Sections

Commentary

Idea to Retire: Internet without policy metrics

Protesters raise their smartphones with screens lit

The convergence of the Internet economy with the telecom world has been a process 20 years in the making. For example, the increased use of cellular data instead of voice calls and text messaging has shifted revenues from the telecom sector to the Internet firms, and some Internet firms such as Google are now able to provide fiber connections that were once upon a time the jurisdiction of the telecom sector.  However, comparative benchmarks for measuring what the Internet delivers to users are still focused on the traditional engineering metrics of access, speed, and delay instead of policy-relevant measures of types of traffic, origins and destinations, routes, interconnections at exchanges, and the use of private links.

Shortly after the Internet started as a node-neutral network of military computers, there were organizations dedicated to measuring network performance. One longstanding body doing this is the Center for Applied Internet Data Analysis (CAIDA) at the University of California, San Diego.  Their efforts have been focused on collecting data from public nodes and information provided voluntarily by organizations with similar goals around the world.  They have provided measurements of tremendous value to those concerned with tracking the growth and structure of the Internet.

However, those measures do not tell us enough about the Internet, especially with regards to its enormous economic and strategic significance in the first two decades of the twenty-first century.  Since the mid-2000s, Internet exchanges, content delivery networks, and transit contracts undertaken by telecommunications operators have come to define the form of the Internet. The complex relationships between companies such as Google, Akamai, Facebook, Amazon, Netflix, and telecommunications companies bring together at times divergent interests that piece together the critical components for network operation.

Much is known about how the top and bottom tiers of the Internet work; i.e., how content developers produce and use materials and how communications companies deliver services to end users.  However, details about what goes on in between are shrouded in secrecy, and remain commercially sensitive and mysterious to most of us. We cannot know enough about how much of the Internet is transported by Google, or how much of that traffic is created within the frontiers of any one country.  We may be able to guess about the relationship between what we can measure of delivered Internet content, but we know too little about what precedes that to be able to offer economic analyses or policy guidance.

The International Telecommunications Union, a United Nations body, uses its state-level authority to provide some metrics in conjunction with communications companies, but these efforts fall short of any consensus about Internet metrics.  They are also flawed by incompatibilities of measurements, lack of access to significant types of digital traffic, and commercial and national sensitivities. Increasing amounts of digital communication takes place within companies, government agencies, or in other types of private networks through veiled forms of access and through the so-called “dark web.” These are not accounted for systematically and their economic and policy significance is largely ignored.

This situation is aggravated by the relationship between the physical transport mechanisms of cables, network nodes, exchanges and control centers, and the virtual controls provided by Internet protocols (IP). In the IP world, the routing and re-routing of traffic is done through virtual links and networks are changing dynamically in such a way that we require real time information.

The economic significance of the Internet continues to grow even while we fall into greater ignorance about its details.  The transition to the high address capacity of Internet protocol version 6 (IPv6), the growth in the number of connected devices (the so-called “Internet of Things”) and the digitization of everything means that good understanding of the shape and size of the Internet is of paramount importance. 

At the same time, the economic dynamics are constantly under threat not only by nefarious forces such as hackers but also by legitimated trends.  Competitive forces have brought about a commercial landscape of giants, dominant players whose commercial activities range far beyond their core businesses of advertising, social media, communications, shopping, or entertainment into controlling positions of infrastructure and the rules of the game that govern the whole system.  Governments seek to control national systems, or cede control to allied interests.  Already Iran, China, Russia and other countries operate such extensive controls over content and access that it is plausible to refer to the balkanization of the Internet.  Turkey, Cuba, Saudi Arabia, Vietnam, and others operate what Freedom House calls “partly free” or “not free” networks at the national level. 

While many activists concerned with the governance of the Internet argue about nuances of network neutrality, many telecommunications service providers are offering “zero-rated contracts” to users that favor access to selected Internet sites with downgraded quality or restricted access to others.  This recent trend makes it difficult to comply with or even to understand the point of net-neutrality rules and guidelines.  While analysts struggle to understand the significance and the extent of these trends, regulators and the public do not know how to respond to the new ways of business over the Internet.  The associated effects are wide-ranging, from taxation and consumer protection to privacy and security.  They affect how cloud services are likely to develop, how pricing is determined, how traffic is managed and how investors are incentivized. 

A simple table of metrics is needed that would allow anyone in the Internet-telecom sector to have a general sense of what is critical, growing, and changing. These requirements should not be dismissed because of company or governmental sensitivities.  In the pharmaceuticals sector, companies hold many secrets but it is still possible to know the price of a medicine. In the automotive manufacturing sector it is possible to know the basic cost of a component, along with its input costs and output pricing.

Unless we can devise appropriate means to measure, monitor and assess the Internet we put at risk many of the economic and public policy goals that we aspire to.  We risk our ability to devise fair practices that promote access to network resources, to protect the rights of consumers and to foster well-functioning markets.  We also jeopardize the ability of nations to enforce legitimate practices of taxation, to promote good governance using the Internet, and to strive for the broader economic benefits that are associated with it. 

The institutions of Internet governance cannot be effective if they are bound by outmoded approaches to gathering and using evidence.  This lack of transparency, coordination, and data access has allowed persistent asymmetries of regulation to distort market relations among telecommunication network operators, Internet content companies and others. Half-hearted measures will stymie efforts to reduce monopolistic behavior, promote national development goals, and ensure an enduring innovative digital economy.

Read more essays in the Ideas to Retire series here.