Measuring Broadband

This essay is based upon a day-long workshop organized by Kenneth Flamm of the University of Texas at Austin, John B. Horrigan of the Pew Internet & American Life Project, William Lehr of the Massachusetts Institute of Technology, and Sharon Gillett of MIT (at the time of the workshop). The text was written by Amy Friedlander, in collaboration with Kenneth Flamm, John Horrigan, and William Lehr.

This document should be cited as Kenneth Flamm, Friedlander, Amy, Horrigan, John B., Lehr, William. Measuring Broadband: Improving Communications Policymaking through Better Data Collection. (Washington, D.C.: Pew Internet & American Life Project, 2007). The views and opinions expressed herein are those of the authors. They do not necessarily reflect the views and opinions of the Pew Research Center, the Pew Internet & American Life Project, the Massachusetts Institute of Technology, the University of Texas at Austin, the National Science Foundation or the U.S. government.

Questions of vital importance to understanding the information society are difficult to address because of poor data.

Do places with more widely available or higher quality information infrastructure perform better economically than those without?
Does new investment in broadband connections have economic or social payoffs for communities?
Are civic institutions healthier in broadband rich areas – or not?
What portion of the lag in home broadband adoption in rural America is attributable to lack of available infrastructure?
Do those with “second generation broadband,” such as those with fiber to the home, behave differently online than those with “first generation” broadband such as DSL or cable modems?
How should we account for the impact of advanced information networks in measures of productivity and other economic activity?

Imperfect or absent data are are rarely mentioned in policy discussions. Yet the communications policy debate in the United States today is inseparable from debates about the data used to make claims about policy propositions. President Bush articulated in 2004 a goal to have universal and affordable broadband available in the United States by 2007. The way data are collected by government agencies cannot answer questions about whether that goal has been met or not. International organizations – using the imperfect data – report that the United States’ ranking in per capita broadband adoption is lower today than it was a few years ago. This paper argues that the country cannot properly gauge its own progress or know how dire America’s international standing is without good data about broadband adoption, deployment, price, and quality.

In June 2006, researchers from the Pew Internet & American Life Project, the University of Texas at Austin, and the Massachusetts Institute of Technology convened a workshop of likeminded specialists from government, academia, and industry to discuss challenges involving the state of data collection about the deployment and use of communications infrastructure. The workshop’s wide-ranging discussions yielded the following recommendations on the principles that should guide efforts to improve data collection on the deployment and use of communications infrastructure.

Collection of data should be at a sufficiently fine-grained level to permit regional analysis of the impacts of communication technology. Nearly all publicly available data on adoption of communications goods and services are gathered at the national level. More granular data – about adoption patterns among individuals and businesses at the local or regional level – would permit more rigorous analysis of the impacts of information and communications technology on economies and communities. Such data should capture the price users pay for service. With such data, policy makers and researchers would be better able to determine the expected payoffs from encouraging broadband deployment and adoption.
The United States should be able to produce a map showing the availability of infrastructure in the country. The current methods of tracking the availability of highspeed infrastructure relies on providers reporting by 5 digit zip code where they offer service. A provider with one customer in a zip code can report that it provides service in the zip code, which may misleadingly suggest that the entire zip code can get service. Understanding where infrastructure is available is critical for understanding what kind of choices consumers have for service. Workshop participants argued thaqt efforts to improve mapping of infrastructure must be accompanied by the Federal Communications Commission updating the definition of broadband to reflect advances in the nation’s information infrastructure. The current decade-old definition of 200 kilobits per second for broadband is widely viewed as outdated.
Academic researchers, non-profit organizations, the government, and the private sector must work collaboratively to gather data that permits assessment of quality of service and the user experience: The type of internet experience the end-user has at the desktop depends on “last mile” infrastructure availability, users’ awareness of it, and capacity to securely and skillfully take advantage of online connections. Assessing quality of service depends, in part, on computer scientists measuring online data traffic. These online metrics have become more difficult to acquire since the internet backbone was privatized, yet they remain important to properly maintaining an increasingly vital part of the nation’s critical infrastructure. Understanding the user experience also requires social science research into the community and cultural contexts of technology adoption and use.

The 2006 workshop was motivated by the desire of researchers to have better data with which to study the social, economic, and policy consequences of dissemination of information and communications technologies. At the same time, workshop participants recognized the sensitivity endemic in collecting data on commercial activity. Companies understandably do not want to make available to the public data that might reveal proprietary or strategically important information.

Yet the rewards from confronting those challenges and improving data collection are great. Policymakers would have a better understanding of the social and economic consequences of investments in communication infrastructure they may have under consideration. Economists in the public and private sector could better understand how new information and communication technology affect productivity. And with a clearer understanding of user behavior, planners in government agencies at the state, local, and federal level could more effectively design electronic service delivery applications for citizens.

Since the workshop, both the U.S. House of Representative and the Senate have held hearings on bills designed to improve data collection on broadband infrastructure. These bills represent valuable first steps in addressing this issue. Continued dialogue between researchers and policymakers is needed to develop data collection practices in the United States that will allow for informed deliberations on communications policy.