For my first post as a new member of the faculty of the University of Colorado, it seems appropriate to discuss the state of the U.S. broadband market before this weekend’s Silicon Flatirons annual broadband conference.
The broadband state of the Union has improved dramatically. Most Americans now have access to broadband speeds that were only just dreamed of a few years ago. This improvement has primarily been driven by cable companies, who have been deploying a new technology (called DOCSIS 3.0), which tremendously increases the capacity of their wires. They are now able to sell services with speeds of 50 to 100 mbps, many times faster than the 4 mbps that the FCC says is necessary to stream HD movies. In a recent filing, the cable companies reported that they have deployed this technology to more than 75 percent of American households, and if all goes as planned they will soon reach 90 percent.
You would think that this good news has relieved the fears of commentators who have been worried that the country was falling behind in broadband deployment. But this good news has caused a whole new round of consternation. In recent articles, Michigan professor Susan Crawford has predicted that the cable companies will soon gain a complete monopoly in the U.S. broadband market. She acknowledges only one potential area of completion: the 15 percent of the country where Verizon has deployed its FiOS fiber service all the way to customers’ homes: “Where Verizon FiOS service exists, there will be competition with cable Internet access service providers for high-speed Internet access at speeds that are necessary to carry out real-time video conferencing or watch high-definition video. Where FiOS is not installed, there will not be any competition, and consumers will have just one provider to choose from: their local cable monopoly.”
While Professor Crawford raises a legitimate issue, she overgeneralizes. The situation she describes—with only cable broadband providing the ability to stream HD movies and use video conferencing—will apply to at most a third of the country. The situation for more than half of households will be much better: they will have a choice between at least two broadband providers delivering speeds capable of providing all services consumers currently use, including streaming HD video and video conferencing. The top 20 percent will have a choice between the new cable speeds and even faster fiber-to-the-home service from a telecom company. Verizon’s FiOS service now passes 16.5 million premises, and it plans to expand the service to 18 million premises. When FiOS is added to municipal deployments and fiber builds by independent telecoms, more than 20 percent of Americans will have access to fiber-to-the-home services.
Another 30 percent (and growing) of Americans have access to fiber-to-the-node services from a telecom. AT&T’s Uverse service now passes more than 30 million households, CenturyLink’s fiber-to-the-node services pass another 5.7 million, and other mid-sized carriers are deploying similar services. These deployments currently pass more than 30 percent of households, but it is reasonable to anticipate that the economics will not support deployment past 40 percent of households. AT&T has recently announced that its deployment is “largely complete.”
These deployments allow the ISPs to provide broadband at up to 40 mbps. CenturyLink sells service at up to 40 mbps. AT&T and Windstream sell service up to 24 mbps (they dedicate bandwidth to IPTV). These speeds are more than enough to handle the needs of the vast majority of households. They are more than enough to handle HD video. I have a 12 mbps service from CenturyLink, which provides speeds above 10 mbps almost every time I run a speed test. Last weekend in Boulder, we had one of our periodic 2-foot snowstorms. My wife and I spent the time catching up on movies and TV shows on Netflix and Apple TV. Most of the video we watched was HD, and not once was the CenturyLink broadband service too slow to keep up.
The bottom line is that these fiber-to-the-node services are fast enough to handle any and all services consumers currently use. Most consumers don’t need speeds higher than 7 to 10 mbps. And you don’t have to take my word for it; consumers are proving it. When given the choice of speeds, most opt for speeds of 7 to 12 mbps to save money. (A graduate student in our program researched this issue last semester, and every telecom and cable company he contacted admitted that customers are choosing the lower speeds.)
It is possible that content companies will create new services that take advantage of the higher speeds that are now available. If the demand does develop, the fiber-to-the-node deployments can be adapted to provide higher speeds. The next stage of investment will be pushing fiber further into neighborhoods to reach speeds of up to 100 mbps. BT has been deploying these services, and it now offers speeds up to 80 mbps. To reach these speeds, the telecom companies will drive fiber to a 4-block area, rather than the 300 homes that current node deployments serve. The telecom companies will be able to make these deployments without increasing total investment. As expanding the fiber-to-the-node footprint becomes uneconomic, the telecoms will shift their wireline investment dollars to increasing the speeds of fiber-to-the-node areas. And BT is demonstrating that fiber-to-the-cabinet can be converted to fiber-to-the-home. It is offering to deploy fiber from the cabinet directly to homes to customers who are willing to pay.
So approximately 60 percent of households will have a choice between broadband services that far exceed their speed needs. But what about the other 40 percent? I will cover them in tomorrow’s post.