Today (Dec. 4), the Circuit Court of Appeals in Washington will hear oral arguments challenging the Federal Communications Commission’s new Internet regulations. As you’ll recall, the FCC in February reversed long-standing policy and said it would govern the Internet using the Communications Act’s 80-year old telephone rules, known as Title II. Especially surprising was the FCC’s choice to subject mobile broadband to the old rules. Until the final rule was revealed, no FCC net neutrality proposal had included wireless, and for very good reason, as we shall see.
The policy upheaval was difficult to justify given the extraordinary success of the U.S. Internet economy. Under a two-decade-old, bipartisan consensus of Internet freedom, the U.S. created most of the world’s cutting-edge Internet technologies and content platforms. Residential broadband expanded to reach nearly 100 million homes, while mobile subscriptions grew to some 330 million, more than the American population. The U.S. jumped to the lead in search, social networking, cloud computing, and mobile hardware and apps. U.S. Internet traffic, a rough measure of consumer value, expanded from one terabyte per month to more than 20 million TB per month. And today, the market value of just seven American information technology companies — Apple, Google, Microsoft, Facebook, Amazon, Intel and Oracle — totals $2.5 trillion.
If the FCC’s reversal was unusual as a matter of policy — if it ain’t broke, don’t fix it — it was equally questionable as a matter of law.
In the Telecom Act of 1996, Congress issued an explicit finding, that the “Internet and other interactive computer services have flourished, to the benefit of all Americans, with a minimum of government regulation.” Congress also said the policy of the United States is “to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation.”
Congress classified Internet access as a mostly unregulated Title I “information service,” clearly segregating it from the old “telecom services” of Title II. In 1993, Congress had given mobile Internet access an additional layer of protection. Under Section 332, it said all mobile services besides traditional telephony were not Title II common carriage telecom services, but information services under Title I.
Until February, mobile broadband had always been classified as an information service, and even the most aggressive regulators thought it was beyond Title II’s reach. And for very good reason: Wireless is special. Wired broadband is no easy feat. But wireless is technically more tricky, it is capacity-constrained, and it operates in a shared medium, with many users competing for limited capacity. Airborne wireless signals must contend with trees, buildings and other wireless signals in their attempt to reach mobile devices traveling at 70 miles an hour down the highway. With wires, moreover, we can light up another fiber optic cable if need be. But mobile is constrained by the amount of commercially available wireless spectrum, which has been in short supply and where the future pipeline is uncertain.
Mobile networks thus employ a number of technical and commercial methods to deliver robust and economical services. Wireless relies on differentiated signaling and packet prioritization to make sure users get their fair share of capacity and to ensure various applications work properly. FaceTime video calling, for example, is more high-maintenance than email. In addition, because of capacity constraints, you see lots of experimentation with new business models like “zero rating,” where a Spotify, Pandora or Amazon Video in effect pays for the data its service consumes. Title II, however, places these technical and business model differentiators under suspicion.
My American Enterprise Institute colleague Richard Bennett makes a crucial point in a new paper. One big promise of the Internet was “convergence” — the idea that we might deliver numerous and varied applications and services over a common interconnected, interoperable infrastructure. In the past we needed a telephone network to deliver voice, a radio network to deliver audio, and a TV network to deliver video, for example. The Internet, by contrast, can in principle accommodate nearly any type of content. But this does not mean every data packet is treated the same or that it should be. Rather, huge efficiencies can be gained when a common infrastructure delivers diverse apps and services with varying bandwidth, latency, and priority characteristics.
Over the last several decades, the Internet Protocol (IP) was designed with Type of Service, IntServ and DiffServ capabilities so that, as more services were added to the Internet, we could distinguish among them and prioritize them. And yet the aggressive version of net neutrality embodied in Title II seeks to ban many integrated and differential services. Because emerging applications like high-definition video conferencing, telemedicine, virtual reality and real-time gaming will require more differentiation, not less, and because 4G and 5G mobile depend more on differentiated traffic than previous networking platforms, Title II’s effort to homogenize all traffic, applications and business models is especially ill suited to wireless. It will slow mobile innovation.
Wireless is doubly important for the economy. Wireless is also doubly vulnerable to bureaucratic meddling that would interfere with technical and commercial innovation. Fortunately, wireless is doubly protected by statute. The courts will likely strike down the FCC’s ill-advised and illegal rules, with a double denunciation for wireless meddling.
Bret Swanson is president of Entropy Economics LLC and a visiting fellow at the American Enterprise Institute’s Center for Internet, Communications, and Technology Policy.