FCC approves new method for tracking broadband's reach

WASHINGTON--As expected, federal regulators on Wednesday voted to overhaul the way they measure how widely broadband is available across the United States.

For years, the Federal Communications Commission has been drawing up reports on the state of U.S. Internet access availability based on methodology that considers 200 kilobits per second (Kbps) service to be "high speed"--and such access to be widely available even in ZIP codes that may, in reality, house only one connection.

The decision to move away from that methodology is potentially significant. Critics, both inside and outside the agency, have charged that the inadequacy of data that the FCC collects semiannually from Internet service providers hinders both the government's ability to set smart pro-broadband policies and could slow investment on the technology side. It could also help federal regulators determine whether the United States is really as far behind in broadband penetration as some international studies have suggested during the past few years.

"This is really the first step toward the national broadband strategy that we so desperately need."

--Jonathan Adelstein, commissioner, FCC

If not for good government data, "our economy would come to a screeching halt," said Commissioner Michael Copps, a Democrat. For example, manufacturers depend on unemployment and gross domestic product figures to set their production targets, and schools and hospitals rely on U.S. Census numbers to project demand for their services, he said.

"When companies and investors put money into e-commerce or voice over Internet Protocol or Internet video...they need to know what kind of broadband infrastructure America actually has," Copps said.

Democratic Commissioner Jonathan Adelstein said, "This is really the first step toward the national broadband strategy that we so desperately need."

Despite his support for the new data collection method, FCC Chairman Kevin Martin said he believes the United States has made incredible strides in broadband deployment since he joined the commission in 2001, with the number of lines growing from 9 million to more than 100 million. Still, he acknowledged, "there is certainly more work to be done."

The FCC, as is typical, won't release the full text of the changes it adopted for a few weeks, but here's a rundown of major components described at Wednesday's meeting:

• 200Kbps speeds are no longer considered "broadband." Until this point, the FCC has considered any service that produces 200Kbps speeds in the upload or download direction to be "high speed." With Wednesday's vote, that methodology is no more. Now, 768Kbps, which is the entry-level speed offered by major DSL providers like Verizon, will be considered the low end of "basic broadband," a range that extends to under 1.5Mbps.
• Broadband service speeds will have to be reported both for uploads and downloads. Previously the FCC had six big categories of broadband speeds, and they effectively only tracked download speeds. Now the agency says it will require reporting on upload speeds. Pro-regulatory advocacy groups like Free Press say that's a necessary step in part because of Comcast's admitted throttling of peer-to-peer file-sharing uploads.
• Upload and download speeds will have to be reported in a more specific way. At the moment, the broadband speeds most commonly offered by cable and telephone companies are lumped into two major categories: those between 200Kbps and 2.5Mbps, and those between 2.5Mbps and 10Mbps. The FCC's new rules would require them to be broken down further, in an attempt to address charges that the current buckets have the potential to overstate the number of high-end subscriptions and understate the number of low-end subscriptions. Those new tiers will be: 1) 200Kbps to 768Kbps ("first generation data"); 2) 768Kbps to 1.5Mbps ("basic broadband"); 3) 1.5Mbps to 3Mbps; 4) 3Mbps to 6Mbps; and 5) 6Mbps and above.
• ISPs will be required to report numbers of subscribers, and at the census-block level. Under the current methodology, ISPs report only the number of ZIP codes in which they have at least one subscriber, and they report numbers of lines nationwide. Now they'll have to report the number of subscribers in each census tract they serve, broken down by speed tier. The FCC decided to use census tracts because researchers may be able to use other demographic statistics collected by the U.S. Census, such as age and income level, to gain insight about what drives broadband penetration rates.
• ISPs will not have to report the prices they charge....yet. Democratic commissioners and liberal consumer advocacy groups had argued such a step is necessary to give consumers an idea of the value they're getting for their money--and to compare U.S. prices to those for comparable services abroad. Copps said on Wednesday that he continues to believe it's a "mistake" to omit that requirement, and Adelstein also voiced concern. But a majority of the commissioners opted to push that decision off until another time and gather more comments.

Each of the five commissioners voted in favor of adopting the order, although some attached reservations about some portions of the rules. Adelstein said he would have liked to see the commission require that ISPs distinguish between residential and business customers when doing their reporting. Republican Commissioner Robert McDowell said he was concerned that some of the definitions contained in the rules--particularly that of broadband--could have negative long-term effects.

"Government cannot outguess the genius of free markets, nor should it try," McDowell said.

Representatives from the cable and telephone industry had advised the commission against making major changes to its data collection methods. They said they would not be able to comment on the FCC's vote Wednesday until after reviewing the full text of the order.

The old method's last gasp
In an ironic twist, at the same meeting, the commissioners narrowly voted to adopt the FCC's latest report about the state of American broadband deployment--except based on the old methodology that they went on to revamp. Because of that, Copps and Adelstein ripped apart the report and said they couldn't support its conclusions. (Martin, McDowell, and Republican Deborah Tate voted for adoption of the document.)

The report (PDF), which covers the first half of 2007, concluded that "broadband services are currently being deployed to all Americans in a reasonable and timely fashion."

High-speed lines--meaning, mind you, capable of 200Kbps or greater data transfer speeds--grew from 82 million to 100 million lines during that time, the FCC said. Its report also found that an Internet service provider reported having at least one connection in 99 percent of the country's ZIP codes, and that 99 percent of the American population lives in those ZIP codes.

Copps, for one, called the ZIP code methodology "stunningly meaningless."

"I'm happy we're starting to change our benchmarks," he said, "but my goodness, how late in the day it is."

The FCC's actions drew mixed reviews from groups who have been pressing for better broadband data and Net neutrality rules.

Gigi Sohn, the president of Public Knowledge, one such group, commended the FCC's new data collection plan, although she said she would have preferred to see price data included and information about residential and commercial customers separated. She also deemed it a "mystery" that the FCC also chose to issue the broadband availability report "when, mere moments later, the Commission admitted the inadequacy of the information."