The Future of Fiber Is Oversold

Over the last few years, the inexpedient belief has emerged in broadband policy circles around the world that fiber optic broadband coverage must be a goal in and of itself and that any broadband that is not based on fiber cables is deficient. This is akin to saying that mobility is important, but only SUVs are adequate while sedans are deficient. The focus should not be on the particular physical medium through which bits move but on capacity and latency of networks.

To be sure, many countries have high levels of fiber deployment. OECD data from June 2022 shows fiber subscriptions as a share of total subscriptions have increased dramatically as a whole—up from 22 percent five years earlier. South Korea has the highest fiber penetration rate at 87.3 percent of fixed connections; the OECD average is 35.9 percent. At 19.9 percent of connections, U.S. fiber penetration ranks eighth from last among OECD countries.

Taken out of context, the data appears to suggest that the United States is failing to achieve an important goal. But at a time when even the fixed and wireless marketplaces are closer to convergence than ever and a good Internet connection can be run through a variety of mediums, outsized focus on coverage by individual technologies is more distraction than education.

All non-fiber networks are not created equal, and U.S. broadband networks are fast, expansive, and diverse. The majority of the country is passed by high-speed wired infrastructure. In fact, the U.S. average speed ranks above that of South Korea—which is number one for fiber penetration—in Ookla’s latest fixed speedtest results. Fixed wireless networks are also capable of speeds that rival high-quality wired networks’, and the majority of the country has wireless coverage.

There are also plenty of factors other than network technology that can influence speed. DOCSIS standards, which set the speed at which a modem brings the Internet into a house, have been updated roughly twice per decade since their debut, and the DOCSIS 4.0 industry standard—with a 10 Gbps downstream capacity—was issued in 2019. The speed at which users experience Internet running on even the fastest network is still limited by the quality of the modem and in-house Wi-Fi, and these deserve just as much attention as network technology does.

All of these technologies can be used to close a much more important gap than one predicated on an inflated need for fiber: the digital divide. Households without any form of Internet connection suffer obvious disadvantages. Ensuring every household has access to a quality Internet connection is a clear priority in a way that maximizing fiber penetration is not. In this sense, fiber is just one of many tools available, and it shouldn’t always be our first choice given the costs of its use.

But thirst for fiber has run roughshod over more expansive goals, and the U.S. government has signaled its preferences with forthcoming infrastructure funding distributed according to guidelines that prioritize fiber. This prioritization threatens to reduce the number of new connections that the funding can create.

Fiber is fast and, in a vacuum, more speed is better. However, virtually all everyday applications demand considerably slower speeds than fiber allows. As long as there remain households that are totally unconnected, prioritizing one technology—and excluding others—whittles down the toolbox of potential solutions. Moreover, “future-proof” is not as absolute as it sounds when fiber networks have an expected lifespan of only 20–40 years, and a new broadband technology could materialize at any point that renders all existing technologies obsolete.

Building end-to-end fiber is not what really matters when it comes to equipping society for the digital age. Connecting people with adequate speeds is. Fiber is one of many tools for accomplishing that, but its concomitant costs mean that it is not always the right choice.