Growth and Change, Centre for Business and Economic Research (CBER), forthcoming 2012. Jitendra Parajuli, Ph.D. Candidate, George Mason University, School of Public Policy, Kingsley E. Haynes, University Professor and Hazel Endowed Chair at the School of Public Policy, George Mason University.
The authors would like to thank the editor, three anonymous reviewers, and the participants of the 57th Annual North American Meetings of the Regional Science Association International, Denver, 2010 and the 3rd Annual Competition and Regulation in Network Industries, Brussels, 2010 for their valuable feedback on this paper. The authors express their appreciation to the U.S. Department of Agriculture (Award No. 2008-55401-04487) and U.S. Department of Commerce/Economic Development Agency (Grant No. 99-07-13862) for support of this activity. The authors are fully responsible for all analysis and policy interpretations associated with this work.
Data envelopment analysis is used to assess the efficiency of broadband Internet adoption and use in the U.S. Analyses at the state level from 2005 through 2007 suggest that broadband adoption and use is not equally efficient across the U.S. states. Although the instantaneous efficiencies are relatively high for many states, the changes in total factor productivity suggest that the efficiency of broadband adoption and use still increases over time. Moreover, it is observed that efficiency values are often spatially autocorrelated, which suggests spatial dependency from spillovers or interstate competition. This leads to a particular spatial diffusion pattern in broadband adoption. Although states have different strategies in support of broadband expansion they need to identify their shortcomings and use the appropriate mix of inputs (and outputs) to improve their efficiencies.