Bandwidth is Political: Reachability in the Public Internet

Abstract

The global public Internet faces a growing but little studied threat from the use of intrusive traffic management practices by both wholesale and retail Internet service providers. Unlike research concerned with bandwidth and traffic growth, this study shifts the risk analysis away from capacity issues to focus on performance standards for interconnection and data reachability. The long-term health of the Internet is framed in terms of “data reachability” – the principle that any end-user can reach any part of the Internet without encountering arbitrary actions on the part of a network operator that might block or degrade transmission. Risks to reachability are framed in terms of both systematic traffic management practices and “de-peering,” a more aggressive tactic practised by Tier-1 network operators to resolve disputes or punish rivals. De-peering is examined as an extension of retail network management practices that include the growing use of deep packet inspection (DPI) technology for traffic-shaping. De-peering can also be viewed as a close relative of Net Neutrality, to the extent that both concepts reflect arbitrary practices that interfere with the reliable flow of data packets across the Internet. In jurisdictional terms, however, de-peering poses a qualitatively different set of risks to stakeholders and end-users, as well as qualitatively different challenges to policymakers. It is argued here that risks to data unreachability represent the next stage in debates about the health and sustainability of the global Internet. The study includes a detailed examination of the development of the Internet’s enabling technologies; the evolution of telecommunications regulation in Canada and the United States, and its impact on Internet governance; and an analysis of the role played by commercialization and privatization in the growth of risks to data reachability.