Carriers Say FCC Must Not Allow Outsiders Access to Key Broadband Data

"The comments filed in this proceeding overwhelmingly oppose Free Press’ Request for access to individual providers’ Form 477 Data,” Verizon said in reply comments. “As the comments filed by numerous broadband providers and industry trade associations explain, denying Free Press’ request is the right answer in light of the longstanding Commission and court precedent concluding that reporting providers’ Form 477 data should be protected because it is confidential and commercially sensitive.”

Only the New America Foundation sided with Free Press in the initial comment round, Verizon said. But New America’s comments “do not offer any legitimate basis for departing from the longstanding Commission and court precedent concluding that Form 477 data should be protected,” Verizon said: “Although New America Foundation claims that the Commission could impose ‘limitations in terms of how the data are used’ and researchers would only be ‘able to leave with the aggregate statistical analyses,’ these measures would not eliminate the potential for competitive harm.”

Free Press cited language in the National Broadband Plan. The plan says, in the FCC’s words, that the commission “should have a general policy of making the data it collects available to the public, including via the Internet in a broadband data Depository,” except when the data are competitively sensitive or protected by copyright, the group said. “Free Press’ request is the embodiment of this recommendation,” Free Press said. “We are asking the Commission for a protective order to make Form 477 data currently deemed as confidential available for research purposes, as the Commission has already done in numerous other proceedings, including the National Broadband Plan itself. Free Press has signed numerous other protective orders, and has completely adhered to all terms for confidential treatment."

Sprint Nextel also opposed release of the data, saying “the data, which Free Press seeks to examine and which the Commission has consistently safeguarded from disclosure, are highly confidential and disclosure would be competitively harmful."

"The confidential information contained in the FCC Form 477 reports must be protected,” said a group of small wireless carriers represented by the Blooston law firm. The carriers said no good would come from allowing public interest groups to view the data. “While Free Press, Media Alliance and New America may be frustrated by, and find fault with, the Commission’s own internal analyses of the Form 477 data, there is no indication that Free Press’ analysis will be better."

Comments were also due this week on a proposal by the FCC Office of Strategic Planning & Policy Analysis for a 10,000-household testing and measurement project on residential fixed broadband to be done by SamKnows Ltd. U.K. regulator Ofcom used the company for a similar project in its country.

The Open Internet Coalition said collection of such data is critical. “Currently … there is a lack of reliable, up-to-date, and readily-accessible information about the actual performance of broadband service offerings available in the marketplace,” the group said. “Broadband measurement is an essential tool for filling this information gap. Along with empowering users, broadband measurement is critical for advancing network research and facilitating the continuing evolution and development of the Internet."

AT&T said it generally supports the measurement proposal but a few changes are needed. The company said the FCC should release additional details about the M-Labs infrastructure to be used for the measurements, so it can be evaluated by those interested. SamKnows is a member of the Measurement Labs (M-Labs) research consortium and plans to use the lab during tests. “Second, while SamKnows’ plan to use self-selected individuals who volunteer for the measurement project may introduce certain biases into the data, those biases likely can be mitigated through appropriate statistical analysis techniques,” AT&T said. “We would encourage SamKnows to disclose the specific techniques it intends to use so they can be properly vetted by interested parties. Third, when reporting the results of its measurements, we would encourage SamKnows to publish, at a minimum, the mean value, the standard error and the confidence interval associated with each measurement."

Verizon also supported the pilot project but also asked for changes. “To ensure that the results are statistically valid, SamKnows should make certain that the test participants reflect (to the extent possible with the limited size of the test panel) the broadband user population in key respects, such as technical issues like local loop length (for DSL) and more subjective issues like satisfaction with a broadband provider,” Verizon said. SamKnows should also be required to control for variables that could influence test results, such as the file sizes to be used for testing, the locations to which tests for websites will be run, and the effect of other in-home Internet traffic, it said. “Third, the reporting of test results should fairly portray service performance across technology platforms and be meaningful to consumers, including ensuring that any measurements that include public Internet performance are not attributed to specific providers that do not control all of the network segments tested,” Verizon said.

USTelecom said the summary that the commission offered in its request for comments is “a good first step” in describing SamKnows’ proposed methodology but leaves many technical details unspecified. “The devil … is in the details,” the group said. “Because wireline broadband providers use different broadband technology platforms, have different network architectures, employ various content delivery methods such as caching, content distribution networks, and peering, and penetrate into the in-home network to varying degrees, it is important that SamKnows provides further details as to how it will ensure that its testing architecture, methodology, and data-reporting are fair and comparable across broadband providers, networks, and technologies.”

TIA said the results would be more useful if SamKnows is required to identify the part of the broadband network that each provider is responsible for and isolate tests to its part in a way that statistically models the consumer experience for a service provider and service tier. “Although this is a challenging task, it would aid in differentiating across providers and service tiers by acknowledging that providers use different broadband technology platforms and architectures,” TIA said.