For most of the past decade, the FCC has failed almost utterly to correctly track broadband penetration levels in the United States. In addition to considering a 200kbps connection broadband, the agency has consistently stated that if a zip code has just one home or business with broadband service, that zip code is "wired" for broadband. Obviously, this method has made broadband penetration look better than many of us know it actually is -- particularly in rural markets. That makes mega-ISPs happy, so reform has been a long time coming.

Even FCC Commissioners agree. "Finding one high-speed subscriber in a zip code and counting it as service available throughout is not a credible way to proceed," stated Commissioner Copps back in 2003. The GAO also issued several reports (pdf) that found fault with the FCC's methods. Independent parties that tried to get accurate data were rebuffed by the FCC and incumbent operators.

It's now almost 2008, and a House subcommittee today gave initial approval for The Broadband Census of America Act, a bill by Representative Edward J. Markey (D-MA). The bill (original discussion draft pdf) initially stated that if a connection wasn't 2Mbps, it technically wasn't broadband. That provision has since been removed in a compromise with Republicans.

It will, however, still eliminate the FCC's zip code metric, instead tasking the NTIA with giving grants to communities in order to -- get this -- actually head into the American wilds and see who has broadband.

The cable industry's largest trade group, the NCTA, has come out in favor of the bill -- because they likely feel that improved broadband mapping will highlight their faster speeds and somewhat superior rural neighborhood broadband penetration when compared to DSL.

That rural penetration is thanks, in part, to the local franchise system the phone industry has been successfully eliminating -- in order to remove build-out requirements as they enter the TV market. Expect the phone industry to vehemently oppose the bill, lest it highlight the deployment limitations of their next-gen deployment efforts (FiOS, U-Verse). If it's any good, it should highlight limited DOCSIS 3.0 deployment as well.

"This is a very consumer-friendly mapping function and 'demand-side' identification that the high tech and telecommunications industry also supports," Markey says in a statement on his website. For now, the bill moves forward, with a bipartisan understanding that the FCC's methods have been flawed. Of course, knowing you have a problem is only the first step on the long road to recovery.