These days, a high-speed internet connection is like electricity or running water — a convenience of modern life that most Americans take for granted. But, just as the light switch and the tap only work because of huge historic investments in the power grid and municipal water and sewage systems; broadband internet depends on an expensive terrestrial infrastructure. These huge investments haven’t been spread evenly. In many parts of rural America, the broadband infrastructure just isn’t there — at least not on the ground. “Without access to broadband, entire communities are increasingly left behind in today’s information-driven economy,” announced the American Broadband Initiative (ABI) in February.

The ABI is the Trump Administration’s latest addition to a long string of U.S. policy efforts aimed to connect rural America with essential services. In the 1930s, as the electric grid and the telephone system were being built out in the United States, it quickly became apparent to economists that there was no business case for service carriers to run their wires to every last halt and hamlet. In response, President Franklin Roosevelt signed legislation in 1936 to create the Rural Electrification Administration (REA), which provided loans and grants to rural electric and telephone companies to get rural residents connected. It was a policy decision that the federal government should step in where the market failed and subsidize the big carriers to provide service where it would otherwise be unprofitable to do so. In the 21st century, the internet has replaced the phone as the medium of connectivity, but there is still no business case for connecting remote rural communities. The successor to the REA, the U.S. Department of Agriculture’s Rural Utilities Service doles out about $800 million annually in loans and grants to companies bringing terrestrial high-speed internet services to rural communities. But that is an order of magnitude below the annual $8.8 billion that the U.S. Federal Communications Commission (FCC) raises through its Roosevelt-era Universal Service Fee — more than half of which is designed to be spent subsidizing high speed internet for Americans in isolated rural areas.

“We live in a divided nation when it comes to broadband access,” U.S. Rep. Mike Doyle, a Democrat from Pennsylvania, told a congressional hearing last year. “All too often people living in urban areas are the digital haves, where as those living in rural areas are being left behind with few or no choices, higher prices, and lower speeds.” He might have been thinking of Linefork, Kentucky, deep in the foothills of the Appalachian mountains. In communities like this, geography and demography — an isolated or remote location and a sparse population — conspire to make it an uneconomic proposition for the large terrestrial providers to lay cable. And there are a lot of such places. Across the U.S., the FCC estimates that just over 30 percent of rural residents and a total of more than 24 million Americans lack access to terrestrial broadband — which the agency defines as 25 Megabits-per-second (Mbps) download and 3 Mbps upload speeds.

Terrestrial internet speeds are basically a function of how far the consumer is from a fiber connection, says Larry Thompson, Chief Executive Officer (CEO) of broadband engineering consultancy Vantage Point Solutions and a member of the FCC’s Broadband Deployment Advisory Committee, a panel that provides the agency with recommendations on expanding access to high speed internet. “That last piece, between the fiber and the consumer, is usually where your bottleneck is. On the fiber, your capacity is practically unlimited.”

In most big cities and many inner suburbs, fiber connections reach, if not into the home, at least to every block. Consumers routinely have access to speeds up to 1 Gbps. But those speeds fall off as the signal makes longer trips from the fiber to the home using legacy coaxial cable or — worse — telephone infrastructure, says Thompson. Customers in thinly populated areas who are further away from urban centers will experience slower terrestrial internet service.

Without high speed internet, many rural Americans are effectively shut out from the digital economy — and that increasingly means the economy as a whole. It’s no coincidence that the U.S. counties with the lowest broadband availability often suffer from the highest unemployment rates, according to a Microsoft analysis of U.S Bureau of Labor Statistics data published in December 2018. For the satellite industry, that represents an opportunity to do well by doing good, according to Mike Cook, Executive Vice President (EVP) of North American sales and marketing for Hughes — the world’s largest provider of satellite-based internet service. The company says it has almost two-thirds of the nearly two million subscribers in the U.S. “Our market is the unserved and the underserved,” Cook told Via Satellite, “the people in communities not served by cable.”

The great advantage satellite providers have over terrestrial competitors in reaching those out of the way places, as FCC Chairman Ajit Pai told Via Satellite last year, is that ”the topography of the Earth doesn’t matter to a satellite.” Running fiber to remote communities is “a very expensive proposition,” says Thompson. Cooperatives and other small rural providers “do a pretty good job” of providing broadband, but for the large, profit-driven telcos, “there is absolutely no business case for serving those rural communities, especially given the lower incomes that prevail there. The end-user revenue just can’t support the cost of building out there,” says Thompson.

A decade ago, the FCC estimated that it would cost $350 billion to bring fiber optic cable, to every home in the country. A more recent study from the Boston Consulting Group put the price of universal fiber connectivity at between $45 billion and $65 billion. Either way, “It’s completely uneconomic to build out cable to everyone,” says Hughes’ Cook.

For the past two decades, satellite internet providers have been viewed as a last resort to customers in rural communities. James Baller, founder and president of the U.S. Broadband Coalition, says that a satellite’s technological limitations have prevented operators from providing services comparable to those available in more densely populated areas. Satellite Industry Association (SIA) President Tom Stroup says that the industry has spent “tens of billions of dollars to innovate and increase broadband connectivity in the U.S. and across the globe.”

That investment has begun to bear fruit in the past few years as a new generation of GEO satellites have come on line. Hughes launched the Jupiter-2/ EchoStar 19 satellite at the end of 2016. Their main U.S. competitor, Viasat, followed with the launch of Viasat-2 in June 2017. This latest generation of satellites utilizes spot beams that deploys multiple, tightly focused beams, each fixed on a particular geographical area. “It’s a way of reusing the spectrum over and over again,” Viasat President and Chief Operating Officer (COO) Rick Baldridge explains. That matters because spot beams multiply the throughput capability of the satellite — increasing the speeds that satellite providers can offer their subscribers. It’s an order of magnitude increase, rather than the incremental improvements that can be achieved by boosting the efficiency of spectrum utilization, or improving coding algorithms.

The new spot beam technology “makes your [internet] service feel like your neighbor’s who can get cable,” boasts Rick Baldridge. As a result of this new capability, the satellite industry was, for the first time, able to report to the FCC that it offered internet service to U.S. consumers that met the agency’s 25 Mbps download/3 Mbps upload broadband benchmark. Hughes and Viasat’s next generation of satellites, Jupiter-3 and Viasat-3 respectively, are slated for launch in 2021. Once in orbit, they will offer U.S. consumers 100 Mbps-plus internet speeds—the same speeds terrestrial providers offer customers in major urban areas.

Last year, the FCC licensed nine companies to operate internet service from Non-Geostationary Orbit (NGSO) satellites. SpaceX plans this year to start launching nearly 12,000 satellites, in two Ka- and Ku-band constellations. The first 4,425 satellites will orbit at around 1,200 km (745 miles). The second, consisting of 7,518 V-band satellites orbiting at about 340 km (210 miles). The company says its Starlink internet service will offer “fiber-like” speeds to subscribers, with latency of between 25 and 35 milliseconds.

However, SpaceX is up against a very tough clock. FCC rules state that the company will have to launch at least half of its constellation of Ka- and Ku-band satellites by March 2024, or else lose the right to put the remainder into orbit at all. If SpaceX makes that first deadline, the company will then have another three years to launch the rest. The same tight timetable applies to the second constellation, which the agency greenlighted in November.

At the time this article is being written, OneWeb has plans to launch the first six of its 650 LEO satellites at the end of February on a Soyuz rocket from the Kourou spaceport in French Guiana. Regular monthly launches will then start in September 2019. The other seven constellations licensed by the FCC last year include a 140-strong system that Kepler Communications wants to use to offer global internet connectivity for Internet-of-Things (IOT) devices, as well as a 117-satellite constellation from Telesat and a 78-satellite LeoSat constellation — both designed to provide high-speed, low-latency broadband internet in the U.S. market. LEO constellation satellite internet providers hope to achieve what Nokia did for mobile telephony in the mid 1990s — make services cheap and ubiquitous.

Cook’s company, Hughes, is an investor in OneWeb and a technology partner. He thinks the new LEO systems will coexist with GEO satellites. Baldridge agrees, stating that Viasat also doesn’t view LEO constellations as a competitive threat. In fact, Viasat is building advanced phased array antennas for the LEO constellations. To ensure that there is always a satellite overhead for subscribers, Baldridge says that LEO constellations have to be “fairly evenly distributed around the world, but the people are fairly concentrated,” which means an inefficient use of resources as the satellites pass over oceans or other uninhabited areas.

Public awareness hasn’t caught up with these advances in technology, says Baldridge. “There’s definitely a lagging perception. When we get in front of [regulators] and actually show them what we’re doing, they’re surprised. That tells you that the market doesn’t really understand this yet.”

Stroup points out that in most services for which consumers use their internet — things like streaming video, web-surfing and email — latency doesn’t matter. “The vast majority of internet-based consumer services are not reliant on low-latency service,” he says. In any case, Baldridge says that a new generation of satellites offering consumers internet service from much lower orbits will have concomitantly lower latency — comparable with mobile broadband services like 4G Long Term Evolution (LTE).

Counting the Digitally Deprived

The digital divide is real, but how many people live on the other side? How many Americans lack access to high speed internet today? The short answer: It’s hard to know exactly, and the way the government currently counts them is widely acknowledged to be inadequate.

Section 706 of the 1996 Telecommunications Act requires the FCC to report annually on whether “advanced telecommunications capability” is being deployed to “all Americans in a reasonable and timely fashion.”

The act defines that as “high-speed, switched broadband [service] that enables users to originate and receive high-quality voice, data, graphics, and video telecommunications using any technology.”

From the act’s passage until the Obama administration, the FCC defined broadband service as anything exceeding download speeds of 200 Kbps. In 2010, the agency adopted a minimum speed of 4 Mbps download and 1 Mbps upload.

In 2015, the agency rewrote its standards again, adopting the current 25/3 Mbps speed minimum. But it collects data about connection speeds that Americans have access to across the country in several tiers — which can result in a bewildering variety of figures.

And all the data is based on self-reporting from providers, which critics charge is self-serving.

In its 2018 Broadband Deployment Report, the FCC included some satellite services in its 25/3 tier data for the first time, concluding that about 14 million Americans — fewer than five percent of the population and the lowest number ever — mostly in rural areas, lacked access to broadband services.

“The FCC has arbitrarily set a broadband standard that makes it appear as though [Americans] have much better service than we do,” said Larry Thompson, CEO of broadband engineering consultancy Vantage Point Solutions and a member of the FCC’s Broadband Deployment Advisory Committee. Moreover, the data is all based on self-reporting by providers, who submit annual figures to the FCC, classifying their service speeds as “up to” a certain number.

“The FCC relies on self-serving data submitted by [providers],” James Baller tells Via Satellite, “and it has no sound mechanisms to confirm their accuracy.” Baller, an attorney, was one of the early advocates for wider broadband access a decade ago, and was founder and president of the U.S. Broadband Coalition. Moreover, Baller adds, “in many rural areas, infrastructure is so bad that even if speeds ‘up to’ 25/3 Mbps are theoretically available, they will rarely achieve that standard, and certainly not on a sustainable basis.”

More importantly, say critics, the FCC counts an entire census block as having broadband access as long as any provider reports even a single address in it as served. In sparsely populated rural areas, a single census block can cover many square miles and there’s no guarantee that everyone in it could get service, even if a provider is reaching one of their neighbors. VS