Communications - November 2009

Vviewpoints
DOI: 10.1145/1592761.1592771
economic and
Business dimensions
The Broadband Price
is Not right
Developing an effective pricing index is essential to understanding
the value of broadband connectivity.

HoW MuCh shouLd subscrib- ers pay for broadband? Without a sound economic benchmark, it is difficult o tell if the old expression “whatever the market will bear” is too high. We need a good benchmark, but we do not have one.

Why do we need one? Consider recent experience. In September 2001, approximately 45 million U.S. households accessed the Internet through a dial-up connection, while only 10 million used a broadband connection. By March 2006, the situation was moving to the opposite: Approximately 47 million households (and growing) had broadband connections, while 34 million (and declining) used dial-up connections. According to the latest survey of the Pew Internet and American Life Project, in April, 2009, less than 10% of U.S. households had dial-up Internet connections, and 63% of U.S. households had broadband.

iLLUStration by gLUeKit

What happened to the price of broadband during and after that deployment? In our study, “The Broadband Bonus: Accounting for Broadband Internet’s Impact on U.S. GDP,” http://www.nber.org/papers/ w14758, Ryan McDevitt and I inquired into broadband’s value. We approached this question from a novel angle. We asked: “How fast would prices have to come down to reflect the

value created from the replacement of dial-up access with broadband?”

What was so novel about that question? While answering that question we strictly followed the standard procedures used by the U.S. Bureau of Labor Statistics (BLS) to estimate the Consumer Price Index (CPI).

We had several reasons for doing this. First, the consumer price index is the primary measure of inflation in the U.S.—for example, the Social Security Administration uses it to adjust its checks. However, economists have long suspected the CPI index contains biases. Many years ago those suspi-

cions were confirmed in integrated circuits and personal computers, as well as in cellular telephony. Accordingly, measurement procedures in those markets changed.

Should broadband change too? It is a big and open question because nobody has yet estimated the size of the bias (if any). Governments in the developed world use procedures similar to those used in the U.S. If U.S. statistics contain a bias, it is also likely in other countries.