UNDERSTANDING INTERCONNECTION IN THE INTERNET

Networks are a key component of much of our basic infrastructure, including the net­works of roads, water, electricity, and telecommunications that are used ubiquitously in our daily lives.

To reduce total costs, it is usually desirable for access nodes to be connected via a hierarchy of network switching nodes, each of which is connected to multiple access or network nodes. Thus, a fabric of small local roads or local electric power distribution lines connects homes to highways or high-power distribution facilities to allow traffic to be aggregated and routed via shared facilities towards its final destinations. Similarly, telecommunications networks are comprised of local switches - or in packet networks, routers - that collect the traffic from access nodes and route it to tandem switches, which are in turn connected to other local or tandem switches in a hierarchy that allows traffic to be routed across town, across the state, or across the globe.

A common feature of such networks is that they allow a large number of source/ destination nodes to share network resources economically to take advantage of the fact that demands are not perfectly correlated in time¹ and of the scale economies that are common with network technologies.² Also, larger networks that provide more options for routing traffic between a larger number of source and destination nodes are typi­cally more valuable to each subscriber, or in economic terms, exhibit positive network externalities.³ The presence of shared resources and benefits that characterizes networks poses a number of challenges for network economics.

The cost recovery challenge is further complicated because much of the investment is long-lived and must be put in place in advance of the demand that it is intended to serve. In telecommunications networks like the Internet, the rapid pace of change introduces significant market, technical, and policy uncertainty and asymmetries that further com­plicate the capacity provisioning and cost recovery challenges. Furthermore, most big networks are actually networks of networks, comprised of multiple smaller networks that are interconnected and that are (at least partially) independently controlled. The focus of the discussion here is on the interconnection between these networks and how those are used to route both dollars (for cost recovery and value transfer) and traffic.⁴

Finally, in all of the examples cited above (roads, electricity, water, and telecommu­nications), the networks are viewed as basic infrastructure by society since they provide services that we use ubiquitously in the course of our social and economic lives. As a con­sequence, there is a government interest in ensuring that we have affordable and universal access to these basic infrastructures. In many cases, we rely on public utilities (water, electricity, roads) to provide the basic network infrastructure that is paid for by a mix of taxes (direct subsidies) and subscription fees. Moreover, because of the public interest in basic infrastructure, these networks are subject to significant regulatory oversight, and regulating interconnection is a key focus of network regulatory policy.

In telecommunications networks, there is a long history of interconnection regulations that established both the terms and prices for interconnection in ways that often embed­ded significant implicit and explicit subsidies. For example, in the USA, telecommunica­tions network providers are required to provide interconnection services to each other to ensure connectivity. The terms and rates differ depending on the type of provider and by type of interconnection. As policy-makers have sought to transition toward competition, the legacy of asymmetric and above-cost interconnection charges has posed a significant obstacle to efficient competition, and efforts have been made to rationalize interconnec­tion rules to leave more scope for market forces and to eliminate implicit subsidies and asymmetries. Collectively, these rules are sometimes referred to as ‘intercarrier compen­sation’ policies to highlight their role in managing cash flows between service providers as part of the interconnection challenge. The desire to cut the Gordian Knot of inefficient interconnection policies provided some of the motivation for proposals to adopt ‘bill- and-keep’ interconnection rules that would set intercarrier payments to zero.⁵

In contrast to legacy telephone networks, interconnection in the Internet has been unregulated. This is largely a consequence of the Internet’s evolution as an application that ran on top of the regulated public switched telephone network (PSTN). From an economic perspective, this means that the decision of whether to interconnect, and if so, how to interconnect, was left to the ISPs.

From a technical perspective, the Internet is an end-to-end (e2e) packet delivery network that routes packets from source to destination Internet Protocol (IP) addresses. This often requires packets to traverse multiple networks. These networks, identified as autonomous systems (AS), control a range of IP addresses that they manage and that they are responsible for routing to.⁶ Most AS are associated with ISPs.

There are two basic ways in which ISPs may interconnect: at Internet exchanges where multiple ISPs interconnect, and directly via negotiated bilateral agreements. The focus of this chapter is on the bilateral interconnection agreements via which most traffic is exchanged.⁷ In the early days of the commercialization of the Internet, two sorts of bilat­eral interconnection agreements among ISPs emerged. One was a traditional customer­supplier arrangement, in which one ISP purchased transit service from another, perhaps larger, ISP. An ISP that offers traditional transit service agrees to provide access to the entire Internet for its customers. The other arrangement was peering, in which two ISPs that each had traffic for the other agreed to interconnect to exchange that traffic directly. A peering arrangement between two ISPs does not give either ISP access to the entire Internet via the other; normally each ISP exchanges with the other ISP only traffic that is local to the region of that ISP and its customers. In other words, a peering agree­ment implies a routing restriction with respect to the traffic exchanged; the only traffic exchanged originates from the source ISP and its customers and terminates in the desti­nation ISP and its customers.

In contrast to a transit arrangement, which is a commercial arrangement between a buyer and a seller in which monetary payments flow from the buyer to the seller as com­pensation for services rendered, peering (as the name might suggest) has been viewed as an interconnection among approximate equals, with value to both parties and no a priori obvious direction for monetary payments to flow.

Earlier academic analyses of Internet interconnection focused on these two types of contractual arrangements based on a hierarchical identification of ISPs: transit, a vertical contract in which smaller ISPs serving end users purchased services from larger backbone ISPs; and peering, a horizontal contract between core or backbone ISPs to exchange traffic under a bill-and-keep or revenue-neutral arrangement.⁸ A focus of this earlier lit­erature was on investigating the incentives of backbone ISPs to peer and the implications for competition. For a number of years, these two options - transit and revenue-neutral peering - captured the set of expectations among parties that negotiated interconnec­tions; they represented an informal norm or bargaining regime. But like many such informal norms, revenue-neutral peering has been breaking down slowly as the various parties to potential peering relationships no longer see the simplicity of the balanced- value approximation as serving their needs. It is being replaced (as has happened in other venues like the bargaining over international trade agreements) with a period of more unconstrained bilateral negotiation among the parties. There are a number of reasons why this seems to be happening:⁹

• In the past, many ISPs were more or less similar. There were small ISPs, and larger ISPs with larger footprints, of course, but many ISPs had the same mix of custom­ers.

• Networks that serve different sorts of customers may have very different internal cost structures. A residential broadband provider, with an extensive ‘outside plant’ of fiber, HFC or copper pairs, may have a much higher cost (measured in dollars per megabit) compared to an ISP that serves only business customers with typically much higher peak data rate connections.¹⁰

• Furthermore, with the continued evolution of the Internet (e.g., introduction of routing protocol enhancements¹¹) and the rise of multi-homing, wherein end users and ISPs connect to multiple ISPs thereby expanding the range of routing options, it is increasingly feasible to implement more granular traffic treatments as part of interconnection agreements.

In light of these changes, it has become clear to many ISPs in the USA that they can reduce transit payments if they are able to negotiate peering arrangements to exchange traffic directly. However, in these new circumstances, it is possible that one party or the other regards the legacy approach of revenue-neutral peering as unacceptable, and thus other approaches to interconnection emerge, including the option of paid peering or partial transit. These latter models represent blends of the transitional pure transit or peering models that involve payments from one ISP to another and that restrict the range of addresses to which traffic may be delivered. In paid peering, one ISP pays the other ISP to accept the peering traffic, whereas in partial transit, the ISP that is paying for transit is only purchasing delivery to a subset of the rest of the Internet. These con­tracts are privately negotiated and are not standardized. The payment terms, the traffic (volume, address) restrictions, and other interconnection terms may be negotiated on a case-by-case basis.

16.2.1 High Volume Content and its Providers

A particular focus of policymakers’ attention and for this chapter is the sub-case of interconnection between residential broadband ISPs such as Comcast and AT&T (access networks) and networks that deliver high-volume commercial content such as Level 3 and Cogent. During the first half of 2013, 62 percent of peak-time traffic was due to real-time entertainment, with Netflix accounting for 32 percent and YouTube from Google accounting for 17 percent.¹² In 2011, Netflix content was delivered by three com­mercial operators: Akamai, Limelight and Level 3.¹³ So these three providers plus Google originate and control a significant majority of all real-time entertainment content and peak-time residential download traffic. Collectively, we will call these content delivery networks, in contrast to access networks. A cost-effective connection between a content delivery network and an access network will often be of advantage to both, because of the high volumes of data being transferred.

Negotiation between a content delivery network and an access network about direct connection is at heart a peering negotiation, because it would normally imply the same routing restriction as we described above. Specifically, the content delivery network is normally only trying to get access to the customers of the access network. But because of the high volumes of traffic potentially involved and the fact that the traffic is highly asymmetric (from content to access), negotiation about payment may be commonplace. The carriage of this high-volume traffic will generate costs for access networks, which they will naturally attempt to recover; at the same time one might argue that access to the content creates value for the access network’s customers. Based on anecdotal evidence, it appears that the most common outcome is for content delivery networks to make payments to access networks. We are interested in understanding both the direction and magnitude of these sorts of payments.

Interconnection between content delivery and access networks has attracted sig­nificant attention in recent years. From a business point of view, these interconnection arrangements are of considerable interest because of the high volumes of data being exchanged, and the implications of this high volume for internal costs. Industry observ­ers and regulators are also interested, because the high content-related value of this data raises questions as to whether any actor will have sufficient market power to benefit by extracting rents derived from the value of the content, not its delivery.

The desire to separate concerns about content and the delivery or ‘conduit’ of that content has a long regulatory history. Traditional telecommunications regulation focused on regulating ‘conduits’ and the need to ensure common-carriage access to basic telecommunications services, for which public utility regulation was long justi­fied by the concern that telephone networks were a natural monopoly.¹⁴ On the other hand, regulation of content was manifest in broadcast and programming regulation and media cross-ownership rules.¹⁵ The regulation of cable television services, where investor interests in content and conduit services were conjoined, led to its own silo of regulation in the United States. With the rise of the Internet as the ‘new PSTN’, policy­makers are faced with the question of how much of the traditional regulatory model for telecommunications ought to be mapped over to the new world of the broadband Internet. In their efforts to craft a framework that is more technology-agnostic (i.e., less focused on whether the underlying infrastructure has evolved from a telephone or cable television network, or supports fixed or mobile services), policy-makers have opted for increased reliance on market forces and more light-handed regula­tory approaches.¹⁶ This debate continues in its current incarnation as the debate over network neutrality.¹⁷

One notable component of this debate is policy-makers’ concerns that ISPs might engage in discriminatory network or traffic management practices that may interfere with competition, or worse, limit end users’ access to content or applications of their choice. The implicit presumption is that access ISPs have market power¹⁸ and may seek to use such market power to earn monopoly profits from over-charging end users or other participants in the value chain, such as application service providers or content provid­ers. The focus of our concern here is on the potential for extracting excess payments from content providers and content delivery networks. To the extent that any discrimination might distort competition in content markets, the potential for discriminatory practices is especially worrisome to policy-makers for the reasons noted above.

In this chapter, we will maintain a clean distinction between two sorts of payments that flow among parties. Transport payments are payments among parties that cover the internal costs related to the delivery of flows of data. Content payments are payments that relate to the value and costs of the commercial content itself, not its transport.¹⁹ While this distinction may not always be precise, we believe that it is an acceptable simplification for the purpose of this chapter.²⁰

Traditionally, there has been no content-related component in residential customers’ broadband access payments. If the consumer wants to receive fee-based commercial content, such as Netflix or The Wall Street Journal, the relevant content payments flow directly from the consumer to the content provider. However, this is starting to break down in small but suggestive ways. ESPN3, an online source of streaming sports content, has negotiated an arrangement with certain broadband providers in which they purchase bulk access to ESPN3 content for all their broadband customers, which implies that the cost of this is recovered by a part of the monthly bill that each customer pays. Right now, providers’ per customer content payment related to ESPN3 is about $0.10 a month²¹ but other content arrangements may follow.

Similarly, there has been no content-related component in the historical pricing of ISP interconnection. Transit is a service that provides access to any end point on the network, not specific high-value content; to date and for the most part, this seems to be a competitive, commodity business.²² However, the special case of intercon­nection between content and access networks raises the possibility that there might be a content-related component to the negotiated payments. One of the goals of this chapter is to propose an approach for determining whether payments between parties are likely to include such a content-related component. But we stress that payment from a content delivery network to an access network (e.g., a form of paid peering) does not automatically imply that there is a content-related component in the payment. The payment may be only to cover the transport-related payment that the access network has successfully negotiated to recover some of its internal costs for delivering the content to end users.

16.3