Internet Architecture and Innovation

by Barbara van Schewick



Internet Architecture, Innovation and Network Neutrality

“The original architecture of the Internet fostered innovation. It is now changing in ways that are bad for innovation. Regulators need to intervene to protect the beneficial environment for innovation that the architecture of the Internet originally created.”

These claims have formed the basis for calls for regulatory intervention since the late 1990s. Today, they are at the core of the network neutrality debate, the debate over whether governments should establish rules limiting the extent to which network providers can interfere with the applications and content on their networks.

While intuitively appealing, the notion that the Internet’s internal structure has contributed to the explosion of new applications is not obvious from an economic perspective. In Internet Architecture and Innovation, Barbara van Schewick carefully evaluates whether and why these claims are true, using insights from economics, management science, engineering, networking and law.

She demonstrates:

  • how the Internet’s internal structure, or architecture, has fostered innovation in the past;
  • why this engine of innovation is under threat;
  • why the “market” alone won’t protect Internet innovation;
  • and which features of the Internet’s architecture we need to preserve so that the Internet continues to serve as an engine of innovation in the future.

van Schewick puts the international calls for network neutrality regulation on a solid theoretical foundation. In the process, she addresses many questions that are hotly debated in the network neutrality debate: Does it matter whether users can decide freely how to use the network? Do network providers have an incentive to exclude applications and content from their networks, and does competition remove this incentive? How important is innovators’ ability to innovate at low costs? (Get a PDF overview of questions).

The End-to-End Arguments and the Architecture of the Internet

“The end-to-end principle is one of the fundamental design principles on which the Internet was based.”

Most network engineers agree with this statement; many legal academics don’t. Beyond that, even network engineers tend to disagree. Often, both proponents and opponents of a specific technical solution point to the end-to-end arguments to justify their view.

In Internet Architecture and Innovation, Barbara van Schewick uncovers the source of this confusion. There is, she argues, no single version of the end-to-end arguments, but two different ones that embody different rules for architectural design.

The first version, which she calls the “narrow version,” was presented by Jerome Saltzer, David Reed, and David Clark in the 1981 paper in which the end-to-end arguments were first named and identified as a design principle; the second version, which she calls the “broad version,” is the focus of later papers by these authors. The difference between the two versions is not immediately apparent, and Saltzer, Reed and Clark never explicitly drew attention to the change in definition. There are, however, as van Schewick demonstrates, real differences in scope, content, and validity that make it necessary to distinguish between the two versions. At the same time, the silent coexistence of two different design principles under the same name explains some of the confusion that surrounds the end-to-end arguments.

Tracing specific design decisions to the end-to-end arguments, van Schewick shows how the two versions of the end-to-end arguments shaped the original architecture of the Internet. She also discusses many of the questions that often come up in discussions of the end-to-end arguments: Do the end-to-end arguments prohibit quality of service? Do they make it impossible to make the network more secure? How do they relate to current architectural developments such as the evolution of applications towards a more distributed structure and the proliferation of middleboxes? And should the end-to-end arguments continue to guide the evolution of the Internet in the future? (Get a PDF overview of the technical aspects of the end-to-end arguments addressed by the book).

Architecture and Economics

“Architecture is politics.”

Mitch Kapor’s insight, made popular by Lawrence Lessig’s Code and Other Laws of Cyberspace, has become central to many discussions of Internet policy.

“Architecture is quality.”

That the architecture of a complex system determines whether the actual system will be able to exhibit the desired qualities is a central insight in the field of software architecture and product architecture.

“Architecture is economics.”

Building on work that explores the economic implications of the architecture of physical products, Internet Architecture and Innovation focuses on a consequence of architectural decisions that we are just beginning to understand: The architecture of a complex system fundamentally affects the economic system that drives its development, production, and use. It influences which actors will be able to develop, produce, or use the system, what they do, and how they interact.

Understanding how architectures influence economic systems and are influenced by them, van Schewick argues, will be critical for businesses, law and public policy. From the book’s introduction: “These insights open new opportunities for businesses, for law, and for public policy. Businesses may want to engage in “strategic design” by creating architectures that shape the competitive environment in their favor. In the future, being able to design architectures that further a firm’s strategic interests or knowing how to evaluate other firms’ architectural strategies and react to them may be as important to a firm’s success in the marketplace as a firm’s ability to engage in more conventional forms of competition.

For law and public policy, the economic impact of architecture seems to be empowering and challenging at the same time. Traditionally, policy makers have used the law to bring about desired economic effects. Architecture may provide an alternative way of influencing economic systems. Apart from using architecture to realize their own economic goals, policy makers may have to constrain the extent to which private actors can use architecture to further their private economic interests. This is particularly relevant to communications policy, a field in which certain architectures may seriously restrict regulators’ ability to regulate at a later stage. Moreover, as communications networks continue to permeate more and more sectors of the economy, the negative effects of an architecture that strongly favors a few economic actors may be particularly long-term and severe.

To exploit the effect of architecture on the economic system in practice, to design architectures that further our interests, or to understand what other people’s architectures may mean for us, we need to understand exactly how an architecture influences the economic system and what features of an architecture we must tweak to create a specific economic effect. This book is a step toward that goal.”

Readers interested in the approach to “architecture and economics” (see box I.1) advanced by the book can follow a specific thread through the book. From the introduction: “Within the framework described in chapter 1, the book focuses on the effect of one constraint (architecture) on one specific activity (innovation), and sets aside consideration of other factors (such as the effect of non-architectural constraints (i.e. laws, norms or prices), or the mechanisms by which architectures are influenced by economic systems) that are relevant within the framework. This thread ties together the portions of the book that touch on these other factors (see table I.3). In doing so, it complements the detailed analysis (in chapters 4-8) of how architectures—particularly that of the Internet—affect innovation.”