New York State

What we know about rural broadband access

Senator Chuck Schumer is criticizing the Federal Communications Commission for relying on internet providers to self-report their service speeds, which critics say affords companies to opportunity to lie, claiming to provide better service than they actually do.

Cables being embedded into the ground to provide a broadband connection in a rural area.

Cables being embedded into the ground to provide a broadband connection in a rural area. Amy Johansson/Shutterstock

Broadband access across the country is improving, the digital divide between urban and rural America has narrowed substantially and the number of residents lacking high-speed internet access is falling. That’s according to a new report from the Federal Communications Commission, but some officials and lawmakers – including Senate Minority Leader Chuck Schumer – are skeptical of the FCC’s claims.

At a press conference on Sunday, Schumer criticized the agency after the release of its 2019 Broadband Deployment Report, calling for more scrutiny of the data. He and other critics think the rosy picture it paints doesn’t accurately reflect the situation on the ground. “While we live in an era of faster and faster, the reality of internet speed across New York is that it may move more like molasses than lightning,” Schumer said.

Much of the criticism of the report focuses on the method and sources for obtaining data on broadband access. The FCC relies on internet providers to self-report their service speeds, which critics say affords companies the opportunity to lie, claiming to provide better service than they actually do. “It's kind of a scandal that's hiding in plain sight,” said Tim Wu, a professor at Columbia Law School who advised former New York Attorney General Eric Schneiderman on technology, internet and competition policy and enforcement. “Everybody knows this.”

The FCC reports that the number of Americans without access to high-speed broadband has dropped from 26.1 million at the end of 2016 to 21.3 million at the end of 2017, marking of drop of more than 18%. It also found that 4.3 million Americans who gained access live in rural areas. The minimum figure that the FCC categorizes as “high-speed” is download speeds of 25 megabits per second and uploads of 3 Mbps. But according to a study conducted by Microsoft, the population of those who are still not using internet at those speeds is much higher – 162.8 million across the country, according to Microsoft’s internal data from 2018. Microsoft uses the same download speed benchmark as the FCC to define “high-speed.”

Microsoft’s data does not measure how many people have access to high-speed broadband, as the FCC does, but how many people are using high-speed internet. The number of people who use broadband internet and the number of people who have access to it are not equivalent statistics. But Microsoft argues that the gap between the two figures suggests that the FCC’s picture of broadband access is at least somewhat overstated. “These significant discrepancies across nearly all counties in all 50 states indicates there is a problem with the accuracy of the access data reported by the FCC,” Microsoft’s chief data analytics officer John Kahan told the technology news outlet TechCrunch earlier this year.

Schumer called attention to discrepancies between the two reports in New York, specifically. “The FCC has said internet speeds are up to standards in its report. But Microsoft did its own report and it shows that over four and a half million New Yorkers and Long Islanders are not getting the speed on the internet that the carriers say they’re getting,” he said on Sunday, referring to Microsoft data that shows more than 4.5 million on New York City and Long Island combined are not accessing high-speed data. Across the state, Microsoft’s data says that 8.7 million people are not accessing internet at broadband speeds, but the FCC’s new report says that the number of people in New York without access to high-speed internet is closer to 300,000.

According to the FCC report, 87.1% of New Yorkers in rural areas have access to high-speed internet, compared to 99.9% in urban areas. Some of the worse-off counties include Hamilton County, where 21.6% have high-speed internet access, and Allegany County, with 72.9%. Data from Microsoft on how many people in Hamilton County use the internet at broadband speeds is roughly in line with the FCC’s figure on how many people have access to it. In Allegany County, however, the FCC data shows that roughly 12,000 people lack access to high-speed broadband, while Microsoft’s data indicates that roughly 34,000 people aren’t using high-speed broadband.

It’s not just the source of the data used in the FCC’s broadband deployment report that critics complain about, but the methodology. The FCC divides coverage maps into census blocks created by the U.S. Census Bureau, and if an internet provider shows that at least one person in each block can connect to high-speed internet, the whole block is counted as covered. A single census block can include anywhere from 600 to 3,000 people or between 240 and 1,200 housing units. A “covered” census block could theoretically contain one household with high-speed internet access and 1,199 without it. “You can always cut the cake in different ways and it gives a different result, it makes things look different ways,” Wu said.

The possibility that internet service providers might misreport data to the FCC is not just theoretical. When a preview of the FCC’s 2019 Broadband Deployment Report was released earlier this year, the media advocacy group Free Press found that conclusions in the report were based on erroneous data. An internet service company called BarrierFree claimed to service roughly 62 million people in the Northeast – including coverage of every single census block in New York – which would make BarrierFree the nation’s fourth-largest internet service provider in terms of population coverage. But Free Press found that just months before making this claim, the company wasn’t serving a single census block in the country. The FCC report released last month is adjusted to reflect that BarrierFree’s claim was erroneous, but the commission has nonetheless drawn criticism for failing to detect the error in the first place. BarrierFree did not respond to a request for comment.

The FCC’s two Democratic commissioners have criticized the latest report and the conclusion made by Chairman Ajit Pai that this data shows that broadband is being deployed in a reasonable and timely fashion. Asked about the criticisms, a spokesman for the Republican-majority FCC pointed to a report from December 2018 which shows that for most major broadband providers, measured broadband speeds were 100% or better than advertised speeds during peak hours. “The FCC’s Measuring Broadband America Program conducts ongoing, rigorous tests of broadband speeds received by thousands of consumers who represent a sample set of the population nationwide who subscribe to Internet service from providers serving 80% of the residential marketplace,” the spokesman wrote over email.

There’s no shortage of criticism over the fact that the FCC’s broadband deployment uses self-reported data from internet service providers – which they do with a form called Form 477 – and even the FCC has acknowledged its flaws. “We recognize that the Form 477 data collection is not perfect, and the commission has an open proceeding considering ways to improve the accuracy and granularity of that data collection,” the recent report reads.

But there’s not necessarily an obvious alternative to the current system, given the difficulty of testing and verifying internet access data across all parts of the country. “They don't necessarily have the resources, or even the desire, to want to spend all their time trying to figure things out,” Wu said of the FCC. “The reality on the ground is extremely hard to find out. It takes a huge effort to find out what's really going on.”

That doesn’t mean others haven’t tried to crack down on how internet companies report or advertise their speeds. New York’s attorneys general have made a habit it of it. In 2015, Schneiderman launched a probe into the accuracy of advertised internet speeds and whether three of the state’s biggest providers – Verizon, Time Warner Cable and Cablevision at the time – were misleading their customers.

Then, last December, then-state Attorney General Barbara Underwood reached a $174.2 million settlement with Charter, finding that the company, which initially operated as Time Warner Cable, failed to provide customers the fast and reliable service it advertised. In addition, Underwood also reached an agreement at the time with four other internet service providers – Altice, Frontier, RCN and Verizon – to institute reforms including substantiating their speed claims and advertising them accurately. Representatives from Charter and Comcast declined to comment on the FCC report or its reporting method, and Verizon did not return a request for comment.

Gov. Andrew Cuomo also launched his Broadband for All initiative in 2015, committing $500 million in subsidies to private companies to build out broadband infrastructure in underserved and unserved areas, with the goal of statewide broadband by 2018. Nearly 2.4 million residents have gained broadband access since, but the some of the deadlines have been pushed back to 2021. One obstacle was cleared when the state reached an agreement with Charter for the company to remain in the state and provide high-speed internet to 145,000 residences and businesses upstate. Cuomo’s office did not respond to a request for comment.

Despite these state-level efforts at holding providers accountable, there’s still little clarity about what the actual state of broadband access is on the ground. “Because the FCC doesn’t particularly seem to be too interested in the public interest, it's all these stop-gap measures – private companies like Microsoft, state AGs, Congress getting involved,” Wu said. “It's hard to get a real picture of what's going on, so the opportunity for fraud is there, and people take it.”