Why The FCC Broadband Plan Underwhelms
Media reports of the FCC broadband plan (pdf) that was sent to Congress today were fawning. Almost all contain this quote from FCC Chair Julius Genacowski, which describes the plan to connect 100 million U.S. households (~85 percent) to 100 Mbps high speed broadband by 2020 as
a 21st-century roadmap to spur economic growth and investment, create jobs, educate our children, protect our citizens and engage in our democracy.
But does this plan meet that lofty claim? No. (Read on if you want to argue with me.)
Why do I say the plan is not as advertised?
Broadband Is Not A Synonym For “High Speed”
But that’s how the AP reporter (read on dead trees) repeatedly referenced today’s plan, which was leaked via executive summary on Monday. Understand that the FCC is using a technical definition: “if it’s not dial-up, then it must be broadband.” But reporters have gulped this Kool-aid without pausing to think or to advise readers that all broadband is not created equal and they have treated the word like a synonym for “high speed” access. Which it isn’t.
Last summer, John Shepler of T1Rex.com warned:
If you’ve been eagerly anticipating FTTF (Fiber to the Farmstead), you may find the definition of broadband a bit disheartening. The bar has been set at a breathtaking speed of 768 Kbps download and 200 Kbps upload to meet the criteria of “broadband” Internet service.
In 2007 — three years ago — the average advertised broadband speed in Japan was 93.6 Mbps. Genacowski is saying that the U.S. is going to get to 100 Mbp …to match Japan in 2007 … in 10 years. Ten years! Moreover, even though Japan’s average advertised broadband speed was an order of magnitude greater than the U.S. three years ago, the cost was one-quarter of the U.S. A year later, the U.S. had made no inroads:
What are some other examples of national goals?
- Australia: 100 Mpbs to 90 percent households by 2018 (two years and ~5 percent ahead of U.S. plan)
- Finland: 100 Mbps in every household by 2016 (four years and ~15 percent ahead of the U.S. plan)
- Singapore: next generation Internet to all households by 2013 (seven years and ~15 percent ahead of U.S. plan)
- South Korea: 1 Gbps by 2014 (six years and an order of magnitude ahead of U.S. plan)
Don’t give me “but they’re small” response! Australia isn’t. And “small” means less resources! Moreover, Finland is pretty darn rural and mountainous (and frigid) and Australia is extremely rural.
Did any major news story (NYT, AP, Reuters) put the 100 Mbps goal into this type of global context? Not that I saw.
The plan also calls for having gigabit Internet access in every community by outfitting a public institution –schools, hospitals, military installations. Laudable but late. The University of Washington, for example, has had gigibit Internet since 2005. Gigibit Internet is what the Google challenge is all about. And in 2005, Hong Kong was offering gigabit Internet access to one-third of its households.
For those of us who don’t keep these arcane numbers in easy-to-access memory:
- One kilobit per second (Kbps or kbps) is 1,000 bits per second (bps). Dial-up modems are limited to 56 kilobits per second.
- One megabit per second (Mbps) is 1,000 Kbps or 1,000,000 bps. Basic cable is 768 Kbps to 1.5 Mbps. Fiber optic (FiOS) is 15-25 Mbps.
- One gigabit per second (Gbps) is 1,000 Mbps or 1,000,000 Kbps or 1,000,000,000 bps
As an aside, Apple has been shipping computers with one gigabit per second ethernet capability since about 2004.
If Genacowski had mapped out this plan 10 years ago, I’d say it was lofty and bold and visionary. Today? It’s politically expedient and modest. By this I mean it is designed to make the fewest possible entrenched (incumbent) interests unhappy while providing sweet soundbites. And it ignores how engineers have led us to a 50 percent increase in access speed each year, for the past two decades.
Yes, broadcast TV owners are going to whine about losing airwaves (public spectrum). Let them whine: the airwaves are owned by the public. TV owners have made big bucks for the 70+ years that they’ve been broadcasting on the publicly-owned spectrum. (The first experimental TV broadcast in the U.S. was in 1928.) In 10 years, it’s not unreasonable to think that more “TV” will be “received” via internet protocol (IP) than “through the air.” If this spectrum can be used for WiMax (more on that in a moment), that’s in the larger public interest.
Average access speed has increased about 1,000x in the past decade, from about 5 kbps to about 5 Mbps* (that is 5 kbps to 5,000 kbps). In 1998, two-thirds of the respondents in the Georgia Tech WWW survey connected with dial up modems at speeds between 28.8 and 56 kbps. About 1-in-5 American households accessed the Internet then; a decade later, that number was about 3-in-5 households. The FCC says it’s now about 4-in-5 households.
Jakob Nielsen has charted connectivity for a “high end” user for the past two decades; by his calculation, connectivity speeds have increased 50 percent a year. That’s a compound growth rate of 57x over a 10-year period. All of this bandwidth expansion has come about in the absence of a federal broadband policy and despite some bad law regarding competition.
Yet with this new public policy (and the federal purse), FCC Chair Genacowski promises that we’re going to increase from 5 Mbps to 100 Mbps in the next 10 years, less than half the growth rate of the past two decades. This is supposed to be “a 21st-century roadmap”?
Early days of Internet connectivity for U.S. households rested on existing infrastructure: copper telephone wire. The first “high speed” connections for the home (DSL and its siblings) also relied on existing copper infrastructure. The faster “cable” connections relied on existing infrastructure. The (relatively) big jump in speeds available via FiOS are the result of new infrastructure.
If we wanted to implement 100 Mbps connectivity for every home today, then this would truly be a massive infrastructure project because the “last mile” connecting homes to the Internet backbone would have to be fiber optic cable. Today’s “wireless” solution to the last mile problem is WiMax, but real-world throughput today is less than 5 Mbps. However, technologists have been coaxing more speed out of existing infrastructure (and airwaves) for at least 20 years. I’ve seen nothing that suggests this engineering advancement has hit a brick wall. Therefore, the plan’s goal appears far too conservative.
Another thing that helped spur speed developments in copper: the telephone companies were forced to share their copper networks with all comers. Speakeasy.net, for example, was a leader in the DSL field because of this ruling, and their upstartedness spurred telecos to action. Because of Speakeasy, I was able to get DSL years before Qwest (it was still US West then) was able to deliver service to my townhouse in Bellevue WA.
Today, consumers would be better off if Speakeasy could also provide service over existing cable and fiber lines. However, the cable industry successfully lobbied Congress, which made them exempt from having to open up their publicly-sanctioned monopoly to competition. I’ve seen nothing to suggest that the FCC is arguing for opening cable and fiber infrastructure to the same sort of competition as telephone copper. This is one of the places where we truly need a public policy initiative. The FCC plan calls for Congress to change other laws, why not this one?
We had a digital divide of sorts in the early part of the 20th century re telephony and electricity to rural America. There are systems and organizations in place that could leverage those earlier successes. (The feds loaned money to cooperatives, which built the wire and paid back their loans; for-profit corporations weren’t interested in the business and ignored the loan offers.) Connectivity in rural areas is a real issue, especially rural areas without long lines of sight (ie, mountainous and hilly areas). We should look to Japan (all of those islands) and Norway (all of those mountains) for hints on how to deploy to rural areas, and we need think seriously about broadband over power lines (which is MIA in the FCC report). But we must also turn that process over to the locals, not try to manage it from inside the beltway.
Summary Of Plan Shortcomings
- The 10-year plan looks more like the Japan of today and is less bold than the Hong Kong of yesterday. Thus it falsely positions the U.S. as a leader or the plan as particularly visionary.
- The plan is politically expedient, ie, makes few incumbents uncomfortable while sounding good to the uniformed (which seems to include too many journalists).
- It ignores the natural progression in bandwidth that comes from engineers coaxing more bits through existing infrastructure, be it wired or airwaves.
- It ignores the very real need for competition on existing physical infrastructure such as cable and fiber.
* Extrapolation from Georgia Tech study for 5 kbps; 5 Mbps is what is being reported today from FCC, implicitly with the 100 Mbps = 20 times faster than most home connections today.