Thursday, November 1, 2007

E-mail Story     Print     Feedback

FCC Inks Another Broadband Report: Too Little Too Late

The FCC has released its semi-annual report on broadband penetration in the United States, mostly to the guffaws of an industry that by now either totally ignores or seriously criticizes the agency's concepts of how to do such a job.

As things stand today, the FCC counts 200 Kb/s as broadband, and if one house in an entire zip code gets DSL or cable broadband, it imagines that entire area is broadband- enabled.

The result is that the FCC thinks that broadband is available to 99 percent of the United States.

Even FCC Chairman Kevin Martin, after bearing years of criticism for the commission's outmoded concepts of surveying broadband, finally seems to understand there's a problem, and he's begun circulating a "reform" proposal among fellow commissioners to change the reporting methods. The catch: Martin's proposal falls well-short of proposals being floated in both the House and Senate. TelecomWeb news break thinks what Martin is doing is really trying to head off any really meaningful reporting - the kind of thing that would disclose that the United States really has no broadband policy, and that's the reason that, in fewer than five years, the country has fallen to anywhere from 15^th to 19^th, and maybe worse, in the world in terms of broadband, depending on whose metrics you believe.

The last little goody? The FCC's new numbers are for Dec. 31, 2006. Once again, it's taken the government 10 months to crank up the old adding machines, turning out numbers that are more in the nature of historical curiosities than they are data companies in the industry can use as they plan their businesses.

For what its worth, the FCC said that, as of the last day of last year, there were 82.5 million "high speed lines" in the United States, with 58.2 million residential users, up 61 percent from a year earlier. The FCC's definition of high speed is at least 200 Kb/s either upstream or downstream. As far as is discernible by TelecomWeb news break, the FCC's definition is carefully crafted not based on technology, but in order to make sure that even the slowest 256 Kb/s service offered by telcos can be counted as broadband.

Of those 58.2 million households, 53.6 percent were cable broadband; while 39.1 percent were ADSL connections; 0.2 percent were symmetric SDSL or traditional wireline connections; 1.3 percent were fiber connections to the end-user premises; and 5.8 percent used other types of technology, including satellite, terrestrial fixed or mobile wireless (on a licensed or unlicensed basis) and electric power line, the FCC said.

The agency also reported 59.5 million "advanced service" lines, up 36 percent year over year, of which 53.5 million were residential. "Advanced service" means that the speed in both directions is at least 200 Kb/s. In TelecomWeb news break's estimate, that still doesn't tell us who has broadband. The FCC did say that 63 percent of the advanced service lines had at least 2.5 Mb/s service in at least one direction. That's something of a help, but it's not clear how many of those are business lines and how many are residential. There's no count on how many subscribers can get broadband fast enough to support IPTV.

As for availability, the above-mentioned metric that broadband is available to at least one house in 99 percent of the zip codes in the United States quickly proves to be a useless waste of taxpayer money to even count. Perhaps of more interest is that broadband was available to only 79 percent of the households to whom incumbent LECs could provide local telephone service as of Dec. 31, 2006. "High-speed cable modem service" - which the FCC doesn't define - was available to 96 percent of the households to which cable operators could provide TV service. Making the numbers even more inscrutable, there's no tally by the FCC of what part of that 79 percent could get "real" broadband, which many define as megabit-class.

As for Martin's proposal for revamping the FCC's data collection methods, he's suggesting setting up five tiers of broadband to be surveyed: 200 Kb/s -768 Kb/s will be considered first-generation broadband; 768 Kb/s-1.5 Mb/s is to be considered basic broadband; 1.5-3 Mb/s will be high-speed service; 3 Mb/s - 6 Mb/s will be called robust service; and anything more than 6 Mb/s will be called premium.

Martin's also finally suggesting the FCC actually count how many people in a zip code get broadband, perhaps even getting more granular than that, considering that some rural zip codes cover hundreds of square miles. There's no indication, though, that Martin so far sees any problem with the FCC taking 10 months to add up the numbers.