Yesterday, AT&T filed comments with the Office of Management and Budget on the FCC’s woefully deficient analysis of the burdens associated with the so-called enhanced transparency requirements adopted in the 2015 Open Internet Order (OIO). The Commission’s analysis evinces a complete disregard for its responsibilities under the Paperwork Reduction Act. The FCC has not specifically identified the things Internet service providers (ISPs) must do to comply with the new transparency requirements; it has not separately estimated the burden of each requirement; it has not explained the benefits that would justify these requirements; and its lowball estimate of the overall costs is absurd on its face.
By way of background, the PRA requires agencies like the FCC to minimize the burden of required data collections. The FCC must obtain approval from OMB before any such collection can take effect. In this case, the FCC has had ample time to undertake a thorough analysis of the paperwork burdens of the OIO, yet has submitted a superficial, conclusory and slipshod analysis that OMB should reject. The treatment of just one of the new requirements, disclosure of packet loss, demonstrates the inadequacy of the FCC’s analysis.
To comply with the new transparency requirements, wireless providers will have to measure something they have not been required to measure or report previously (packet loss), in geographic areas where they do not currently take any similar performance measurements and may not have previously measured, and during undefined “peak hour” time periods. For AT&T, compliance with this requirement would cost far more than the FCC is estimating. Indeed, it could very well cost AT&T alone more than what the FCC has estimated for the entire industry to comply with all of the FCC’s transparency requirements because it could require extensive additional drive testing at a cost of many millions of dollars each year. Yet the FCC has estimated that the burden to the entire industry of all its transparency requirements, included those adopted in 2010, is only $640,000.
The FCC goes on to suggest that wireless providers may be able to meet these requirements by relying on data collected through the agency’s Measuring Broadband America program. The FCC neglects to say, however, that the mobile MBA program is unlikely to collect a statistically significant quantity of data. Nor does the FCC even mention that in order to get their mobile MBA data, mobile providers would have to pay $180,000 to the FCC’s vendor. Needless to say, that $180,000 per provider appears nowhere in the FCC’s burden estimates.
Finally, the FCC has provided no serious explanation of the practical benefits to consumers of measurement and disclosure of packet loss. Packet loss, which depends on, among other things, router buffer size, is unlikely to be of any value to consumers, and attempts to reduce it could actually impair service quality. And the OIO didn’t even attempt to explain the benefits of disclosing packet loss.
As this example shows, the FCC’s PRA analysis, which took the Commission more than a year to complete, does not pass the straight-face test. We hope OMB will reject it.