Efficient and super-efficient use of broadband access by the US states

Abstract

In an earlier study, Parajuli and Haynes (Growth and Change 43:590-614, 2012) used Data Envelopment Analysis (DEA) to assess efficiency of broadband utilization across US states. They found that a number of states in the USA assumed an efficiency score of one for broadband adoption and use. While this is not unusual, the commonly used DEA methods-the Charnes, Coopers, and Rhodes and the Banker, Charnes, and Cooper extension-do not rank relative efficient decision-making units (DMUs) across the efficient frontier. The super-efficiency estimation presented here is one method that overcomes this inherent limitation and allows for ranking the efficient DMUs. This paper uses a super-efficiency method to rank states that were efficient in broadband adoption and use in the USA from 2005 through 2007.