America's Spectrum Gamble
The world is building the infrastructure for the digital economy. The elements of this infrastructure are well known: fiber-optic cables, satellites, cloud services, software tools (such as artificial intelligence), and 5G networks. Some call it the next industrial revolution for a reason, since connected digital technologies will be the source of future wealth and power. Progress in building this infrastructure is crucial for the United States’ future, because the country now has a powerful competitor in the digital economy with a strong technology base and very different views of how the world should work.
Leadership in technology is a key part of this new competition. In most technologies, the United States is doing well. Its companies remain leaders in semiconductors, and the CHIPS and Science Act is strengthening manufacturing in the United States. In software, U.S. companies lead the market in cloud services and in making advanced software tools (a better name for artificial intelligence). The United States is strong in quantum technologies, although other countries, including China, also have strengths. The U.S. education system could do better at training Americans in science, technology, engineering, and mathematics (STEM), but immigration compensates for this. For telecommunications, the backbone of the connectivity needed for the next industrial regulation, the United States has done well in developing and deploying 5G, but it now risks falling behind.
This risk to national security is created by gridlock in spectrum allocations. Spectrum is crucial because the digital future will be wireless—connecting not just cell phones but commercial applications like transportation, hospitals, and factories. This is enterprise connectivity, not consumers streaming videos. The wrong decisions on spectrum will shrink the market for U.S. technology and make U.S. companies less efficient in producing goods and services than their foreign competitors.
The area where the United States lags behind the rest of the world is in allocating spectrum for licensed, full-power use. 5G spectrum is spread among low-, medium-, and high-frequency bands. Each spectrum band has different characteristics for coverage and penetration, and medium frequency (mid-band) offers the best mix of capabilities. The United States has done well in allocating low- and high-frequency bands for 5G, but not in mid-band, the band that the rest of the world is choosing.
As other countries allocate mid-band spectrum for licensed commercial use for 5G, the United States is left with less and different spectrum allocated to 5G. To use an automotive metaphor, if the rest of the world chooses right-hand drive and the United States insists on left-hand drive, it will sell fewer cars in the global market. The result is the United States will fall behind first in 5G spectrum, then in supplying the global 5G market, and finally in finding ways to gain economic advantage from 5G applications. Other nations (particularly China, since it is strong in both devices and software apps) will gain economic advantage over the United States, and America’s technological leadership will be put at risk. The decision by the United States to take a different path for spectrum allocation is a giant gamble to see if other U.S. tech strengths can compensate for the misstep.
The backdrop for this decision is China, Huawei, and 5G. Until the United States raised objections, Huawei was set to dominate the 5G infrastructure. It is on course to do so again, thanks to decisions on spectrum allocation. Letting Huawei succeed would be commercially damaging and create real security risks for the United States. A world connected by China is not a good outcome, but the United States may be on track to make this happen.
This is a complicated situation because telecommunications technology is changing as the field is reshaped by 5G, ORAN, and IP-based networks. 5G and later-generation networks will provide increased data transfers and reliability, and data has become a key element to produce goods and services in industries ranging from manufacturing to finance. Open Radio Access Network (ORAN) technology has the potential to provide wireless connectivity at lower cost and reduce supply chain risk. IP-based networks (and cloud services) also increase reliability and lower costs and are the core of the global internet. The likely direction for the future of networks and telecommunications will be IP-based networks using 5G and ORAN technologies over mid-band frequencies. The country that gets the right set of policies will have an economic advantage.
The fault lines for spectrum allocations are formed by government decisions on the balance between commercial and government allocations, and between licensed and unlicensed allocations. The United States has preferred unlicensed and government spectrum use while the rest of world has chosen to rely on licensed, full-power, commercial use. Given the global nature of the telecommunications market, the United States by itself does not have the heft to shift the axis in its direction. The United States is an outlier, gambling that its researchers and companies can create new technological solutions to allow spectrum to be shared among many different users.
Spectrum Island
The United States, China, and others have very different spectrum strategies. For mid-band spectrum (the most useful for 5G), the average allocation for licensed spectrum among 15 industrial nations is 355 megahertz (MHz). The United States ranks thirteenth with 270 MHz; China, with 460 MHz allocated, ranks fourth(after Japan, the United Kingdom, and France). In contrast, the United States ranks first for unlicensed spectrum (with 1,905 MHz allocated). China ranks last for unlicensed spectrum, with 325 MHz. The global average is 1,005 MHz. The dilemma with unlicensed spectrum is that there are issues with interference, reliability, and range that limit its commercial applications. The United States allocates a significant portion of the spectrum for government use, more than any other country, particularly in the mid-band range. U.S. government users have been allocated about 60 percent of the mid-band spectrum, compared to only 5 percent for licensed commercial users.
This skewed distribution reflects historical allocations from a time when spectrum demands were different and less competitive. Spectrum did not become an economic and political issue until the 1990s, with the universal adoption of mobile phones. The cell phone was the first of many mobile spectrum-using devices to enter widespread use, and it immediately created competition among various claimants for spectrum allocation. The problem has grown more acute as the demand for digital connectivity has skyrocketed and as mobile applications move beyond telephony into industrial and commercial uses.
The United States was a pioneer in spectrum use. This now creates a potential barrier for economic growth and tech innovation because of legacy allocations. The largest consumers of spectrum resources before the 2000s were in government. Government incumbents, unsurprisingly, are reluctant to move from the spectrum they have been allocated for decades given potential expense and inconvenience. If the United States was the only country to use spectrum this would not be as great a problem, but digitization is global and the competition to supply the digital market is intense. An easy summary would be that the United States leads in government and unlicensed use, but not in commercial wireless networks or business uses, which provide the greatest return to the economy.
Companies will build devices to use specific spectrum bands for commercial purposes—these are allocated by a global UN negotiation process known as the World Radiocommunication Conferences (WRC). If the United States does not build those devices, they will be built by China, Europe, and others. Digital infrastructure around the globe will use spectrum allocated to commercial purposes, using devices built by China and Europe. If the United States chooses not to match spectrum allocated everywhere else for commercial purposes, it will be economically handicapped in any competition.
Some limited forms of spectrum sharing exist today and are increasingly important. The hope is that these can be improved quickly enough to provide technological solutions for spectrum scarcity. However, the competitive timeline puts the spectrum issue on an accelerated trajectory. There are three years until the next WRC further locks in allocations (the United States was isolated at the last WRC), while the timeline for dynamic spectrum sharing is likely to be at least five to seven years for successful testing and deployment. There is also risk to resilience in the face of opponent electronic warfare, if the United States relies on military systems not intended for the now-crowded global spectrum environment. China will not necessarily dominate the manufacture of spectrum-using devices (and the markets that grow from it), but the United States will certainly not if it continues its current course.
The Gamble
Given its history of spectrum use in the twentieth century, the problem for the United States has always been how to reallocate excess spectrum from government use to commercial use. One approach developed by previous administrations, starting with the Bush administration in 2003, was a formula that worked well. Federal agencies would identify spectrum for reallocation. The Federal Communications Commission (FCC) would use its auction authorities to sell this spectrum. A portion of the funds generated by auction would go to a Spectrum Relocation Fund that agencies would use to cover the costs of the move. As with any bureaucratic endeavor, there are complicating details, but in essence, this smoothed the path for moving to twenty-first-century spectrum allocations that better served national security and the economy. Unfortunately, this formula of using auctions to compensate for spectrum reallocation is broken, but this points to several actions Congress and the administration can take.
Clearing spectrum for commercial use has always been politically difficult. The amount of spectrum available for any use is fixed, and the story of spectrum technologies has been cramming more users into the same frequency while providing the same quality of service. One possible solution to spectrum scarcity is to improve technologies that allow for greater spectrum sharing. Spectrum sharing is currently not adequate for the needs of the large scale, commercial networks that future economic growth will require, but an effective technology for spectrum sharing would obviate the need for difficult political decisions while keeping the United States competitive. There are both positive and negative precedents for this approach.
On the positive side, it is usually a bad idea to bet against technology (although the decision of other leading countries not to wait for improved spectrum sharing is troubling). Improvement in artificial intelligence tools could make spectrum sharing easier. An earlier telecom technology called Code Division Multiple Access (CDMA) was a 1990s digital cellular technology that allowed multiple users to communicate over the same frequency band simultaneously. CDMA has been largely replaced, first by LTE and now by 5G, but it shows that the concept of sharing based on advanced technologies is possible and, in combination with artificial intelligence tools, could offer greatly increased spectrum sharing sometime in the next decade.
The converse is that improved spectrum sharing has also appealed to earlier administrations, yet it did not arrive in a timely fashion. A 2003 CSIS Spectrum Commission report that helped shape the Bush administration’s spectrum policy stated, “the same technologies that create exploding demand can provide the solution to this problem, by letting more devices use spectrum than is now possible.” It has been two decades, however, and progress in creating these sharing technologies and creating a business case has been slow.
A recent effort involved Citizens Broadband Radio Service (CBRS), an initiative of the FCC to enable shared access to spectrum for both commercial and federal users. However, CBRS is lightly used, while spectrum made available for licensed commercial use at the same time has been deployed extensively. This can be explained by some of the challenges with spectrum sharing that technology may not be able to address. These include licenses with small geographic scope or limited timescales that remove the economic incentives and certainty that companies need to invest and build. However, companies are unwilling to take the risk of investing heavily in a shared resource given the economic risks of increased investment uncertainty and high transaction costs.
The policy problem for the United States is how to free up spectrum for use now, to remain competitive, while preparing to expand spectrum sharing in the future. Dynamic spectrum sharing is not ready for prime time, but it could likely be workable by the end of the decade, given probable advances in research, testing, and artificial intelligence tools.
A program of research and testing could be built on the National Spectrum Strategy Implementation Plan. A precedent for an expanded effort could be Semiconductor Manufacturing Technology (SEMATECH). The group of chip companies, created in 1986, was originally funded by the Defense Advanced Research Projects Agency (DARPA) but required participation and financial support from semiconductor companies in exchange for access to research. Requiring financial support ensured companies have skin in the game.
The fundamentals of a mutually beneficial “trade” for spectrum has two parts. It involves identifying and reallocating spectrum for auction now in exchange for cooperation to build dynamic spectrum sharing later. Three senators have introduced legislation that offers a path to fix the United States’ spectrum problem. The Spectrum Pipeline Act of 2024 calls for the National Telecommunications and Information Administration (NTIA) and the FCC to identify 2,500 MHz now allocated to federal use for reallocation to non-federal use. The Spectrum Pipeline Act would also restore FCC auction authority, as would other legislation being drafted in the House and Senate. The 2,500 MHz benchmark is best seen as an opening bid rather than a firm demand. The bill allocates half of this spectrum to commercial licenses acquired under auction.
The Spectrum Pipeline Act
In exchange for reallocation of federal spectrum, the leading companies using spectrum would commit to support an expanded research initiative for dynamic spectrum sharing, perhaps under the umbrella of a collaborative effort between DARPA (with programs like the Spectrum Collaboration Challenge or Shared Spectrum Access for Radar and Communications) and the National Institute of Standards and Technology’s (NIST) Institute for Telecommunication Sciences (ITS). In essence, this builds on proposed legislation. Support for research would be an additional cost on top of the auction price, but it could ensure the participation of actual commercial users in dynamic spectrum-sharing research. Auctions could be held incrementally, with the release of additional tranches of spectrum for auction conditioned on participation in the research effort. A new spectrum policy should focus on three primary goals:
- Restore FCC auction authority.
- Create public-private entities for spectrum-sharing research (modeled on SEMATECH).
- Identify spectrum for reallocation (perhaps in tranches) based on the Spectrum Pipeline Act.
There is no doubt that the future demands on spectrum will require effective sharing. But there should be no doubt that competition with China requires making licensed mid-band spectrum available now. This would also hedge the bet on when improved spectrum-sharing technology will become available. Even more than earlier infrastructure, digital infrastructure is tightly interconnected. Lag in one area and other areas suffer as well, slowing the entire economy. The United States is on track to trip up, and while breathless claims of a race with China can be overheated, China and other countries are taking a different approach when it comes to 5G that appears likely to give them an advantage.
The next generation of telecommunication networks will have the same transformative effect on the economy and national security as electricity or the internet in the twentieth century, if the United States can seize the 5G opportunity. Right now, it is not on a path to do so. Whether the United States leads or lags in this transformation remains to be seen. The countries that do best in deploying digital infrastructure will be economic powerhouses. Policymakers know that their decisions on spectrum allocations create real risk to national security, innovation, and economic growth. The United States has gambled alone on technologies in the past and won. With the right public and private support, spectrum sharing could be another winning gamble. But the wise investor tries to hedge their bets. Auction now, share later is the safest approach in a race the United States cannot afford to lose.
James A. Lewis is senior vice president and director of the Strategic Technologies Program at the Center for Strategic and International Studies in Washington, D.C.