High-speed internet access has developed rapidly in the last decade and is increasingly viewed as essential infrastructure for our global information economy. For example, as recently as mid-2000 there were only 4.1 million broadband lines in the United States and only 3.2 million of these were residential lines. Thus, in mid-2000 less than one household in thirty could access the internet at a download speed of 200 kbps or greater. Six years later, the number of broadband lines, excluding mobile wireless connections, had soared to more than 53.5 million, 49 million of which were in residences. Residential penetration had therefore risen to nearly 50 percent by the middle of last year. (If mobile wireless connections are included, total U.S. broadband lines had risen to more than 64.6 million lines.)
While most communications sector analysts concur that the ability to deliver broadband communications is a critical feature of the modern global communications infrastructure, there is limited recent empirical research on the economic effects of broadband. In particular, much of the available research is now several years old or refers to the benefits of the Internet generally or more broadly of the "digital economy" rather than to the broadband telecommunications infrastructure per se.
This study provides new estimates of the effects of broadband penetration on both output and employment, in the aggregate and by sector, using state level data. We estimate these benefits by using FCC data on broadband penetration for the lower 48 states over the 2003-05 period, controlling for a variety of other factors that also could account for the growth in output and employment during this time. Although the FCC's definition of broadband is broader than we would like - since it includes all connections of 200 Kbps and faster at a time when broadband speeds are routinely greater than 1 Mbps - the FCC penetration data are the most comprehensive and reliable source of such information currently available.