans, ISPs and their customers have to contend with the real world.  The internet is a limited capacity network, and performance characteristics like bandwidth, latency, and jitter are scarce resources that need to be allocated, ideally in a way that promotes competition and maximizes value to consumers.  This is not an abstract point.  High latency or jitter means not just choppy video calls and lagging video games but also unreliable control of physical systems like drones, cars, and industrial machinery.  Conversely, high latency has little to no perceivable effect on web browsing or video streaming.  So, it benefits the public for ISPs to be able to offer special low latency routing and make sure that only applications that benefit from it receive it. 

Don’t take my word for it. Professor Tim Wu himself, in the very same 2003 paper in which he coined the term “network neutrality,” conceded that treating all internet traffic equally would not benefit consumers. Tim Wu, Network Neutrality, Broadband Discrimination, 2 J. on Telecomm. & High Tech. L. 141 (2003). 
  He asked his readers to “[c]onsider that it doesn't matter whether an email arrives now or a few milliseconds later.  But it certainly matters for applications that want to carry voice or video.”  Id. at 148.
  He counsels that “[d]elivering the full possible range of applications either requires an impracticable upgrade of the entire network, or some tolerance of close vertical relationships” through which different kinds of traffic could be handled differently.  Id. at 149.
 

This seems to fly in the face of the tired net neutrality catchphrase: “no blocking, no throttling, and no paid prioritization.”  A Commission mandate to not throttle traffic would prohibit ISPs from prioritizing the delivery of video call data over web browsing data, even when the user’s subjective web browsing experience would be unaffected.  And even if a ban on throttling were sufficiently caveated, a ban on paid prioritization would mean that ISPs must determine by diktat which kinds of traffic are more worthy than others, and somehow enforce that in the face of perverse incentives for everyone to mark all of their traffic as being of utmost importance.  It would be like not allowing the post office to charge more for overnight delivery, but instead requiring it to determine for itself which kinds of packages are important and to trust senders to not always claim that their box contains perishable medicine.  Indeed, to distinguish between different kinds of traffic is the opposite of “neutrality,” it is discrimination.  But it’s desirable discrimination.  The kind of discrimination that allows you to pay more for your package to arrive overnight instead of in two weeks. 

This is why every time the Commission has attempted to impose net neutrality, it has had to create various exceptions.  Inevitably, the exceptions are so expansive that they swallow the rule, so narrow that they fail to accommodate necessary traffic differentiation practices, or so vague that they stunt action through legal uncertainty rather than proscription.  In any case, “no blocking, no throttling, and no paid prioritization” is reduced to an empty catchphrase, and the exceptions themselves become the real rules.  It’s no coincidence that this is exactly where Professor Wu went wrong in his 2003 paper.  He proposed a law prohibiting broadband providers from “impos[ing] restrictions on the use of an Internet connection” and then lists six exceptions that, at best, provide no actionable principles by which to guide conduct, or at worst, countenance the Commission and ISPs sitting as central planners of the Internet and decreeing which kinds of traffic are more worthy than others.  Id. at 166-7.
 
 
It's been over 20 years since Professor Wu wrote his article, but it might as well have been yesterday.  It is a testament to just how much net neutrality has become an object of political piety, instead of sober analysis, that the order we approve today is hardly more sophisticated and nuanced than Professor Wu’s first cut at the topic in 2003.  

The central exception that this order relies upon is “reasonable network management practices.” The order defines a “network management practice” as one that has a primarily technical, rather than business, justification.  And it says such practices are reasonable only if they are primarily aimed at a “legitimate network purpose.  Safeguard and Securing the Open Internet and Restoring Internet Freedom, WC Docket Nos. 23-320, 17-108, Declaratory Ruling, Order, Report and Order, and Order on Reconsideration (April 25, 2024) (“2024 Open Internet Order”) at ¶ 568.
”  Those legitimate network purposes include security, blocking unwanted traffic, and alleviating congestion.  But the order casts skepticism on the legitimacy of congestion-alleviating measures that are not agnostic to “source, destination, content, application or service.   Id. at ¶ 575.
”  With all of these vague definitions and skepticism, it’s not clear that there is room for prioritizing the latency of real-time applications. 

The escape valve for the rigidity of being unable to use market mechanisms for reasonable network management is yet another exception, the exclusion of non-BIAS data services from the rules. But the order can’t even tell us what a non-BIAS data service is, and refuses to give a definition.  It says non-BIAS data services are typically only used to reach a small number of internet endpoints, provide application-level services, and use “network management to isolate [their] capacity from that used by broadband Internet access services.”  But just in case that definition was not unhelpful enough, the order makes clear that a service could meet all of these criteria but still not be deemed a non-BIAS data service.   Id. at ¶ 195. 
  It goes on to say that, to make sure this exception is not used to avoid the net neutrality rules, the Commission will be skeptical of any services that could function just as well (how well is just as well?) over the open internet.  And in a further hit to any sense of certainty this exception might impart, the order says that these services should not “have a negative effect on the performance of BIAS . . . or the capacity available for BIAS over time” (emphasis added).   Id. at ¶ 197. 
  But the whole point of this exception is supposed to be that traffic from a specialized service (like a remote-controlled surgery) will be routed more quickly and reliably, at the expense of other traffic (like web browsing). 
 
The vitality of this exception is further cast into doubt by language in the order stating that the Commission is “likely to find that connectivity used for video conferencing offered to consumers would evade the protections we establish for BIAS if the video-conferencing provider is paying the BIAS provider for prioritized delivery.   Id.  
”  But how else, other than through payment, is an ISP supposed to tell apart different kinds of video conferencing, like virtual reality video conferencing?  And how is it supposed to weigh prioritization of those against other prioritized use cases altogether?  Without a market mechanism, the ISP and Commission have to sit as central planner and decide what’s important, instead of allowing consumers and businesses tell them with the price they are willing to pay. 

Amazingly, somehow concerned that these inadequate and vague exceptions might still provide too much freedom of action for ISPs, the order imposes a general conduct standard, under which the Commission reserves the right to find an operator in violation of net neutrality rules even if it didn’t violate them.  The general conduct standard prohibits any person from “unreasonably interfer[ing]” with end users’ ability to use the internet as they wish and edge providers ability to offer the services they want to.  But just in case that vague formulation goes too far, it too has its own exception, for “reasonable network management.” 2024 Open Internet Order at ¶ 516.
 

To the extent these rules guide conduct, they do it by making it very legally risky for an ISP to invest in innovative internet routing offerings.  Perhaps if the broadband industry were plagued with anti-competitive behavior, then the benefits of discouraging and punishing such conduct would outweigh the risks of overly stringent or vague regulations stifling investment and innovation, but remarkably, this order identifies no major instances of anticompetitive conduct by ISPs in their internet routing decisions in the last decade.  

The action we take today is to offer a solution in search of a problem.  It’s hard to imagine a more brazen violation of our duty under the Administrative Procedures Act to refrain from arbitrary and capricious rulemaking.  Further, Section 10 of the Communications Act itself requires the Commission to forbear from applying any part of Title II that is unnecessary for preventing wrongdoing, protecting consumers, or otherwise protecting the public interest.  This order fails that test throughout.

Even more significant than the misguided adoption of net neutrality rules is the declaratory ruling’s imposition of Title II of the Communications Act on the broadband industry.  The Commission is flip-flopping, yet again, on whether broadband is a “telecommunications service” subject to regulation under Parts I and II of Title II. It is often said that a service is either an “information service” or a “telecommunications service” under the Communications Act, but this is not strictly true. First, there is a third category, “common carriers”—defined as anyone “engaged as a common carrier for hire[] in interstate or foreign communication by wire or radio”—that are subject to Part I but not Part II of Title II.  Notably, and unlike the definition of “telecommunications,” the definition of “common carrier” does not contain the caveat that the transmission must be “without change in the form or content of the information as sent and received.”  All telecommunications services are common carrier services, but the opposite is not true.  Second, nowhere does the Act say that an information service cannot also be a common carrier service.  The significance of this is that when properly addressing what services are subject to Title II, we need to consider the possibility that a service might be properly classified as a common carrier service despite not meeting the definition of a telecommunications service.  In that case, it would be subject to (most of) Part I of Title II, but not Part II.  It is likely, for example, that online communications systems like email and instant messaging are properly classified as common carriage services under the Act, since they hold themselves as open to the public for carriage of communication by wire or radio.  Supporting this, Section 2 of the Act contemplates the existence of common carriers that are “engaged in foreign or interstate communication solely through physical connection with the facilities of another carrier…” and makes only Sections 201 through 205 of the Act applicable to them. 
 

This reclassification is how the Commission gives itself the power to impose net neutrality rules, but also much more.  Parts I and II of Title II combine to create one of the most comprehensive suites of regulatory authority known to any agency in this country.  With this reclassification, the Commission could claim the power to require prior authorization for the deployment of every new inch of fiber, every new service offering, every contract, every interconnection agreement, and every change in price in the broadband industry.  The Commission could force ISPs to allow their competitors to use their infrastructure.  It could even require ISPs to continue offering service indefinitely in unprofitable areas, essentially seizing their private property.  Instead of asking what the Commission can do under Title II, the better question is what it cannot do.  The sobering answer: not much. 

For now at least, the Commission is forbearing from many of these authorities, but that should be of little comfort to the broadband industry.  The Commission is always one notice and comment period away from reversing course on any of these forbearances and imposing stringent regulations whenever it sees fit.  Take rate regulation.  My colleagues have all promised Congress they have no plans to impose rate regulations under Section 201, but plans can change quickly.  It’s not hard to imagine my Democratic colleagues turning to rate regulation if the Affordable Connectivity Program (ACP) is not reauthorized by Congress or if the federal courts strike down the Digital Discrimination Order, two Commission programs that operate as de facto rate regulation already.

Is the reclassification of broadband as a Title II service legal?  Commissioner Carr, and many commenters, make a compelling case that it is not.  On the other hand, proponents of applying Title II to ISPs point to the D.C. Circuit upholding the 2015 Open Internet Order, to supportive dicta from the same court when it considered the 2017 Restoring Internet Freedom Order, and to Justice Scalia’s dissent in the BrandX case, in which the Supreme Court upheld the classification of broadband as not a Title II service.  As those decisions show, the federal courts have let the Commission have it either way, as an exercise of Chevron deference.

But if the Supreme Court abolishes Chevron deference this summer, as many expect, then they will face a new question when they consider challenges to the present order.  Instead of being able to say that the Commission’s interpretation is reasonable enough in the face of some statutory ambiguity, they will have to determine whether our declaratory reasoning reflects the best read of the law.  I urge the courts to take this opportunity to settle once and for all whether ISPs are subject to Title II—either as “common carriers” or “telecommunications carriers”—and put an end to this embarrassing and harmful regulatory whiplash.  The legitimacy of the Commission is called into question if after every change of administration, it reverses course on a question so fundamental as whether broadband is a common carriage service.  This uncertainty also robs regulated companies and their customers of the confidence in their rights and responsibilities that the rule of law is supposed to provide them, the confidence that forms the cornerstone of investment and innovation. 
 
I want to specifically note that if the Court dispenses with Chevron deference, the reasonableness of the Commission’s position on whether this or that Title II authority is necessary to serve the public interest becomes irrelevant to the question of whether the Title II reclassification is legal.  After all, the political merits or demerits of a law do not usually bear on its interpretation.  But these forbearance and non-forbearance decisions do have to meet the criteria of Section 10 of the Act, which as I already explained, require the Commission forebear from applying any part of Title II unless it needs those authorities to prevent wrongful practices, protect consumers, or otherwise serve the public interest. Today’s decision not to forbear from various sections of Title II repeatedly fails to meet this standard. 

One thing is clear: nothing in the law required the Commission to take this action today.  The D.C. Circuit upheld the 2017 Restoring Internet Freedom Order, in which the Commission ruled that broadband was not subject to Title II of the Act.  The Supreme Court declined to hear an appeal from that case.  We would be on the firmest of legal ground to keep that order in effect.  Given the demonstrated lack of necessity for Title II authority over broadband, we would have been wise to do so.  

The 2015 Open Internet Order dispensed with the traditional justification of Title II BIAS classification under the “local monopoly” theory.  This pivot was clearly warranted because, by 2015, broadband internet was already becoming multimodal; cable and fiber were beginning to be more effectively supplemented or replaced by satellite and fixed wireless, and the local monopoly theory was going to be more and more a mere relic of the Bell System landline era as time went on.  The 2015 Commission correctly called this trend.  We have seen low-latency, high-speed satellite and fixed wireless explode in uptake since, both domestically and abroad.

This created a problem for the 2015 Commission.  If the local monopoly theory is off the table, what new basis could be found to justify a Title II BIAS classification order?  The 2015 Commission decided upon, and the present one has also adopted, a novel “gatekeeper theory” Protecting and Promoting the Open Internet, WC Docket No. 14-28, Report and Order on Remand, Declaratory Ruling, and Order, 30 FCC Rcd 5601 (2015) at ¶ 78-84.

 that places Title II obligations upon any intermediary between a consumer and a service by characterizing such “gatekeeper” as a terminating monopoly even in the absence of market power analysis. See Lawrence J. Spivak, What are the Bounds of FCC Authority over Broadband Service Providers? A Review of the Recent Case Law, Vol. 18 No. 7, J. of Internet L., p. 1, 28 at fn. 88 (January 2015).



This may, however, prove too much. Gatekeepers that otherwise do not resemble terminating monopolies abound on the Internet.  For example, app stores and platforms have been described as “gatekeepers” See, e.g., John Bergmayer, Public Knowledge, Tending the Garden: How to Ensure that App Stores Put Users First (June 2020) at p. 8 (https://publicknowledge.org/policy/tending-the-garden-how-to-ensure-that-app-stores-put-users-first/).
 and leverage their intermediary capacities to generate much of their revenue.  With the last mile out of the equation, does the FCC have a principled reason to exclude platforms from gatekeeper analysis, or are we just going after BIAS because it’s a quadrennial tradition?

While we waste our time and saddle a well-functioning industry with unnecessary rules and investment-stifling legal uncertainty, this Commission continues to give a free pass to the real villains of the free and open internet: large edge providers. There is no bigger threat to free speech in this country than the edge providers who have anointed themselves the arbiters of which ideas are allowed to be expressed and which are not.  Every day, they abuse their market power, and positions as gatekeepers in the digital marketplace, to pick and choose who is allowed to speak and who is not.  They keep armies of advocates on their payroll whose sole task is to prevent Americans from sharing and reading about dissenting views on issues of national importance.  And they maintain their market power through anti-competitive practices like refusing interconnection with competing platforms.

We need to put an end to these abuses.  All options should be on the table.  I am calling for a thorough inquiry into the Commission’s potential authority over social media and other internet companies.  Whether we can find the powers we need in Title II of the Communications Act, in Section 230, or in other existing or new legislation, no stone should be left unturned. 

Sadly, this order today is part of the edge providers’ agenda to insulate their abusive monopolies from competition.  Major ISPs are among the few companies well-positioned to challenge the market power of Google and Facebook in the advertising market, or of Amazon and Microsoft in the cloud hosting business.  But internet platform companies have successfully convinced the Commission to make such competition illegal.  Themselves unrestrained by any comprehensive federal privacy laws, these edge providers are aggressively lobbying the Commission to adopt privacy rules that ban ISPs from running the exact same kind of advertising networks that they themselves do.  And net neutrality rules will make it very difficult for ISPs to use their physical facilities across the country to build competitors to cloud services like AWS and Azure, lest they be accused of prioritizing their own traffic or violating the general conduct standard.

For all of these reasons, I dissent.


2