The Faster, the Better? The Economic Return on Internet Speed

21 Pages Posted: 27 Mar 2016 Last revised: 13 Aug 2016

See all articles by Yang Bai

Yang Bai

Pennsylvania State University, Donald P. Bellisario College of Communications, Students

Date Written: March 26, 2016


Background and Purpose of the Proposed Study: The relationship between technological advancement and economic growth has long been attracting the interest of economists, policy makers and technology scholars. Recently, much attention has been given to the next-generation, super-fast broadband Internet. Many governments and international organizations regard the high-speed broadband Internet as the new engine for economic growth and has made the improvement of broadband availability, especially that of the fast broadband, a priority (FCC, 2010; UNTAND, 2011). Their enthusiasm is not unfounded. Empirical studies, in general, do show a significantly positive causal relationship between broadband and economic growth (Cronin, Parker, Colleran & Gold, 1991; Koutrompis, 2009; McMahon, Thomas & Kaylor, 2012; Whitacre, Gallardo & Strover, 2014). However, contrary to this “broadband optimism” (Kwak, Skoric, Williams & Poor, 2004, p. 424), some evidence of a diminishing return of Internet investment has been found: as we keep improving the broadband technology and, thus, increasing the data transmission speed, the social and economic benefits, though still rising, has become less and less significant (Grimes, Ren & Stevens, 2009; Howell & Grimes, 2010). This study will explore the issue of the economic return of Internet speed. The question I seek to answer is: as the broadband Internet with increasingly higher speed is deployed, does the economic gain which is supposed to follow rise accordingly? Rationale and Contribution of the Proposed Study: Empirical studies on the relationship between broadband diffusion and economic growth are abundant; however, most of them treat broadband Internet as one type of technology without considering the possible existence of differential effects of different broadband technologies, characterized by their various speed. Though, at the level of individuals and firms, scholars have discovered evidence of diminishing return of broadband speed (Kwak, Skoric, Williams & Poor, 2004; Grimes, Ren & Stevens, 2009), the nature of the micro level analysis may blind us to the externality and overall social value created by the high-speed broadband. Therefore, an econometric study analyzing the differential effects on the economic development of various types of broadband characterized by their data transmission speed will be conducted.

Method and Data: An empirical study on how the availability of broadband of different speed tiers affects the local economy, i.e. the gross output of each state in the U.S., will be conducted using appropriate panel data approach. The estimated gross state output for the year 2011, 2012, 2013 and 2014 are collected from Interactive Data, Regional Economic Account of Bureau of Economic Analysis. Labor force data has been obtained from the Regional and State Unemployment published by Bureau of Labor Statistics annually. Since the state-level stock of capital is not directly available, this study employs the method developed Yamarik (2013) to apportion the national capital stock, published by the Bureau of Economic Analysis each year to each state. The data on the availability of broadband by tiers of speed for 2011, 2012, 2013 and 2014 can be retrieved from National Broadband Map.

Keywords: Broadband, Internet Speed, Economic Development, Panel Data

Suggested Citation

Bai, Yang, The Faster, the Better? The Economic Return on Internet Speed (March 26, 2016). TPRC 44: The 44th Research Conference on Communication, Information and Internet Policy 2016. Available at SSRN:

Yang Bai (Contact Author)

Pennsylvania State University, Donald P. Bellisario College of Communications, Students ( email )

University Park, PA 16802
United States

Register to save articles to
your library


Paper statistics

Abstract Views
PlumX Metrics