Network neutrality is in jeopardy -- just not in the way you might have heard.
At this moment, with the exception of one company subject to merger conditions, broadband access providers in the United States aren't legally prohibited from blocking competitors' content, arbitrarily degrading unaffiliated services, favoring their own content artificially or offering prioritization deals solely to favored partners. Any of these practices would cause damage to the innovative dynamism and openness of the Internet. And there's reason to think that, without regulatory oversight, those harms might be realized. After all, they have been before.
Federal Communications Commission Chairman Tom Wheeler recently circulated a proposal to his colleagues that reportedly would ban all of these practices. (As of this writing, the proposal is not public, although Wheeler gave a general overview in a blog post.) The proposal reportedly would establish ongoing FCC review of agreements between broadband and content providers, to ensure that better performance is offered on reasonable terms. And it would beef up transparency obligations to shine a light on attempts to create an artificial slow lane or manipulate network management anti-competitively. It would do all this on a legal foundation recently validated in court. (The agency's two prior attempts in this area were overturned on jurisdictional grounds.)
In short, the FCC is apparently considering a proposal to make a form of network neutrality the law of the land. So why is there an explosion of outrage that the FCC is gutting, killing or abandoning net neutrality?
The answer has to do with that old saw about the perfect and the good. The FCC proposal doesn't prohibit one category of business agreements that could be abused. And it doesn't use the simplest and most expansive legal theory. That has network neutrality advocates up in arms. On the other hand, Wheeler's approach has something significant going for it: It might actually stick. In the political morass of today's Washington, that factor shouldn't be minimized. And it could be an effective model for protecting ongoing Internet-based innovation.
The concept of network neutrality (or in the FCC's preferred parlance, the open Internet) draws upon technical and economic insights about the amazing success of the Internet as a catalyst for innovation, free speech and economic growth. The basic idea is that network operators should not discriminate against certain content, applications, services, or devices. The significance of this principle for protecting innovation and tomorrow's upstarts is increasingly evident -- the growth of Internet companies over the last twenty attests to its importance. In the words of Internet Protocol co-creator Vint Cerf, network neutrality means "the creators of new Internet content and services need not seek permission from carriers or pay special fees to be seen online."
Net neutrality highlights vitally important aspects of the way the Internet works today. Other than the FCC's recent, failed attempts, however, it has never been a legal mandate, let alone a technical requirement. Thanks to the extraordinary work of scholars and public interest advocates over the past decade, the basic principle of network neutrality is now embraced by the technology industry, the FCC, and even many broadband providers.
To be sure, there is still disagreement at the edges about how to implement network neutrality: How should the FCC balance the benefits of network neutrality against the risks of excessive involvement in business decisions, and the potential to chill capital investment in infrastructure? How exactly should it write and enforce the rules to ensure that tomorrow's upstarts have the necessary freedom to innovate without permission?
In implementing network neutrality, some differentiation of traffic must be allowed on the Internet, even encouraged. Network operators should be able to block malware and denial of service attacks, for example. And companies like Akamai and Limelight should be able to offer content distribution network (CDN) functionality to improve delivery of certain content. There are many thousands of interconnection agreements among network operators, backed by massive amounts of traffic engineering, caching overlays, and other mechanisms to meet business and technical parameters. If the Internet story is one of a neutral network, it's also one of private firms negotiating arrangements and managing their networks as they see fit.
Network neutrality advocates acknowledge all this. They propose various rules that distinguish reasonable from unreasonable discrimination, just as the FCC is doing. So the real question is whether Chairman Wheeler's proposed rule is so much worse than what came before. Under what we understand the FCC proposal to be, access providers can't block, can't degrade, can't arbitrarily favor certain applications, and can't favor their own traffic. If you read the past decade of network neutrality literature, those are the dangers usually mentioned. So what's everyone so mad about?
The major change in the new proposal concerns so-called paid prioritization agreements. In other words, the new rules appear to allow a broadband provider to offer content providers the option of faster or more reliable delivery for a supplemental fee. Under the old rules, the FCC didn't prohibit such deals, but said it was skeptical they would meet its discrimination test.
In the new proposal, the FCC appears to mandate that paid prioritization offerings be "commercially reasonable." This requirement presumably would insist on the availability on the same terms to all, with the FCC reviewing such offerings on a case-by-case basis. Such a requirement might also include the condition that any paid prioritization offerings are only reasonable when the broadband provider offers a sufficiently robust level of non-prioritized broadband. (Comcast is already subject to such a requirement through 2018 under its NBC Universal merger conditions.) And we should emphasize that "might" is the operative word here because we are talking about a proposal (or even sets of proposals) that have yet to be approved by a majority of the FCC Commissioners, and put out for public review, comment, and discussion. Nothing has been decided yet, in other words.
Calling paid prioritization "discrimination" is a matter of semantics; one wouldn't say that FedEx discriminates because it offers the choice of one-day and two-day package delivery. Indeed, the traditional "Title II common carrier" model (which some network neutrality advocates favor) has allowed different levels of service -- paid prioritization in other words -- as long as the prioritized service level was available to all comers.
As for the FCC's apparent proposal, it does not encourage or require paid prioritization. At most, the proposal would allow some commercial offerings -- subject to negotiation between the two firms -- to allow for a higher level of service. This same opportunity exists today in other parts of the Internet ecosystem, including Internet backbone networks and content delivery networks. It's unclear whether paid prioritization will succeed as a business model in the broadband context, however. Network operators and engineers have been pitching "differentiated quality of service" techniques on the Internet for nearly twenty years, with little to show for it. And if broadband providers offer adequate levels of non-prioritized bandwidth, such offerings won't prevent tomorrow's upstarts from taking root on the Internet, which is the core concern that network neutrality needs to protect.
Saying the FCC action will "force companies to pay tolls" or "create a two-tier Internet" makes it seem as though companies such as Netflix and Google currently use the Internet for free. They don't. They pay access providers; they pay intermediaries called transit providers; they pay CDNs; and they pay to build or buy their own infrastructure. Some pay more than others. Big players like Microsoft, Amazon, Google, Facebook, and Apple spend billions every year to speed the performance of their services to end users. They would like to pay less, and network operators would like to charge more; that's the way business negotiations work.
Unlike the original focus of network neutrality, which was network operators favoring their services over others, the objection to Wheeler's proposal is that it favors big players like Netflix and Amazon over startups who can't afford to pay for priority delivery. That's a legitimate concern, but new and small providers are always at a disadvantage compared to their larger and more established competitors. The real policy concern -- and the one network neutrality advocates should analyze once they see the details of the proposal -- is whether the playing field will allow for new innovators to enter and compete against today's incumbents.
The reality of today's Internet environment is that building a business costs money. As one example, YouTube had to sell to Google partly because its explosive growth meant skyrocketing bandwidth bills, which the startup couldn't cover itself. For policymakers, therefore, the relevant question is whether tomorrow's YouTube will face insurmountable obstacles in getting started and launched. As long as innovators can still use the same baseline connections as before (and such connections remain adequate to support new business models),they need not "seek permission from carriers or pay special fees to be seen online," to use Cerf's language.
Are there downsides and risks to Chairman Wheeler's rule? No question, and there should be a vigorous debate on the details of the proposal in the public comment period. Most notably, startup investment could be chilled if entrepreneurs and venture capitalists decide it's not worth entering a market where paid prioritization exists -- let alone necessary to obtain access to adequate levels of bandwidth or service quality. And content providers may feel compelled to pay the supplemental fees if network operators starve the regular Internet pipe and don't provide adequate levels of basic broadband service, making it structurally deficient as a means to reach customers.
These concerns, however, are ones that the FCC's proposed rules could address -- whether as proposed or as modified after public comment. The FCC could conclude, for example, that deliberately degrading basic connectivity runs afoul of the relevant non-discrimination and transparency requirements. Or the FCC could conclude that offering paid prioritization options depends on an adequate level of broadband service to those using the regular Internet pipe. It's simply premature to conclude based on a few leaked highlights of the Chairman's proposal that the FCC has abandoned its commitment to network neutrality.
The effectiveness of the FCC proposal in protecting the open Internet thus depends on how it's enforced, which is an often overlooked point in these discussions. The dirty little secret is that enforcement is a key challenge for every variant of network neutrality, even the most expansive version under "reclassification" of broadband as a regulated Title II telecommunications service. The rules get all the attention, but the real question is what the FCC does with them.
It should also be emphasized that the reason network neutrality is important to begin with is that broadband access in the U.S. is a heavily concentrated market, especially for fixed high-speed connections. So the importance of the FCC doing what it can to encourage more competition in that marketplace should not be lost. The FCC can promote a more competitive broadband market, for example, by preempting excessive restrictions on municipal networks and fees for access to poles and conduits; making more wireless spectrum available on both a licensed and unlicensed basis; managing the transition from the legacy public switched telephone network to an all-Internet Protocol communications world; and supporting disruptive entrants. It's looking at all those areas now.
The surest way to stop progress towards real broadband competition is if the FCC's work grinds to a halt in a miasma of political and legal opposition. That's quite possible if it reverses a decade of precedent and reclassifies broadband. And that would be just the beginning of the process.
How to defend and implement network neutrality is not as simple as banning all forms of paid prioritization. After all, the FCC's old rule left the door open to such service offerings, and even the reclassified Title II framework traditionally allowed different levels of service as long as they were offered to all comers. What really matters is ensuring that the broadband environment continues to provide space for tomorrow's innovators to develop new, disruptive offerings. When the FCC releases the proposed rules for comment, we should all focus on that criterion to evaluate whether they are sufficient and effective.
At this moment, with the exception of one company subject to merger conditions, broadband access providers in the United States aren't legally prohibited from blocking competitors' content, arbitrarily degrading unaffiliated services, favoring their own content artificially or offering prioritization deals solely to favored partners. Any of these practices would cause damage to the innovative dynamism and openness of the Internet. And there's reason to think that, without regulatory oversight, those harms might be realized. After all, they have been before.
Federal Communications Commission Chairman Tom Wheeler recently circulated a proposal to his colleagues that reportedly would ban all of these practices. (As of this writing, the proposal is not public, although Wheeler gave a general overview in a blog post.) The proposal reportedly would establish ongoing FCC review of agreements between broadband and content providers, to ensure that better performance is offered on reasonable terms. And it would beef up transparency obligations to shine a light on attempts to create an artificial slow lane or manipulate network management anti-competitively. It would do all this on a legal foundation recently validated in court. (The agency's two prior attempts in this area were overturned on jurisdictional grounds.)
In short, the FCC is apparently considering a proposal to make a form of network neutrality the law of the land. So why is there an explosion of outrage that the FCC is gutting, killing or abandoning net neutrality?
The answer has to do with that old saw about the perfect and the good. The FCC proposal doesn't prohibit one category of business agreements that could be abused. And it doesn't use the simplest and most expansive legal theory. That has network neutrality advocates up in arms. On the other hand, Wheeler's approach has something significant going for it: It might actually stick. In the political morass of today's Washington, that factor shouldn't be minimized. And it could be an effective model for protecting ongoing Internet-based innovation.
The concept of network neutrality (or in the FCC's preferred parlance, the open Internet) draws upon technical and economic insights about the amazing success of the Internet as a catalyst for innovation, free speech and economic growth. The basic idea is that network operators should not discriminate against certain content, applications, services, or devices. The significance of this principle for protecting innovation and tomorrow's upstarts is increasingly evident -- the growth of Internet companies over the last twenty attests to its importance. In the words of Internet Protocol co-creator Vint Cerf, network neutrality means "the creators of new Internet content and services need not seek permission from carriers or pay special fees to be seen online."
Net neutrality highlights vitally important aspects of the way the Internet works today. Other than the FCC's recent, failed attempts, however, it has never been a legal mandate, let alone a technical requirement. Thanks to the extraordinary work of scholars and public interest advocates over the past decade, the basic principle of network neutrality is now embraced by the technology industry, the FCC, and even many broadband providers.
To be sure, there is still disagreement at the edges about how to implement network neutrality: How should the FCC balance the benefits of network neutrality against the risks of excessive involvement in business decisions, and the potential to chill capital investment in infrastructure? How exactly should it write and enforce the rules to ensure that tomorrow's upstarts have the necessary freedom to innovate without permission?
In implementing network neutrality, some differentiation of traffic must be allowed on the Internet, even encouraged. Network operators should be able to block malware and denial of service attacks, for example. And companies like Akamai and Limelight should be able to offer content distribution network (CDN) functionality to improve delivery of certain content. There are many thousands of interconnection agreements among network operators, backed by massive amounts of traffic engineering, caching overlays, and other mechanisms to meet business and technical parameters. If the Internet story is one of a neutral network, it's also one of private firms negotiating arrangements and managing their networks as they see fit.
Network neutrality advocates acknowledge all this. They propose various rules that distinguish reasonable from unreasonable discrimination, just as the FCC is doing. So the real question is whether Chairman Wheeler's proposed rule is so much worse than what came before. Under what we understand the FCC proposal to be, access providers can't block, can't degrade, can't arbitrarily favor certain applications, and can't favor their own traffic. If you read the past decade of network neutrality literature, those are the dangers usually mentioned. So what's everyone so mad about?
The major change in the new proposal concerns so-called paid prioritization agreements. In other words, the new rules appear to allow a broadband provider to offer content providers the option of faster or more reliable delivery for a supplemental fee. Under the old rules, the FCC didn't prohibit such deals, but said it was skeptical they would meet its discrimination test.
In the new proposal, the FCC appears to mandate that paid prioritization offerings be "commercially reasonable." This requirement presumably would insist on the availability on the same terms to all, with the FCC reviewing such offerings on a case-by-case basis. Such a requirement might also include the condition that any paid prioritization offerings are only reasonable when the broadband provider offers a sufficiently robust level of non-prioritized broadband. (Comcast is already subject to such a requirement through 2018 under its NBC Universal merger conditions.) And we should emphasize that "might" is the operative word here because we are talking about a proposal (or even sets of proposals) that have yet to be approved by a majority of the FCC Commissioners, and put out for public review, comment, and discussion. Nothing has been decided yet, in other words.
Calling paid prioritization "discrimination" is a matter of semantics; one wouldn't say that FedEx discriminates because it offers the choice of one-day and two-day package delivery. Indeed, the traditional "Title II common carrier" model (which some network neutrality advocates favor) has allowed different levels of service -- paid prioritization in other words -- as long as the prioritized service level was available to all comers.
As for the FCC's apparent proposal, it does not encourage or require paid prioritization. At most, the proposal would allow some commercial offerings -- subject to negotiation between the two firms -- to allow for a higher level of service. This same opportunity exists today in other parts of the Internet ecosystem, including Internet backbone networks and content delivery networks. It's unclear whether paid prioritization will succeed as a business model in the broadband context, however. Network operators and engineers have been pitching "differentiated quality of service" techniques on the Internet for nearly twenty years, with little to show for it. And if broadband providers offer adequate levels of non-prioritized bandwidth, such offerings won't prevent tomorrow's upstarts from taking root on the Internet, which is the core concern that network neutrality needs to protect.
Saying the FCC action will "force companies to pay tolls" or "create a two-tier Internet" makes it seem as though companies such as Netflix and Google currently use the Internet for free. They don't. They pay access providers; they pay intermediaries called transit providers; they pay CDNs; and they pay to build or buy their own infrastructure. Some pay more than others. Big players like Microsoft, Amazon, Google, Facebook, and Apple spend billions every year to speed the performance of their services to end users. They would like to pay less, and network operators would like to charge more; that's the way business negotiations work.
Unlike the original focus of network neutrality, which was network operators favoring their services over others, the objection to Wheeler's proposal is that it favors big players like Netflix and Amazon over startups who can't afford to pay for priority delivery. That's a legitimate concern, but new and small providers are always at a disadvantage compared to their larger and more established competitors. The real policy concern -- and the one network neutrality advocates should analyze once they see the details of the proposal -- is whether the playing field will allow for new innovators to enter and compete against today's incumbents.
The reality of today's Internet environment is that building a business costs money. As one example, YouTube had to sell to Google partly because its explosive growth meant skyrocketing bandwidth bills, which the startup couldn't cover itself. For policymakers, therefore, the relevant question is whether tomorrow's YouTube will face insurmountable obstacles in getting started and launched. As long as innovators can still use the same baseline connections as before (and such connections remain adequate to support new business models),they need not "seek permission from carriers or pay special fees to be seen online," to use Cerf's language.
Are there downsides and risks to Chairman Wheeler's rule? No question, and there should be a vigorous debate on the details of the proposal in the public comment period. Most notably, startup investment could be chilled if entrepreneurs and venture capitalists decide it's not worth entering a market where paid prioritization exists -- let alone necessary to obtain access to adequate levels of bandwidth or service quality. And content providers may feel compelled to pay the supplemental fees if network operators starve the regular Internet pipe and don't provide adequate levels of basic broadband service, making it structurally deficient as a means to reach customers.
These concerns, however, are ones that the FCC's proposed rules could address -- whether as proposed or as modified after public comment. The FCC could conclude, for example, that deliberately degrading basic connectivity runs afoul of the relevant non-discrimination and transparency requirements. Or the FCC could conclude that offering paid prioritization options depends on an adequate level of broadband service to those using the regular Internet pipe. It's simply premature to conclude based on a few leaked highlights of the Chairman's proposal that the FCC has abandoned its commitment to network neutrality.
The effectiveness of the FCC proposal in protecting the open Internet thus depends on how it's enforced, which is an often overlooked point in these discussions. The dirty little secret is that enforcement is a key challenge for every variant of network neutrality, even the most expansive version under "reclassification" of broadband as a regulated Title II telecommunications service. The rules get all the attention, but the real question is what the FCC does with them.
It should also be emphasized that the reason network neutrality is important to begin with is that broadband access in the U.S. is a heavily concentrated market, especially for fixed high-speed connections. So the importance of the FCC doing what it can to encourage more competition in that marketplace should not be lost. The FCC can promote a more competitive broadband market, for example, by preempting excessive restrictions on municipal networks and fees for access to poles and conduits; making more wireless spectrum available on both a licensed and unlicensed basis; managing the transition from the legacy public switched telephone network to an all-Internet Protocol communications world; and supporting disruptive entrants. It's looking at all those areas now.
The surest way to stop progress towards real broadband competition is if the FCC's work grinds to a halt in a miasma of political and legal opposition. That's quite possible if it reverses a decade of precedent and reclassifies broadband. And that would be just the beginning of the process.
How to defend and implement network neutrality is not as simple as banning all forms of paid prioritization. After all, the FCC's old rule left the door open to such service offerings, and even the reclassified Title II framework traditionally allowed different levels of service as long as they were offered to all comers. What really matters is ensuring that the broadband environment continues to provide space for tomorrow's innovators to develop new, disruptive offerings. When the FCC releases the proposed rules for comment, we should all focus on that criterion to evaluate whether they are sufficient and effective.