Warning – Extremely Long and Wonky Post Ahead.
No Federal Communications Commission Chair in modern history has so dramatically altered the wireless market in such a short time, and with such direct exercise of regulatory power, as Brendan Carr. Chairman Carr eliminated Echostar (the fourth largest wireless carrier) as a facilities-based wireless competitor, forcing it to sell its spectrum to AT&T and to SpaceX. He also approved the sale by UScellular (the fifth largest wireless carrier) of its network and a significant chunk of its spectrum to T-Mobile – a deal which the Department of Justice described as a “pivotal moment” in the wireless industry that would cement an “oligopoly” of the “Big Three” (T-Mobile, Verizon, and AT&T).
Chairman Carr has publicly explained his theory of spectrum management in a set of Orders and speeches. First, he argues that using spectrum efficiently requires amassing concentration of spectrum in the hands of a very limited number of companies. Where the Justice Department saw the concentration of spectrum from the UScellular/T-Mobile transaction as a blow to competition with potentially terrible consequences for consumer welfare, the FCC approval Order emphasizes that “we are charged with managing spectrum – a scarce and valuable public resource – in a manner that promotes its highest and best use.” Chairman Carr likewise justified his use of the FCC’s regulatory authority to force Echostar to sell its spectrum on the grounds that Echostar was not using its spectrum to capacity (a measure the FCC has never before used to take spectrum away from a functioning network actively trying to build a customer base). After Echostar announced the sales, he wrote to inform Echostar CEO Charlie Ergen that he “appreciated your responses, engagement, and information provided since my original letter” and therefore instructed staff to terminate their various proceedings against Echostar.
But Chairman Carr argues that he has not abandoned competition. Rather, in an about-face on 25 years of FCC reliance on “facilities-based” competition, the UScellular/T-Mobile Order appears to consider network operators dependent on leasing spectrum (such as cable operators) competitors to the national networks that control the spectrum. Additionally, in a recent speech, Chairman Carr boasted that satellite providers (particularly SpaceX) will provide “a new wave of innovation and growth in this sector” through developing “direct-to-satellite” (D2D) technology that “promises a fundamental shift in how we think about competition, connectivity, and convergence.”
While shocking from a traditional Republican (or even a modern Democrat) perspective, Chairman Carr fits well within the Trump administration’s peculiar mix of traditional Republican commitment to deregulation with adoption of the “national champion” strategy usually associated with “command-and-control” economies like China. In wireless, this includes a throwback to the 1980s, when the FCC gave free licenses to AT&T (after the AT&T break-up, the monopoly “Baby Bells”) and one randomly selected competitor to start the mobile industry rather than rely on competition, with some of the non-facilities-based strategies of the 1990s. Really, the only element of this strategy consistent with any 21st century FCC policy is relying on unproven, barely deployed technology and nascent competitors to provide “innovation” that will create future robust competition and thus justify deregulation today. (Which some of us uncharitably compare to budgeting based on your expected Powerball winnings.)
Nevertheless, to paraphrase Patrick Henry, if this be competition policy, make the most of it. If Chairman Carr expects consumers to benefit from returning to the old days when the FCC, rather than the competitive market, determined the “highest best use” of spectrum, he should then embrace the regulatory tools the FCC used to promote consumer welfare and competition in the 1980s and 1990s. We also recommend a serious expansion of 21st century spectrum sharing – just in case this national champion/spectrum consolidation thing doesn’t work out and we end up with the Department of Justice’s dystopian Big Three oligopoly.
How and Why Did Chairman Carr Radically Change FCC Policy in Less Than a Year?
The short version is that he eliminated both the fourth and fifth largest wireless carriers and completely changed the FCC’s underlying philosophy of the “public interest” which previously focused on facilities-based competition (where every competitor owns and operators its own network, including the spectrum holdings) to focusing on “spectrum efficiency” (by which he apparently means intensity of use). Furthermore, while Chairman Carr has merely presided over the sale and dismemberment of the fifth largest provider, UScellular, he actively used FCC investigations and threats to force Echostar (owner of Boost, the fourth largest competitor) to sell off its spectrum to AT&T (further increasing spectrum consolidation) and SpaceX (which represents a bet on new, future technologies to someday, maybe provide competition).
I’m going to put in a plug for this Public Knowledge white paper from a few years ago, which goes into detail on the history of spectrum policy, where this whole “highest best use” and “spectrum efficiency” language comes from, what worked and what didn’t, and recommendations for a better world. If I don’t include a specific link to a source, I’m pulling from there.
To understand just how radically things have changed, I need to review some history. Prior to the early 1990s, the FCC (like every other government in the world) assigned each spectrum license with fairly strict instructions on how you could use it, and assigned it to whoever they thought could best fulfill the purpose for which they created the license. This did not always work out well. In fact, it tended to work out fairly poorly. Why? Because it was a long, expensive process of proving to the FCC that not only would your new thing not interfere with existing uses, but that it was also actually sufficiently useful to be worthy of a spectrum allocation. For reasons I won’t elaborate on here, the FCC did not have a particular talent for magically predicting the future. The agency did not particularly prioritize competition when assigning licenses.
As the pendulum of policy swung toward “markets solve everything,” people increasingly got into the work of Nobel Prize-winning economist Ronald Coase – specifically, a paper he wrote in 1959 called “The Federal Communications Commission.” That paper proposed the radical idea that the FCC should simply auction licenses rather than set specific rules for how to use them (other than technical rules to avoid interference) or trying to select specific licensees based on who the FCC thought would make best use of the license – including letting people freely buy and sell licenses rather than doing a review (as required by statute) to see if the transfer served the public interest. Why? Coase proposed the following: (a) All methods of distribution are equally likely to be inefficient, since, as Yogi Berra is supposed to have said: “It’s tough to make predictions, especially about the future.” (b) Market mechanisms provide the best way to resolve inefficiency over time, since spending money provides incentive to return profit on the asset over time, and someone bidding on something should have some idea how they plan to use it and extract value. (c) If the person guesses wrong about their ability to extract value from the asset, that person can easily correct the mistake by selling the asset, whereas relying on the government to realize it made a mistake, reclaim the license, and try again is extremely difficult/impossible as a practical matter. This means that the market is not only the best means for initial distribution, but also the best way to cure any inefficiency.
In his article, Coase acknowledged there were a bunch of potential problems, including the concern that the wealthiest providers would monopolize licenses (and therefore the market) over time. Coase’s response was basically: “Let’s deal with that problem if/when it happens.” Well, it happened, in part because Coase didn’t have the benefit of understanding network economics, or any actual engineering knowledge about how you need to establish rules that drive use cases as part of creating licenses. Despite the risks, as we moved into the 1990s, lots of people wanted to use auctions to distribute spectrum licenses – but a number of people worried about how to make sure we could have a competitive market.
Congress authorized spectrum auctions in 1993 (and, because they generate lots of revenue, made them mandatory as a way of choosing between competing applicants in 1997, then started mandating auctions no matter what for designated spectrum). The first FCC Chair to design auctions, Democrat Reed Hundt, believed strongly in competition and wanted to make sure it happened. He did this by setting a strict cap on how much spectrum in a market any one company could hold. This ensured no one company could actually provision itself, and would need to make deals with other providers (“roaming agreements”) to provide decent nationwide service. Because everyone had to negotiate with everyone else, everyone had incentive to negotiate in good faith (and a sufficient number of market participants to negotiate with).
Fast forward to the Bush administration, and Republican Michael Powell becomes Chairman. He believed that only “facilities-based competition” – competition where every competitor owns their own networks (including spectrum) – was genuine competition, and competition that relied on regulatory interventions – like Hundt’s spectrum caps – was bad. I will skip the application of this to wireline and just focus on wireless. Then-Chairman Powell eliminated the spectrum cap and replaced it with a softer “spectrum screen” designed to ensure at least three players in every market, but providing carriers enough spectrum to self-provision. Unsurprisingly (to some of us at least), the market promptly consolidated.
Facilities-Based v. Non-Facilities Based
A brief aside on what “facilities-based” and “non-facilities based” mean in the telecom context, particularly wireless. Briefly (and again, highly simplified), to compete with and incumbent network, you can either build out a brand new physical network or lease capacity at a wholesale rate from an already built out network. Normally, we don’t require a business to open up its facilities to a competitor and require the already established business to lease out important elements – except as an antitrust remedy. But telecommunications (and a handful of other industries I won’t trouble you with) have an unusual quality. It costs a lot of money to build out the network before you can even start taking customers. It also takes a long time, and the incumbent can prepare and thwart competition in a bunch of ways. As a result, few (if any) people want to invest in building rival networks. So the primary way of getting competition is to require the incumbent to lease out capacity. This can be seen either as a temporary bridge to building out a competing physical network or as a permanent arrangement to create competition.
In wireless, companies that operate their own physical networks are “mobile network operators,” while those that lease wholesale capacity from an existing network and resell it are “mobile virtual network operators.” A rare case where a term in telecommunications jargon actually makes sense. So naturally, we have to confuse things by dealing with different flavors of MVNO agreement. Wireless providers that have a limited physical network use MVNO agreements outside of their footprint that we call “roaming agreements.” Cable operators use WiFi and their existing cable footprint, supplemented with MVNO agreements inside their territory and roaming agreements outside it.
A switch to facilities-based competition as the only form of “real” competition meant that the market had to consolidate. To effectively serve the entire country, companies need more spectrum. That means consolidation. And, for reasons I explain in the next section, once a small number of wireless carriers gain a substantial lead in spectrum licenses, it becomes extremely difficult for other carriers to catch up without some kind of regulatory intervention.
The Efficiency vs. Competition Argument
Since Chairman Powell, the FCC has relied on this theory of facilities-based competition. The primary difference between Democrats and Republicans has been the extent to which Democrats are willing to use the FCC’s regulatory power to ensure a minimum of four national providers. Why four? Traditional antitrust theory says that four providers is a moderately concentrated market, while less than that is “highly concentrated.” I will skip a lot of antitrust theory here involving network effects and conscious parallelism/tacit collusion, which makes this more complicated. Suffice it to say that the Department of Justice has had an explicit policy of maintaining at least four national wireless carriers since the attempt by AT&T to purchase T-Mobile. And even when the FCC, under Chairman Ajit Pai, seemed willing to simply let the T-Mobile/Sprint deal go through, the Department of Justice insisted on trying to set up a fourth facilities-based competitor by requiring various spin-offs to DISH (now Echostar) of Sprint’s pre-paid service, Boost Mobile, and enough spectrum to at least make a go of it.
Despite the decision to rely on facilities-based competition, Republicans wanted to make sure that we still had enough competition in the marketplace to protect consumers without the need for additional regulation. While I don’t agree with the theory, the general orthodoxy starting in the 1990s has been that, absent market failure, competition generally protects consumers by forcing carriers to offer good service at reasonable prices or lose customers to competitors. Regulators should either ensure sufficient competition, or regulate to limit the market power of the dominant firms. (Again, I am not saying I agree with this. Regulation is often necessary even in the presence of competition.) You can see that the Department of Justice still believes in this when it decries the end of four-firm competition and the creation of the “Big Three oligopoly.”
But the FCC looks at spectrum management through two lenses. In addition to competition, the FCC also wants to make sure spectrum is being used to its maximum potential, often described as “maximizing efficiency” or even “putting spectrum to its highest best use.” As the Public Knowledge white paper I referenced above observed, however, the problem with “spectrum efficiency” is that no one really had a good definition of it. But Chairman Carr appears to think that he has a definition – intensity of use. At least, that is how he has justified threatening to take Echostar’s spectrum away unless they sell it off to someone who will use it “better,” and the justification for letting T-Mobile (and presumably AT&T and Verizon when the agency gets around to approving it) buy UScellular’s spectrum.
In a few short months, Chairman Carr eliminated both the fourth and the fifth largest carriers, apparently on the theory that it is more important to have only three national carriers that use the spectrum “efficiently” (meaning intensely) than to have more than three (traditional) competitors. And, as the Department of Justice observed in its statement on UScellular/T-Mobile, we had already reached a point where it was effectively impossible for a competing wireless carrier to acquire enough spectrum to compete effectively even if they were to win every upcoming auction.
It appears, therefore, that we not only have the highest level of concentration in the wireless market since former FCC Chairman Reed Hundt designed spectrum auctions to produce competition in the 1990s, but we will also stay that way as a matter of physics and not just economics. Absent spectrum divestitures (which are not going to happen), three shall be the number of thy national wireless providers, and the number of thy national wireless providers shall be three. Thou shalt not have four – and five is right out.
Over a decade ago, I wrote about the conflict between “spectrum efficiency” and competition. See here, here and here. Briefly (and as the Department of Justice notes in its UScellular/T-mobile statement), the bigger the wireless company, the easier it is for it to extract value from the spectrum due to economies of scale. It also gains competitive advantages because – unlike most resources – access to high-power spectrum licenses capable of supporting a quality mobile network is a zero-sum game. So as the winning carriers get bigger, they are better able to win licenses at auction or buy them in the secondary market. Absent any sort of regulatory intervention (either by antitrust or by FCC spectrum cap), the largest companies will inevitably wipe out rivals by being more efficient users of spectrum. Or, in other words, spectrum competition requires a level of spectrum inefficiency, since one carrier with all the spectrum would see the most intense use.
If viewed solely through the lens of spectrum efficiency without regard for market power, Chairman Carr’s strategy appears totally logical. The firm with all the customers uses the spectrum most intensively, and is in the best position to achieve economies of scale that keep it efficient. Mind you, on the consumer side, without effective competition, there is no reason to pass these efficiencies on to consumers. Hence the previous idea that you needed a minimum number of competitors even if that means some trade off in spectrum use efficiency.
What Should the FCC Do Under the New Policy To Protect Consumers?
Our previous answer to the question of “what do you do with a problem like an oligopoly” is either to break it up or to tightly regulate it. If neither of these seems attractive, you need to change the definition of competition or count on some form of future competition developing. Not surprisingly, Chairman Carr has opted for this third, and sadly traditional, option: Pin hopes on potential future competitors while lowering the bar for measuring success.
The problem with allowing consolidation today based on the emergence of competitors tomorrow is that – especially without regulatory intervention to protect emerging competitors – tomorrow never comes. This makes Chairman Carr’s exuberance over the potential for satellite “direct to cellular” (D2C) or more broadly “direct to device” (D2D) extremely dubious. Not that we won’t see more D2D generally, or that SpaceX won’t be able to extract value from buying Echostar’s spectrum. But it seems highly unlikely, given the technical and spectrum capacity constraints, that SpaceX can create an independent wireless mobile service in the manner Chairman Carr appears to envision.
What about cable operators? While they have shown some success at capturing wireless customers, they rely primarily on their broadband footprint where they can leverage their existing infrastructure, which imposes real geographic limits. They also (at least at the moment) only sell their wireless service in a bundle with their landline service, limiting their potential customer base further (and at a time when most major cable broadband providers are losing customers). Finally, they lack licensed spectrum and need to rely on MVNO agreements. Because the FCC does not regulate MVNO agreements, the incumbents generally set terms that make it difficult to compete with them.
Could a regional player become a major competitor? If UScellular – the largest remaining regional provider – couldn’t do it, it seems unlikely any other regional provider could. The Department of Justice concluded that UScellular could not hope to expand beyond its rural, regional footprint because of the shortage of spectrum. How about Boost Mobile, which will continue to operate independently by using AT&T’s (and hopefully SpaceX’s if the tech works out) spectrum, but maintaining its own network? Again, this requires betting against the odds in a hostile environment where the Big Three have almost all the existing customers as well as almost all the existing spectrum.
So if we want a competitive market so that carriers are forced to pass on the efficiencies to consumers, and we don’t want monopoly (in this case, oligopoly) retail price regulation, the FCC needs to have a strategy to replace “facilities-based competition” with four national providers.
I’m not going to discuss the pros and cons, including the potential statutory barriers, to these suggestions. I also don’t think these are necessarily the only approaches the FCC could take beyond magical thinking that the agency can simply wish competitors into existence. Rather, proposals either: (a) make more spectrum available to competitors; or, (b) make it easier for customers to switch between providers; while, (c) creating the framework for possible new entrants and working with the handful of existing players in the market who might still grow into real competitors.
A. Wholesale Access to The Big Three’s Spectrum.
Creating competition going forward requires abandoning the theory of facilities competition entirely – not simply where convenient – and embracing parties that rely on others for spectrum access or even network access. Traditionally, we call these “mobile virtual network operators.” MVNOs traditionally lease not just access to the spectrum, but access to the underlying network of a facilities-based mobile provider. MVNOs have existed since the mobile industry began to seriously expand in the 1990s. It mirrors the first efforts at competition in the wireline world in the 1980s, where the FCC (and later Congress in the 1996 Act) required the local monopoly telephone company (“incumbent local exchange carrier” or “ILEC”) to release components of its network to rivals (the FCC in the 1980s did this for enterprise services and information services, such as the dial-up Internet. Congress in the 1996 Act expanded it to would-be providers of local services.) Chairman Powell killed most of this in the early 00s as part of the insistence that only facilities competition was “real” competition, and that forcing carriers to lease facilities at a rate low enough to support competition was regulatory arbitrage and blatant interference in the free market.
MVNOs provide thriving competition in some places, such as Europe, which never adopted the idea that only facilities-based competition was “real.” Where national governments have created rules explicitly to support MVNOs, such as access to facilities-based providers that allows for competition on price and unique service offerings, MVNOs flourish. Until now, the U.S. has eschewed this approach under its “facilities-based competition only” ideology. But it could take steps to affirmatively promote the usefulness of wholesale access to the spectrum. This would not only further competition, but increase the efficiency of spectrum use – the purported reason of the Commission for its recent actions.
In the United States, the inability to get access to spectrum/networks on terms that would let MVNOs offer competing services at lower prices have kept MVNOs niche players for low-income or otherwise price-conscious customers. Through participating in the FCC’s Lifeline program and offering a smaller bundle of services with more aggressive bandwidth caps, MVNOs can offer a low-cost option with lower-margins than the Big Three generally find attractive. But even here, we have seen a distressing trend of the largest facilities providers buying up the largest MVNOs, limiting potential MVNO competition even further.
The FCC could take steps to make spectrum access more affordable for rivals, and prevent any obstacles imposed by contract from offering directly competing services. This could allow the “hybrid” network operators such as cable and Echostar to offer genuinely competitive wireless products on a national basis. It could also potentially allow the remaining regional players to avoid UScellular’s fate.
B. Make it Easier To Switch Networks/Enhance Contract Transparency.
Another way to promote competition is to make it easier for customers to switch from one network to another. The FCC has an open rulemaking on whether to impose on mobile operators a standard rule on unlocking cell phones. Right now, the cell phone networks have varying policies and can use various tricks (such as keeping you on hold indefinitely and bouncing around their phone tree when you ask to unlock your phone) to make phone unlocking hard enough to undermine competition. Indeed, the fact that rules for unlocking vary from carrier to carrier creates confusion and imposes another barrier to switching.
There are other ways the FCC could force providers to make it easier for customers. For example, the broadband, texting and voice part of every wireless service has different rules on what the carrier must disclose to the customer, and how to present that information. This makes comparison shopping extremely difficult.
The easier for customers to switch between providers, the more competition you get. This was the whole point of T-Mobile’s “uncarrier” strategy. The fact that T-Mobile ultimately had to buy Sprint to reach the same size as AT&T and Verizon demonstrates that even the most determined and sustained marketing effort by a competing carrier can only go so far. If the FCC wants to see real competition, it needs to make it possible for consumers to vote with their feet.
C. New Models of Spectrum Sharing.
I’m not going to try to explain unlicensed spectrum (which most of you know from its most popular applications, Wi-Fi and Bluetooth) or the Citizens Broadband Radio Service (CBRS). For purposes of this exercise, I will simply observe that these two methods of sharing spectrum (and thus, massively increasing overall efficiency as measured by intensity of use) have become important elements of cable operator wireless networks. Expanding the availability of these spectrum sharing models, or inventing new ones, creates new opportunities for would-be carriers blocked from traditional high-power licensed spectrum.
Satellite operators also rely on shared spectrum, but using a different model. As Chairman Carr mentions in his speech on how spectrum fits with his “build agenda,” the FCC currently has before it an open proceeding on how to create “spectrum abundance” for satellites. If SpaceX (or other satellite providers) need access to terrestrial spectrum to offer services that compete with, rather than supplement, the Big Three’s mobile offerings, then the Commission should consider how the history of satellite spectrum sharing can apply to terrestrial uses.
Conclusion
Chairman Carr has completely abandoned the underlying theory by which the FCC regulated the wireless industry – facilities-based competition – and embraced spectrum efficiency as the primary goal of FCC spectrum management. In service of this ideology, he has directly intervened in the wireless market at an unprecedented level. Over the direct objections of the Department of Justice, he has presided over the dismemberment of the fifth national wireless carrier without any conditions to protect competition or consumers. Meanwhile, he has forced the fourth largest national carrier, that the Department of Justice took such pains to create after the T-Mobile/Sprint merger, to sell off its spectrum licenses.
But if spectrum efficiency is a genuine public interest strategy rather than an excuse for consolidation leading to an unregulated wireless oligopoly, the FCC must take affirmative action to make competition possible in a way consistent with Chairman Carr’s theory of spectrum efficiency. As discussed above, there are rules and policies the FCC can adopt which will, over time, create the possibility for competitors that are not traditional facilities-based competitors. Having demonstrated a willingness to use the power of the FCC to eliminate competitors, it would be the height of hypocrisy to refuse to use the same regulatory tools to promote spectrally efficient competition and protect consumers.