The FCC's data on broadband penetration in the U.S. is tainted by efforts to protect the interests of incumbents— and to shield the FCC itself from criticism.

by Alex Goldman ISP-Planet Associate Editor [August 15, 2003]

On June 10, 2003, the FCC released data on broadband deployment in the U.S. [.doc] [.pdf] The data, or lack thereof, was not much discussed because that was the same week that Congress intervened on the FCC's decision regarding media ownership—which received much print.

The FCC claims the data shows that broadband deployment is accelerating, which, coincidentally, implies that FCC policies are good policies.

Problems with the data start at the most basic level, with the definition of broadband itself.

This is broadband?
The FCC defines "high-speed lines" as "200 kilobits per second" (kbps) in at least one direction, while "advanced services lines" are those that "provide services at speeds exceeding 200 kbps in both directions."

At the current industry standard ADSL rate of 728 kbps download and 128 kbps upload, users can expect voice and even low quality video to function. At 200 kbps in one direction—what the FCC calls broadband —is merely four times the speed of dialup (and less than that with dialup acceleration).

Although the press has paid little attention, the generally poor quality of the data does worry FCC Commissioner Michael Copps, who last week voiced numerous objections [.doc] [.pdf] to its release. Besides the problem with the definition of broadband (Copps says the definition has been called "so 1997"), there are problems with almost every other aspect of the data.

Copps raised fundamental objections. For example, he pointed out "Finding one high-speed subscriber in a zip code and counting it as service available throughout is not a credible way to proceed. We must dig deeper."

The FCC has data broken down by zip code, but the data the FCC released to the public was not nearly so granular. The FCC published state-by-state data, implying that there is "competition" in a state if more than three providers offer broadband in any portion of that state.

Don't bother us with the details
The data shows number of providers and number of lines in three categories of broadband: ADSL, coaxial, and "other." For any state that has three or fewer providers, the FCC published an asterisk and withheld the data.

The FCC claims it is "protecting the confidentiality" of companies that report subscriber data in areas where there are only one to three broadband providers. (It claims that incumbents in monopoly areas might suffer a competitive disadvantage if their actual subscriber numbers were disclosed or could be inferred.)

But the FCC could have withheld the line information while publishing the number of providers in monopoly areas. By failing to disclose the number of areas served by only one broadband provider, the FCC is hiding data that might be used to criticize it and its policies.

The FCC did acknowledge that ILECs own 95 percent of the ADSL market, but then rationalized that monopoly by saying, "when all technologies are considered, ILECs provided about 36 percent of high-speed connections to end-user customers."

The FCC should publish the percentage of broadband served by all monopolies, whether those monopolies are ILECs or cable companies. Any other statement of the numbers obscures the true power of the monopolies and falsely inflates the amount of competition present in the U.S. communications market.

It could be even worse
If the FCC uses this flawed data as the basis for its policy decisions— instead of more accurate data it has chosen not to publish—it will certainly make flawed decisions.

Copps noted, for example, that one bright spot in broadband deployment is that "a lot of municipal governments are moving ahead and deploying broadband where the private sector has been reluctant to go."

In fact, such municipal deployments have been opposed by the monopolies—using means fair and foul—perhaps because they reveal as false the claims of those monopoly companies that are insisting they need subsidies to deploy broadband. These deployments call into question claims by the FCC that it is encouraging broadband deployment (encouraging municipal deployments is not on the FCC agenda).

If the FCC re-examines its data, its policies might change. It might have to define broadband based on the applications for which broadband can be used. It might have to make it easier for municipalities to deploy broadband. It might even have to try harder to encourage competition in DSL in the states that have "one to three" providers (Delaware, Hawaii, Mississppi, Rhode Island, West Virginia, and Wyoming) and in zip codes (data not released) where there is no competition.

But the FCC doesn't have to do anything. Consider this: if FCC policies fail to bring affordable broadband to the United States, who will be held accountable? The answer to that one is: nobody will be held accountable. Nobody will be held accountable for the flawed data, either.

— End