I recently wrote the following paper as a final for my Antitrust and IP course. While not the best paper ever written, I discuss a couple of elements that Google seems to be prepared to use in invoking antitrust to protect net neutrality. Hope you find it interesting! If you'd like to see the original paper - drop me a line at william(at)finkel.com.
Warning, this is a long post so I cut out much of the introduction!
Footnotes didn't transfer when I put this up, but I can provide citations if needed.
Any feedback is greatly appreciated!
Using Antitrust to Ensure Network Neutrality
How Local Broadband Monopolies Exist
While there is no per se monopoly over access to the internet, there does exist a government sanctioned (created) local monopolies over broadband (high speed) access in the form of cable and telephone operators. At current time, the only way that the vast majority of American internet users have to get access to the internet is through their local cable providers or telephone operators. Under the Telecommunications Act, the FCC was given the right to label different types of operators under the labels (silos) of telecommunications service providers and information service providers for the purpose of deregulation. For those designated as telecommunications service providers, there was a requirement that they open their lines up to competition, resulting in the ability for telephone customers to choose from a variety of providers to offer them phone service, even though their lines are privately owned. Meanwhile, those designated as information service providers (as in cable operators) were allowed to keep their networks private as their value was seen to be their content not their network.
For the purposes of high speed internet, the FCC designated ISPs as information service providers. While this designation resulted in a lot of debate over the nature of the internet (do consumers purchase internet access for the content they are given - like a cable company- or for access to a general network - like a phone company?), the Supreme Court held that these questions were largely academic. In the Supreme Court's recent ruling on Federal Communications Commission v. Brand X case they determined that the accuracy of the classifications was immaterial to the case as long as the legislature had determined that the FCC had the right to designate these classifications - the courts did not have the right to question the classifications if they were done within reason (as per Chevron ). By giving the cable operators the right to maintain their own private ISP, the government granted an extension of their cable television monopoly to high speed internet. Beyond granting a simple monopoly, the government also opened the door to our fears over monopolies, namely, they allowed the cable companies to adopt measures that would have been rejected had there been competition but could now be tried in anti-competitive environment. In an environment of pure competition, the ISPs would never be discussing limiting network neutrality as a competitor would step in and offer neutrality as a selling point. However, in the absence of competition, the ISPs are free to maximize profits at the expense of performance, in terms of the sites accessible from their network.
ISPs have a two-fold opposition to network neutrality. Their primary complaint (likely because it is more sympathetic and legally sound) is that the obligations of having a neutral network force them to carry the bandwidth of other sites slowing their network and stifling their own innovation. Their second claim is that an open network allows other companies to piggyback on the costs of maintaining a network, and to profit off of that relationship. To deal with their opposition, the ISPs have two main options: 1) limit traffic from other ISPs, which would more likely slow traffic rather than stop it or 2) block certain types of high bandwidth sites and programs altogether. Either program would likely include some sort of allowance for paid relationships whereby sites could get higher performance in exchange for compensation.
Potential Legislative Controls
In late 2005, the FCC issued a set of guidelines in a policy statement `to encourage broadband deployment and preserve and promote the open and interconnected nature of public Internet' (1) consumers are entitled to access the lawful Internet content of their choice; (2) consumers are entitled to run applications and services of their choice, subject to the needs of law enforcement; (3) consumers are entitled to connect their choice of legal devices that do not harm the network; and (4) consumers are entitled to competition among network providers, application and service providers, and content providers. These guidelines were not passed as rules but merely as benchmarks of behavior and have been used by the FCC in at least one instance to ensure network neutrality (Madison River , discussed below). The problem is, as guidelines rather than rules their enforcement may well be arbitrary, and given the FCC's reclassification of broadband - they seem to have relinquished the right to enforce these guidelines. Additionally, FCC guidelines do not give rise to a private action should the FCC fail to act. In, Trinko v. Verizon, Trinko claimed that Verizon did not perform their obligations under the 1996 Telecommunications Act to allow competitors to use their phone lines by not helping AT&T to use its support system in aiding their clients. Hailing the precedent of Colgate the court found that it was not the company's obligation to help competitors to work around its existing monopoly and the mere presence of the Telecom Act did not mean that all acts of Verizon had to be anticompetitive - just that they must follow the letter of the agreement. Colgate gives good notice of what the telecoms will claim in defending their actions - they are under no obligation to aid competitors.
Currently, the federal government has been meeting to discuss legislation to ensure network neutrality. As always, a variety of factors will go into any governmental decision to regulate. One large concern on all sides is the unintended consequences of legislation and given the internet's history of self-regulation and unfettered development there is serious reluctance to start building a federal structure. The question of what form federal legislation might take is very undefined and is unlikely to be resolved in the near future.
Eliminating Network Neutrality
If the ISPs engage in limiting traffic as a means to ensure network quality - the question of intent and exclusionary tactics will surface. While the ISPs are entitled to take steps to ensure the quality of their network, those steps must be measured against the consequences that the actions will have upon the users of its networks and the anticompetitive consequences of such actions. Taken on its face, the question will be - the ISPs stated goal in changing its rules (whether to ensure the quality of its network or to ensure they get paid for their resource) could have been accomplished without being so damaging to the competition.
If the goal is to maintain the speed of the network we would quickly degenerate into a very tech heavy discussion of whether and how to best design a network. The reason why this claim is unlikely to be used is that if the goal is to improve the performance of the network - the less restrictive means would be to pass the decision to the consumer about where to allocate their own broadband access rather than making the decision for them. By making the decision in the form of pay-to-play agreements with individual websites, the ISPs will essentially limit the ability for small companies to emerge that can't afford their contracts and limit the ability of consumers to decide what services they want. In many situations, these small companies may not be direct competitors of the ISPs today, but will have the potential to become competitors in the near future in all sorts of media, while slowing their access to the customers will preclude them from ever challenging.
However, there is no guarantee that the consumers have a right to choose once they opt to join the network. The most troubling potential is that with prioritizing certain websites this could give the ISPs the ability to prioritize their own sites. A good example would be the potential for misuse with AOL/Time Warner. Time Warner Cable offers broadband internet service to a large number of consumers through cable internet. They also are the owners of a large network of websites under the AOL moniker. If Time Warner were to begin prioritizing which websites their subscribers can use (or in what volume), there is the potential for abuse. What makes this danger more likely is the fact that, due to the silo system of classification, the true nature of ISPs has been been misclassified as information services. As an information service, the ISPs claim that their service is in the tools they provide online not in gaining access to the internet. Since their business is in providing tools, they would claim that in prioritizing their own network they are not bundling products to create a new monopoly, but rather enhancing their core business.
The other option available to ISPs would be to block certain types of applications altogether. Currently, the highest bandwidth users are streaming content - whether games, video, music, or telephony. What makes this situation most interesting is that the most offensive (in terms of bandwidth consumption) sites are also the sites that are after the same market as the ISPs core businesses (telephony and television). As such, there is an unstated content-driven opposition to the high bandwidth sites, as well as the size driven opposition. For this reason, most claims about program blocking would arise under Section 2 of the Sherman Act and make for interesting cases.
There have already been several cases of ISPs blocking or restricting access to prevent competition or stifle innovation. AT&T previously had a user agreement prohibiting WiFi as a form of theft , while Madison River actively blocked its users from using its broadband services for Voice over Internet Protocol .
Madison River as a Case Study
A good example of where antitrust litigation might have been more effective than regulation is the consent decree issued by the FCC and Madison River. Madison River Telephone Company, LLC is a telephone company that also offers broadband internet service. In early 2005, the FCC launched an inquiry into whether Madison River had violated section 201(b) of the Telecommunications Act. Section 201(b) states that `all charges, practices, classifications, or regulations for and in conjunction with ... telecommunications service, shall be just and reasonable, and any such charge, practice, classification, or regulation that is unjust or unreasonable is declared to be unlawful'. The specific charges were that Madison River had configured its servers so as to not permit its internet subscribers to use a service known as Voice Over Internet Protocol (VOIP), which is essentially an internet telephone. In order to avoid a drawn out proceeding, the FCC and Madison River reached an agreement whereby there was a consent decree and order issued and Madison River paid a fee for the transgression.
For advocates of network neutrality, this case can be seen as a failure in three ways. First, the settlement was for $15,000, which is hardly a large sum for a telecommunications company and pales in comparison with the treble damages that would have been appropriate in an private antitrust suit. Another problem is that the settlement went to the FCC rather than the wronged parties, which in this case consisted of the VOIP providers that were blocked from a market and the end users who were forced to pay Madison River's phone prices rather than the market price that the VOIP would have offered. Additionally, the FCC appears to have made this low settlement due to the fact that the viability of their order was not clear.
At the time of the settlement, the FCC's silo system was a bit more convoluted than it is now. Cable operators who offered broadband were classified as information service providers while telephone companies who offered DSL were classified as telecommunications services. As such, telephone companies were subject to Section 201(b) while the cable companies were exempt. In the wake of the Brand X ruling, where the Supreme Court validated the FCC's designation of cable operators as information services, the FCC has revised its classification of the phone companies to information services for the purpose of internet access. Due to the reclassification, the FCC could no longer bring a charge in the manner it did and likely knew that its own regulatory system would be called into question had it not settled. However, the reclassification has likely opened the door to antitrust resolution of this issue in future situations.
When a regulatory system existed to ensure that 201(b) was followed, a private litigant may not have had a claim as there have been precedents of the court ceding enforcement to the FCC. An antitrust suit against Madison River could have taken two different forms (as mentioned above), there would have been the right of action by a specific VOIP provider who had been wronged by having a market closed to them and there would have been a class action liability by the Madison River customers who paid inflated prices. Both cases would have focused on Section 2 of the Sherman Act, where a company has acted to restrict or control market forces, however the class action suit would be harder to bring as the standing of the parties is not always clear. The VOIP provider's suit would have focused on the purpose and scope of Madison River's actions.
Madison River would likely claim that their actions were taken to maintain the viability of their network - however this claim is rather thin. Using the least restrictive means test, the court would look at the fact that the company specifically blocked the ports that were used by VOIP service rather than an overarching restraint on traffic. As the restraints that were used specifically impacted a competitor to Madison River's main business, their actions can be clearly seen as an attempt to use one business (the ISP) to maintain a separate monopoly (the telephone company). As such, they would likely have lost this case, with damages (and deterrent effect to other companies) far exceeding those from their FCC settlement (which has since been rendered moot).
The actions taken by Madison River (which should be noted is a rather small provider) are exactly the kind of actions that network neutrality advocates fear. VOIP is a nascent technology that is cost-effective and reliable, yet undermines an existing monopoly. For current telephony companies that offer broadband services, VOIP would be among the first targets to go should they reject network neutrality.
Potential Litigation - Television and Telephony
There are two potential fields where we could see antitrust proceedings arise, given the core monopolies of the ISPs and broadband's threats to those monopolies. These fields are television and telephony.
Due to the recent deregulation of telephone ISPs, the FCC has opened up the potential for the telephone operators to enter the television market. AT&T has expressed a strong interest in entering this market and as such has developed plans to develop a higher speed data network (called Project Lightspeed), which reportedly will involve rolling out a lot of new fiber and switches. One of `Project Lightspeed's' main goals is to help AT&T distribute IPTV (Internet Protocol Television), a product that they see seriously competing with the existing cable television operators.
AT&T is claiming that the costs of developing this network are the reason that they must start charging network users for the right to higher speeds. They also make the claim that if everyone can use the higher speed `lanes' in their service then all data will be slowed down. However, if AT&T is allowed to use the higher speed lanes for their IPTV while competitors or potential competitors are forced to pay for access to these lines we are looking at a fairly clear antitrust violation.
A good potential litigant would be Apple Computers . Apple makes revenue through several businesses, but of most interest here is their iTunes store, which sells TV shows and short videos for download. The iTunes Store is a huge business and would be seriously endangered if there were to be faster alternatives. If AT&T were to charge for the right to access the highspeed `lane' in their new network, Apple would be faced with the decision of whether to pay the fees or suffer a loss of business. The potential antitrust suit would arise from the fact that AT&T's IPTV product would stand to benefit from the higher speed network, at the expense of it's potential competitors.
As AT&T progressed with their tiered program, Apple would have to seek an injunction against implementation of the program. In seeking injunctive relief, Apple would have to demonstrate that AT&T had a monopoly power in the field and that they have engaged in exclusionary conduct. The market power could be easily proved on a geographic basis as there are dozens of municipalities where AT&T's broadband DSL is the only broadband service available. This would establish AT&T as a monopolist.
Next Apple would have to define AT&T's behavior as exclusionary. Using the Court of Appeals' standard laid out in the Microsoft case, the plaintiff must show that the act has an `anticompetitive effect' and that any pro-competitive benefit of the act is outweighed by the anticompetitive harm of the act. This is where the analysis could get particularly difficult for the courts, as they will be forced to assess an unknown quantity (IPTV) and its positive economic consequences against the negative economic impact of essentially having Apple subsidize the product. As was discussed in United States v. Grinnell Corporation, there is a higher bar to bringing a suit to prevent establishment of a monopoly as opposed to maintenance, "the willful acquisition or maintenance of a power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident."
Furthermore, the establishment of IPTV as a monopoly unto itself would be refuted by AT&T, as they will point to the existing cable television as the true monopoly.
Perhaps a more clear cut way to assess the problem would be a situation where the monopolist was trying to maintain a monopoly, rather than develop a new monopoly. This would be more apropos of Time Warner. AT&T's `Project Lightspeed' is an innovation (higher speed network) made with the specific intent of creating a new market. Any charges associated with access to the bandwidth of `Project Lightspeed' could be explained as necessary to the procompetitive innovation. Time Warner, however, is already in the business of distributing television. If they were to face off with Apple in a similar antitrust case, the analysis would differ.
If Time Warner were to introduce a tiered access program, they would not be doing so to expand into a new market. They would (in theory) be enhancing an existing service, and charging a fee to use these enhanced services. However, by charging a fee within their broadband monopoly, they would essentially be creating a hurdle that content providers (such as Apple) would have to pass over that their other monopoly (cable television) would not be required to surpass. This case would seem to be a pretty clear violation of Section 2 of the Sherman Act.
Assuming that Apple were to receive injunctive relief in either case, this would not permanently solve (nor protect) all websites from tiered access. The protection that Apple would receive would be due to its status as a direct competitor in television services. As such, the injunction it received would only protect direct competitors of the broadband companies and might only apply on a network-by-network basis. For instance, Apple also sells music through its iTunes store. If AT&T didn't sell any music through `Project Lightspeed', then theoretically they could charge Apple for higher speed access to their music content.
One reason why a few strategic antitrust suits might succeed at derailing all of the tiered access program is that the stated need for tiered access is that the fastest broadband is a limited supply and if there isn't a way to discriminate among who gets to use it, then it gets devalued as all who use it are slowed down. If this is the case, then the injunctions will essentially eliminate the value of the tiered access - as the high speed `lane' will be accessible to all who are potentially effected in an anticompetitive manner slowing down the speed of the high speed `lane', and devaluating it as a service.
Another field in which these antitrust suits might gain traction is in VOIP cases such as the Madison River case. This field holds a lot of potential for antitrust litigation, because (akin to Time Warner maintaining a monopoly over television) the monopolists do not seem to be innovating in the field of telephony so there is not the same claim that AT&T had of creating a new monopoly. Instead, there is just the maintenance of an existing monopoly, a much less persuasive argument.
Legal Network Restrictions
While Madison River is a case of ISP violation of the Sherman Act, there do not seem to be many examples of ISPs restricting traffic while staying within the law. There is a gray area that exists around blacklisting of domains and will surely come to the legal system as ISPs become more aggressive in defending their networks.
Another means of restricting traffic that might work would be hourly restrictions so that during high traffic times of the day, sites don't monopolize an ISPs network. While this does appear to be a less restrictive means of ensuring good service, there is also the potential for problems here.
If an ISP is offering a video service hosted within its own servers then it would not be subject to traffic restrictions, while a competitor from outside of the network would be subject to said restrictions. The key question here would be whether the courts determine that this is a form of tying that treats the ISPs service more favorable to the competitor, or whether they consider all internet services offered by the ISP as part of their core service. This question arises due to the FCC's flawed classification system and perhaps points at the need for an overhaul of the Telecommunications Act of 1996 to include a silo for ISPs.
Short of that happening, the most legally sound way for the ISPs to maintain the integrity of their networks without going afoul of antitrust law would perhaps be a system where they charge the end user for their traffic usage.
Conclusion
While much uncertainty exists as to federally mandated network neutrality, the ISPs must tread lightly as they explore their options in creating a new system. The case of Madison River is exactly what the ISPs are believed to desire out of a selective network, and is exactly where they cross the line into restraint of trade. There are definitely legitimate cases to be made on behalf of the ISPs regarding a desire to safeguard their networks, yet they must be wary not to use the power they have under the Telecommunications Act to not violate the Sherman Act, and stifle their competitors - as such interference leaves them open to antitrust litigation.
While the exact shape and scope of a tiered network remains to be determined, there are numerous red flags that the prospect of tiered access raise. From using the internet to create new monopolies, to using it to protect their existing monopolies, the media conglomerates have amassed unprecedented power, and their diversity is now also becoming a liability. Perhaps more than just a network neutrality issue, this analysis shows the inherent antitrust problems that are bound to arise from the conglomeration of media entities and the abdication of regulatory duties by the Congress and the FCC. This may be a case where deregulation's harmful effects on citizens, will be counterbalanced by the protections afforded by the Sherman Act.