In a recent report, the Federal Communications Commission revealed a worrisome paradox in America’s cellular networks: they reach nearly everyone, but not nearly everywhere. While 95 percent of the U.S. population can receive 5G service from at least one carrier, only 69 percent of the nation's road miles and 39 percent of its land area have such coverage. In rural America, the numbers fall to 60 percent and 38 percent. This is the rural-mobile divide. It is not about whether a household can obtain broadband at home. It is about losing mobile service while driving, farming, hiking, fighting a wildfire or responding to an emergency.

The conventional solution is to build more cell towers. Some are still needed. But extending terrestrial networks to every remote road, ravine and mountaintop would be economically irrational. The FCC has approved a plan to provide up to $9 billion to close the remaining gaps in America’s 5G cellular networks. But instead of treating every dead zone as a construction problem, we should consider whether the last cell towers should instead be placed in orbit. Technology advances now offer a promising alternative to cover this last and most expensive frontier.

Direct-to-device satellite service allows an ordinary compatible cellphone to connect directly to a satellite when no terrestrial tower is available. The satellite effectively becomes a cell tower in space. No satellite phone, dish or other specialized equipment is required. All three national carriers see satellites becoming an integral part of mobile networks. T-Mobile already offers this service through Starlink. The service is included with some T-Mobile plans and sold for $10 a month per line to other eligible users, including subscribers of competing carriers. AT&T and Verizon have their own satellite partnerships.

And in May, all three carriers announced plans for a joint venture to expand direct-to-device coverage, pool spectrum and make it easier for satellite operators to reach their customers. Direct-to-device satellite service doesn’t replace terrestrial 5G. Its capacity is limited, and initial offerings emphasize messaging, emergency communications and selected low-bandwidth applications. Voice and more capable data services are in development. But even basic connectivity can transform a dead zone into a place where someone can summon help, report an accident or tell family members they are safe.

The question is how to deploy that capability quickly and affordably. The answer begins with the FCC's 5G Fund for Rural America. Subsidizing cellular services in areas where private investment alone won’t support deployment was originally proposed in the 2010 National Broadband Plan. The FCC established the 5G Fund in 2020, making available up to $9 billion in Universal Service Fund support over 10 years for mobile service in high-cost areas. The existing rules require fund recipients to deliver conventional 5G speeds that today’s satellite service cannot yet provide.

But before committing to a decade of conventional construction, the agency should create a direct-to-device tier to buy basic mobile coverage in places that towers cannot economically reach, then let competing terrestrial-satellite partnerships bid to provide it. Instead of prescribing the infrastructure carriers must build, the FCC should buy the coverage it wants and let providers compete over how best to deliver it. How would it work? The FCC could ask carriers to bid for the minimum subsidy they need to provide satellite coverage in designated unserved areas without charging the premium.

Carriers would receive support only in return for meeting clear requirements for availability, reliability, emergency communications and compatibility with widely used handsets.

That condition would advance the Communications Act's promise that consumers in rural and high-cost areas should have access to communications services that are "reasonably comparable" to those available in urban areas, at "reasonably comparable" rates. “Reasonably comparable” doesn’t require identical service or prices. Courts allow the FCC to balance comparability with affordability and responsible stewardship of the Universal Service Fund. Here, the balance is clear. Spending enormous sums to extend a terrestrial network over the most challenging terrain would drain resources from places where towers remain economically viable. Leaving that terrain uncovered, on the other hand, perpetuates the rural divide.

And charging those who most often encounter dead zones an additional $120 a year makes those least well served pay extra for basic connectivity. The FCC’s rules require recipients of 5G Fund support to offer rural service at rates reasonably comparable to urban rates. Offering the same rates, terms and conditions in urban and rural areas can satisfy that rule. So, extending the requirement to subsidized direct-to-device coverage would be an evolution of existing universal-service policy, not a new system of price regulation.

Universal-service policy has always adapted to changes in network technology and economics. direct-to-device changes both. The last cell tower needed to close the rural divide no longer has to be built on the side of a mountain. It can be deployed in space, using a federal fund already authorized to bring mobile service to remote locations the market cannot economically reach. Blair Levin is a former chief of staff of the Federal Communications Commission and directed the writing of the 2010 National Broadband Plan. Larry Downes is the co-author of “Big Bang Disruption” and other books on disruptive innovation.