Today it is hardly possible to imagine an internet without its biggest players. Google’s market share in search is somewhere between 70 percent and 90 percent (depending on whether services like YouTube and Google Maps are included). Facebook has more than two billion active users. Netflix alone is responsible for more than a third of all internet traffic in North America during peak times. Given this level of concentration, the notion of a “free and open” internet now seems almost quaint. Most information transmitted online today is sieved through a few large search engines, social media networks, and streaming video services. These firms may collectively be referred to as “edge companies”: they provide online content or services over a network, but they do not own the network themselves. To earn revenue, they either require users to subscribe or they sell ads on their platforms, often tailored to users who typically have to establish an identifying account to use the service.
Although happy bromides about our new interconnectedness prevailed during most of the last two decades, the dominant edge companies now face increasing scrutiny. These companies have made their users compulsive—connected, but remote. They have dislocated commerce that once would have been predicated on local, face-to-face interactions. They have hypocritically placated censorious regimes to get a foothold in large marketplaces. They have contributed to fake news. They may have knitted a global platform, but it is one that has paradoxically given rise to nationalism and various other strains of identity politics. This article could be written—as many articles have been written—on the wayward society these edge companies have had a hand in creating.
Yet perhaps the most significant fact of online life today is that the users logging on to edge companies are not really at the table; in many respects, they are part of the menu. Characteristic of these edge companies’ business models is that users divulge large amounts of individual data to them, whether through a search query, or by populating a profile with personal information, or by writing an email transmitted and hosted for no charge on a server. These firms have achieved their stunning market capitalizations because users have shared these details about their life, work, and location. In doing so, internet users have made it possible for the real consumers of these companies—everyone from the cruise line advertising on your Gmail, to the Amazon e-store owner slinging printer cartridges, to the cabbie-turned-Uber-driver taking you on a ride downtown—to vend their information, wares, or services through the edge companies.
Who is the consumer? And what is the product? These basic questions are essential to any discussion of a rather more mundane but important topic: How do we pay for the underlying infrastructure, the broadband network, which supports all of this? If we who pay for a subscription for broadband service are, in fact, the users who have created the market value of Google and Facebook, then why are Google and Facebook not the ones “subscribing” to the internet to reach us? Why are they not paying for the networks that make their businesses possible?
The Broadband Networks
What one ends up seeing on the internet may be functionally controlled by gatekeepers—the edge providers—but the networks themselves are the property of mostly big, but not as big, firms. They provide a service which has come to be known as broadband. (As the word implies, broadband is faster than its information-transmitting precursor, dial-up, and defined in practice as connections offering download speeds of anywhere from three to one thousand megabits per second.) Broadband does not describe the physical medium by which the service is provided. Fiber optic cable is currently the dominant choice for the broadband infrastructure backbone, outcompeting digital subscriber line (DSL) and cable modem. Wireless and satellite services also provide broadband, although the former ultimately relies on a wired backbone and the latter provides speeds too low for many consumers to consider it a real source of broadband. The technological antecedents of fiber are indicated by most broadband firms’ origins—typically, as the “phone” or “cable” company. But those distinctions have blurred, and the legacy phone and cable companies are effectively all in the broadband business now. The lobby for cable companies, the National Cable and Telecommunications Association, rebranded two years ago and is now “NCTA—The Internet & Television Association.” USTelecom, the trade association for what used to be phone companies, now features the tagline “The Broadband Association” on its website.
The investment in these broadband networks has been massive. Total capital plowed into broadband technologies totaled $76 billion in 2016, according to USTelecom. This figure sits near the average of annual broadband capital spending during the last two decades, though perhaps the relative significance of this spending can only be fully appreciated when put into perspective with other infrastructure investment. By comparison, in 2016, the federal government requested $51 billion from Congress to spend on the national highway system, and actual federal spending on roads has been less than that annually. Meanwhile, the nation’s fifty-two largest electric and natural gas utilities reported a collective $100 billion and $14 billion in capital spending, respectively, in that same year. Cities and towns spent $67 billion on their water utilities in 2015. Another $2 billion was kicked in by privately owned water utilities. Thus, by almost any measure, broadband is among the most capital-intensive infrastructure-based services in the United States—which, given how ethereal the internet may seem, is not always widely understood.
Broadband networks in the United States are unusual in many ways. Private roads are the exception, and not the rule, in the national highway system, as are private water utilities. Energy utilities are for the most part owned by investors, but they are also usually state-franchised monopolies and have their rates set by government regulators. In contrast, broadband networks are neither government-owned, nor do they possess a legal right to monopolize a particular territory, nor do they have their rates set by regulation.1 Unlike all the other public-owned or public-regulated infrastructure in the United States, the return on capital investment, for broadband companies, is dependent on their ability to obtain subscriber revenues in the open market. And competition among capital-intensive networks, it turns out, is not easy.
Deploying broadband to any individual customer requires substantial fixed costs. The presence of two providers in a particular area affords customers a choice—but not one that necessarily results in systemically lower costs. Two overlapping networks, which must both be constantly available to most customers in a particular area, require about twice the amount of investment, less whatever we may assume would be wasted when there is a single company extracting monopoly rents (attended by the laziness and lack of innovation which characterizes firms which needn’t compete for their customers’ business). Some argue, optimistically, that this deduction for monopoly laziness actually causes two rival, cost-cutting networks to ultimately cost less than one. The better argument is that customers are only given a choice of products (wireless data, for example, and not just a wireline connection) because competition caused one firm to leapfrog its competitor by offering something utterly novel. Such an innovation may not have occurred if there had been only a single firm focused, myopically, on the provision of plain old hook-me-up-to-the-jack internet. In any case, and for the moment excluding the possibility of a future technological leapfrog that will make broadband categorically obsolete, we may nonetheless conclude that for the service now popularly demanded—high-speed broadband—three broadband networks in the same place likely cost more than two, and four more than three, and so forth.
There is an important corollary to these principles of broadband economics. Once deployed, each network can usually accommodate—for the fixed costs which it has already incurred—additional customers at little additional cost. Moreover, once the network is built, customer location makes little difference. It does not really cost the broadband provider more for Farmer Fred to monitor his cattle stock pen via webcam from his home office in Wibaux, Montana, than it does for Broker Blake to execute a trade on a stock exchange from his downtown cubicle on West Street. Costs do exist at the margin, but not as a function of throughput data at any given moment. Instead, Farmer Fred is less likely, in the first instance, to have broadband than a dial-up product; upgrading from copper to fiber will incur lots of fixed costs, spread over relatively few people in a rural area. Broker Blake, meanwhile, might want a physical reroute of his office’s traffic so that it gets to the exchange’s server a millisecond earlier than others’ traffic; that, too, will incur capital costs.
If you’re a customer who, like many, has an old coaxial cable hookup protruding idly from your wall, while getting service from another provider, it may cost the idling, rival firm nothing more than a stamp per month to serve you instead, and then only if you do not opt for electronic billing. It may cost such a firm nothing more than five thousand stamps, and perhaps another customer service representative, to take on an additional five thousand customers. The marginal cost of serving additional customers on these networks tends to be quite low. And in a low-marginal-cost environment, both the customer who pays $10 and his neighbor who pays $100 are probably covering those costs and then some. A firm facing these economics will therefore do everything it can to keep its customers—thus the rewards that await consumers who call their cable company credibly threatening to switch providers—even if that firm may appear to take its customers for granted to begin with. It is a game of subscriber-base-as-tribe, with a provider-subscriber relationship that is stickier than the relationship a consumer has to vendors of many other products in the market. Each firm realizes that, given the network’s economics, only a handful of competitors can afford to compete at scale, and each must engage in a grueling slog to attract new customers away from other providers with special enticements. At the same time, each provider hopes that its existing customers are too lackadaisical to switch to better deals and will remain on the network long enough to cover the system’s fixed costs over time.
Network costs do rise significantly when a broadband service provider needs to add capacity, however. This could happen because its existing customers have begun to transmit, cumulatively or at the same peak times, more and more data. Or the firm has recruited more subscribers in a location than the network had been right-sized for when first deployed. Or a provider geographically expands its network to reach new potential subscribers. Or a network replaces one medium for providing broadband with a technology that transmits data faster or more efficiently. If the broadband provider does not invest capital to expand capacity in such conditions, some unpleasant and familiar consumer effects will ensue. Not having enough fiber to transmit aggregate data at peak times will result in a slowdown or throttling of people’s online applications; hence the pixelation that occurs on streaming services at certain times. New people joining a network might increase the overall social utility its users experience—a provider might allow users on the same network to call and text limitlessly—but if too many join, the system can become constrained as commuters all attempt to stream the latest dross to their mobile devices on the subway home. Or if a rival network makes a big spend on new infrastructure, or simply makes a large ad buy, consumers might fear they are missing out on a better product. They may begin leaving their current firm if it does not spend enough to make the same type of splash.
Given these dynamics, broadband providers have taken on enormous debt to afford the table stakes of this particular competition. The telecommunications sector, as many investment analysts still quaintly define the broadband sector, is the most highly leveraged in the United States, other than financial services. Its collective debt-to-assets ratio is close to 50 percent. (Some of the marginal firms have debt that is below investment grade, a situation almost inconceivable for other utility-like companies; Sprint alone had $34 billion of sub-investment-grade debt in early 2016.) Indeed, broadband companies have such large debt loads that they represent one of the few sectors for which the 2017 Tax Cuts and Jobs Act’s negative effect—the partial loss of interest deductibility—actually may outweigh the reduction in corporate income taxes. They are also very much at risk of a downswing if interest rates rise substantially.
The big edge companies, on the other hand, have none of these parlous corporate finances. They have extremely low debt loads, and more cash than they know what to do with. In early 2018, Google had a debt-to-equity ratio of a mere 7.3 percent, and was sitting on nearly $102 billion in cash and other liquid assets; Verizon at the same time had only $13 billion in the bank, and it was certainly flusher than many broadband-provider rivals. If one were to judge innovation by the inverse measure of the amount of money a firm has sitting on the sidelines, uninvested, the verdict would starkly differ from the consensus that Silicon Valley is the font of rapid, unstoppable change. Notably, however, Google has one line of business which is a clear exception to its Midas touch: broadband service. Its Google Fiber entered the market with a splash before retreating dramatically in the last two years. Today, it serves only a handful of markets, and public reports indicate that the parent company’s leadership takes a dim view of its prospects.
Most policy disputes concerning broadband regulation should be understood as a contest between the big edge companies that ride through the fiber and the broadband providers that own it. The former are inextricably dependent upon the latter’s physical platforms; any limitations on the edge companies’ ability to operate over these networks threatens the edge companies’ value. As a result, they will advocate for policies which restrict broadband providers to the role of owning and operating infrastructure, while also seeking to prevent network operators from disadvantaging edge company content in any way. Meanwhile, broadband providers could conceivably rival the dominance of edge companies through vertical integration if they were able to leverage their networks to earn additional revenue from the sharing of information that they produce or curate—reportedly one motivation behind AT&T and Time Warner’s recent merger. The broadband providers will fight for this perquisite.
The vehement debate over net neutrality is an especially telling example of this dynamic.2 Net neutrality simply means that broadband providers should be regulated as the phone company once was: a common carrier of traffic, not permitted to prefer one actor’s data over others’, or to block any site’s data, or to “throttle” or slow it down relative to other data. Blocking, throttling, and fast lanes sound pernicious in a free and open internet, but the internet is not so freewheeling as it once was: the edge companies control so much market share that the retail experience of the typical user is already heavily intermediated—the user mostly sees what the edge companies would like the user to see, read, and buy online. Thus it is ironic that Silicon Valley is so enamored of net neutrality. The edge companies’ business models almost inevitably cut against the grain of “neutral” corporate behavior, as when Google blocked YouTube from Amazon’s streaming media devices in late 2017, or when the community standards of ubiquitous social networks are only selectively enforced.
Putting aside the sloganeering around net neutrality, it represents a clear policy departure from the way broadband providers are treated: recall that they do not receive the benefits typically bestowed on public utilities, such as official monopoly status or regulated rates. Net neutrality policies effectively regulate broadband providers as public utilities, holding that they should only be allowed to compete with one another around certain, cabined elements of broadband service, as the government defines it. And this redefinition of the potential role of broadband providers is not what these firms signed up for. It is at odds with their unique identity as risk-takers—what makes them distinct from energy utilities (which government frequently co-opts for policy ends, and which in turn are happy to be co-opted, so long as they obtain a regulated return) and from government-owned enterprises which provide a public-utility service.
To understand what is really being advocated by net neutrality, it is helpful to take its premises to their logical conclusion. In a net-neutrality world, competition between broadband providers aims toward a single function: the fastest, most interconnected network at the lowest cost, so that data can flow as unconstrained as possible and as inexpensively as possible. If that’s the desired end state, then there really is little point in having multiple broadband networks. Maintaining multiple networks is likely only to increase costs through uneconomic duplication, as discussed above, or to tempt providers into finding new, subtler ways of privileging their network, such as blocking the interconnection of other networks to their own. Some countries have chosen to comprehensively treat broadband as a public utility, but are honest about it. Net neutrality’s posture is pretextual in this regard.
The fundamental question raised by debates over net neutrality is thus whether we should move to regulating broadband as a fully-fledged public utility or to allow broadband providers a full range of motion in their business model. There are good reasons to argue for the latter. First, to the degree that competition in broadband is really sustainable, providers must be allowed to benefit from other sources of revenue or the prospect of growth in other product areas. We do not know what the next technological “leapfrog” will be, but we do know from past experience that having government rigidly ordain the permissible boundaries of a service can end up stifling innovation. In an earlier mode of telecommunications, government regulators were persuaded that the telephone system should be owned and maintained by AT&T. It was only when courts and the Federal Communications Commission rejected this argument that a market for more diverse services and devices grew, and the consumer experience became richer. Now we face a situation where, if net neutrality were to be adopted, we would close the door to as-yet-unknown possibilities for innovation on the part of rivalrous broadband providers. As noted before, other than causing firms to sharpen their pencils on operating and capital costs, this is really the only dividend that competition in the sector typically pays. Second, we should all worry about the social consequences of the dominant edge companies. But for those who doubt the wisdom of government regulation as a vehicle to rein them in, the best policy option is to let their only plausible rivals, the broadband providers, have a go at knocking them from their perch.
Net neutrality is founded on certain legitimate policy concerns about the consumer experience—concerns which are simply misdirected. Its advocates should be more preoccupied with the behavior of the actual intermediators of content, the edge companies, and less so with network infrastructure companies.3 The broadband providers, if anything, have shown too uncreative a preoccupation with building their network rather than leveraging it in a more diversified business model. Compared to the control exerted by the edge companies, the risks of more expansive broadband delivery models seem comparatively mild. In any case, consumers profoundly resist when, for example, network television stations fail to reach agreements to show certain content, such as the local football team’s games. There is every reason to believe that consumers would abandon a broadband provider that acts too smugly in giving preferential treatment to certain content. The paradigm of allowing the broadband sector, unconstrained by rate regulation or public ownership, to experiment at its own risk with an alternative business model should be encouraged, and not forbidden, by government regulators.
What would this experimentation look like? Already, some content is free for consumers of certain broadband providers even when it would ordinarily be available only on a paid-subscription basis.4 In such a relationship, the broadband provider is often paying the content-provider revenues for this product. Imagine a world where the tables were turned, and—realizing that the curator of content is really as much a consumer as a typical user—broadband providers began to charge Google, Facebook, or Amazon fees for their data traffic. Or to charge Netflix in those situations where the streaming video service’s data traffic consistently bumped up against the capacity constraints of a system, causing a slowdown that affected everyone. In such a world, edge companies could choose either to absorb the fee, charge more for advertising, or rethink their business models. Perhaps a dominant edge company would find this behavior on the part of a broadband provider so intolerable that the edge company would choose to play hardball and black out its service on that network in the expectation that consumers would punish the broadband company. Or edge companies might even decide to overbuild a new broadband network in an uncooperative provider’s heartland.
Net neutrality advocates will point to these scenarios as examples of the very ills they seek to avoid—and which broadband providers, seeking to immunize themselves, have also disavowed. But such a situation would not nearly be as unfair as it seems. Making streaming services bear the costs of incremental capacity to alleviate network constraints puts the onus on the actor responsible for the cost—the streaming service—to figure out how to remedy the situation. It could do so either by paying additional fees to the broadband provider, or by using a creative approach, such as siting its servers behind the constraint and closer to the consumer demand, thus alleviating network congestion. Either way, it is fairer that Netflix should bear this cost rather than each and every broadband consumer, who may pay the same rate to the network provider regardless of the amount of data they receive or transmit.
This type of pricing structure, where broadband providers would charge users on both sides of the data exchange, is actually similar to how the pricing of voice telecommunications worked when the service first became widespread. At that time, it seemed absurd that a residential telephone user should pay the same as an insurance broker on Main Street. The actual fixed costs might have been the same to reach either, but the social utility those players received from the public telephone network differed dramatically. After all, the insurance broker’s business was substantially founded on his ability to receive and place calls, while for the residential user the telephone might just be a new convenience. Prices were set accordingly. While net neutrality advocates fixate on the idea that this telephone network carried all calls in a nondiscriminatory manner, they miss the point that the network was paid for by those who obtained the greatest value from its existence.
The likely alternative to this kind of differential pricing structure for broadband is simply more generic and ham-fisted products offered by broadband providers, such as the all-you-can-eat or supersized data plan. But those offerings do not match the physical reality of the system, because you can always eat too much. Thus the fine print on many mobile data contracts permits carriers to throttle your speeds—albeit in a content-neutral way, mind you—if you have used too much data in a month or if the network is congested.
The real-world results of this content-neutral throttling have often been truly stupid. In August 2018, for example, in the midst of some of the largest wildfires in California’s history, Verizon throttled the data of the Santa Clara County fire department, reportedly slowing download speeds from fifty megabits per second to thirty kilobits per second after the fire department had used up its data allotment for the month. By the logic of net neutrality, any further data use—whether for data-intensive locational mapping of fires or the streaming of a droll cat video—had to be subject to the same throttling. We can test the policy logic of net neutrality in such circumstances with a simple hypothetical: How much data was being transmitted for the purpose of streaming pornography to Verizon mobile devices during the days and weeks that the fire department was responding to these fires? Would it be wrong for Verizon to throttle this traffic in such a circumstance, in order to maintain speeds for emergency services?
Yet when these events occurred, the fire department and California policymakers complained, ridiculously, that Verizon was violating the principles of net neutrality. Indeed, Verizon was doing just the opposite. Besides, this claim was almost wantonly hypocritical. The fire department would not tolerate lookie-loos at the scene of a disaster when emergency traffic needs to pass; Verizon should offer a fast lane to data traffic that has a high social value. The only reason carriers have not done so more aggressively is because of the net neutrality shibboleth.
Simply put, some data is more important than others. Even net neutrality advocates tacitly acknowledge this. For instance, Montana’s Democratic governor, Steve Bullock, issued an executive order decreeing that any broadband provider seeking to do business with the state of Montana would have to adopt net neutrality principles. At the same time, however, he had his own Department of Administration issue a request for proposals that required respondents to explain how the state’s data traffic would be “prioritized” in an emergency situation. As the governor found out, the slogan of net neutrality is a good one until it runs up against obvious practicalities.
None of this is to deny that broadband is important and worthy of protection. It is not equivalent to safe drinking water or reliable electricity supply, but it is now difficult to conceive of a world without broadband, and most citizens would regard it as being a necessary, utility-like service. The idea of the internet as an open place where any legal content can be accessed is likewise important, even if the idea that all data should be treated in a neutral manner is dubious. But in order to maintain a fast and reliable internet, we need to encourage the necessary capital investment in technologies that deliver it. Broadband providers unconstrained by either a regulation or policy expectation of net neutrality could start to rebalance the lopsided financial realities of the sector, wherein the ethereal products of the edge providers have catapulted them to the heavens, while some of the owners of this miraculous network are just one missed interest payment away from significant financial distress. Such a rebalancing would ensure the internet’s availability for its highest-value and most socially important uses. If other uses are crowded out to the detriment of the public, or if competition fails because the sector is increasingly consolidated, that is the moment for Congress, the Department of Justice, or the FCC to act—and not before, when these fears are largely imagined and when the competition between edge companies and broadband providers is not given room to play out.
Funding for Universal Broadband Access
If broadband’s economics are something like those of a public utility—and if there is also an advantage to keeping its private capitalization and competitive characteristics intact—then there are at least two circumstances where government intervention is useful: where no broadband exists at all, or where only one incumbent is present to provide it.
There are still many such places with these traits, at least in terms of square mileage, throughout the United States. These rural areas are usually supported by the Universal Service Fund, which Congress established and the FCC administers. Today, about $4.5 billion is contributed annually to the deployment and maintenance of broadband in these places, along with a contribution to low-income consumers and schools, libraries, and hospitals which brings the aggregate total to a little more than $8 billion.
This subsidy program was created to support basic voice telephony. The ability to place a 911 call or to contact one’s neighbors or do business by phone was considered, if not a right, then at least something for which public expenditures were warranted. Congress has established by law that “consumers in all regions of the Nation . . . should have access to telecommunications and information services . . . that are reasonably comparable to those services provided in urban areas at rates that are reasonably comparable to rates charged for similar services in urban areas.”5 Two decades ago, the subsidies provided for this purpose existed solely for the sake of voice telephony. But those subsidies began to be used, at first slowly, and by now almost everywhere and always, to install technologies that also supported broadband. By and by, the FCC disposed of the official fiction that this program exists only for voice; through a series of regulatory orders, the agency explicitly directed the program’s revenues to broadband deployment.
The results of this program have been stunning. Much of rural America has high-speed internet. Rural schools offer simulcast courses which students otherwise could never hope to attend. Telemedicine through high-speed broadband has increased levels of care significantly. Ranchers participate in cattle auctions via streaming service. If one considers broadband a truly meaningful conduit to the good life in modern society, and not merely a portal for entertainment, then these programs have delivered for rural America. They have not been without their waste and inefficiency—the FCC has iteratively reformed them to direct subsidies in a more efficient manner—but it is not an overstatement to say that rural America would be a shadow of itself without them.
Even as the FCC has modernized this program to focus on broadband, however, its funding source has remained the same: a tax on certain telephone calls. And it is no small tax. In the fourth quarter of 2018, there will be a 20.1 percent tax on those phone calls. Broadband subsidies continue to grow, but the revenue base which funds them has fallen to just over $50 billion from over $75 billion ten years ago, when, correspondingly, the tax rate was much lower, about 11 percent. In short, U.S. broadband policy is funded by a tax on something consumers are using less and less, requiring ever higher tax rates just to generate the same amount of revenue. Last quarter’s taxation represents the highest ever level of tax on this product, and one of the highest levels of sales tax that exists anywhere in America. To make matters worse, the tax falls heavily on legacy phone companies, such as CenturyLink, which have service areas so sprawling that they represent a kind of Ottoman Empire of broadband providers, and which are the weakest link in terms of universal broadband deployment. We are clearly taxing the wrong people.
On the other hand, edge companies, the principal beneficiaries of greater broadband interconnectedness, are virtually unaffected by this tax. They also, as we have seen, frequently cause broadband providers to make investments to expand capacity. Few policy changes would be more just, therefore, than to legally require these edge companies to contribute to the universal deployment of broadband. If one were to measure the data transmitted from edge companies over the network of each subsidized carrier, that would be a reasonable heuristic for defining a new tax base to fund broadband deployment.
Moreover, if the rules were set up in such a way that edge companies would actually be paying for broadband deployment, then (and only then) it would make sense to require that the broadband provider carry their traffic neutrally. If such changes were implemented, then broadband providers should be prohibited from extracting further revenue from edge companies after those companies have financed subsidies to pay for network deployment. This framework would provide broadband providers with a clear choice: make the market work and remain unconstrained; or acknowledge a market failure and receive subsidies subject to a much more onerous mode of utility-style rate and performance regulation.
The Future of the Broadband Industry
The United States faces a critical choice that will determine the future of internet policy. Either it turns its broadband industry into a regulated public utility, in which case a single broadband monopoly, or individual regional monopolies, will eventually emerge in a competition that exists only around scale. Or it allows broadband providers to experiment, to find out what consumers truly demand from their experience online. Edge companies, seemingly invincible until recently, would be wise to pursue a grand bargain with their host organism. The public policies they typically advance—alluring to the kids though they may be—are ultimately unfair and stack the competition in their favor. The American public only has use for innovators-turned-monopolists for so long, until they stop looking cool and start looking creepy.
Policymakers, meanwhile, should keep in mind that broadband is a public service, but that it does not need to be regulated like most utilities. What once seemed like two different sectors, telephone and cable, have already become one. At this point it would be prudent to keep the experiment going, let competition between broadband providers and edge companies play out, and then assess its value for the American public.
This article originally appeared in American Affairs Volume II, Number 4 (Winter 2018): 87–102.
2 The Obama-era Federal Communications Commission adopted “net neutrality” through its Preserving the Open Internet Order (2011), which the U.S. Circuit Court of Appeals for the District of Columbia vacated in Verizon v. FCC (2014), ruling that the agency did not have legal authority to regulate an “information service” extensively. The court suggested that if the FCC simply redefined broadband as a “telecommunications service,” the FCC would be within its rights to widely regulate it. Subsequently, the FCC did just that, opening a new administrative rulemaking and issuing the Protecting and Promoting the Open Internet Order (2015). This ruling left broadband awkwardly subject to statutory provisions adopted by Congress in 1934 for the purpose of regulating voice telephony. Like its precursor, this order was litigated against by industry. Meanwhile, Donald Trump was elected. His pick for FCC chairman, Ajit Pai, has led the Commission to reverse the 2015 order in the Restoring Internet Freedom Order (2018). It follows an allow-but-disclose model for conduct the Democratic FCC would have prohibited. As of this writing, a handful of states, most notably California, have passed laws that purport to impose net neutrality on broadband providers, although these state actions are likely preempted by federal law.
3 If net neutrality does become official government policy of the United States, then the policy itself should apply neutrally, both to the broadband providers who control the physical infrastructure and to edge providers who control the retail interface of the internet to consumers.
4 An even more extreme version of net neutrality would forbid even these arrangements. California’s net neutrality statute, signed into law in late September 2018, forbids “zero rating,” where broadband providers whose plans include limits on total data do not charge certain content against those data caps.
5 47 U.S.C. § 254(b)(3).