Abstract
Network neutrality for the Internet has been a fiercely debated topic for more than 15 years. Arguably spurred by recent regulatory actions in the US and the tremendous ecosystem evolution as characterized by the emergence and growing importance of global cloud platforms as well as the Internet of Things, the debate has recently seen a resurgence. This paper describes the origins and development of the regulatory stance towards network neutrality in the US and the EU. Against the background of the diverse evolutionary forces that shape the ecosystem, we examine network neutrality regulation through the lens of network economics. In doing so, we describe a series of challenges and misconceptions associated with current regulations and expound the requirement for a market-driven understanding of network neutrality.
1 Introduction
Network neutrality regulations impose rules for the management and pricing of traffic by broadband access service providers. The debate about network neutrality is long-lasting and has touched on a wide range of issues. While it has been characterized by fierce controversy, it is primarily driven by concerns that gatekeeping broadband access service providers might discriminate opportunistically against unaffiliated providers of content and applications or else attempt to stifle opposing opinions or political views (e.g. Bauer and Obar 2014). The debate, along with the associated narrative, was initially framed by lawyers and technologists (Wu 2002, 2003; Yoo 2005, 2006; Lessig and McChesney 2006; Clark 2007; van Schewick 2007, 2010). It comes as a surprise that economists entered the arena only after a significant delay.[1] While this delay has shaped the early stages of the debate, it certainly impacted its course over time.
In the absence of a clear or generally accepted definition of network neutrality (e.g. Krämer et al. 2013; Greenstein et al. 2016), the issue as such presents a moving target, especially due to the rapid evolution of the Internet ecosystem. While the debate has been shaped by the emergence of technologies that enable broadband access service providers to (opaquely) introduce traffic differentiations,[2] its complexity has arguably increased in light of emerging challenges associated with 5G and the Internet of Things (IoT), but also due to the rise of hypergiants like Google, Amazon, or Facebook that operate their own private cloud platforms. The recent repeal of a rather stringent network neutrality regime in the US by the Federal Communications Commission (FCC) instated a fundamental change in the regulatory stance towards network neutrality in the US (FCC 2018). It also established a divergence between the regulatory approaches towards network neutrality taken in the US and the EU. This change has reinvigorated heated and highly politicized debates and has attracted input by formerly less active stakeholders like the Internet Society (2019).
An appreciation of the tremendous evolution in the Internet ecosystem is essential for a meaningful network economic assessment of the network neutrality debate. Today, broadband networks clearly provide the infrastructural basis for Internet access. While the Internet is part of a broadband-centric ecosystem that is characterized by diverse innovation processes, broadband networks have evolved into general-purpose platforms. They enable the delivery of a broadening and evolving range of content, applications, and use cases via the same physical local network infrastructures. The rivalry between an evolving range of Internet-based and non-Internet-based applications for the same local infrastructure resources implies diverse demands for broadband access services. Providers of broadband access services are thus confronted with increasingly complex and dynamically changing capacity allocation problems. Questions regarding the justification and impact of network neutrality regulations arise. Do network neutrality regulations unduly restrict the entrepreneurial freedom that is necessary to achieve economically optimal capacity allocations based on entrepreneurial Quality of Service (QoS) differentiations?[3] Do they distort ecosystem evolution? Also, do they impair innovational complementarities (e.g. Bresnahan and Trajtenberg 1995; Bresnahan 2010; Bauer and Knieps 2018) which can be conceptualized as bidirectional cross-layer innovation spillovers? And, more generally, what is the “public Internet” (Lehr et al. 2018)?
Ecosystem evolution also has implications for the power balance within the ecosystem. The ecosystem has undergone a seismic shift in the way it is interconnected, and also how, and by whom, traffic is delivered. It has become more densely interconnected. Traffic is exchanged, processed, and delivered increasingly local, i.e. close to end-users or things. Content Delivery Networks (CDNs) already deliver a majority of Internet traffic (e.g. Stocker et al. 2017; Cisco 2019; Labovitz 2019). Additionally, a (private) “cloud ecosystem” is emerging. As these developments call for a reassessment of the conventional wisdom about hierarchies, competition, bargaining positions, and revenue sharing conventions in the Internet ecosystem, the role of broadband access service providers must be reassessed (e.g. Bender et al. 2017; Wohlfahrt et al. 2018; Stocker 2019). Moreover, new tussles in cyberspace between cloud providers and Internet Service Providers (ISPs) (or broadband access service providers) may emerge. Naturally, these players have conflicting views about the role broadband access networks should play in the future Internet (Huston 2017; Peterson et al. 2019).
In this paper, we conduct a network economic analysis to examine network neutrality regulations. In doing so, we place a particular focus on innovation and economic efficiency. The remainder of this article is structured as follows. In order to inform discussions of the regulatory status quo, Section 2 outlines the many dimensions of ecosystem evolution. In doing so, we aim at elucidating trends that have shaped the range of market players in the ecosystem, their business models and business relationships, and have altered how content and applications are delivered. Since these trends have also shaped the capacity allocation problems with which broadband access service providers are confronted, we consider an appreciation of the current state of ecosystem evolution as essential for meaningful (re)assessments of public policies and regulations concerning the Internet in general and network neutrality regulation in particular. Section 3 provides an overview of network neutrality regulations in the US and the EU. The origins and basic dimensions of network neutrality are described, and an overview of the respective regulatory histories is provided. Section 4 then describes the need for entrepreneurial traffic management by broadband access service providers and outlines the concept of a market-driven network neutrality. We explain how this concept implies an adaptive state of non-discrimination between an evolving range of content, applications, and use cases, thus enabling economically efficient capacity allocations in broadband networks and undistorted ecosystem evolution. A discussion and critical appraisal of the regulatory status quo are provided in Section 5. Beyond critically assessing the economic justification for such a regulatory intervention, we describe the essential role of innovational complementarities for a workable innovation ecosystem. In Section 6, we provide conclusions.[4]
2 Broadband Evolution and the Growing Complexity of the Internet Ecosystem
Since the commercialization and privatization of the Internet more than two decades ago, it has developed into a central platform for communications. It presents a focal point of innovational activity that lies at the core of a rapidly evolving ecosystem.[5] However, modifications to the Internet’s basic suite of protocols, and subsequent Internet-wide deployments are fraught with difficulty. Non-trivial coordination problems and tussles between different and often competing market players have resulted in an impaired ability to evolve (e.g. Clark et al. 2005; Peterson et al. 2019; see also Anderson et al. 2005; Handley 2006).[6] With the growing adoption of broadband, a widening and increasingly diverse range of Internet-based content and applications became available and created a growing mismatch between the performance available from the “basic Internet” and the emerging demands for differentiated content delivery (e.g. Leighton 2009; Nygren et al. 2010). Strategies based on the over-provisioning of network capacities were not sufficient. Beyond being economically inefficient (e.g. Berger-Kögler and Kruse 2011; Knieps and Stocker 2016), capacity alone can neither resolve all performance issues (e.g. Ou 2008: p. 27) nor service or business coordination problems or other tussles. It is therefore unsurprising that performance limitations and a lack of differentiation at the basic Internet’s traffic service level have spurred the development of an evolving range of technologically complementary innovations.
The architecture of the Internet is layered and compositional. Therefore, traffic differentiation can typically be implemented at different layers by different entities (e.g. Bauer and Knieps 2018; Zave and Rexford 2019; see also Claffy and Clark 2014, 2016). Cloud, content, or CDN providers can thus augment the capabilities of the basic Internet without requiring modifications to the underlying infrastructure or suite of protocols. As the extent to which these innovations can enhance delivery performance and introduce traffic (and price) differentiation, however, critically hinges on the performance of traffic delivery via the basic Internet, some of these innovations aim at reducing the reliance on Internet-based traffic delivery (e.g. Knieps and Stocker 2016; Stocker et al. 2017).
A range of mechanisms can be implemented in endpoints (e.g. servers) and enable optimizations of the end-user experience in the consumption of content and applications over the Internet.[7] Other forms of traffic differentiation are based on mechanisms that strategically modify the location where (i) traffic is exchanged, (ii) content and applications are stored, and (iii) data is processed.[8] Respective mechanisms make use of geographically distributed interconnection points[9] and (highly) distributed architectures of servers that are strategically positioned across the Internet near to end-users and things.[10] By reducing the physical and virtual distance data packets need to travel between communicating endpoints, delivery performance can be improved (e.g. Stocker et al. 2017). On the one hand, reducing the physical distance data packets need to travel reduces both propagation delays and improves the performance of the (distance-sensitive) Transmission Control Protocol (TCP), the Internet’s dominant transport layer protocol (e.g. Leighton 2009; Smith 2009). On the other hand, reducing the virtual distance means that the number of interconnection points that need to be traversed between communicating endpoints, and thus the number of entities involved, are reduced. Performance bottlenecks at interconnection points and problems associated with service and business coordination can thus be reduced. Thus, both delivery performance and control over traffic can be increased. Some approaches are based on the deployment of servers that deliver content and applications or perform local processing of data within ISP (or broadband access service provider) networks. Thus, relevant communications constitute intra-network or private communications.[11] Interconnections become “performance-irrelevant” (e.g. Stocker et al. 2017; Stocker 2019). Generally, these mechanisms imply a trend towards traffic localization.
In the meantime, a growing number of cloud and CDN providers like Google and Akamai have deployed private wide area networks (WANs) to connect globally distributed servers (e.g. Kaufmann 2018; Google 2019). Similarly, some providers of interconnection points connect geographically dispersed interconnection facilities via private WANs (e.g. Nomikos et al. 2018). These entities, which are not ISPs, now act as traffic service providers. By delivering traffic via their own backbone infrastructures, they can bypass Internet-based traffic delivery, thus reducing their reliance on the “public Internet.” The emerging cloud ecosystem emphasizes trends towards traffic localization and indicates developments towards ecosystem fragmentation. Not only have they flattened traditional hierarchies, but they have also changed competition, business models and business relationships (Bender et al. 2017; Wohlfahrt et al. 2018; Böttger et al. 2019; Claffy and Clark 2019; Stocker 2019). Changing industry structures, bargaining positions, and revenue sharing approaches, but also emerging tensions and tussles between ISPs and cloud providers must be recognized. As these players have conflicting visions about the role of broadband access networks in the ecosystem (Huston 2017; Peterson et al. 2019), their views regarding network neutrality are also likely opposed.
Taking into account these innovative means for traffic differentiation, why do we need traffic differentiation in broadband access networks after all? The answer is simple. These innovations do not provide perfect substitutes for traffic differentiation at the level of traffic services by broadband access service providers. Insofar as they rely on traffic delivery via the public Internet, their capabilities are limited (e.g. Knieps and Stocker 2016). With this in mind, what are the relevant capacity allocation problems with which broadband access service providers are confronted? Important insights can be obtained from a general, and perhaps obvious observation: in contrast to legacy and purpose-built infrastructures like the Public Switched Telephone Network (PSTN) or the Cable TV (CATV) network, broadband access networks are not purpose-built. They are not optimized to deliver a specific or narrow set of services. Instead, they present general-purpose platforms that spur innovation and provide the infrastructural basis for the delivery of a widening and rapidly evolving range of content, applications, and use cases that may or may not be considered part of the public Internet (e.g. Bauer and Knieps 2018; Lehr et al. 2018; Stocker and Whalley 2019).
As a strong dynamism characterizes the application sphere, the demands for broadband access services are becoming more complex and diverse. For instance, the phasing-out of legacy voice telephony, the increasing smartification of objects and networks in the context of the IoT, but also Tactile Internet applications or virtual and augmented reality broaden the range of applications and use cases considerably. Increasing mobility requirements further add to the complexity in demands. Specific challenges arise due to applications that require high bandwidth or ultra-reliable communications at low latencies. Some IoT applications like Intelligent Transportation Systems (ITS) for self-driving cars require explicit and deterministic performance guarantees regarding traffic delivery (e.g. ITU-R 2015: pp. 11–12; Knieps 2017). Considering the growing requirements for a flexible and agile means to meet these emerging requirements, the orchestrated and synergistic use of a set of technologically complementary mechanisms gains importance. In this mix of mechanisms, local big data processing within edge clouds will likely play an essential role to ensure ultra-reliable low latency communications. However, whether delivered via the public Internet or not, the delivery of deterministic performance guarantees requires complementary traffic differentiations within broadband access networks.
3 Network Neutrality in the US and the EU
3.1 The Origins of the Debate
The origins of the network neutrality debate are closely tied to strong efforts at deregulation of broadband access in the US (e.g. Bauer and Obar 2014). Especially the market inception and early success of cable-based broadband access, as well as the ensuing “technological duopoly” of DSL-based and cable-based access technologies, motivated a reassessment of the regulatory approach towards broadband. Sparked by an “anomaly” marked by the individual evolution and subsequent convergence of formerly purpose-built infrastructures, regulatory challenges arose.[12] When cable broadband emerged, it was exempt from the rather rigorous Title II regulations that applied to DSL-based broadband Internet access, which was classified under the heading of “telecommunications services.” In order to achieve a harmonized and more symmetric regulatory regime for broadband, the FCC opted for a path of deregulation. From 2002 to 2005, they took a stepwise approach, eventually (re)classifying DSL-based and cable-based Broadband Internet Access Services (BIAS) as “information services.” As a consequence, BIAS were subject to less rigorous regulations (e.g. Bauer 2005; Hazlett and Caliskan 2008; Gilroy 2019).[13] While this fundamental shift in the stance towards the regulation of broadband was reflective of the availability of two technologically different and competing access platforms, it has given rise to concerns that widely unregulated broadband access providers may selectively discriminate against (unaffiliated) third-party providers of content and applications. Corresponding (de)regulatory efforts can thus be considered to have sparked the network neutrality debate.
Before we turn our attention to the history of network neutrality regulations in the US and the EU, we would like to explore a bit more the origins of the debate. In this context, it is essential to note that the term “network neutrality,” as well as the intellectual foundation of the associated narrative, was coined by law scholar Tim Wu (2002, 2003). Relating non-discrimination to open access regulations, Wu conceptualized network neutrality as a (set of) principle(s) to ensure non-discrimination between content and applications on the Internet (Wu 2003: pp. 169–170). He considered “[t]he argument for network neutrality … as a concrete expression of a system of belief about innovation” (Wu 2003: p. 145) and argued that “the promotion of network neutrality” is animated by an interest in preserving and enabling “Darwinian competition among every conceivable use of the Internet” (Wu 2003: p. 142). Central to the network neutrality principle is the recognition of the potentially detrimental effects of a strict (or “truly neutral”) traffic regime and justification for and desirability of certain types of traffic differentiation (Wu 2003: p. 167).[14] This insight is emphasized by Wu’s assertion that his network neutrality “principle represents ultimately an effort to develop forbidden and permissible grounds for discrimination in broadband usage restrictions” (Wu 2003: p. 168, emphasis in original).
Regulatory frameworks for legacy telecommunications focused on mandating incumbent firms to provide (third-party) service providers non-discriminatory access to their infrastructures. Network neutrality shifts the focal point of regulatory intervention to a set of rules as to how broadband access service providers treat traffic within their networks (e.g. Wu 2002: p. 5). It reflects concerns regarding the technical ability and incentives of broadband access service providers to actively interfere with the traffic management decisions made by communicating edges in an adverse and opportunistic fashion (e.g. Frischmann and van Schewick 2007). At its core, network neutrality revolves around principles for determining whether a specific network management practice is permissible or should be forbidden – or, in regulatory jargon, whether a practice is “reasonable” or “unreasonable” (see Section 3.3). As much as openness and neutrality can hardly be considered undesirable by regulators, the network neutrality debate reflects conflicting views as to what non-discrimination means, what a neutral Internet is, and how it should be achieved. Relevant “systems of belief” vary considerably. Not only does this reveal the inherently subjective nature of network neutrality as a normative proposition, but it also helps to explain how opposing stakeholder views have created a long-lasting and fierce debate that is both highly polarized and politicized.[15]
Interpretations of network neutrality range from strict regimes, which favor the regulatory imposition of egalitarian traffic rules in which all packets must be treated equally, to regimes in which broadband access service providers are free to engage in contractual agreements that specify any price and traffic differentiation (e.g. Bauer and Obar 2014: Table 1 at p. 4; Cave and Vogelsang 2015: p. 85; Gans 2015). Apart from the varying scope of discussion touching on political issues such as free speech, network neutrality has become a regulatory concept that determines the autonomy and competencies of broadband Internet access service providers in managing the delivery and pricing of traffic over the Internet.
As we will describe in more detail below, four things are essential to note. First, network neutrality regulations introduce ex ante traffic regulations. Second, they do not prevent all forms of differentiation based on network management, but rather define the scope of permissible deviations from best-effort-based network management. Third, network neutrality regulations apply exclusively to broadband access service providers. They do not apply to edge providers like CDN providers, cloud providers, or other content providers. Fourth, network neutrality regulations impose traffic rules that apply only to broadband Internet access services (BIAS) that are considered part of the “public Internet” and not to (non-Internet) traffic that relies on other broadband access services.
3.2 Different Dimensions of Network Neutrality Regulations
Before we describe the regulatory implementations of network neutrality in the US and the EU, we want to start by describing typical dimensions of such regulations. One significant dimension is sector-specific consumer protection measures/transparency regulations. These measures aim at reducing problems of asymmetric information by imposing obligations for broadband access service providers to transparently disclose their network management practices. This dimension has been viewed rather uncontroversial in the debates about network neutrality. Transparency constitutes an essential precondition for informed decision-making by market participants and thus workable competition. Moreover, transparency is necessary to effectively detect discriminatory behavior and to audit compliance with regulations.
Another significant dimension is two-tiered traffic regulations. Broadband access services constitute an input for the provision of a diverse and rapidly evolving range of content, applications, and use cases (see Section 2). While these services share the same broadband network resources, not all access services are used to carry Internet traffic. Access services might be used to deliver services like facilities-based voice over IP, proprietary IPTV services, or IoT-related applications that might be considered non-Internet services. Instead of imposing symmetric rules for all broadband access services, network neutrality regulations impose rules that affect only a subset of traffic delivered via general-purpose broadband networks (Knieps and Stocker 2015, 2016). Two-tiered traffic regulations instate a regulatory market split – a bifurcation, establishing two separate and asymmetrically regulated “walled gardens” (see also Hazlett and Wright 2012; Stocker 2015). They impose ex ante traffic regulations for broadband access services for the “public Internet,”[16] not for those broadband access services that are considered private or otherwise outside the public Internet.[17] The underlying delineation of broadband access services does not result in a market-driven fashion. Instead, it is solely determined by a regulatory market split that distinguishes between two different types of broadband access services as a basis for two-tiered traffic regulations (Knieps and Stocker 2015, 2016), the primary distinguishing criterion being their reach, i.e. whether they provide access to (virtually) all endpoints connected to the Internet or only to a closed subset (e.g. as is the case with proprietary IPTV services) (see Section 3.3). Figure 1 provides a high-level summary of Sections 3.1 and 3.2.
Network Neutrality Regulation – A Schematic Overview.
Source: Authors.
3.3 Network Neutrality Regulations in the US and the EU: A Brief Overview
In order to describe the path that led to the currently divergent regulatory stances towards network neutrality in the US and the EU, we briefly provide an overview of the genesis of these entities’ respective regulations below. In doing so, we aim not to provide a comprehensive and exhaustive overview, but instead one that emphasizes specific milestones which we consider foundational for understanding the regulatory status quo and the relevant dimensions of network neutrality regulation as introduced in Section 3.2.
3.3.1 Network Neutrality Regulation in the US
In accordance with Wu’s notion of network neutrality as a non-discrimination principle and sparked by the deregulation approaches of broadband access in the US, the FCC took initial steps towards tackling network neutrality in 2004. They described a set of principles for “Preserving Internet Freedom” (Powell 2004a,b), based on which they formulated an Internet Policy Statement that was issued in August 2005. The statement explained the FCC’s mission by stating that they should “ensure that providers of telecommunications for Internet access or Internet Protocol-enabled (IP-enabled) services are operated in a neutral manner” (FCC 2005c: para. 4 at p. 3, emphasis added) and introduced four guiding principles intended to ensure that such a state could be achieved.[18] In subsequent years, existing concerns related to the stepwise deregulation of broadband access were further intensified by cases of anticompetitive behavior by broadband access providers. In an early well-known case, Madison River, a regional telephone company and DSL-based broadband access provider, was found to be blocking rival VoIP services with the intention of preventing a decrease in the revenues they obtained from legacy telephony services (FCC 2005a,b; Lee and Wu 2009: Footnote 1 at p. 61; Brennan 2017: p. 476). In 2007, cable-based broadband access provider Comcast was found to be sabotaging traffic belonging to the peer-to-peer application BitTorrent in an opaque way, i.e. without disclosing their practices to their customers (e.g. Svensson 2007).
The FCC (2008: para. 1 at p. 1, emphasis added) condemned this behavior by Comcast and issued a ruling finding “that the company’s discriminatory and arbitrary practice unduly squelches the dynamic benefits of an open and accessible Internet and does not constitute reasonable network management.” This statement revealed the FCC’s differentiated approach towards network neutrality – their decision was based on a delineation between reasonable and thus permissible forms of network management and unreasonable forms of network management that the FCC would consider harmful and like to see forbidden. Besides, the ruling revealed consideration of another aspect, transparency, as revealed by the FCC’s assertion that “Comcast’s failure to disclose the company’s practice to its customers has compounded the harm” (FCC 2008: para. 1 at p. 1). The FCC order was challenged, brought before the court, and eventually vacated in 2010. The court ruling justified their decision mainly based on the FCC’s lack of authority to intervene in the entrepreneurial decision-making of broadband access service providers and the way they manage their networks – note that these services were classified as “information services” (see Gilroy 2019: pp. 3–4).
The Comcast case was seminal in tracing the path towards the regulatory status quo insofar as it motivated another piece of legislation, the Open Internet Order (hereafter “OIO”). While not changing the classification of broadband Internet access services as an “information service,” the OIO was based upon three pillars: (i) transparency, (ii) no blocking, and (iii) no unreasonable discrimination (FCC 2010: para. 1 at p. 2).[19] In addition to the recognition that end-user control and transparency constitute “touchstones of reasonableness” (FCC 2010: para. 6 at p. 4), the FCC imposed an asymmetric, two-tiered traffic regulation which was based on a regulatory market split between BIAS and specialized services, with traffic regulations exclusively applying to BIAS (FCC 2010: paras. 112–114 at pp. 61–62). Notably, the “no unreasonable discrimination” standard (FCC 2010: paras. 76–77 at pp. 43–45) codified a regulatory delineation concept to distinguish between reasonable and unreasonable network management practices, seeking to identify and determine the scope for permissible deviations from an egalitarian traffic regime via reasonable network management practices (FCC 2010: paras. 80–92 at pp. 47–52). A regulatory bifurcation was introduced through the regulatory market split. While the reach of the different service types in terms of endpoints was used as the main distinguishing criterion, a separate “garden” was established for the provision of specialized services. The private/specialized service garden was thus separated from the BIAS/“public Internet” garden both from a logical and regulatory perspective. The FCC argued, that this should only be allowed if the provision of specialized services was not in conflict with the benefits of an open Internet (FCC 2010: paras. 112–114 at pp. 61–62). After being challenged yet another time, the OIO was vacated in 2014, again justified mainly on the grounds of a lack of authority to intervene in the entrepreneurial freedom of BIAS providers to introduce network management practices (Knieps and Stocker 2015: p. 47; Gilroy 2019: pp. 5–6).
When considering these rulings, it is unsurprising that after being defeated twice in court, the FCC sought to address the persistent problem of its lack of regulatory authority. Corresponding efforts culminated in the adoption of the Open Internet Order (hereafter “OIO II”) in 2015 (FCC 2015b). The OIO II contained different provisions that enhanced the FCC’s regulatory authority. Apart from enhanced transparency rules, the reclassification of BIAS (both fixed and mobile)[20] as telecommunications services under Title II provided the FCC with the authority to impose “utility-style” regulations and represented a core element of the order (FCC 2015a,b). While the asymmetric, two-tiered traffic regulation of the OIO and a regulatory market split were maintained – even though services that had been previously termed “specialized services” were now termed “non-BIAS data services” – the OIO II imposed a rigorous traffic regime for BIAS based on a general conduct (or antidiscrimination) rule and three bright-line rules: (i) no blocking (FCC 2015b: paras. 111–118 at pp. 47–51), (ii) no throttling (FCC 2015b: paras. 119–124 at pp. 51–53), and (iii) no paid prioritization (FCC 2015b: paras. 125–132 at pp. 53–58). Responding to series of peering disputes between edge providers (e.g. large content providers or CDN providers) and BIAS providers that involved alleged cases of throttling, the FCC announced to consider “commercial arrangements for the exchange of traffic with a broadband Internet access provider …on a case-by-case basis” (FCC 2015b: para. 29 at p. 10).[21]
While the OIO II granted the FCC the authority to impose heavy-handed utility-style regulations and thus to impose traffic rules that defined the permissible scope of network management practices, the OIO II and the classification of BIAS as a telecommunications service was soon to be reversed. This time, however, it was not due to a court ruling. Instead, it was caused by a shift from a Democratic to a Republican administration, which, in turn, subsequently reversed the 3-2 partisan divide within the FCC that had previously implied a Democratic, pro-network neutrality position (e.g. Hazlett and Wright 2017: pp. 504–505; Kang 2017). In 2017, a vote along party lines led to the adoption of the Restore Internet Freedom Order (hereafter “2017 Order”) that was released in 2018 (FCC 2018). The 2017 Order introduced significant changes to the regulatory stance and imposed a more light-handed regime. Significantly, a vital element of the 2017 Order was the introduction and reliance on enhanced transparency regulations.[22] Whereas the regulatory market split was upheld, the extent of the asymmetry between the two gardens was reduced. BIAS were (re)classified as an information service such that the FCC’s authority to impose utility-style regulations was removed.[23] The 2017 Order is based on the conviction that “the transparency rule we adopt, in combination with the state of broadband Internet access service competition and the antitrust and consumer protection laws, obviates the need for conduct rules by achieving comparable benefits at lower cost” (FCC 2018: para. 239 at p. 140).[24] Since it was found “that consumers have means to take remedial action if an ISP engages in behavior inconsistent with an open Internet” (FCC 2018: para. 4 at p. 3), traffic regulations via the general conduct rule, as well as the three bright-line rules, were not part of the 2017 Order. Considering this dramatic shift in the FCC’s regulatory stance towards network neutrality, it is not surprising that the 2017 Order provoked public outcry as well as local (e.g. initiatives to impose network neutrality laws at the state level as in California) and congressional activity (i.e. bills that would reverse the 2017 Order). So far, however, the 2017 Order seems to have survived these challenges (e.g. Gilroy 2019; Harrison 2019).
3.3.2 Network Neutrality Regulation in the EU
While debates and regulatory efforts were undertaken relatively early in the US, the EU remained largely in a state of restraint, emphasizing the differences in the respective approaches towards the regulation of broadband access infrastructures.[25] Whereas the US decided to take an approach that favors facility-based (or network-based) competition and strongly deregulated upstream infrastructures as described above, the situation was fundamentally different in the EU, where the implementation of rigorous broadband regulations enabling service-based competition was preferred (e.g. Cave and Crocioni 2007; Carter et al. 2008: pp. 38–39; Vogelsang 2013: p. 227).[26] It is thus not surprising that the first significant effort at establishing network neutrality as a policy principle was more generic than the US approach. In 2009, the European Commission issued a declaration on network neutrality that was appended to Directive 2009/140/EC and laid out enhanced transparency obligations as well as “powers for national regulatory authorities to prevent the degradation of services and the hindering or slowing down of traffic over public networks” (EC 2009b: p. 69, emphasis added). Transparency obligations were introduced in the Articles 20(1)(b) and 21 of Directive 2009/136/EC (EC 2009a: pp. 23–25), while Article 22 of Directive 2009/136/EC contained provisions that endow national regulators with the authority to impose “safety net” minimum QoS levels (EC 2009a: p. 25). In subsequent years, a series of consultations on network neutrality were issued by the European Commission and the BEREC (e.g. EC 2010; BEREC 2012a,b, c; see also Marsden 2013: pp. 4–8). Incentive for the introduction of network neutrality on an EU-wide basis was arguably provided by local initiatives of national regulators in the Netherlands and Slovenia, who passed network neutrality laws in 2012, and the prospect of regulatory fragmentation that would have been antagonistic to the idea of a harmonized, EU-wide Digital Single Market (e.g. Marcus 2016: pp. 265–270).
In 2013, the European Commission issued a proposal for a regulation that included network neutrality regulations (EC 2013). After a “trialogue” between the Commission, the Council, and the Parliament that involved modifications and amendments to the proposal, a consensus was successfully reached in 2015 and the Regulation (EU) 2015/2120 (EU 2015) could be adopted. It went into effect on April 30, 2016. Core elements of the regulation are comprehensive transparency obligations (Articles 4 and 5); a “universal service-style” safety net provision that allows national regulators to impose minimum QoS levels for Internet access services as described in Article 5(1); and an asymmetric, two-tiered traffic regulation based on a regulatory market split (Article 3). Similar to the US, a distinction is made between Internet Access Services (IAS) that can be considered as providing access to the “public Internet” and “services other than internet access services which are optimised for specific content, applications or services, or a combination thereof, where the optimisation is necessary in order to meet requirements of the content, applications or services for a specific level of quality” (EU 2015: p. 9). Rather rigorous traffic regulations apply exclusively to IAS, leaving room for reasonable network management practices. Even though the same rules do not apply to “other services,” their provision is conditional on meeting two requirements, detailed below.
On the one hand, they must not negatively impact the availability or general quality of Internet access services. On the other hand, they must not be used to circumvent traffic regulations through strategic labeling of services (EU 2015: p. 9). What is important to note is that the regulation includes a component that introduces some dynamism. To ensure that the regulation is applied in a consistent and harmonious way across the Member States, Article 5(3) of Regulation (EU) 2015/2120 delegates the authority to interpret the regulation and to develop detailed guidelines for its implementation to BEREC (EU 2015: p. 10; BEREC 2016: para. 1 at p. 3). Following a public consultation, a first version of the guidelines was published on August 30, 2016 (BEREC 2016). Following an additional public consultation that is slated to take place in the third quarter of 2019, a second version of the guidelines is expected to be published in the first quarter of 2020 (BEREC 2018: p. 3). Also worth mentioning is Directive (EU) 2018/1972 (EU 2018), establishing the European Electronic Communications Code (EECC), adopted in December 2018. Significantly, the EECC introduces a broadband universal service which can be considered a “safety net,” imposing minimum service requirements for IAS that must be met by relevant providers.[27]
4 Entrepreneurial Traffic Management and Market-Driven Network Neutrality
The analytical framing of the problem of optimal capacity allocation in broadband networks has its roots in a long tradition of analyses of congestion and scarcity in network infrastructures. The starting point is the insight that if the usage of network capacities results in a situation of scarcity, an economic allocation problem arises. The economically efficient solution to such a problem is based on the well-established principle of opportunity costs[28], which in turn informs incentive-compatible pricing, and investment decisions that maximize social welfare.[29] When examining allocation problems in broadband networks, insights can be gained from the field of transportation economics, where the relevance of interdependencies between (short run) social optimal congestion charges and (long run) optimal investment dimensions is well established (e.g. Mohring and Harwitz 1962).
Corresponding economic models yield important insights. First, they suggest that congestion externalities indicate the opportunity costs of infrastructure usage. Second, they imply that it is economically unjustified, and thus economically inefficient, to expand infrastructure to such an extent as to eliminate congestion externalities. Complete elimination of congestion is considered undesirable; it would imply inefficient resource usage via over-provisioning of infrastructure capacities and would lead to a (transient) state of perfect non-rivalry. Instead, socially optimal, and thus economically efficient, solutions create states of socially optimal levels of congestion and counterbalance the marginal benefits of capacity extension due to the reduction of congestion externalities over the whole lifetime of the capacity with the marginal cost of capacity extension. These well-established principles of optimal capacity allocation in networks provide the proper analytical basis for market-driven price and QoS differentiations in general-purpose broadband networks. Achieving economically efficient capacity allocation is based on the demand-driven entrepreneurial search for innovative traffic architectures, as well as for corresponding configurations and traffic differentiation strategies. Endogenous pricing schemes reflect the opportunity costs of different types of network usage in a responsive fashion.
In the case of a single traffic class without QoS differentiation, only the level of congestion is relevant for the opportunity costs of an additional data packet. Optimal prices thus constitute optimal congestion fees reflecting the level of congestion. Users who are not prepared to pay the congestion fee are excluded from data packet transmission. Optimal usage-dependent prices (short-run problem) and the optimal choice of bandwidth capacity (long-run problem) are derived simultaneously (MacKie-Mason and Varian 1995).[30] Economically optimal capacity allocations for general-purpose broadband networks are based on entrepreneurial search processes for market-driven price and QoS differentiations. The realization of economically efficient capacity allocation relies on agile and general traffic architectures. These architectures ensure the technical feasibility of required QoS differentiations via entrepreneurial traffic management. In other words, they are capable of reflecting and adapting to the increasingly complex and dynamically changing demands for traffic delivery. The opportunity costs of network usage depend on the specific choice of traffic architecture. That is, they vary between different types of network usage based on various bandwidth reservation and packet prioritization strategies that are used to establish a hierarchy of traffic classes. Bandwidth reservations and strategies of logical separation are typically used to establish traffic classes providing maximum QoS control and deterministic end-to-end performance guarantees. These traffic classes can be designed and customized to meet the stringent requirements of highly QoS-sensitive applications. Packet prioritization strategies can be used to establish traffic classes that provide stochastic QoS guarantees and are capable of meeting the requirements of applications that are less sensitive to delay, jitter, or packet loss.[31] Finally, a best effort traffic class provides a basic network service comparable to the traditional resource-sharing paradigm of the Internet (Knieps 2015: pp. 738–741). Economic analyses of price and QoS differentiation strategies for a hierarchy of deterministic as well as stochastic QoS guarantees are provided in Knieps and Stocker (2016) and Knieps (2017).
These articles describe how workably competitive markets for broadband access services yield economically optimal capacity allocations. Since traffic delivery within general-purpose broadband all-IP networks is application-agnostic, optimal capacity allocations imply a dynamic and adaptive state of non-discrimination – a state that is best described as de facto or market-driven network neutrality. The essential feature of such market-driven network neutrality is an entrepreneurial search for traffic management such that there are no incentives for broadband access service providers to discriminate between the range of possible network applications they deliver. Optimal differentiation strategies and thus, desirable states of non-discrimination, require that all applications can be charged according to the opportunity costs of the required QoS capacities of traffic delivery (Knieps 2011: p. 25).
This has profound implications. Market-driven network neutrality presents an inherently diverse and adaptive proposition. It does not present a static, one-size-fits-all concept that legally prescribes a single set of rules on how broadband access service providers manage and price traffic. Instead, it embraces the fact that economically optimal differentiation strategies, which in turn yield states of non-discrimination, may vary between different providers and are changing over time. Market-driven network neutrality enables broadband access service providers to meet the diverse demands with which they are, and will be, confronted efficiently. It thus permits for an evolving range of price and QoS differentiations that (i) yield efficient solutions to the complex and dynamically changing allocation problems that arise in general-purpose broadband networks, and (ii) renders broadband an agile and versatile platform that enables a functional innovation ecosystem.
Figure 2 illustrates the link between the requirement for a generalized traffic architecture and the implementation of economically efficient price and QoS differentiations.
Market-Driven Price and QoS Differentiation.
Source: Authors.
5 A Critical Appraisal of the Regulatory Status Quo
Adopting a network economic perspective, network neutrality regulations can be critically appraised with a particular focus on the role of innovation within QoS traffic architectures. Specifically, we would like to describe some fallacies and misconceptions associated with the network neutrality debate. Taking a series of regulatory fallacies pointed out by Knieps and Zenhäusern (2008), Knieps (2011: pp. 25–26), and Knieps and Stocker (2015) as a starting point, we will point to additional misconceptions and challenges that are primarily associated with ecosystem dynamics and evolution. As we will expound, distortions of innovational activity might divert the ecosystem’s evolution path and arguably fuel tendencies towards ecosystem fragmentation, thus creating and perpetuating a bifurcation into a public Internet ecosystem and a separately evolving, non-public part of the ecosystem.
5.1 Discrimination and Dirt Roads: Why Regulate?
The narrative that underlies current regulations is that broadband access service providers might behave in a discriminatory and thus anticompetitive fashion. Thus, in the absence of competitive forces, so the narrative goes, a proactive regulatory intervention is inevitable; only then can undesirable behavior be prevented. From the perspective of communications law, ex ante regulatory obligations should only be applied to firms with significant market power. Ex ante regulations are considered necessary in cases where anticompetitive behavior can be reliably anticipated, and competition or antitrust law cannot provide required remedies (e.g. Frieden 2015; EU 2018: Recital 163 at p. 66). The economic literature on network neutrality regulation typically suggests that ex post regulations should be preferred over ex ante regulations as social welfare effects of network neutrality regulations are ambiguous and depend upon circumstances (Jamison 2019).
Now, adopting a normative network economic perspective, the disaggregated regulatory approach suggests that public interventions in the network management of broadband access service providers are only justified if there is a monopolistic bottleneck situation, which results if the supply side of a market is characterized by cost subadditivities (that imply a natural monopoly situation) and irreversible costs (Knieps and Zenhäusern 2008: pp. 128–130). In such market scenarios, a market is dominated by the owner of a monopolistic bottleneck component who has network-specific market power. This grants the relevant firm an insurmountable advantage: it can effectively deter market entry. Thus, it can behave in an anticompetitive fashion and opportunistically extract monopolistic rents. In the absence of competition, disaggregated regulation of this network-specific market power becomes inevitable. Knieps and Zenhäusern (2008: pp. 128–130) and Knieps and Stocker (2015: p. 50) have examined network neutrality regulations from the perspective of disaggregated market power regulation. They have shown that the markets for broadband access services are not characterized by monopolistic bottleneck situations. It is interesting to note that the shift from specialized local and long-distance telecommunications infrastructures to general-purpose broadband infrastructures does not change the need for a disaggregated market power regulation approach. If broadband infrastructures possess the characteristics of a monopolistic bottleneck, network-specific market power must be tackled at its roots via appropriate ex ante remedies. Since complementary network service markets for broadband access services are not characterized by monopolistic bottlenecks, there is no network economic justification for ex ante market power regulations in these markets. Network neutrality regulations that deal with public interventions in the network management of providers of competitive broadband access service markets are not justified. Instead, they unduly create regulatory distortions of otherwise competitive network service markets.
A narrative that emphasizes the role of broadband access service providers as gatekeepers has become known as the “dirt road argument.” It suggests that, if allowed to engage freely in price and traffic differentiations, broadband access service providers are incentivized to degrade the performance of a basic best-effort-based service class that is free of charge. This is done to such an extent that customers are willing to avoid this “dirt road” service class and migrate to and pay for higher traffic classes. Thus, the argument goes, broadband access service providers can increase their profits (e.g. Lessig and McChesney 2006; Greenstein et al. 2016). Sidak and Teece (2010) vehemently oppose this reasoning and broadly criticize the argument. They conclude that “invocations of Lessig’s dirt road metaphor and the related theory of vertical foreclosure are rhetorically powerful but economically irrelevant, as neither theory accords with the factual realities of the broadband marketplace” (Sidak and Teece 2010: p. 594). This is supported by Knieps (2011: p. 35), who describes how such ad hoc discriminations neglect the fact that capacity is allocated endogenously between different service classes according to customer preferences for heterogeneous QoS levels and performance guarantees. It follows that ex ante regulatory interventions can only be justified if the gatekeeping broadband access service provider owns a monopolistic bottleneck facility.
While the dirt road argument can be considered a crucial driver of the regulatory arguments for splitting the markets for broadband access services and establishing differently regulated walled gardens, in the recent FCC’s 2017 Order, the dirt road fallacy is comprehensively discussed and considered rather skeptically:
“We reject assertions that allowing paid prioritization would lead ISPs to create artificial scarcity on their networks by neglecting or downgrading non-paid traffic. This argument has been strongly criticized as having ‘no support in economic theory that such incentives exist or are sufficiently strong as to outweigh countervailing incentives.”’ (FCC 2018: para. 258 at p. 152, footnote omitted)
As pointed out by Knieps and Stocker (2015: p. 50), market-driven evolution of innovative price and QoS differentiation strategies is of particular relevance for enabling a dynamic and adaptive state of non-discrimination and economic efficiency which can be conceptualized as market-driven network neutrality.
Another source of regulatory intervention might be related to the existence of information problems. If broadband access service providers have an information advantage which they can abuse in an opportunistic fashion, consumer protection measures and corresponding transparency regulations are the proper regulatory response, not traffic regulations. Insofar as the regulations in the US and the EU emphasize the role of transparency, this can be supported from a network economic perspective.
5.2 Regulatory Market Splits and Ignored Innovational Complementarities
Regulatory ad hoc case-by-case delineation between reasonable and unreasonable traffic management as pursued by the FCC creates large regulatory uncertainties. From this perspective, a clear-cut and transparent regulatory market split between a regulated TCP/IP-based “best effort” public Internet on one side, and an unregulated active traffic-managed innovative “specialized service” network on the other, seems attractive at first glance. Such a regulatory market split is typically combined with a broad version of the dirt road fallacy and regulatory requirements such that the provision of specialized services must have no negative impact on public Internet services. This regulatory market split concept became the leading regulatory framework first in the US and, since then, also in the EU (Knieps and Stocker 2015). From a network economic perspective, serious concerns arise against this regulatory market split approach because it conflicts with a market-driven network neutrality for general-purpose broadband infrastructures. Based on the opportunity costs of network usage, rivalry in consumption for scarce traffic capacities exists within the entire market for broadband access services. A regulatory market split in best-effort Internet traffic services and specialized services disturbs entrepreneurial incentives for pricing and investment decisions and therefore distorts the entrepreneurial search process for innovations.
In the context of the broadband ecosystem, innovational complementarities emphasize the role of bidirectional innovational spillovers between vertically related layers and how these evolve separately and interactively, thus “jointly” shaping the ecosystem’s evolution path. In contrast to the mechanics of such dispersed innovational activity, network neutrality regulations constrain the entrepreneurial freedom to innovate at the level of broadband access services and consequently imply a preference for “non-network-based” innovations (e.g. Lee and Wu 2009; Bauer and Knieps 2018). In addition to static inefficiencies, these regulations induce rigidities at the level of broadband access services that, intentionally or not, substantially distort and divert innovational activity. These distortions can be conceptualized as regulation-induced cross-layer inefficiencies and occur where network neutrality regulations apply. While this emphasizes the inherently selective nature of network neutrality regulations due to two-tiered traffic regulations, it also underscores the dynamic dimension of economic inefficiencies. Regulatory market splits and, subsequently, asymmetric regulatory regimes are based on artificially separated markets. They create and perpetuate a bifurcated ecosystem consisting of two separately evolving walled gardens.
Network neutrality has been analyzed in the context of different states of ecosystem evolution and network environments. In this respect, a distinction between three different phases is useful. Phase 1 describes a scenario in which the Internet was mainly operated on top of legacy voice telephony networks and was dominated by best-effort narrowband applications with rather homogenous delivery requirements. Phase 2 describes a scenario that is dominated by broadband Internet access technologies. CDNs deliver a majority of Internet traffic and traffic is increasingly localized. Although broadband has gained importance, parallel legacy infrastructures are still available for providing legacy voice telephony and cable TV services. In the meantime, however, we have reached the beginning of phase 3. In this phase, the network neutrality debate is confronted with the complex requirements of the future IoT, 5G networks, and network slicing (Bauer and Knieps 2018).
During phase 1, network neutrality was not an issue. In the early stages of the commercialized and narrowband-dominated Internet, traffic delivery via networks that treated packets equally appeared as the “natural way” to provide services like e-mail or file transfers. Endpoint-based differentiations were generally sufficient to meet the demands for differentiated traffic delivery. The transition towards broadband networks, and thereby the shift to phase 2, resulted in a strong bias in the economic analysis of network neutrality regulations. Many models were based on M/M/1-based queuing theory or bandwidth partitioning approaches, thus implying a narrow focus on specific traffic architectures (e.g. Choi and Kim 2010; Economides and Hermalin 2012; Krämer and Wiewiorra 2012; Reggiani and Valletti 2016; Choi et al. 2018; Easley et al. 2018). Even though these models enable to examine innovations in networks supporting stochastic service quality, they do not permit to analyze the role of deterministic quality guarantees that are becoming increasingly important in phase 3. Complex and dynamically changing demands for access services imply a growing need for more general traffic architectures and an extension of the analytical focus to examine the entire range of access services. With the emergence of applications and use cases that require ultra-reliable and high bandwidth broadband communications at ultra-low latencies, an open space of network-based innovations and agile means for capacity reservation and prioritization strategies are essential. They enable to meet the demands for heterogeneous deterministic guarantees enabling the delivery of application services and use cases with highly stringent connectivity demands.
In phase 3, all-IP broadband network infrastructures endowed with agile and versatile traffic logistics provide the general-purpose technologies for future IoT application services as well as conventional communication services (Bauer and Knieps 2018: pp. 177–180). Incentive-compatible capacity allocations are based on relevant price- and QoS differentiation models (Knieps and Stocker 2016; Knieps 2017). Since the economically efficient solution to complex capacity allocation problems critically hinges on entrepreneurial freedom and flexibility to adapt to the diverse and dynamically changing demands that arise, network neutrality regulations that effectively constrain entrepreneurial freedom, thus preventing optimal capacity allocation and distorting innovational activity, must be critically appraised.
6 Conclusions
In this paper, we examined network neutrality regulations through the lens of network economics. Against the background of a highly dynamic and rapidly evolving ecosystem, we described and assessed network neutrality regulations in the US and the EU. We explained how traffic regulations that interfere with the entrepreneurial search for optimal capacity allocation lack an economically sound justification and reasoned that they represent the wrong path. A specific emphasis was placed on the analysis of the effects of two-tiered traffic regimes based on artificial market splits induced by network neutrality regulations. We described how these cause and perpetuate static and dynamic inefficiencies. Significantly, traffic regulations that apply to those broadband access services that are considered part of the public Internet unduly constrain the entrepreneurial search for optimal capacity allocation based on price and QoS differentiations. They prevent economically optimal allocation based on integrated optimizations that can embrace the complexity and dynamism of rivalry scenarios that arise on the basis of an evolving range of content, applications, and use cases in the consumption of the same general-purpose broadband access capacities. Innovational activity might also be diverted as traffic regulations constrain the market-driven search by broadband access service providers for a set of permissible innovations via reasonable network management.
Since the delineation between reasonable and unreasonable network management practices is vague and presents a moving target, regulatory uncertainty arises. More generic sources of regulatory uncertainty are associated with the political economy of the debate. Arguably, emerging tussles between providers of private clouds and CDNs on the one side and ISPs or (broadband access service providers) on the other are not only crucially determining the evolution path of the ecosystem, but also constitute a significant source of conflict that fuels the current network neutrality debate. In addition to opposing stakeholder interests that explain the polarization of the debate, the debate is also highly politicized. The latter has been showcased in the US by the reversal of the OIO II following the change in administration and the reversed majorities within the FCC (e.g. Hazlett and Wright 2017). In the EU, political efforts to safeguard the Digital Single Market and to prevent regulatory fragmentation arguably presented a driving force for the introduction of network neutrality regulation (Marcus 2016: pp. 265–270). Thus, beyond a regulatory preference for upstream investments and non-network-based (downstream) innovations, corresponding uncertainties might further distort investment and innovation incentives. These distortions imply that innovational complementarities must remain (partially) unexploited. Rigidities that diminish the Internet’s ability to evolve create undesirable path dependencies and establish a bifurcated and fragmented ecosystem that is characterized by differently regulated and separately evolving walled gardens.
When considering the recent revival of the debate in the US, but also the current reassessment of the regulatory status quo in the EU, it will be interesting to see how regulators approach the challenges and growing pressure to reconcile technological innovation – as exemplified by the rise of the IoT and resulting in trends towards traffic localization – with their understanding of network neutrality. As we described above, meeting the emerging challenges in an economically efficient fashion is based on a more adaptive institutional approach that stems from a market-driven understanding of network neutrality. Network-specific market power should be regulated at its roots as long as broadband infrastructures possess the characteristic of monopolistic bottlenecks. Regarding broadband access service markets, transparency regulations and ex post competition policy should be favored over the imposition of ex ante traffic regulations.
Funding source: Federal Ministry of Education and Research of Germany (BMBF)
Award Identifier / Grant number: 16DII111
Funding statement: The authors gratefully acknowledge the constructive and very helpful comments by the editor, Wolfgang Briglauer, and two anonymous referees. Volker Stocker would like to acknowledge that this work has been funded by the Federal Ministry of Education and Research of Germany (BMBF) under grant no. 16DII111 (“Deutsches Internet-Institut”).
References
Abbate, J. (1999) Inventing the Internet, Cambridge, MA and London, UK: MIT Press.Search in Google Scholar
Ai, Y., M. Peng and K. Zhang (2018) “Edge Computing Technologies for Internet of Things: A Primer”, Digital Communications and Networks, 4(2):77–86.10.1016/j.dcan.2017.07.001Search in Google Scholar
Anderson, T., L. Peterson, S. Shenker and J. Turner (2005) “Overcoming the Internet Impasse through Virtualization,” Computer, 38(4):34–41.10.1109/MC.2005.136Search in Google Scholar
Antonello, R., S. Fernandes, C. Kamienski, D. Sadok, J. Kelner, I. Gódor, G. Szabó and T. Westholm (2012) “Deep Packet Inspection Tools and Techniques in Commodity Platforms: Challenges and Trends,” Journal of Network and Computer Applications, 35(6):1863–1878.10.1016/j.jnca.2012.07.010Search in Google Scholar
Ash, G., A. Morton, M. Dolly, P. Tarapore, C. Dvorak and Y. El Mghazli (2010) “Y.1541-QOSM: Model for Networks using Y.1541 Quality-of-Service Classes,” RFC 5976. available: https://tools.ietf.org/html/rfc5976 [accessed 6 August 2019].10.17487/rfc5976Search in Google Scholar
Bauer, J. M. (2005) “Unbundling Policy in the United States Players, Outcomes and Effects,” Communications & Strategies, 57(1):59–82.Search in Google Scholar
Bauer, J. M. and G. Knieps (2018) “Complementary Innovation and Network Neutrality,” Telecommunications Policy, 42(2):172–183.10.1016/j.telpol.2017.11.006Search in Google Scholar
Bauer, J. M. and J. A. Obar (2014) “Reconciling Political and Economic Goals in the Net Neutrality Debate,” The Information Society, 30(1):1–19.10.1080/01972243.2013.856362Search in Google Scholar
Baumgartner, J. (2014) “Netflix to Pai: ‘Open Connect Is Not a Fast Lane’: Responds to FCC Commissioner’s Concern About Private CDN Program,” Multichannel NEWS, 11 December. available: http://www.multichannel.com/news/tv-apps/netflix-pai-open-connect-not-fast-lane/386236 [accessed 12 July 2019].Search in Google Scholar
Begen, A. C. and C. Timmerer (2017) “Adaptive Streaming of Traditional and Omnidirectional Media,” ACM SIGCOMM Tutorial, 21–25 August. Los Angeles, CA, available: http://conferences.sigcomm.org/sigcomm/2017/files/tutorial-adaptive-streaming.pdf [accessed 12 July 2019].Search in Google Scholar
Bender, B., S. Surdock and A. Dhamdhere (2017) “The Internet is Flat: Revisited – A Small Transit Provider Case Study,” Presentation at NANOG 70, 5–7 June. Bellevue, WA, available: https://pc.nanog.org/static/published/meetings/NANOG2019/1833/20180921_Bender_The_Internet_Is_v2.pdf [accessed 12 July 2019].Search in Google Scholar
BEREC (Body of European Regulators for Electronic Communications) (2012a). Differentiation Practices and Related Competition Issues in the Scope of Net Neutrality. Draft Report for Public Consultation, BoR (12) 31, 29 May.Search in Google Scholar
BEREC (Body of European Regulators for Electronic Communications) (2012b). BEREC Guidelines for Quality of Service in the Scope of Net Neutrality. Draft Report for Public Consultation, BoR (12) 32, 29 May.Search in Google Scholar
BEREC (Body of European Regulators for Electronic Communications) (2012c). An Assessment of IP-Interconnection in the Context of Net Neutrality, Draft Report for Public Consultation. BoR (12) 33, 29 May.Search in Google Scholar
BEREC (Body of European Regulators for Electronic Communications) (2016). BEREC Guidelines on the Implementation by National Regulators of European Net Neutrality Rules, Guidelines. BoR (16) 127, 30 August.Search in Google Scholar
BEREC (Body of European Regulators for Electronic Communications) (2018). BEREC Report on the Outcome of the Consultation on the Evaluation of the Application of Regulation (EU) 2015/2120 and the BEREC Net Neutrality Guidelines. Report, BoR (18) 245, 6 December.Search in Google Scholar
Berger-Kögler, U. and J. Kruse (2011) “Net Neutrality Regulation of the Internet?” International Journal of Management and Network Economics, 2(1):3–23.10.1504/IJMNE.2011.042577Search in Google Scholar
Böttger, T., F. Cuadrado, G. Tyson, I. Castro and S. Uhlig (2018) “Open Connect Everywhere: A Glimpse at the Internet Ecosystem through the Lense of the Netflix CDN,” ACM SIGCOMM Computer Communication Review, 48(1):28–34.10.1145/3211852.3211857Search in Google Scholar
Böttger, T., G. Antichi, E. L. Fernandes, R. di Lallo, M. Bruyere, S. Uhlig and I. Castro (2019) “Shaping the Internet: 10 Years of IXP Growth,” arXiv e-prints, 8 July. available: https://arxiv.org/abs/1810.10963 [accessed 12 July 2019].Search in Google Scholar
Brennan, T. J. (2017) “The Post-Internet Order Broadband Sector: Lessons from the Pre-Open Internet Order Experience,” Review of Industrial Organization, 50(4):469–486.10.1007/s11151-016-9551-ySearch in Google Scholar
Bresnahan, T. (2010) “General Purpose Technologies.” In: (Hall, B. H. and N. Rosenberg, eds.) Handbook of the Economics of Innovation: Volume 2. Amsterdam et al.: North-Holland, pp. 761–791.10.1016/S0169-7218(10)02002-2Search in Google Scholar
Bresnahan, T. and M. Trajtenberg (1995) “General Purpose Technologies ‘Engines of growth’?” Journal of Econometrics, 65(1):83–108.10.3386/w4148Search in Google Scholar
Briscoe, B. (2007) “Flow Rate Fairness: Dismantling a Religion,” ACM SIGCOMM Computer Communication Review, 37(2):65–74.10.1145/1232919.1232926Search in Google Scholar
Briscoe, B. (2008) “A Fairer, Faster Internet,” IEEE Spectrum, 45(12):42–47.10.1109/MSPEC.2008.4687368Search in Google Scholar
Briscoe, B. and S. Rudkin (2005) “Commercial Models for IP Quality of Service Interconnect,” BT Technology Journal, 23(2):171–195.10.1007/s10550-005-0015-4Search in Google Scholar
Brodkin, J. (2014) “Netflix Performance on Verizon and Comcast has been Dropping for Months: Latest Netflix Data Shows some ISPs Struggling, while Google Fiber Soars,” ArsTechnica, 10 February, available: https://arstechnica.com/information-technology/2014/02/netflix-performance-on-verizon-and-comcast-has-been-dropping-for-months/ [accessed 07 June 2019].Search in Google Scholar
Carter, K., J. S. Marcus and C. Wernick (2008) Network Neutrality: Implications for Europe. WIK Discussion Paper No. 314, December, Bad Honnef: WIK.Search in Google Scholar
Cave, M. and P. Crocioni (2007) “Does Europe Need Network Neutrality Rules?” International Journal of Communication, 1:669–679.Search in Google Scholar
Cave, M. and I. Vogelsang (2015) “Net Neutrality: an E.U./U.S. Comparison,” Competition Policy International, 11(1):85–95.Search in Google Scholar
Choffnes, D., P. Gill and A. Mislove (2017) “An Empirical Evaluation of Deployed DPI Middleboxes and Their Implications for Policymakers,” Paper presented at the 45thResearch Conference on Communication, Information and Internet Policy (TPRC), 7–9 September. Arlington, VA, available: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2941535 [accessed 07 June 2019].Search in Google Scholar
Choi, J. P. and B.-C. Kim (2010) “Net Neutrality and Investment Incentives,” RAND Journal of Economics, 41(3):446–471.10.1111/j.1756-2171.2010.00107.xSearch in Google Scholar
Choi, J. P., D.-S. Jeon and B.-C. Kim (2018) “Net Neutrality, Network Capacity, and Innovation at the Edges,” Journal of Industrial Economics, 66(1):172–204.10.1111/joie.12161Search in Google Scholar
Cisco (2019) “Cisco Visual Networking Index: Forecast and Trends, 2017–2022,” available: https://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/white-paper-c11-741490.pdf [accessed 9 July 2019].Search in Google Scholar
Claffy, kc and D. D. Clark (2014) “Platform Models for Sustainable Internet Regulation,” Journal of Information Policy, 4:463–488.10.5325/jinfopoli.4.2014.0463Search in Google Scholar
Claffy, kc and D. D. Clark (2016) “Adding Enhanced Services to the Internet: Lessons from History,” Journal of Information Policy, 6:206–251.10.5325/jinfopoli.6.2016.0206Search in Google Scholar
Claffy, kc and D. D. Clark (2019) “Workshop in Internet Economics (WIE2018) Final Report,” ACM SIGCOMM Computer Communication Review, 49(2):25–30.10.1145/3336937.3336941Search in Google Scholar
Clark, D. D. (2007) “Network Neutrality: Words of Power and 800-Pound Gorillas,” International Journal of Communication, 1:701–708.Search in Google Scholar
Clark, D. D., J. Wroclawski, K. Sollins and R. Braden (2005) “Tussle in Cyberspace: Defining Tomorrow’s Internet,” IEEE/ACM Transactions on Networking, 13(3):462–475.10.1145/633025.633059Search in Google Scholar
Coase, R. H. (1960) “The Problem of Social Cost,” The Journal of Law and Economics, 3:1–44.10.1086/466560Search in Google Scholar
Crowcroft, J. (2011) “The Affordance of Asymmetry or a Rendezvous with the Random?” Communications & Convergence Review, 3(1):40–56.Search in Google Scholar
Dischinger, M., M. Marcon, S. Guha, K. P. Gummadi, R. Mahajan and S. Saroiu (2010) “Glasnost: Enabling End Users to Detect Traffic Differentiation,” Proceedings of the 7thUSENIX conference on Networked systems design and implementation (NSDI ‘10), 28–30 April. San Jose, CA. available: https://ratul.org/papers/nsdi2010-glasnost.pdf [accessed 07 June 2019].Search in Google Scholar
Easley, R., H. Guo and J. Krämer (2018) “Research Commentary–From Net Neutrality to Data Neutrality: A Techno-Economic Framework and Research Agenda,” Information Systems Research, 29(2):253–272.10.1287/isre.2017.0740Search in Google Scholar
EC (European Commission) (2009a) “Directive 2009/136/EC of the European Parliament and of the Council of 25 November 2009 Amending Directive 2002/22/EC on Universal Service and Users’ Rights Relating to Electronic Communications Networks and Services, Directive 2002/58/EC Concerning the Processing of Personal Data and the Protection of Privacy in the Electronic Communications Sector and Regulation (EC) No 2006/2004 on Cooperation Between National Authorities Responsible for the Enforcement of Consumer Protection Laws,” Official Journal of the European Union, L337:11–36. 18 December, Brussels, Belgium.Search in Google Scholar
C (European Commission) (2009b) “Directive 2009/140/EC of the European Parliament and of the Council of 25 November 2009 Amending Directives 2002/21/EC on a Common Regulatory Framework for Electronic Communications Networks and Services, 2002/19/EC on Access to, and Interconnection of, Electronic Communications Networks and Associated Facilities, and 2002/20/EC on the Authorisation of Electronic Communications Networks and Services,” Official Journal of the European Union, L337:37–69. 18 December, Brussels, Belgium.Search in Google Scholar
EC (European Commission) (2010) “Digital Agenda: Commission Launches Consultation on Net Neutrality,” IP/10/860, 30 June, Brussels, Belgium.Search in Google Scholar
EC (European Commission) (2013) Proposal for a Regulation of the European Parliament and of the Council Laying Down Measures Concerning the European Single Market for Electronic Communications and to Achieve a Connected Continent, and Amending Directives 2002/20/EC, 2002/21/EC and 2002/22/EC and Regulations (EC) No 1211/2009 and (EU) No 531/2012. COM(2013) 627 final, 11 September, Brussels, Belgium.Search in Google Scholar
Economides, N. and B. E. Hermalin (2012) “The Economics of Network Neutrality,” RAND Journal of Economics, 43(4):602–629.10.1111/1756-2171.12001Search in Google Scholar
EU (European Union) (2015) “Regulation (EU) 2015/2120 of the European Parliament and of the Council of 25 November 2015 Laying Down Measures Concerning Open Internet Access and Amending Directive 2002/22/EC on Universal Service and Users’ Rights Relating to Electronic Communications Networks and Services and Regulation (EU) No 531/2012 on Roaming on Public Mobile Communications Networks within the Union,” Official Journal of the European Union, L310:1–18. 26 November, Brussels, Belgium.Search in Google Scholar
EU (European Union) (2018) “Directive (EU) 2018/1972 of the European Parliament and of the Council of 11 December 2018 establishing the European Electronic Communications Code (Recast),” Official Journal of the European Union, L321:36–214. 17 December, Brussels, Belgium.Search in Google Scholar
Faulhaber, G. R. (2011) “Economics of Net Neutrality: A Review,” Communications & Convergence Review, 3(1):7–25.Search in Google Scholar
FCC (Federal Communications Commission) (2005a) In the Matter of Madison River Communications, LLC and Affiliated Companies. Order, File No. EB-05-IH-0110 (DA 05–543), adopted: 3 March, released: 3 March, Washington, D.C.: Federal Communications Commission.Search in Google Scholar
FCC (Federal Communications Commission) (2005b) In the Matter of Madison River Communications, LLC and Affiliated Companies. Consent Decree, File No. EB-05-IH-0110 (DA 05–543), Washington, D.C.: Federal Communications Commission.Search in Google Scholar
FCC (Federal Communications Commission) (2005c) In the Matters of Appropriate Framework for Broadband Access to the Internet over Wireline Facilities et al. Policy Statement, CC Docket No. 02–33 et al. (FCC 05–151), adopted: 5 August, released: 23 September. Washington, D.C.: Federal Communications Commission.Search in Google Scholar
FCC (Federal Communications Commission) (2008) In the Matters of Formal Complaint of Free Press and Public Knowledge against Comcast Corporation for Secretly Degrading Peer-to-Peer Applications; Broadband Industry Practices Petition of Free Press et al. for Declaratory Ruling that Degrading an Internet Application Violates the FCC’s Internet Policy Statement and Does Not Meet an Exception for “Reasonable Network Management”. Memorandum Opinion and Order, File No. EB-08-IH-1518, WC Docket No. 07–52 (FCC 08–183), adopted: 1 August, released: 20 August. Washington, D.C.: Federal Communications Commission.Search in Google Scholar
FCC (Federal Communications Commission) (2010) In the Matter of Preserving the Open Internet; Broadband Industry Practices. Report and Order, GN Docket No. 09–191, WC Docket No. 07–52 (FCC 10–201), adopted: 21 December, released: 23 December. Washington, D.C.: Federal Communications Commission.Search in Google Scholar
FCC (Federal Communications Commission) (2015a) “Fact Sheet: Chairman Wheeler Proposes New Rules for Protecting the Open Internet,” 4 February. available: https://docs.fcc.gov/public/attachments/DOC-331869A1.pdf [accessed 07 June 2019].Search in Google Scholar
FCC (Federal Communications Commission) (2015b) In the Matter of Protecting and Promoting the Open Internet. Report and Order on Remand, Declaratory Ruling, and Order, GN Docket No. 14–28 (FCC 15–24), adopted: 26 February, released: 12 March. Washington, D.C.: Federal Communications Commission.Search in Google Scholar
FCC (Federal Communications Commission) (2018) In the Matter of Restoring Internet Freedom. Declaratory Ruling, Report and Order, and Order, WC Docket No. 17–108 (FCC 17–166), adopted: 14 December, released: 4 January. Washington, D.C.: Federal Communications Commission.Search in Google Scholar
Frieden, R. (2015) “Ex Ante versus Ex Post Approaches to Network Neutrality: A Comparative Assessment,” Berkeley Technology Law Journal, 30(2):1561–1612.Search in Google Scholar
Frischmann, B. M. and B. van Schewick (2007) “Network Neutrality and the Economics of an Information Superhighway: A Reply to Professor Yoo,” Jurimetrics, 47(4):383–428.Search in Google Scholar
Gans, J. S. (2015) “Weak versus Strong Net Neutrality,” Journal of Regulatory Economics, 47(2):183–200.10.3386/w20160Search in Google Scholar
Gilroy, A. A. (2019) The Net Neutrality Debate: Access to Broadband Networks. Congressional Research Service Report, 15 April. available: https://fas.org/sgp/crs/misc/R40616.pdf [accessed 07 June 2019].Search in Google Scholar
Google (2019) Our Infrastructure. available: https://peering.google.com/#/infrastructure [accessed 10 July 2019].Search in Google Scholar
Greenstein, S. M. (2015) How the Internet Became Commercial: Innovation, Privatization, and the Birth of a New Network. Princeton and Oxford: Princeton University Press.10.2307/j.ctvc777ggSearch in Google Scholar
Greenstein, S. M., M. Peitz, and T. M. Valletti (2016) “Net Neutrality: A Fast Lane to Understanding the Trade-Offs,” Journal of Economic Perspectives, 30( 2):127–150.10.3386/w21950Search in Google Scholar
Handley, M. (2006) “Why the Internet Only Just Works,” BT Technology Journal, 24(3):119–129.10.1007/s10550-006-0084-zSearch in Google Scholar
Harrison, S. (2019) “Why Net Neutrality Advocates Remain Optimistic,” WIRED, 1 June. available: https://www.wired.com/story/why-net-neutrality-advocates-remain-optimistic/ [accessed 07 June 2019].Search in Google Scholar
Hazlett, T. W. and A. Caliskan (2008) “Natural Experiments in U.S. Broadband Regulation,” Review of Network Economics, 7(4):460–480.10.2202/1446-9022.1158Search in Google Scholar
Hazlett, T. W. and J. D. Wright (2012) “The Law and Economics of Network Neutrality,” Indiana Law Review, 45(3):767–840.10.2139/ssrn.1937287Search in Google Scholar
Hazlett, T. W. and J. D. Wright (2017) “The Effect of Regulation on Broadband Markets: Evaluating the Empirical Evidence in the FCC’s 2015 “Open Internet” Order,” Review of Industrial Organization, 50(4):487–507.10.1007/s11151-016-9556-6Search in Google Scholar
Hotelling, H. (1938) “The General Welfare in Relation to Problems of Taxation and of Railway and Utility Rates,” Econometrica, 6(3):242–269.10.1007/978-1-4613-8905-7_10Search in Google Scholar
Huston, G. (2017) “The Rise and Rise of Content Distribution Networks,” Presentation at the 8th Workshop on Internet Economics (WIE 2017), 13–14 December. La Jolla, CA, available: http://www.caida.org/workshops/wie/1712/slides/wie1712_ghuston.pdf [accessed 07 June 2019].Search in Google Scholar
Internet Society (2019) “Net Neutrality Legislation: A Framework for Consensus,” 29 May. available: https://www.internetsociety.org/resources/doc/2019/net-neutrality-legislation-a-framework-for-consensus/ [accessed 07 July 2019].Search in Google Scholar
ITU-R (2015) “IMT Vision – Framework and Overall Objectives of the Future Development of IMT for 2020 and Beyond,” ITU-R Recommendation M.2083-0, 09/2015.Search in Google Scholar
Jamison, M. (2019) “Net Neutrality Policies and Regulation in the United States,” Review of Network Economics. 17(3):151–173.10.1515/rne-2018-0041Search in Google Scholar
Kang, C. (2017) “F.C.C. Repeals Net Neutrality Rules,” New York Times, 14 December. available: https://www.nytimes.com/2017/12/14/technology/net-neutrality-repeal-vote.html [accessed 07 June 2019].Search in Google Scholar
Kaufmann, C. (2018) “ICN – Akamai’s Backbone,” Presentation at LINX Meeting 101, 22 May. London, UK, available: https://www.linx.net/wp-content/uploads/LINX101-Akamai-ICN-ChristianKaufmann.pdf [accessed 12 July 2019].Search in Google Scholar
Knieps, G. (2011) “Network Neutrality and the Evolution of the Internet,” International Journal of Management and Network Economics, 2(1):24–38.10.1504/IJMNE.2011.042578Search in Google Scholar
Knieps, G. (2015) “Entrepreneurial Traffic Management and the Internet Engineering Task Force,” Journal of Competition Law & Economics, 11(3):727–745.10.1093/joclec/nhv018Search in Google Scholar
Knieps, G. (2017) “Internet of Things, Future Networks and the Economics of Virtual Networks,” Competition and Regulation in Network Industries, 18(3–4):240–255.10.1177/1783591718784398Search in Google Scholar
Knieps, G. and V. Stocker (2015) “Network Neutrality Regulation: The Fallacies of Regulatory Market Splits,” Intereconomics, 50(1):46–51.10.1007/s10272-015-0524-4Search in Google Scholar
Knieps, G. and V. Stocker (2016) “Price and QoS Differentiations in All-IP Networks,” International Journal of Management and Network Economics, 3(4):317–335.10.1504/IJMNE.2016.079877Search in Google Scholar
Knieps, G. and P. Zenhäusern (2008) “The Fallacies of Network Neutrality Regulation,” Competition and Regulation in Network Industries, 9(2):119–134.10.1177/178359170800900202Search in Google Scholar
Krämer, J. and L. Wiewiorra (2012) “Network Neutrality and Congestion Sensitive Content Providers: Implications for Content Variety, Broadband Investment, and Regulation,” Information Systems Research, 23(4):1303–1321.10.1287/isre.1120.0420Search in Google Scholar
Krämer, J., L. Wiewiorra and C. Weinhardt (2013) “Net Neutrality: A Progress Report,” Telecommunications Policy, 37(9):794–813.10.1016/j.telpol.2012.08.005Search in Google Scholar
Kroes, N. (2010) Net Neutrality – The Way Forward. SPEECH/10/643, European Commission and European Parliament Summit on ‘The Open Internet and Net Neutrality in Europe’, 11 November. Brussels, Belgium, available: http://europa.eu/rapid/press-release_SPEECH-10-643_en.pdf [accessed 07 June2019].Search in Google Scholar
Labovitz, C. (2019) “Internet Traffic 2009–2019,” Presentation at NANOG 76, 9–12 June. Washington, DC, available: https://pc.nanog.org/static/published/meetings/NANOG76/1972/20190610_Labovitz_Internet_Traffic_2009-2019_v1.pdf [accessed 07 June 2019].Search in Google Scholar
Lee, T. B. (2014) “Comcast’s Deal with Netflix Makes Network Neutrality Obsolete,” Washington Post, 23 February. available: https://www.washingtonpost.com/news/the-switch/wp/2014/02/23/comcasts-deal-with-netflix-makes-network-neutrality-obsolete/?noredirect=on&utm_term=.6f3abed5499f [accessed 07 June 2019].Search in Google Scholar
Lee, R. S. and T. Wu (2009) “Subsidizing Creativity through Network Design: Zero-Pricing and Net Neutrality,” Journal of Economic Perspectives, 23(3):61–76.10.1257/jep.23.3.61Search in Google Scholar
Lehr, W. H., D. D. Clark, S. Bauer, A. Berger and P. Richter (2018) “Whither the Public Internet?” Paper presented at the 46th Research Conference on Communication, Information and Internet Policy (TPRC), 21–22 September. Washington, D.C., available: http://dx.doi.org/10.2139/ssrn.3141969 [accessed 07 June 2019].10.2139/ssrn.3141969Search in Google Scholar
Leighton, T. (2009) “Improving Performance on the Internet,” Communications of the ACM, 52(2):44–51.10.1145/1461928.1461944Search in Google Scholar
Lessig, L. and R. W. McChesney (2006) “No Tolls on the Internet,” Washington Post, June 8. available: http://www.washingtonpost.com/wp-dyn/content/article/2006/06/07/AR2006060702108.html [accessed 07 June 2019].Search in Google Scholar
MacKie-Mason, J. K. and H. R. Varian (1995) “Pricing the Internet.” In: ( Kahin, B. and J. Keller, eds.) Public Access to the Internet. Cambridge, MA and London, UK: MIT Press, pp. 269–314.Search in Google Scholar
Marcus, J. S. (2016) “New Network Neutrality Rules in Europe: Comparisons to those in the U.S.,” Colorado Technology Law Journal, 14(2):259–280.Search in Google Scholar
Marcus, J. S., D. Elixmann and K. R. Carter (2008) The Future of IP Interconnection: Technical, Economic, and Public Policy Aspects. Study for the European Commission, Final Report, 29 January. Bad Honnef: WIK-Consult.Search in Google Scholar
Marsden, C. T. (2013) “Net Neutrality Law: Past Policy, Present Proposals, Future Regulation?” Proceedings of the United Nations Internet Governance Forum: Dynamic Coalition on Network Neutrality, 25 October. Nusa Dua Bali, Indonesia, available: https://ssrn.com/abstract=2335359 [accessed 07 June 2019].Search in Google Scholar
McCann, K., A. Mattei, J. Gledhill and G. Kirson (2012) Technical Evolution of the DTT Platform. An Independent Report by ZetaCast, Commissioned by Ofcom, Version 1.3, available: http://www.zetacast.com/wp-content/uploads/2013/10/Evolution-of-DTT.pdf [accessed 07 June 2019].Search in Google Scholar
Mohring, H. and M. Harwitz (1962) Highway Benefits: An Analytical Approach. Evanston, IL: Northwestern University Press.Search in Google Scholar
Nomikos, G., V. Kotronis, P. Sermpezis, P. Gigis, L. Manassakis, C. Dietzel, S. Konstantaras, X. Dimitropoulos and V. Giotsas (2018) “O Peer, Where Art Thou?: Uncovering Remote Peering Interconnections at IXPs,” Proceedings of the Internet Measurement Conference 2018 (IMC ‘18), 31 October–2 November. Boston, MA, pp. 265–278.10.1145/3278532.3278556Search in Google Scholar
Nygren, E., R. K. Sitaraman and J. Sun (2010) “The Akamai Network: A Platform for High-Performance Internet Applications,” ACM SIGOPS Operating Systems Review, 44(3):2–19.10.1145/1842733.1842736Search in Google Scholar
Ou, G. (2008) Managing Broadband Networks: A Policymaker’s Guide. The Information Technology and Innovation Foundation, December. available: http://www2.itif.org/2008-managing-broadband-policymakers-guide.pdf?_ga=2.198385327.1411444044.1542745432-1969726516.1542745432 [accessed 07 June 2019].Search in Google Scholar
Peterson, L., T. Anderson, S. Katti, N. McKeown, G. Parulkar, J. Rexford, M. Satyanarayanan, O. Sunay and A. Vahdat (2019) “Democratizing the Network Edge,” ACM SIGCOMM Computer Communication Review, 49( 2):31–36.10.1145/3336937.3336942Search in Google Scholar
Powell, M. K. (2004a) “Preserving Internet Freedom: Guiding Principles for the Industry,” Remarks at the Silicon Flatirons Symposium, University of Colorado School of Law, 8 Febuary. Boulder, CO, available: http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-243556A1.pdf [accessed 07 June 2019].Search in Google Scholar
Powell, M. K. (2004b) “Preserving Internet Freedom: Guiding Principles for the Industry,” Journal on Telecommunications & High Technology Law, 3(1):5–22.Search in Google Scholar
Reggiani, C. and T. Valletti (2016) “Net Neutrality and Innovation at the Core and the Edge,” International Journal of Industrial Organization, 45:16–27.10.1016/j.ijindorg.2015.12.005Search in Google Scholar
Satyanarayanan, M. (2017) “The Emergence of Edge Computing,” Computer, 50(1):30–39.10.1109/MC.2017.9Search in Google Scholar
Schuett, F. (2010) “Network Neutrality: A Survey of the Economic Literature,” Review of Network Economics, 9(2):1–15.10.2202/1446-9022.1224Search in Google Scholar
Seufert, M., S. Egger, M. Slanina, T. Zinner, T. Hoßfeld and P. Tran-Gia (2015) “A Survey on Quality of Experience of HTTP Adaptive Streaming,” IEEE Communication Surveys & Tutorials, 17(1):469–492.10.1109/COMST.2014.2360940Search in Google Scholar
Sidak, J. G. and D. J. Teece (2010) “Innovation Spillovers and the “Dirt Road” Fallacy: The Intellectual Bankruptcy of Banning Optional Transactions for Enhanced Delivery over the Internet,” Journal of Competition Law & Economics, 6(3):521–594.10.1093/joclec/nhq003Search in Google Scholar
Smith, J. M. (2009) “Fighting Physics: A Tough Battle,” ACM Queue, 7(3):20–26.10.1145/1538788.1538806Search in Google Scholar
Stocker, V. (2015) “Interconnection and Capacity Allocation for All-IP Networks: Walled Gardens or Full Integration?” Paper presented at the 43rd Research Conference on Communication, Information and Internet Policy (TPRC), 25–27 September. Arlington, VA, available: http://dx.doi.org/10.2139/ssrn.2587833 [accessed 07 June 2019].10.2139/ssrn.2587833Search in Google Scholar
Stocker, V. (2019) “Ecosystem Evolution and End-to-End QoS on the Internet: The (Remaining) Role of Interconnections.” In: ( Knieps, G. and V. Stocker, eds.) The Future of the Internet – Innovation, Integration and Sustainability. Freiburger Studien zur Netzökonomie, Nomos: Baden-Baden, pp. 171–193. Forthcoming.10.5771/9783748902096-171Search in Google Scholar
Stocker, V. and J. Whalley (2019) “Who Replies to Consultations, and What Do They Say? The Case of Broadband Universal Service in the UK,” Telecommunications Policy, published online: 10 June. available: https://www.sciencedirect.com/science/article/abs/pii/S0308596118302192 [accessed 12 July 2019].10.1016/j.telpol.2019.04.007Search in Google Scholar
Stocker, V., G. Smaragdakis, W. H. Lehr and S. Bauer (2017) “The Growing Complexity of Content Delivery Networks: Challenges and Implications for the Internet Ecosystem,” Telecommunications Policy, 41(10):1003–1016.10.1016/j.telpol.2017.02.004Search in Google Scholar
Svensson, P. (2007) “Comcast Blocks Some Internet Traffic,” Washington Post, 19 October. available: http://www.washingtonpost.com/wp-dyn/content/article/2007/10/19/AR2007101900842.html [accessed 12 July 2019].Search in Google Scholar
Trevisan, M., D. Giordano, I. Drago, M. Mellia and M. Munafo (2018) “Five Years at the Edge: Watching Internet from the ISP Network,” Proceedings of the 14th International Conference on emerging Networking EXperiments and Technologies (CoNEXT ‘18), 4–7 December. Heraklion, Greece, pp. 1–12.10.1145/3281411.3281433Search in Google Scholar
van Schewick, B. (2007) “Towards an Economic Framework for Network Neutrality Regulation,” Journal on Telecommunications and High Technolology Law, 5:329–392.Search in Google Scholar
van Schewick, B. (2010) Internet Architecture and Innovation. Cambridge, MA: MIT Press.10.7551/mitpress/7580.001.0001Search in Google Scholar
Vogelsang, I. (2013) “The Endgame of Telecommunications Policy? A Survey,” Review of Economics, 64(3):193–270.10.1515/roe-2013-0301Search in Google Scholar
Wohlfahrt, F., N. Chatzis, C. Dabanoglu, G. Carle and W. Willinger (2018) “Leveraging Interconnections for Performance: The Serving Infrastructure of a large CDN,” Proceedings of the 2018 Conference of the ACM Special Interest Group Data Communication (SIGCOMM ’18), 20–25 August. Budapest, Hungary, pp. 206–220.10.1145/3230543.3230576Search in Google Scholar
Wu, T. (2002) “A Proposal for Network Neutrality,” June. available: http://www.timwu.org/OriginalNNProposal.pdf [accessed 07 June 2019].Search in Google Scholar
Wu, T. (2003) “Network Neutrality, Broadband Discrimination,” Journal on Telecommunication and High Technology Law, 2:141–179.10.2139/ssrn.388863Search in Google Scholar
Yoo, C. S. (2005) “Beyond Network Neutrality,” Harvard Journal of Law and Technology, 19(1):1–77.Search in Google Scholar
Yoo, C. S. (2006) “Network Neutrality and the Economics of Congestion,” The Georgetown Law Journal, 94(6):1847–1908.Search in Google Scholar
Zave, P. and J. Rexford (2019) “The Compositional Architecture of the Internet,” Communications of the ACM, 62(3):78–87.10.1145/3226588Search in Google Scholar
© 2019 Walter de Gruyter GmbH, Berlin/Boston
