International Trade Today is a service of Warren Communications News.
Unfair to AT&T?

Internet Speeds Still Being Delivered As Promised, FCC Finds

Most ISPs are holding steady in the FCC’s third wireline broadband speed report (http://fcc.us/YkWygE). Industry representatives were generally pleased with the results, with the test showing ISPs continuing to meet or exceed advertised speeds even though consumers are migrating to ever faster speed tiers. Frontier Communications got a shout out as the most improved, increasing 13 percentage points from the last reporting period for peak download speeds. But some questioned a specific reporting methodology they said makes AT&T look like it’s not meeting its advertised speed.

Sign up for a free preview to unlock the rest of this article

If your job depends on informed compliance, you need International Trade Today. Delivered every business day and available any time online, only International Trade Today helps you stay current on the increasingly complex international trade regulatory environment.

When the FCC released its first Measuring Broadband America report in 2011, it found that the average ISP delivered 87 percent of advertised download speeds during peak hours. In Friday’s report, which used data collected during September, that number was to 97 percent. The report also said consumers are continuing to subscribe to higher speed services, and the FCC said it plans to “evolve” its testing program to handle higher speeds as they are rolled out. The average subscribed-to speed is 15.6 Mbps, about 20 percent higher than last year, the report said.

Of the 14 ISPs included in the last report, a handful are well below the average 97 percent delivered speed. According to the report, Windstream delivered 81 percent of its advertised speed, Qwest 82 percent, Insight 85 percent, CenturyLink 87 percent and AT&T 87 percent. These ISPs demonstrated either no change from the last report, or a slight drop off. CenturyLink and Qwest are part of the same company.

Windstream’s 81 percent figure was a three percentage point drop overall in delivered speeds during peak hours, but that’s an average across all its offerings. “Performance varies among service tiers,” the report said. It said “Windstream’s 12 Mbps tier delivers 72 percent of its advertised speed” during peak periods, “a low across all ISPs and speed tiers.” Windstream had seen a 200 percent increase in Internet usage over the past two years, and spent $2 billion in network enhancements to help address growing demand for capacity, said Senior Vice President of Government Affairs Eric Einhorn. “We do recognize the need to enhance and expand broadband service so that it is available to all Americans, regardless of where they live.” That’s why the telco is “the largest single participant in the federal broadband stimulus program,” he said in a written statement. “Without appropriate federal support, Windstream simply cannot address all areas of its network."

AT&T’s average peak hour download speed of 87 percent of advertised speed shouldn’t imply that the telco isn’t meeting its advertised speeds, a spokesman said. “AT&T’s consumer wired broadband Internet access services are offered in discrete speed tiers,” he said. “For example, a customer with AT&T U-verse High Speed Internet Max should expect service capability speeds between 6.1 and 12.0 Mbps."

"It does seem to me AT&T’s getting caught in a technicality,” said Stifel Nicolaus analyst David Kaut. “It would be fairer to compare to their average advertised speed.” Kaut said he understands that the FCC is “trying to have a consistent comparison” between providers, and “holding everybody to the upper portion” of their advertised speed. But AT&T is unique among providers in explicitly telling consumers how low they might expect their speeds to go, he said. If the FCC used AT&T’s average advertised speed in its calculation, “in essence AT&T would be coming in a little bit above,” he said.

An FCC official said the methods used in this report are consistent with its practices in the first and second reports, and AT&T’s numbers have been reported in the same way this whole time. AT&T hadn’t complained of any inaccuracies in the results, and it’s being treated like any ISP that promises speeds of “up to” a certain number, the official said.

"I'm not sure why the FCC has to undertake this measurement reporting function, as opposed to Consumers Union or some other private organization,” said Free State Foundation President Randy May. “Especially at a time when government resources are constrained, it just seems the FCC ought not to be taking on tasks that private entities can perform -- assuming there is a demand for the testing.” Regardless, May said he’s “struck by the fact that most operators are delivering very close to, or meeting, advertised speeds; that consumers are migrating to faster speeds; and that satellite broadband service is improving. All good news, which should be acknowledged.”

CenturyLink is “proud” of its “first-class broadband network,” and the report reflects the “substantial investment and on-going improvements that CenturyLink has made in delivering quality broadband services to customers,” a spokesman said. Frontier Executive Vice President-External Affairs Kathleen Abernathy said the report confirms that company’s “ongoing investments are paying off, and will continue to pay dividends for our broadband customers.” The “good news” from a cable perspective is that providers are delivering speeds that consistently meet or exceed advertised numbers,” said NCTA in a blog post (http://xrl.us/bog8r5). Verizon said the report, which found that FiOS delivery consistently outperforms its advertised speed, “reaffirm the results from the past two FCC broadband performance tests” that FiOS is “blazing-fast.”

"I'd love to see something more comprehensive” able to break down the results by service package, as did a report from Akamai, said engineer Thomas Gideon of New America Foundation’s Open Technology Institute’s M-Lab, which used its servers to measure the speeds in the FCC study. Gideon said the report as a whole “seems pretty consistent with the last one.” The FCC had put in place some additional “on-net” servers to act as redundant back-ups to M-Lab, but those weren’t needed, Gideon said. “M-Lab made good on its continuing commitment to improve operations, monitoring and response in service of our entire research community, including SamKnows and the FCC. I see this as validating my view that focusing constructive efforts on improving existing resources is sufficient before considering adding more servers, the operational policies of which were the subject of much debate last year.”