A government-run network is a bad idea.
Editor’s Note: The following piece originally appeared at AEIdeas, a public-policy blog produced by the American Enterprise Institute. It is adapted here with permission.
On Sunday night, Axios published a memo and slides, reportedly obtained from a National Security Council official, suggesting the federal government might want to build its own commercial 5G wireless network. A government 5G network, the memo said, could be like Dwight D. Eisenhower’s interstate highway system but for the information age. On Monday morning, all three Republican commissioners of the Federal Communications Commission (FCC) issued statements opposing government construction or operation of such a large-scale network.
The memo and slides did not appear to show a fully baked, or even half-baked, plan, but the stated motivation was clear: The US was in danger of falling behind China’s supposed lead in technological innovation and falling prey to its cyberwar capabilities. A government network could both accelerate construction — “within three years,” faster than the private sector — and also ensure the security of America’s next-generation information infrastructure.
Just after noon-time Monday, the administration told Recode that the internal document was “outdated,” “stressed it had merely been floated by a staff member,” and said it had no plans to build a government 5G network.
So what was this memo all about? Was it a trial balloon? A warning to American wireless firms to stay away from Chinese network equipment companies? A slapdash effort, on the heels of a triumphant Chinese Davos meeting, to combine technology and infrastructure in a big new policy idea to counter the ascendant Xi Jinping? Or, as the administration seemed to suggest, merely a lame (if ambitious) effort by a junior staffer that never went anywhere?
It’s true that 5G will be a crucial component of our national infrastructure. It will be a part of an upgraded internet, a new platform I’ve called the “exanet.” Where the first four generations of wireless mostly served our communications and digital content needs, the 5G exanet will connect the rest of the economy, integrating the physical industries with the digital world:
The next phase of the internet will thus not only bring virtual and augmented reality (VR/AR) to entertainment and games, but also to education, training, and manufacturing. The next phase will of course deliver more 4K video than ever, but it will also have to connect millions of cars, and tens of billions of cameras, industrial sensors, geolocation tags, and medical devices (in what we often refer to as the internet of Things, or IoT).
The internet was not conceived to be the commercial, social, and industrial system for the entire planet. (The founders, for example, didn’t think too much about security, and the TCP/IP protocol, while in some ways ingenious, is less than ideal for many modern applications.) The internet’s unlikely success is a testament to the network firms that figured out how to massively scale an experimental project and to entrepreneurs who invented so many helpful and bizarre ways to exploit the newfound bandwidth.
If the first several decades of internet were based on interoperability through digital packet switching and expanded capacity via fiber optics and broadband, the next phase will (in addition to continual capacity additions) focus on ubiquity, latency, reliability, application diversity, and security.
Unlike highways, however, lots of private companies are already building this information infrastructure. Since 2010, US network firms invested some $200 billion in mobile networks, and they plan to invest at least that amount into 5G over the coming years.
The 5G exanet will connect the rest of the economy, integrating physical industries with the digital world.
The memo was not wrong to emphasize security. It’s far from clear, however, that a government network would be more secure than a private one. Remember, China’s theft of millions of sensitive employment files from the federal government’s Office of Personnel Management in 2014-2015 was one of the biggest hacks of all time.
It’s also true that the federal government can help accelerate 5G, but not by taking over construction. Big Dig, anyone? Instead, it can help by expanding the amount of commercially available spectrum and streamlining siting rules for the millions of small cells the new network requires. The FCC is already well down this path, with its Spectrum Frontiers program, for example, which eyes auctions for high-frequency airwaves in the not-too-distant future.
This silly memo has at least one happy irony: The two Democratic FCC commissioners, Mignon Clyburn and Jessica Rosenworcel, also acknowledged that government-run communications networks are bad ideas.
READ MORE:
White House Floats Socialism for the Internet
American Socialists Misunderstand Socialism
— Bret Swanson is a visiting fellow at the American Enterprise Institute, the president of Entropy Economics LLC, and a scholar at the U.S. Chamber of Commerce Foundation.