After a new technology is introduced to the market, there is usually a predictable decrease in price as it becomes more common. Laptops experienced precipitous price drops during the past decade. Digital cameras, personal computers, and computer chips all followed similar steep declines in price. Has the price of broadband Internet followed the same model? Shane Greenstein decided to look into it.
Since there are no public data on what has happened to broadband prices over the last decade, Shane Greenstein, a professor of management and strategy at the Kellogg School of Management, and his co-author Ryan McDevitt, an assistant professor of economics and management at the University of Rochester and a graduate of Northwestern University, analyzed the contracts of 1,500 DSL and cable service providers from 2004 to 2009. They found evidence of only a very small price drop, between 3 and 10 percent, nothing like the rates of price decrease that characterize the rest of the electronic world.
It might seem like the cost of obtaining access to broadband Internet service would be prohibitive for many, but Greenstein notes that cable television wires already pass by more than 95 percent of US homes, while 75 percent of homes are close enough to a telephone switch for a DSL provision. “Any place with a population above 50,000 is not going to have a problem getting service,” Greenstein says.
Greenstein says many observers believe broadband prices have stagnated largely due to the concentrated market structure. Other factors have also concentrated broadband supply, including economies of scale and regulatory rules. In most urban markets, only two wireline providers supply the vast majority of homes, and the remainder are served by a range of wireless Internet providers. Revenue from homes makes up 70 to 80 percent of revenue in wireline Internet access market, while business demand makes up the rest.
“So if you were in such a market as a supplier, why would you initiate a price war?” Greenstein asks. With no new entries on the market, suppliers can compete by slowly increasing quality but keeping prices the same. According to Greenstein, quality is where providers channel their competitive urges.
Meanwhile, once companies have installed the lines, their costs are far below prices. “At that point, it becomes pure profit,” Greenstein says. A company might spend around $100 per year to “maintain and service” the connection, but people are paying nearly that amount every other month. Greenstein says that it is not surprising that prices were high during the buildout phase in the early and mid-2000s, since the firms were trying to recover their costs. “However, we are approaching the end of the first buildout, so competitive pressures should have led to price drops by now, if there are any. Like many observers, I expected to see prices drop by now, and I am surprised they have not.”
A New Trajectory
At the start of the 21st century, less than 5 percent of households had access to broadband Internet as most used dial-up systems. However, broadband soon achieved access to more homes through cable and telephone lines. Near the end of the decade, optical fiber lines had joined the ranks of broadband Internet providers. The decade started out with broadband representing just 6 percent of the total revenue from Internet services and that figure had grown to 72 percent by 2006. Greenstein and McDevitt decided to base their consumer price index on data from the second half of the decade, from 2004 to 2009, when the use of broadband really exploded.
They began by poring over 1,500 contracts from different years and services, including both standalone agreements and bundled contracts, in which a user combines Internet with cable television and/or phone service. They found that, even though prices stayed relatively constant, the quality of service rose through the years—for example, in 2004 the median cable modem contract price was about $45 with an upload bandwidth of 3000 bits per second, while in 2009 the median contract cost $53 but had an upload bandwidth of 8000 bps.
Greenstein says that a related problem may begin to creep into the system. With only slow quality upgrades and the rampant growth of streaming video online, most observers expect broadband to have multiple bottlenecks in its access networks soon.
The broadband consumer price index created by the Kellogg researchers has some limitations. An official price index would include a weighted average, which would give more share to the biggest providers to reflect the market more accurately. Greenstein and McDevitt were unable to create a weighted index because they lacked confidential data from the Bureau of Labor Statistics about market shares.
The most surprising discovery, Greenstein says, is that national decisions are being made without the type of data that he created in the consumer price index. “As an observer of communications policy in the U.S., I find it shocking sometimes how often government makes decisions by the seat of their pants,” he says. Without real data and statistics, decisions are based solely on who has better arguments—in essence, a debate. A better consumer price index will help produce better decisions for the future of the Internet and its users.
Shane Greenstein blogs about this topic and related subjects in Virulent Word of Mouse.
Greenstein, Shane and Ryan McDevitt. 2011. Evidence of a Modest Price Decline in US Broadband Services. Information Economics and Policy, June, 23(2): 200-211