The U.S. National Broadband Map: Data limitations and implications

The 2011 release of the National Broadband Map (NBM) has generated significant interest from the telecommunications policy community. The NBM is a multiagency effort, including the National Telecommunications Administration (NTIA), the Federal Communications Commission (FCC), fifty U.S. states and five U.S. territories, to collect and disseminate information on broadband provision and quality of service for the United States. From a geographic perspective, the NBM represents a marked departure from previous broadband data efforts. Specifically, instead of disseminating FCC Form 477 data on providers at the ZIP code or Census tract level, the NBM reports provision information at the Census block level—the smallest geographic unit in which the Census bureau tabulates survey information. While this increased level of geographic data resolution is a welcome change, there are several notable limitations to these data that are important to consider when conducting spatial econometric analysis for public policy evaluation. With this in mind, the purpose of this paper is twofold. First, this paper explores the salient characteristics of Census block geographic base files, highlighting their strengths and weaknesses as summary units of spatial analysis. This also includes a brief discussion on how the NBM treats geographically large blocks and their use of road segments for aggregating provider data. Second, this paper examines the impacts of empty blocks (i.e. blocks that have no household information associated with them) for the spatial analysis of broadband. Finally, this paper provides a short overview of how these data limitations can impact public policy evaluation and provides a blueprint for improving the National Broadband Map.

► The U.S. National Broadband Map is evaluated for spatial data accuracy.
► Characteristics of Census blocks and their geographic structure are reviewed.
► The impacts of zero population blocks on broadband availability estimates are discussed.
► Implications of data uncertainty for spatial-econometric analysis and policy evaluation are detailed.
► A blueprint for improving the National Broadband Map is outlined.