Beyond Broadband Access: Developing Data-Based Information Policy Strategies 9780823252084

Citation preview

Beyond Broadband Access

DONALD MCGANNON COMMUNICATION RESEARCH CENTER’S E V E R E T T C . PA R K E R B O O K S E R I E S

SERIES EDITOR: PHILIP M. NAPOLI

This series seeks to publish research that can inform the work of policy makers, policy advocates, scholars, and students as they grapple with a rapidly changing communications environment and the variety of policy issues arising within it. The series employs a broadly defined notion of communications policy, in that it considers not only scholarship addressing specific policy issues and processes but also more broadly focused communications scholarship that has direct implications for policy making. editorial board Patricia Aufdherheide, American University Ellen Goodman, Rutgers University School of Law, Camden Allen Hammond, Santa Clara University School of Law Robert B. Horwitz, University of California at San Diego Robert W. McChesney, University of Illinois Jorge Schement, Rutgers University, New Brunswick

Edited by Richard D. Taylor and Amit M. Schejter

Beyond Broadband Access developing data-based information policy strategies

f o r d h a m u n i v e rs i t y p re s s

. n e w y o rk . 2 0 1 3

Copyright © 2013 Fordham University Press All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means—electronic, mechanical, photocopy, recording, or any other—except for brief quotations in printed reviews, without the prior permission of the publisher. Fordham University Press has no responsibility for the persistence or accuracy of URLs for external or third-party Internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. Fordham University Press also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Library of Congress Cataloging-in-Publication Data is available from the publisher. Printed in the United States of America 15 14 13 5 4 3 2 1 First edition

CONTENTS

Acknowledgments Introduction: Numbers That Matter Richard D. Taylor and Amit M. Schejter PART I:

1.

2.

3. 4. 5.

1

theory

Beyond Broadband Access: What Do We Need to Measure and How Do We Measure It? Catherine Middleton Understanding Digital Gaps: A Quartet of Empirical Methodologies Bin Zhang and Richard D. Taylor

9

23

Broadband Microfoundations: The Need for Traffic Data Steven Bauer, David Clark, and William Lehr

51

Adoption Factors of Ubiquitous Broadband Sangwon Lee and Justin S. Brown

69

Data and Modeling Challenges in International Comparisons Johannes M. Bauer and Sungjoong Kim

6.

vii

Data, Policy, and Democracy Jorge Reina Schement

88 103

7.

“Rulers of thousands, hundreds, fifties, and tens”: Does Democracy Count? Amit M. Schejter

113

the use and abuse of data in information policy making

PART II:

8.

9. 10.

11.

12.

vi

PhD Heal Thyself: In Search of Evidence-based Research for Evidence-based Policy Eli Noam

131

Case Studies in Results-Driven Decision Making at the FCC Rob Frieden

143

The Determinants of Disconnectedness: Understanding US Broadband Unavailability Kenneth Flamm

158

European Broadband Spending: Implications of Input-Output Analysis and Opportunity Costs Ibrahim Kholilul Rohman and Erik Bohlin

189

Using Data for Policy Development: Designing a Universal Service Fund for Tanzania Heather E. Hudson

209

Notes

229

Bibliography

269

List of Contributors

297

Index

305

contents

ACKNOWLEDGMENTS

This book is based on selected edited papers that were first delivered at a workshop, Beyond Broadband Access: Data-Based Information Policy for a New Administration, held at the New America Foundation in Washington, D.C., on September 22–24, 2009. We gratefully acknowledge the support and collaboration of many people in making both the workshop and this book possible. We would like to express our deep appreciation to Sascha Meinrath, director of the New America Foundation’s Open Technology Initiative, and his staff at the Foundation for providing the facilities and personnel support for the program. We also wish to acknowledge the other co-organizers of the workshop: Professor Johannes Bauer, codirector of the Quello Center for Telecommunication Management and Law at Michigan State University; Professor Jorge Reina Schement, dean of the School of Communication and Information at Rutgers University; and Professor Bin Zhang of the School of Economics and Management at Beijing University of Posts and Telecommunications. We thank them and their institutions and organizations for helping make the workshop the success that it was. The workshop would not have been possible without the generous support of its diverse coalition of sponsors: AT&T, Google, the Media Democracy Fund, the Social Science Research Council, T-Mobile, and Verizon. We would like to acknowledge and thank them for providing the material support and resources that made both the program and the wide-range of attendees possible. It is the thoughtful generosity of both

vii

foundations and corporations like theirs that make public scholarship possible, and we hope their example will be an encouragement to others. We owe a great debt to Dr. Benjamin Cramer at the Institute for Information Policy at Pennsylvania State University for his invaluable assistance with matters of style and editing. Special appreciation and gratitude goes to the people of Fordham University Press for their support and guidance throughout this project: our colleague Professor Phil Napoli, director of the Donald McGannon Communication Research Center at Fordham University and the editor of the Everett C. Parker book series; Fredric Nachbaur, director of the Press; Eric Newman, managing editor; and Will Cerbone. Last but not least, we have benefitted from the stimulating, and sometimes challenging, thinking of our colleagues at Penn State University, and from the support of the College of Communications, Dean Doug Anderson, Associate Dean Emeritus John Nichols, and Associate Dean Marie Hardin. In particular, we appreciate the cooperation of the college’s financial officers— Jane Agnelly, Annette Rice, and Dorie Glunt—in keeping our budgets organized and in-line.

viii

acknowledgments

INTRODUCTION: NUMBERS THAT MATTER

ri chard d. taylor and amit m. schejter

In recent years the information policy discourse has been inundated by numbers and metrics, which purportedly describe the “information society” and reflect national levels of such measures as e-readiness and the digital divide. In policy circles, just the ranking among some of these lists has been seen as an impetus for policy development. For example, President Barack Obama, shortly after being elected, stated that “it is unacceptable that the United States ranks fifteenth in the world in broadband adoption.”1 Most approaches involving quantitative “indicators” produce results, which are primarily descriptive and comparative (e.g., which nation has more Internet access, are operators providing services at the level they promise). Too often such studies have been “looking for the penny under the street light”— choosing indicators because data are available—while looking backward to a time when information and communication technologies (ICTs) had limited capabilities. Previously, the mere calculation of penetration levels may have provided sufficient information to direct policies that were solely focused on increasing availability. But in today’s multimodal multimedia ICT environment, what is their value? Looking ahead, are they the numbers that matter? Have past models been loaded for or against certain outcomes? Can the underlying methods be transformed into truly useful policy tools? If so, with what data? For the future, can we find the numbers that matter? What is missing are analyses that are not only descriptive and comparative but also explanatory and predictive, making it possible to understand why things have happened and why they are happening, and to make predictions about what will happen. While not always admitting it, current approaches 1

involve a high degree of subjectivity; it is important to adopt statistical methods that reduce this aspect. However, method must first be guided by theory, and in this field, theory is remarkably lacking. Important policy decisions are being made worldwide about information services to promote innovation, knowledge development, social equity, and democratic values. And while there is broad international consensus that these decisions are improved if they are informed by empirical data, there is no accepted doctrine as to what data really matters. The first step in developing a strategy for building a comprehensive, data-based approach to understanding policy consequences, and for improving policy outcomes through the utilization of meaningful empirical analyses, statistical methods, and the development of new conceptual frameworks, requires enhanced theoretical development. For a long time we have felt that there was a need and a demand for a compilation of good theories that will generate an informed debate about how to formulate questions that lead to testable hypotheses for which the appropriate methods must be deployed. In a world in which information-based power is becoming the dominant paradigm, such empirically based, data-informed policy analysis is a critical tool, and recent activity in policy circles in the United States, the European Union, and China (among others) demonstrates that policy makers are developing and utilizing such tools. This book is an attempt to fill that void. In response to a competitive international call for papers, nearly thirty scholars from around the world— representing a wide range of disciplines, among them economics, law, policy studies, computer science, information science, and communications studies—convened in Washington, DC, in September 2009. They presented studies and papers that focused on two aspects of the debate on broadband policy: different theoretical approaches to data-based communications policy making and case studies demonstrating both the use and abuse of data in the development of information policy. From these we present here twelve contributions that provide a well-rounded and international perspective of both evolving theory and theory implementation. Following the direction of the studies presented at the workshop, the book is divided into two parts. Part I deals with theoretical aspects of measuring information, and with issues that should be considered when designing broadband-focused information policy. Part II demonstrates how data has been both used and abused for argumentation purposes with regards to choices among policy paths and for policy building. In Chapter 1, Catherine Middleton argues that while existing information society measures like the International Telecommunication Union’s ICT Development Index and the World Economic Forum’s Networked Readiness Index provide a useful starting point for comparing national ICT 2

r . d . t ay l o r a n d a . m . s c h e j t e r

indicators, they fail to offer sufficiently detailed metrics upon which to formulate policy related to the development and use of broadband networks. She offers more nuanced approaches to understanding whether, and how, citizens actually benefit from access to broadband technologies, and offers suggestions for the development of new, policy-relevant measures of ICT usage. In Chapter 2, Bin Zhang and Richard Taylor present a series of exercises in using statistical methods to analyze the digital divide, using China as a model, because China has pursued the empirical study of informatization for some time, and has collected extensive relevant data. They demonstrate the use of four approaches—Static and Dynamic Analysis/Analytic Hierarchy Process, Hierarchical Clustering Analysis, Time-Distance Analysis, and Data Envelopment Analysis—showing how they open the door to more fundamental theoretical questions and suggest future study avenues. Steven Bauer, David Clark, and William Lehr claim in Chapter 3 that data focused on availability and on adoption metrics is becoming less informative as broadband penetration rises, and that since the next set of questions will center on the health of the broadband access market, network traffic data will be central to understanding and answering many of these questions. To answer these questions it will be helpful to learn more about the distribution of usage across the user population, the characteristics of users that participate during peak periods of network congestion, and the variance in usage and how it differs by type of user. In their contribution in Chapter 4, Sangwon Lee and Justin Brown demonstrate how the results of regression analysis suggest that network competition between fixed and mobile broadband, platform competition in fixed broadband markets, and the multiple standardization policy in mobile markets, are all significant factors of ubiquitous broadband deployment. While the theoretical reasoning for quantification of connectivity raises methodological challenges, the social justification for government policy promoting access to information and communication technologies involves moral and ethical considerations. The transition from theory to practice is more often than not countryand region-specific, so the value of economic, statistical, or social theories needs to be seen in a local perspective. In Chapter 5, Johannes Bauer and Sungjoong Kim address one particularly challenging problem affecting cross-national comparisons of advanced communication services: the increasing importance of service and price differentiation, particularly the reliance on multipart self-selection tariffs. They propose a new method, the “Low Expenditure Frontier” for putting such comparisons on a more robust basis, and they illustrate it using the case of mobile voice services. The last two contributions in part I discuss the connection between citizenship, democracy and information policy. In Chapter 6, Jorge Reina introduction

3

Schement argues that the roots of information policy can be found in the implied social contract between governments and the governed and identifies a new social contract relevant for the information age. Amit Schejter, in Chapter 7, decries the hypercommercialization of the information network and its control by private interests using the marketplace metaphor. He argues that government policy should adopt a justice-based metaphor instead, and focus on quantifying the amount of democratic opportunity created, after reaching consensus on the definition of democratic participation. Opening the discussion on the use and abuse of data for information policy development, Eli Noam notes the reluctance in policy debates to engage in unbiased data analysis, with parties seeking only results-oriented analysis. He uses the debate on media concentration as an example. He questions whether there is an interest in Congress and the US Federal Communications Commission (FCC) in serious data analysis. He identifies a specific structural impediment in the rule-making process, and recommends a requirement for an “academic impact statement” in rulemakings, which would be an independent literature review of relevant peer-reviewed literature. In Chapter 9, Rob Frieden identifies several instances in which the FCC could have used empirical research and peer review to achieve an accurate measure of whether a telecommunications market operates competitively. However, deregulatory zeal and wishful thinking motivated the FCC to refrain from engaging in rational decision-making. In Chapter 10, Kenneth Flamm reviews recent US broadband data, discusses why the broadband “connectedness” portrayed in the FCC postal code–level data is in some respects problematic, and outlines an approach that can be used to systematically model the determinants of “disconnectedness” and assesses what economic and demographic factors can be identified as important, and statistically significant, in reducing or increasing “disconnectedness.” The final two Chapters discuss cases outside the Americas. In Chapter 11, Ibrahim Kholilul Rohman and Erik Bohlin address a methodology to evaluate the opportunity cost of using government funds for broadband in the context of a future European proposal to publicly fund broadband infrastructures; and in Chapter 12, Heather Hudson examines how data, primarily from existing sources, is being used to plan implementation of a universal service fund in Tanzania. That nation is presented as an example of relying primarily on existing data from multiple sources that could be used in other developing regions where time and/or resources to conduct dedicated field studies may be limited.

... The wide range of issues discussed in part I—from traffic data to democracy measures and from statistical analyses to considerations in comparative 4

r . d . t ay l o r a n d a . m . s c h e j t e r

studies—may seem to some either overwhelming or even counter-productive: If there is so much we need to know and in so many different ways, perhaps the effort of making all that information available may not be worth it. Another criticism of this intellectual effort may be that the complexity of the policy-making process which would arise from adopting some of these theoretical approaches will lead to its professionalization, and result in the exclusion of new, innovative voices that are not necessarily capable of participating in this type of discourse. Indeed, these and other critiques are a welcome aspect of the debate we hope this book creates, because a debate is what is badly needed. Only a rigorous and open debate can lead to an accepted, meaningful formula for planning policy as objectively as possible to serve the needs of society and its members. The alternative is that we encounter more of what is described in part II: data will be misused and abused to advance political or personal interests, or will be a result of incompetence and ignorance. Of the two alternatives, to us the choice is clear. There are many additional aspects of meaningful broadband planning that did not make their way into the contributions in this book. One that comes to mind is the use of mapping and advanced mapping techniques. Another is a discussion of the many documents that are already out there as a result of foundation or think-tank initiatives and that can serve the conversation by bringing well-informed approaches to the table. Yet another may be tapping into the experiences of those who were at the helm of designing communication and information policies as society was transitioning into the information age, and building on their experiences. Indeed, much more can be said. One thing this exercise demonstrates is clear—the haphazard, off-the-cuff, intuitive, and interest-serving approaches of the past should make way to theory and data-driven approaches in the future. We offer a number of well thought-out proposals to pursue this quest. As early as 1976, in his groundbreaking work The Coming of Post-Industrial Society,2 Daniel Bell noted that among the distinguishing characteristics of what he perceived to be a revolutionary change in the structure of society were the centrality of theoretical knowledge, the capability of planning, and the rise of “intellectual technology.” Indeed, it would be an irony of our times if at the culmination of these dramatic changes, the planning for full utilization of their benefits for the sake of all members of society were to be left in the hands of chance.

introduction

5

This page intentionally left blank

CHAPTER 1

Beyond Broadband Access What Do We Need to Measure and How Do We Measure It? ca therine middleton Around the world, claims that broadband infrastructure is central to the development of the knowledge economy are becoming indisputable. Many governments are taking steps to ensure their regulatory environments encourage private sector investment in broadband,1 consistent with Organisation for Economic Cooperation and Development (OECD) recommendations to rely on competition to the maximum extent possible as a means of building broadband infrastructure.2 In instances where the private sector cannot establish a business case for broadband deployment, governments are committing public funds to extend the reach of broadband networks, justified by the widely held belief that broadband access is essential infrastructure for an information society. Investment in broadband infrastructure is premised on the dual assumptions that broadband networks enable the information society and the knowledge economy, and, by providing citizens with access to broadband, citizens will participate in, and reap the benefits of a knowledge-based economy and society.3 However, Preston, and Cawley observe that broadband development is often driven by “supply-side, technology-focused policies” that do not explicitly consider the needs of users.4 Deployment of broadband infrastructure is expected to encourage the creation and uptake of “socially useful” applications, but this is not guaranteed. Indeed, at present there is a gap between the discourses linking broadband deployment with the development of a knowledge-based society and the ability to deliver the desired outcomes. This chapter explores this gap. 9

A central motivation for investing in broadband is that providing citizens with access to broadband connectivity will allow them to engage in the information society. But there are three problematic parts of this statement: 1) access to broadband does not ensure that broadband is used, or that it is useful for the user; 2) all broadband networks are not the same, meaning that the potential benefits of broadband access may not be equal for all broadband users; and 3) there is uncertainty as to exactly how to recognize the broadband-enabled benefits of engagement in the information society. Although there is a great deal of research activity regarding broadband and the information society, this chapter argues that there is a need for better research questions, improved analytical approaches and more sophisticated and wider-ranging data collection in order to fully assess the extent to which broadband networks actually do enable citizens to become participants in the information society. The chapter begins with a consideration of the nature of broadband networks, followed by a discussion of how broadband can enable engagement in the information society. The availability and analysis of data on broadband use is then explored, and suggestions for improved data analysis approaches are offered. The chapter concludes with a brief discussion of the challenges of developing more advanced measures.

all broadband connections are not the same Many claims are made about the benefits of broadband. As more evidence is compiled demonstrating the positive returns on investment in broadband, it is important to consider exactly what is encompassed in this term so as to better understand the type of investment (and resultant infrastructure) that enables positive outcomes. What is available to citizens? National and pan-national (e.g., European Union) statistical agencies do not apply a common definition of broadband, nor do they collect data in a consistent format. Some differences in approach are explained below, highlighting the need to understand speeds and network characteristics. Statistics Canada’s Canadian Internet Use Survey (CIUS) collects data on the type of household Internet connection. Those reporting cable or satellite connections are recorded as having high-speed access. Respondents with telephone connections are asked if their connection is a high-speed connection. No definition of high speed is offered, but it is assumed that respondents in this category would describe their Internet as high speed if they were not using a dial-up connection. In total, the 2009 CIUS data indicate that 92 percent of Canadians with Internet access at home (70 percent of the total population) have a high-speed connection.5 10

c at h e r i n e m i d d l e t o n

The term broadband is not applied to these data by Statistics Canada, but data reported by the CRTC (the Canadian telecommunications regulator) differentiates between “high-speed” and “broadband” networks, using broadband to describe connections with download speeds greater than 1.5 Mbps.6 The CRTC reports that 93 percent of Canadian home Internet users had high-speed Internet connections in 2008, but only 70 percent of these were broadband connections. Taking into account the nonusers, this means that just over 50 percent of Canadian households had Internet connections that provided access at speeds greater than 1.5 Mbps, a number that is quite different than the 70 percent of households with high-speed connections reported by Statistics Canada. In Australia, as of December 2009, 89 percent of Internet subscriptions were for broadband services, defined as providing downloads at speeds greater than 256 Kbps.7 If applying the same minimum as the Canadian data (i.e., > 1.5 Mbps), only 60 percent of subscriptions would be categorized as broadband. But while the Australian Bureau of Statistics considers speeds above 256 Kbps as broadband, the Australian government has embarked on a program to offer speeds of up to 100 Mbps to homes, orders of magnitude faster than speeds currently experienced by many Australians.8 2010 Eurostat data9 show that 64 percent of households have a broadband connection, defined as “connectable to an exchange that has been converted to support xDSL-technology, to a cable network upgraded for Internet traffic, or to other broadband technologies.”10 OECD data show that European countries are now the international broadband leaders,11 thus it is interesting to note that across fifteen European countries not quite two-thirds of households have broadband connectivity. After years of promoting the benefits of the information society, it seems that many citizens are yet to be convinced that they require broadband access. More than 50 percent of Japanese broadband subscribers are connected by fiber networks that can provide download speeds of 100 Mbps or more, and several other OECD countries have extensive fiber uptake.12 In countries with competition among broadband technologies, cable companies are upgrading their networks to offer download speeds of up to 120 Mbps, and telephone companies are increasing the availability of faster DSL connections (with download speeds in the range of 20–50 Mbps).13 Mobile broadband services are also becoming more common, providing users with coverage anywhere served by mobile broadband networks. Although speeds vary enormously, and connectivity is not always reliable, mobile broadband is used by many as a substitute for fixed line service. Examples could be provided from many other countries, but the data above are sufficient to demonstrate that there are many definitions of broadband, and that there are vast differences in the speeds of broadband connections beyond broadband access

11

used by citizens around the world. Variations in connection speeds result in differing experiences of the Internet. Additionally, actual connection speeds often do not match the advertised speeds, with connections frequently slower than advertised. The discrepancies are noted in reports by the Federal Communications Commission in the United States,14 Epitiro in Australia,15 and Ofcom in the United Kingdom,16 and indicate that the problem is not confined to a specific country. There are three other pertinent issues regarding the speed of Internet connections that impact the way that citizens experience broadband networks. The first is that in a few countries, Internet service providers (ISPs) impose strict limits on the volume of data that can be downloaded in a fixed period (often a month).17 These download caps counteract the value of having a high-speed connection, because once the monthly cap has been reached, financial or technical penalties (e.g., reduced download speeds) are imposed on the subscriber. As such, it is important to understand the conditions of access that govern a broadband connection, as speed is not the only factor that can constrain usage. The second concern is that in some countries, independent of download caps, ISPs are known to “throttle” or “shape” (i.e., degrade) certain types of Internet traffic. This is typically done in the guise of network management, but it is argued these practices violate the principles of network neutrality and detract from a citizen’s ability to freely access online content.18 Given the differential traffic shaping practices among ISPs, it is useful to understand how these practices can impinge upon citizens’ Internet activities, as they may impact their ability to fully engage in the information society. Third, the interactive nature of the Internet means that not only do people want to download content and services, they also want to upload their own content and create their own services.19 Residential Internet connections have traditionally been asymmetrical and it is a technical challenge to increase upload speeds on copper and cable networks.20 But as citizens become more engaged with the information society they want fast download speeds to be matched with fast upload speeds. Some providers do offer symmetrical or near symmetrical speeds, but faster upload speeds are usually considered a premium service and cost more. Without knowing the connection speed, whether there is a download cap, whether the connection is subject to traffic management practices, and whether the connection supports symmetrical or near-symmetrical uploads it is difficult to make assumptions as to the capacity of an individual’s broadband network. Why do speed and network quality matter? Most people would instinctively choose a faster speed connection over a slower one, but it is useful to consider the specific benefits that can arise with the adoption and use of faster, 12

c at h e r i n e m i d d l e t o n

higher quality, unrestricted networks. There are numerous reports that describe the applications that are enabled by higher speed connections,21 as well as documents that explore the added functionality offered by faster broadband networks and consider demand for such networks.22 But simply building a higher speed network and making it available to users without imposing download caps or traffic shaping does not ensure that they will be able to benefit from access to it. Thus, the observations made here about the potential uses of higher speed networks are premised on the assumption that the users will have the necessary skills and interest to take up these applications in meaningful ways. One of the most frequently cited benefits of higher capacity networks is the ability to support video applications. These networks will support multiple high-definition television (HDTV) streams into homes, but with more symmetrical network architecture, networks can also be used to support interactive high-resolution video conferencing. Advanced video services can support healthcare, entertainment (e.g., interactive gaming) and educational applications, as well as enabling energy management functionality, and remote monitoring of any location. Speed and quality matter because they have a large influence over what can be done on a particular broadband network. If broadband networks are to deliver benefits to users by enabling certain functionality, then it is essential to understand whether the user’s broadband network can actually deliver such functionality. By assessing users’ needs and recognizing the limitations of lower speed, restricted, asymmetrical broadband networks, it is possible to create an upgrade path to provide improved connectivity and enhanced access to the information society. Assessing broadband connectivity. This section has outlined some of the problems in assuming that the simple availability of a broadband network will enable all of a citizen’s desired online activities. Given the many variations in broadband speeds, access restrictions and network quality, there is a strong case to be made for ensuring that the attributes of broadband networks are well understood by those planning to deploy services over them, and by those responsible for providing broadband connectivity as a means of enabling the information society. Table 1-1 highlights network characteristics that should be considered in order to assess the extent to which a broadband network can meet its users’ needs.23 Ideally, this information could be used to develop a household broadband profile.24 This profile would allow for immediate determination of whether a household has the capacity to engage in specific activities, and identification of gaps to be remedied. The answers to some questions are clear cut (e.g., advertised speed), others are more subjective (e.g., definitions of affordability, which is relevant in understanding whether a citizen can beyond broadband access

13

table 1-1

.

sample components of a household broadband profile

Issue: Broadband Network Characteristics

Information Required

Speed

What are the advertised upload and download speeds? What are the actual upload and download speeds? Do these speeds support all the applications citizens want to use?

Type of connection

What is the access technology (or technologies) in use? Are there technical limitations imposed by the type(s) of connection in use (e.g., will a wireless connection support all necessary applications)? Is the connection upgradable to meet demands for increased bandwidth?

Quality of service

Does the network provider allow for prioritization of specific types of traffic? Does such prioritization meet user and application provider needs?

Service provider

Do service providers’ policies (e.g., with respect to traffic shaping or download caps) negatively impact the user’s experience? Is the connection affordable?

maintain this connection over time) or subject to variation over time (e.g., actual speeds, whether a connection can support particular applications). Developing household broadband profiles is only a first step in the exploration of how citizens can benefit from broadband infrastructure. The profile provides information on the current and future capacity of a household’s broadband connectivity, but capacity is a measure of potential benefit, not of realized benefit. As Gillett et al. observed, in order to benefit from broadband infrastructure, “broadband ha[s] to be used, not just available.”25 The issue of use is central to the “broadband enables the information society” logic. Although it seems obvious that broadband networks must be used to create value, this point is not always explicitly considered. Indeed, many international measures of broadband “use” are actually access measures, and simply measuring use does not address the question of “use for what purpose.” The next two sections address these concerns.

using broadband to participate in the information society It is beyond the scope of this chapter to review the discourses of the information society and the related dialogues that consider the role of information and 14

c at h e r i n e m i d d l e t o n

communication technologies in enabling a knowledge-based society or digital economy.26 There are, for instance, streams of research focusing on conceptualizing the information society,27 technologies and environments that foster ICT to support the information society,28 measuring the information society and its enabling technologies,29 and policy making.30 Additionally, extensive work has been done by governments and NGOs to discuss, define, and measure the information society.31 Despite enormous diversity in research and policy making regarding the information society, there is some common ground in the basic definitions of the concept. In 1994 a commission on Europe and the Global Information Society noted that “Technological progress now enables us to process, store, retrieve and communicate information in whatever form it may take, whether oral, written or visual, unconstrained by distance, time and volume.”32 In the 2006 World Information Society Report, Yoshio Utsumi, the secretary general of the World Summit on the Information Society (WSIS) described “a future Information Society [as one] in which Information and Communication Technologies (ICTs) are available anywhere, anytime and to anyone.”33 The European Commission’s Knowledge Society Web Portal (now defunct, but accessible through archive.org) defined the information society as “a society in which low-cost information and ICT are in general use” and then applied the term knowledge-based society “to stress the fact that the most valuable asset is investment in intangible, human and social capital and that the key factors are knowledge and creativity.”34 In brief, the information society is built on information and communication technologies, and broadband can provide access to them. Many studies offer evidence that broadband connectivity provides economic benefits. The World Bank recently reported that “for every 10-percentage-point increase in the penetration of broadband services, there is an increase in economic growth of 1.3 percentage points.” Van Gaasbeck found that increased broadband use in California is associated with employment growth,35 Fornefeld et al. discuss the ways that broadband spurs growth and innovation in European industries,36 and Greenstein and McDevitt note broadband adoption has a positive impact on US gross domestic product.37 But for the purposes of understanding how broadband facilitates individual engagement in the information society, economic impact studies offer little specific insight. Assertions that broadband enables employment, education, and healthcare are plentiful, but specific studies that consider how individual usage of broadband38 actually results in increased engagement with the information society are less common,39 even when the possibilities for such engagement are clearly articulated.40 Recent work by Kolko suggests that benefits to households as a result of increased broadband beyond broadband access

15

expansion are “ambiguous,” and concludes that “the local economic development benefits of broadband are mixed.”41 One of the reasons for the somewhat uncertain relationship between broadband adoption and improved information society outcomes is that these improved outcomes are not always easily recognized. In other words, it is difficult to know exactly when an individual is actually improving his or her information society “involvement” when using a broadband network. Kolko calls for more research that specifically considers the relationship between investments in broadband infrastructure and tangible benefits for citizens, and assesses how (or whether) broadband uptake can lead to better social outcomes.42 Because there are many disparate broadband applications that can provide different sorts of information society benefits, it will be necessary to develop a range of assessment tools. There are two key requirements for these tools. The first is that they enable recognition of broadband-enabled information society benefits; that is, they make it possible for researchers to identify the precise ways that broadband network use contributes to an individual’s engagement with the information society. The second is that the tools allow for a measure of the extent or size of the benefit. There is some work being done that offers more detailed articulation of the ways in which Internet usage can help people participate in the information society. For instance, the World Bank notes that “in developed countries and urban areas in developing countries, an increasing number of individuals build up social networks through broadband-enabled, peer-to-peer Webbased groups that facilitate economic integration and drive development. Blogs (Web logs, or online diaries), wikis (Web sites where users can contribute and edit content), video sharing sites, and the like allow new, decentralized, and dynamic approaches to capturing and disseminating information that allows individuals to become better prepared for the knowledge economy.”43 This kind of description starts to explain how the Internet fosters information society engagement, but does not consider the extent to which it actually works for various individuals. To consider another example, it is agreed that broadband connections support e-learning. But how is that support actually provided, and how can it be measured? E-learning activities might consist of corresponding with an instructor by email, downloading reading materials and contributing to a chat forum, none of which are sophisticated or bandwidth-intensive activities. It might also consist of high-definition video conferencing between students, interactive simulations conducted by video or in virtual worlds, creating video content to share with other students and streaming video live from field trips. These two descriptions of e-learning demonstrate the complexity inherent in trying to understand how broadband enables e-learning, 16

c at h e r i n e m i d d l e t o n

and, perhaps more importantly, highlight the fact that information society engagement can take place on many levels. Data collected from surveys that pose simple questions like “have you used the Internet at home for . . .” and then provide a list of possible activities like email, banking, searching for work, education, watching TV, and searching for information44 cannot be used to assess the extent of information society benefits gained by these types of Internet usage. Such data do allow for a very basic description of the nature of individuals’ online activities, but can really only be used to understand the relative popularity of applications, not to understand how or whether people benefit from using them. As the e-learning example suggests, developing a detailed questionnaire about the benefits of e-learning requires a sophisticated understanding of its potential and possibilities. A more informative approach would be to gather extensive observational data, with researchers actually observing how people use their broadband connections over time, and allowing them to articulate the benefits in their own ways, rather than by selecting from a stated list of anticipated benefits.45 As such an approach may not be very feasible on a wide scale, use of diaries and personal video recordings (e.g., to describe uses, benefits, and challenges faced in engaging with the information society) could offer a less resourceintensive means of gathering richer data, while still allowing individuals to share their experiences in ways that are meaningful to them. Table 1-2 summarizes the objectives for data collection regarding use of, and benefits from broadband (not just Internet) connectivity. If the overall objective of encouraging the development of broadband capacity is to allow it to be used to foster engagement or participation in the information society, then it is important for policy makers to be able to determine whether, and how this interaction is actually happening. Researchers can help to reveal the ways in which broadband connections are being used by individuals to participate in the information society, and

table 1-2

.

linking broadband use to information society benefits

Issue: Mechanism for Information Required Benefiting from Broadband Network use

Do citizens access services that enable them to participate in the information society? Do citizens accrue benefits from the use of such services? How? What skills are needed to ensure maximum benefit?

beyond broadband access

17

can then identify actions to be taken to encourage greater engagement as appropriate.

understanding measures of internet and broadband use In addition to understanding what sorts of information society activities are undertaken, it is also important to know more about the extent to which specific activities are undertaken (this helps to articulate the extent of benefits realized from a given activity). A problem with much of the current information on broadband usage is that it measures the type of use (what people are doing) without considering frequency or intensity of use (e.g., data record that individuals use e-learning applications, but don’t indicate whether the use is frequent or infrequent, and whether it is extensive or superficial). An additional problem, especially within various international ICT indexes that purport to assess use, is that many measures of use are actually measures of adoption. The reason this is problematic is considered below, followed by a discussion of what can be derived about usage and information society engagement from other data sources. In order to use a technology or service, the first decision is an adoption decision. If a household adopts a broadband connection, this means that the connection is available for use. The fact that the connection has been established means that the household is now considered to be a “broadband household,” and the household’s connection would be recorded by the OECD, the ITU or statistical agencies that track broadband subscriptions. But adoption does not mean that the technology or service is actually being used or will be used in the future, nor does the act of adoption imbue potential users with the skill or capacity to use whatever it is they have adopted. Adoption allows for the possibility of use. While people generally adopt technologies with the intention to use them, there is potential for much variety in actual technology use. Acquiring access to a broadband network is a first step toward participation in the information society, but it may not produce immediate benefits. As such, efforts to understand how having access to a broadband connection can enable the benefits of the information society must be informed by usage data, not adoption data. But to date, a variety of widely used ICT indexes are based on adoption data, limiting their usefulness in informing governments and policy makers about citizens’ readiness to engage with the information society. Information society indicators. An enormous amount of effort has been expended to develop indexes that provide internationally comparative data on various ICT and information society indicators. Measuring the information society is a complicated task, and much progress has been made by the 18

c at h e r i n e m i d d l e t o n

international working parties (comprised of a mixture of representatives of statistical agencies, the United Nations, the OECD, the International Telecommunication Union [ITU], telecommunications policy makers and academics, among others) that do this work. Two high-profile indexes are the ITU ICT Development Index,46 and the World Economic Forum/ INSEAD Networked Readiness Index.47 Both indexes offer data that can be used by policy makers for international comparisons, and for assessing progress within countries over time. But they offer minimal value in assessing the extent to which individuals in particular countries are using their broadband connections to engage in the information society and to achieve personal benefits. Despite assertions that their indexes assess the state of usage of broadband technologies, the data really report on technology adoption (i.e., penetration rates). Waverman and Dasgupta observe that “the literature on mobile communications and telecommunications generally has not, to date, looked in sufficient detail at factors beyond penetration.”48 They, as well as Pepper et al., propose stage models of adoption that do take into account real measures of usage.49 These models demonstrate increasing levels of sophistication of (aggregate) use of communication technologies among populations and offer some commentary on how to encourage greater societal and economic benefits through increased usage of ICTs. Furthering this approach, Waverman and Dasgupta’s Connectivity Scorecard recognizes the importance of what is termed “useful connectivity,” attempting to develop an index that takes into account not just the development of infrastructure, but also the uses of this infrastructure, and investment in developing skills among users, in order to assess the contributions of connectivity to economic growth.50 Pepper et al. identify the “intensive” stage of broadband adoption, at which “e-commerce, e-government services, business collaboration, and social networking, among others, are pervasive and have become an integral part of the social fabric and economy.”51 At this stage, reached when approximately 50 percent of households have broadband access and two-thirds of the population are using the Internet, broadband networks are being used to enable services that foster engagement in the information society. Only 23 countries, with a total population of about 850 million, had reached this intensive use stage by 2009. These countries are at the leading edge of the information society, but if the information society is to be integrated into the lives of everyone in the world, universal access is needed. While reaching 50 percent household broadband penetration, with two-thirds of the population online is an accomplishment, it also highlights the broadband digital divide—any society with 50 percent of its citizens not having broadband access cannot be considered a broadband-enabled information society. beyond broadband access

19

Information society indicators, as currently formulated, do not provide good information on the nature of use of information technologies in the information society. The basic adoption/penetration information provided in these indexes offers value to policy makers and researchers, as does the opportunity for cross-national comparisons, but to really understand the extent to which the goals of an information society, with information access for all, can be achieved by the deployment of broadband and Internet technologies, further information is needed. Additionally, the evidence provided by these indexes and other investigative models suggests that there is much work to be done to reach universal or even near-universal access and use of information society enabling technologies. These two points are explored further in a discussion of other sources of Internet use data. Internet use data. As noted in the previous section, it is challenging to assess the relationship between broadband usage and specific information society benefits. However, there are two ways that the relationship can be explored. The first is by considering basic usage data, on the assumption that if citizens are not using broadband connections, they will not be engaging in the information society. The second is to consider the scope and intensity of users’ activities, as this can identify usage patterns that encourage positive information society outcomes. As discussed above, international ICT indicators (generally based on source data from the OECD and the ITU) are not very helpful in assessing Internet and broadband use. They do show that broadband adoption continues to grow steadily around the globe, meaning that the number of users is increasing, and the potential for more citizens to engage with the information society through broadband connections is growing. Outside the indexes, OECD broadband adoption statistics are widely used, and widely criticized.52 There are other sources of information on national ICT indicators, including national or pan-national statistical agencies (e.g., Australian Bureau of Statistics, Eurostat, Singapore’s Infocomm, Statistics Canada), and nonprofit research consortia (e.g., Pew Internet Project, World Internet Project). Additional data and commentary are available from various consulting firms and other private sector companies like Nielsen.53 Much of the available Internet use data comes from user surveys. Survey responses provide static measures of use, enabling cross-sectional data analysis. These data are helpful in understanding the types of activities favored by Internet users, and activities can be assessed in terms of how they help individuals participate in an information society (e.g., by accessing educational content, government services, healthcare and so on). Although longitudinal tracking of changes in specific individuals’ habits over time is not possible with most surveys, data that is collected on a regular basis can be used to identify trends within a population. For instance, the Canadian Internet Use 20

c at h e r i n e m i d d l e t o n

Survey data shows how use has changed from 2005 to 2007 to 2009, illustrating a decreasing digital divide, and increasing uptake of a wider variety of online activities.54 Kolko does use longitudinal data to consider the impact of broadband adoption, concluding that not all “socially desirable” activities increase when individuals switch from dial-up to broadband services. His findings offer a tangible example of how broadband usage data can be linked to achievement of information society benefits by individuals, and reinforce the observation that broadband connectivity alone does not ensure beneficial outcomes. He also makes the point that assessing the benefits of broadband use is very complicated, and argues that much more research is needed to “help assess how socially or economically desirable various online behaviors are.”55 In addition to assessing the types of activities people conduct online, it is important to understand users’ capacity to fully engage with the applications and services that they are using. Gurstein argues that in order to enable “effective use” of information and communication technologies, individuals must have knowledge, skills, and a supportive environment.56 Hargittai’s work demonstrates that many users are not highly skilled, which could indicate that their ability to benefit from the use of various Internet and broadband applications is decreased.57 Similarly, Middleton et al. suggest that even when broadband adoption rates are high, many Internet users engage in relatively few activities and do not use the Internet particularly intensely.58 To enable better understanding of the extent of benefits that individuals can realize by adopting and using broadband networks, the following information is required: international comparative indicators that measure use, not just adoption of information and communication technologies; detailed data on broadband usage, revealing how often and for what purposes broadband connectivity is used; assessment of users’ skills and capacity to use their broadband connectivity, to allow for full benefits of connectivity to be realized. The key points here are that adoption alone does not guarantee that an individual will engage in the information society, and that variations in extent of use of broadband applications and services will result in variations in benefits accrued to individual users.

conclusion This chapter explores three challenges that arise when trying to understand the ways in which broadband network connectivity, as used by individuals, enables the information society. It identifies research questions and data required to assess the direct impact of broadband development and use. Any research that considers the benefits of investing in broadband connectivity must begin with a description of broadband services actually available to users. beyond broadband access

21

Applications and services that can offer benefits must be identified, and user uptake of such services should be tracked. When examining broadband usage, it is essential to go beyond yes/no measures of adoption, developing models of use and engagement that reflect what citizens do with their broadband connections, how frequently they use the network, and their overall capacity to use broadband networks to achieve positive social and economic outcomes. It is a relatively straightforward exercise to identify general ways in which data collection and measures of broadband impact could be improved. It is more complex to put these ideas into action in a way that will allow for meaningful international comparisons of data, and that will result in tangible conclusions regarding the impact of broadband connectivity in enabling and sustaining an information society. As governments invest in broadband infrastructure, it is very important that they also provide support for assessment of the impact of this investment, explicitly funding research, and collaborating in data analysis. As discussed, such research should build on existing baseline data around user characteristics and adoption patterns, to offer detailed quantitative and qualitative insight on what citizens do with broadband connections, and how this connectivity provides benefits to individuals and to society as a whole. As broadband networks are considered an essential infrastructure in the twenty-first century, improved understanding of how they support individuals in their everyday activities will allow for more effective use of this infrastructure. Research on the specific ways that broadband can and does enable the information society will reveal shortcomings in current approaches. Such research can inform development of improved applications and services, identify appropriate technical characteristics of faster networks, and provide guidance on skills development and digital literacy standards that will ensure individuals have the capacity to engage with applications and services available to them. Billions of dollars will be invested in broadband infrastructure development. This chapter offers some specific suggestions as to how to ensure that a return on this investment is realized. The outcome of the research approach outlined here can inform the development of good public policy regarding broadband deployment and use.

22

c at h e r i n e m i d d l e t o n

CHAPTER 2

Understanding Digital Gaps A Quartet of Empirical Methodologies bin zhang and richard d. taylor Since the 1960s, the field of information studies has had a tradition of trying to understand the role of information in society by measurement, typically by counting things: media, words, bits, and so on. This reflects an intuitive sense that something important is happening. Because of the intangible nature of the subject “information,” however, its role has been hard to grasp. This goal manifests itself in current times most often as the study of the so-called digital divide (or e-readiness). Approaches to this matter have grown more sophisticated over time, and now use complex statistical methodologies to parse huge databases for lessons from the past for potential gains from shaping the future. This chapter provides a detailed exercise in four of those methods, as applied to one country, China, but with the view that their underlying principles may be of broad application and of substantial use to planners and policy makers.1 These are not the only approaches—there are many— indeed, so many that the very idea of finding coherence in the field is challenging. While much progress has been made, it is argued that going forward there needs to be a new way of thinking about this field, and new emphasis on theory and testing as a way of developing analyses that are both explanatory and predictive. This study is the product of an international collaboration (United States and China) to advance the thinking in this field, which in China is broadly referred to as informatization, a concept that resonates with industrialization— a sweeping industrial and social change affecting all aspects of life and society. China has highly prioritized informatization, and created a top-level state 23

council Leading Group to coordinate and promote the concept. The United States has no such grand metaphor, focusing instead on universal service and broadband access, which are much more limited concepts. Continued advancement of theory and practice in this field will be of greatest benefit to whoever advances informatization the most. Our intention with this study is to encourage scholars and policy practitioners to give informatization greater consideration. We chose China as a case study because it has pursued the empirical study of informatization for some time, and has collected extensive relevant data. This chapter provides examples of four methods for analyzing that data, with the hope that their underlying principles may be of general application and useful to policy makers. These methods are: • Static and dynamic analysis/analytic hierarchy process (AHP) • Hierarchical clustering analysis (HCA) • Time distance analysis (TDA) • Data envelopment analysis (DEA) The examples are different in name, but similar in purpose. They have the same starting point, the Informatization Level index, described below. They use different assumptions, and the calculations are different. Their common purpose is to objectively describe and analyze the digital divides within China. Based on the particular situation and the data available, one or several methods may fit better than others. Rarely will one method describe all the relevant considerations. In the examples in this chapter, the results are in some ways different, providing different insights; but also similar, in that their respective conclusions converge. Which method to choose depends both on the researcher’s goals and the philosophy.

case 1: using static and dynamic analysis ⁄ analytic hierarchy process Based on a review and analysis of the structure (inductive/deductive), statistical methods, conclusions, and comparative strengths of twenty-eight existing digital divide index systems, a comprehensive index system was developed for the measurement of regional digital divides in thirty-one regions of China. Using the AHP model, rankings of indicators were obtained from the work of experts using factor analysis, and indicators measuring more effective access were given higher weights. Index weights were then determined for twenty-nine factors by computing a comparison matrix. Using regional data in China from 2002 to 2007, normalized by means of equalization, the 24

b . z h a n g a n d r . d . t ay l o r

index values and ranks of informatization levels for each region for each year were obtained; then mean deviation was used to analyze the changing trends in the digital divides during those years. Attention was focused on five core factors affecting digital divides in China: technology, economy, government, education, and society. Case 1 focuses on the measurement of the digital divide: this refers to the effective access gap between regions in information and communication technology (ICT). Effective access targets the uses of ICTs through which people, organizations, or society can obtain economic and cultural advantages. If people who have access to ICT do nothing but upload or download music online, then their access to ICT does not make sense (is not effective) because they lack the ability or the opportunities or the will to use ICT to broaden their cultural knowledge or better their economic situation. So the importance of effective access is emphasized. There are twenty-nine indicators in this index system, which reflect the main factors affecting the digital divide. Indicators measuring effective access are given higher weights. From this point of view, therefore, AHP has greater flexibility, and under the guidance of experienced experts, becomes a more accurate measurement model for the selected target. In recent years China has increased its level of informatization rapidly. All regions have shown a narrowing trend in their digital divides, yet the real levels of the regions and the gaps between them are not precisely known. Due to the fact that the indicators have different measurement units and orders of magnitude, equalization is used to make indexes dimensionless during the process of static analysis. Thus the mean deviation is used as a reference to describe the digital divide in a dynamic analysis—to observe changes over time, which is the more intuitive way to present the regions’ real informatization levels and the disparities between them and the average level. index system The informatization level (IL) developed in this case is a composite index. Specifically, there are twenty-nine indicators composing the entire framework—seven indicators for Society, ten for Technology, five for Government, four for Economy, and three for Education. Figure 2-1 shows the structure and indicators of the index. determination of index weights Using weighting scores from experienced experts, AHP was used to derive the local and global priorities; the consistencies of hierarchy (overall and local) were also checked. As a result, the weights of every level in the informatization level framework were determined. In accordance with the results u n d e r s ta n d i n g d i g i ta l g a p s

25

Capacity of the communication network

Technology

Coverage of information Communication network building condition

Informatization level

Economy

Percentage of administrative villages with fixed telephone lines Percentage coverage of television (population) Percentage coverage of broadcast (population) Length of optical cable route for access network Length of residential trunk optical cable route Length of long-distance optical cable route per square kilometer

Consume ability of inhabitants

Expenditure of communication per capita Disposable income per capita GDP per capita

Contribution of communication industry

Expenditure of communication per capita

Scientific and educational investment & innovation

Educational funds investment per capita Authorized patents per capita Scientific and technical investment per capita

Government Investment for ICT of the country

Proportion of investment in telecommunications fixed assets as part of total investment of the country Proportion of investment for information facilities as part of total investment in fixed assets of the country

Labor resource

Percentage of employees in the information industry as part of gross employment Percentage of on-campus college students as part of total population

Education Cultural quality of citizens

Application level of ICT Society The Internet local contents

f igure 2-1.

Capacity of mobile switching exchanges per 10,000 people Capacity of local telephone exchanges per 10,000 people Number of public telephones per 1,000 people Penetration of main (fixed) telephone lines

Rate of literacy in over 15-year-olds Fixed long-distance traffic per capita Mobile traffic per capita The amount of the telecommunication service per capita Penetration of the Internet Penetration of mobile phones Number of CN domain name per 10,000 people Number of WWW websites per capita

Hierarchical structure of the informatization level.

of the local priorities, global priorities were computed as well. The complete results will be provided on request. Some of the more noteworthy results are: • The application of ICTs is shown as the key aspect of this index system, and the weight of Penetration of the Internet hit 0.147, which is the highest among all indicators. It suggests that indicators related to the application of the Internet have been stressed, which demonstrates the concept of the digital divide emphasizing effective access. • Besides emphasizing application, this index system indicates that the Economy and Education components have significant influence upon the regions’ informatization levels. This is evidenced by the expenditure of communication per capita and rate of literacy over 26

b . z h a n g a n d r . d . t ay l o r

fifteen years old, both reaching 0.103. It also shows that the digital divide is not just equality of access; to truly and effectively narrow the digital divide we have to improve the citizens’ cultural quality, literacy, awareness of access to information and extent of their ICT application level. • In terms of technology, more attention has to be paid to the coverage of information; from an economic perspective, the study focuses on the consuming capacity of inhabitants; and in the government component, investment for ICT of the country is quite vital. static analysis of the digital divide in china Data normalization. Due to the fact that indicators have different dimensions, the data was normalized within the same indicator by using the following formula (equalization): xij′ = xij / x j ( f x j = , let l xiij′j = 1 + xij ) Using the original data divided by the mean of data of the same variable, the mean of every indicator turns out to be 1, and the standard deviations are the coefficients of variation of the original variables. This method effectively eliminates the impact of dimension and magnitude, and at the same time it retains the variation information of the original data. The greater the variation, the greater will be the influence on the comprehensive analysis. This kind of undimensionalization tries to preserve the variation information through the coefficients of variation of the original variables, not the standard deviation of the original variables, which can save both the comparability and the variation information of the original data. Analysis of index values and ranks. Applying the index system and normalization method discussed above, the index values were developed. The analysis shows that: • Beijing, Shanghai, Guangdong, Zhejiang, and Tianjin, which ranked at the top of the list in recent years, maintained their leading positions during the six years. • Regions that made gradual progress, indicating that the importance of informatization has been recognized step by step, and that scored some achievements in these areas include Jiangsu, Hainan, Shanxi, Jiangxi, Inner Mongolia, and Hena. • Liaoning, Jilin, Hebei, and Yunnan apparently showed deterioration, which reflects that those regions had lower awareness u n d e r s ta n d i n g d i g i ta l g a p s

27

of informatization building compared to other areas, or lack of sustainable development, causing the backward rankings. On the other hand, it implies those regions have more room for improvement in the future. • If we take a closer look at the results for 2007, we find that Beijing and Shanghai had much higher index values than all the other regions, with figures arriving at over 2.7 and 2.5, respectively. They were the top group of all the regions, presenting advanced informatization levels and there was very large disparity between them and other areas. Guangdong, Zhejiang, Tianjin, and Fujian were the next highest group, with index values all above 1.1, which indicates that those regions were in a relatively good position; and there were smaller gaps between the remaining regions, which all pointed at an intermediate or low level. • Group analysis of index values: Table 2-1 shows the results when applying a descriptive statistical analysis to these numbers. Table 2-1 shows that in these six years, the minimum index values increased moderately from 0.57630 in 2002 to 0.67054 in 2007, approaching the mean 1. Meanwhile, the maximum decreased from 3.39502 in 2002 to 2.71674 in 2007, approaching the mean 1 as well. So we can see that the digital divide in China has been narrowing in recent years. But these conclusions only provide rough information about the entire body of data. For a closer look at the index values, we divided the data into ten groups. They are: below 0.5 (the first group), 0.5–0.625 (the second group), 0.625–0.75 (the third group), 0.75–0.875 (the fourth group), 0.875–1 (the fifth group), 1–1.125 (the sixth group), 1.125–1.25 (the seventh group), 1.25–1.375 (the eighth group), 1.375–1.5 (the ninth group), over 1.5 (the tenth group). We classified and applied frequency statistics, so that the following additional conclusions could be reached.

table 2-1.

28

.

descriptive statistics from 2002 to 2007 2002

2003

2004

2005

2006

2007

Maximum

3.39502

3.24164

2.99855

2.88652

2.92703

2.71674

Minimum

0.57630

0.59443

0.60965

0.61940

0.66275

0.67054

Mean

1

1

1

1

1

1

Case Number 31

31

31

31

31

31

b . z h a n g a n d r . d . t ay l o r

The number of regions that can be included in the third group (0.625– 0.75), the fourth group (0.76–0.875), and the fifth group (0.875–1) were the majority of the total thirty-one. Moreover, the numbers grew slightly in the six-year period. All of this shows that the index values of the thirty-one regions tend to be concentrated in the center. There is a trend towards concentration seen in the grouping process, and in order to better support the conclusion by means of data, we computed the standard deviation of the index values for the six years and the result is presented in Figure 2-2. It is manifest that the standard deviation generally declined from 2002 to 2007, especially during the period between 2002 and 2006, when it showed a clear drop, which suggests that the digital divide was controlled during those years. But there was a small increase in the figure from 2006 to 2007, which showed the gap widened slightly. In fact, the Internet and the number of Internet users in China saw rapid growth in 2006, which was mainly due to the dramatic expansion of the Internet and the drastically increasing number of websites, web pages and Internet users in the eastern part of China. This unparalleled situation led to a greater gap between the eastern and western parts of China. So here we need to analyze the digital divide using both horizontal and vertical comparisons. Further analysis shows that the frequency of the 0.625–0.75 group fluctuates drastically, but it stopped at 8 in 2007, which followed a drop compared to 10 in 2002; the figure of the 0.76–0.875 group climbed with less dramatic fluctuation, ending up with the same level in 2007 as it did five years ago (10); however, the number of the 0.875–1 group generally rose in those years, and reached 6 in 2007, which doubled the magnitude in 2002 (see Figure 2-3). Overall, whether through charts or data, we can see the current levels and developing trends of informatization in China, that is, the digital divide in China is narrowing, and the informatization level index is centralizing to 1.

0.60 Standard 0.55 0.50 0.45 0.40 2002

figure 2-2.

2003

2004

2005

2006

2007

Standard deviation of the index values, 2002–2007. u n d e r s ta n d i n g d i g i ta l g a p s

29

12 0.625–0.75

0.75–0.875

0.875–1

10

8

6 4

2

0 2002

fi gure 2-3.

2003

2004

2005

2006

2007

Frequency statistics for three groups, 2002–2007.

dynamic analysis of the digital divides in china In the prior analysis, the index value for the digital divide and the ranks of thirty-one regions were obtained, and equalization methods for detailed analysis were applied. A number of other methods can also be applied to calculate the digital divide. In order to see changes in the digital divide over time, the concept mean deviation (a more intuitive way to show the disparity between the index value of a specific region and the mean value) is referred to, to determine the shape of the digital divide over time as presented below (assume the year has been fixed): n e value(i ) − 1 ∑ Index i = 1 Digital Divide d ( DD D )= n DD stands for Digital Divide; Index value means the index value of informatization level in the No. i region; n equals the total number of regions in our research, which is 31 including provinces, municipalities, autonomous regions and municipalities; 1 stands for the mean of index values. Figure 2-4 shows the changing trend of the digital divide from 2002 to 2007 according to the computed results and the digital divide measurement model. 30

b . z h a n g a n d r . d . t ay l o r

0.400 D 0.300 0.200 2002

figure 2-4.

2003

2004

2005

2006

2007

Changing trend of the digital divide in China, 2002–2007.

The result indicates the different conditions of informatization in China’s thirty-one regions, and the reduction in the digital divide over recent years accordingly; in other words, the gaps between regions are being bridged. Starting at 0.367, the number plunged rapidly from 2002 to 2005, followed by a steady period during the next two years, then leveling off at approximately 0.309. Overall, the result shows a decreasing trend, as seen in Figure 2-5. To better understand the cause of the narrowing trend, it is necessary to analyze the changes from the five key components: Technology, Economy, Government, Education, and Society. • The Society index mainly describes the application of ICT. Its values are much higher than the others over recent years. But at the same time, its decline was very fast until 2006 and 2007 when it began to stabilize. That suggests that instead of the huge differences between regions in the past, now almost all regions have an increasing awareness of the importance of the application of ICT, and have 0.7 0.6 0.5 0.4 0.3 0.2 0.1 2002

2003

2004

Technology Government

figure 2-5.

2005

2006

2007

Economics Education

Digital divide trends comparing five indexed components, 2002– 2007. u n d e r s ta n d i n g d i g i ta l g a p s

31

started using ICT products and services simultaneously. This change, therefore, brought about a progressively reduced gap in ICT application and produced a rising informatization level in China. • The Technology index primarily embodies the ICT infrastructure. The index values show that there was a small divide in this factor, with a slow decrease during those years which followed a similar trend with the average divide. This indicates that the whole country has been paying relatively greater attention to the ICT infrastructure. • The Government index chiefly represents ICT and educational investments from the government. Its values fluctuated more strongly than the other ones, and interestingly its values have grown over the years, meaning that the investment divide of Government seemed to widen between regions as opposed to narrowing. This surprising situation shows that the government formulated national ICT development policies with some strategic considerations. Yet, there are gaps between regions in terms of support and investment for ICT, which is the foundation for further improvement. • The values of Economy and Education were both small and showed a gradual decrease, following the same trend as the average. This suggests that education and the economy were generally less emphasized. All in all, the digital divide in China has dropped from year to year.

case 2: using hierarchical clustering analysis Case 2 adopts the indicators system in case 1. Then an HCA for China’s Digital Divide Index in thirty-one regions was carried out from 2002 to 2007. The results of this approach provide a deeper and more useful understanding of the digital divide in China. Case 2 also takes into consideration the five core factors, that is, Technical, Economic, Governmental, Educational, and Social factors, and ranks each indicator for the thirty-one regions from 2002 to 2007. Based on this, through relevance measure clustering, case 2 identifies twelve different types of factors which influence the digital divide. It analyzes the reasons for rank changes in the digital divide index in some provinces, and suggests policy directions. An advantage of cluster analysis is that it can simplify a situation that might otherwise be too complicated to analyze. There are so many provinces in China that it is necessary to cluster them by their similarities relative to the digital divide. In case 2, hierarchical clustering is used to cluster the thirty-one regions based on their digital divide indexes. Hierarchical clustering cannot only be used to conduct horizontal comparisons among regions, but 32

b . z h a n g a n d r . d . t ay l o r

also vertical time-distance comparisons of classifications and rankings of regions, which can lead to a clearer understanding and awareness of the changes of the digital divide status of regions during the six years studied. The similarity method of hierarchical clustering. The method used to determine the degree of similarity between samples is called the similarity measure. When different types of similarity measures are used, the same sample can be divided into different classes. Therefore, it is important to be clear about the kinds of similarity measures used to classify when we do hierarchical clustering analysis. There are many kinds of similarity measures, which can be divided into two types: relevance measures and distance measures. A distance measure focuses on the comprehensive distance between samples, while a relevance measure focuses on the structural similarity between samples. Explanation of the index system and weight distribution. The same index system is used as in case 1. There are two reasons for reusing that index system. First, it was systematically derived from twenty-eight prior index systems, so it is comprehensive. Second, using different methods on the same index renders the results comparable. The same applies to the weights. So here we also adopt the weight distribution of case 1. the hierarchical clustering process The data of thirty-one regions from the year 2002 to 2007 is used as the sample data, using SPSS to apply the hierarchical clustering method for analysis. Those data are collected from the China Statistical Yearbook of the National Bureau of Statistics of China, the Statistical Yearbook of Communication, Information Statistical Yearbook, Communication Statistics Annual Report, and the Internet Statistics Report” of CNNIC. The procedure of hierarchical clustering is as follows: 1. Data standardization: As the method “mean of 1” can eliminate the effects of the dimension and order of magnitude, while still retaining the information about the degree of variability of the original data, we use this method to standardize data, that is, to make the mean value of data in the scope of 1. The formula is: xij′ = xij x j if x j = 0,then , h xij′ = 1 + xij . 2. Calculate the values of the criterion levels’ indicators—which are Technology, Economy, Government, Education, and Society—and the digital divide index (the higher the value the higher the level of informatization, which means the digital divide between this subject province and the most advanced level is smaller). 3. Cluster the variables according to the digital divide index and get the result of clustering. We use distance measurement to measure the u n d e r s ta n d i n g d i g i ta l g a p s

33

comprehensive distance among samples, which means to select the Squared Euclidean distance in SPSS. We use the between-group linkage (also known as the category average) method to measure the distance between the different categories. 4. We can cluster the variables in accordance with the ranking of five indicators for criterion level, that is, Technology, Economy, Government, Education, and Society. We use Cosine in SPSS, a Relevance Measure method, to measure the degree of structural similarity for samples. We also use the between-group linkage method to measure the distance between the different categories. clustering results and analysis Clustering the data for six years by distance measurement. One hundred and eighty-six digital divide indexes—which were samples calculated by using the data of thirty-one provinces from 2002 to 2006—divided into twentyfour categories, which were then clustered. The provinces were then ranked in accordance with the distance between the samples. Case 2 shows the differences in the level of informatization for all regions in the years from 2002 to 2007. The samples are divided into twenty-four categories and there are digital divides between categories. Beijing and Shanghai are clearly in the leading position in their level of informatization. Guangdong, Tianjin, Zhejiang, and Fujian follow closely. The level of informatization of Gansu, Guizhou, and Anhui are lower, and the rest of the provinces are in the mid-range. The time distance of the digital divide between regions by the category rankings of all 186 samples can be analyzed. For example, if category 6 included “Beijing 3, Shanghai 6,” we could say that the level of ICT development of Shanghai in 2006 had almost reached the level of Beijing in 2003, which means the time-distance between Beijing and Shanghai would be three years. Clustering for each year’s data by distance measure. Based on the distance matrix obtained from cluster analysis by distance measure, we can get the ranking of the informatization levels among regions in each year. The better the ranking the narrower the digital divide; and the worse the ranking the larger the digital divide. According to digital divide indexes, we cluster thirty-one regions into twelve categories, and in accordance with the ranking of regions, three types of the categories can be merged into one class, and eventually thirty-one regions are divided into the following four classes: • First-class regions: smallest digital divide, highest ICT application level. • Second-class regions: smaller digital divide, higher ICT application level. 34

b . z h a n g a n d r . d . t ay l o r

• Third-class regions: medium digital divide, medium ICT application level. • Fourth-class regions: large digital divide, low ICT application level. At the same time, in each class, the regions are divided into three categories, representing three ICT application levels in each region: high, medium, and low. According to the classification table for the digital divide for 2002–2007, the total classification table of those six years was produced. Beijing and Shanghai, as the leading regions in ICT application, have been in the first class, and their digital divide index rankings have stayed first and second for six years. Clustering the data for six years by relevance measure. From 2002 to 2007, ranking for Technology, Economy, Government, Education, and Society indicators for each year among the thirty-one regions were made into a set of data. By relevance measure clustering for this set of data, the thirty-one regions can be divided into a variety of types. In clustering the results using SPSS, twelve types of regions were selected and fine-tuned based on the actual situation. Then the characteristics of each type were summarized. The clustering results can be used as the reference for regional classification. When the characteristics of each type are described, the use of “outstanding” for a certain indicator means the ranking of that indicator is relatively higher than others, and “performs evenly” means the ranking of that indicator is a little lower than the “outstanding” indicator. When a certain indicator’s ranking is “relatively weak,” it means that indicator’s ranking is much lower than the “outstanding” indicator; the specific difference between them depends on the outcome of the clustering. Reasons for changes in digital divide ranking for regions in 2002–2007. We consolidated all of the rankings data to do the cluster analysis. There was interaction between different years’ data, so it was not useful for making policy based on the performance of each year. Because of this, relevance measure clustering was done separately for each criterion level indicator of thirty-one regions. Since showing all the results would require too much space, only one region, Hainan, is analyzed here as an example to show the rationale of ranking changes and to suggest policy implications. Hainan. There were fluctuations in Hainan’s digital divide index rankings, but not large. As to the classification by distance measure clustering, Hainan showed a steady increase in the third-class regions. Hainan stayed in the low end of third-class regions from 2002 to 2004, in the middle in 2005 and 2007, and in the high end in 2006. The data show that from 2002 to 2003, Hainan’s digital divide index ranking decreased eight positions. There was a slight increase in the u n d e r s ta n d i n g d i g i ta l g a p s

35

Education indicator in 2003, but the small increase did not compensate for the impact of the decline in other indicators’ rankings, especially the Government and Economy indicator rankings’ sharp decline, which was the main reason for the decline of Hainan’s digital divide index ranking. Although the ranking of its Technology indicator dropped, compared to other indicators’ rankings, the Technology indicator ranking was still better, so Hainan was advanced by technology. During 2003 and 2005, Hainan belonged to the “promoted by Government” type. The substantial increase of the Government indicator ranking and the slight increase of the Technology and Economy indicator rankings in 2004 prompted the digital divide index ranking of Hainan to sharply rebound seven positions. From 2004 to 2005, despite its Government indicator ranking dropping one position, Hainan still belonged to the “promoted by Government” type. Its Society indicator ranking’s substantial increase made its digital divide index ranking rise one position. Since the Society indicator refers primarily to the ICT application level, we can say that a significant increase in the level of ICT applications was the main reason for the increase of the digital divide index ranking. From 2005 to 2007 Hainan belonged to the type “promoted by Government and Society.” Although its Technology indicator ranking had a sharp decline in 2005–2006, the strong ranking of Government and Society indicators and the increase of other indicators’ rankings, especially the significant increase in its Economy indicator ranking, made Hainan’s digital divide index ranking rise two positions. The important influence of economic factors on Hainan’s digital divide index ranking started to appear. From 2006 to 2007 the Government indicator ranking increased slightly, and the Economy indicator ranking continued to show a significant increase. The economic factors had become the most important factors to enhance Hainan’s digital divide index ranking. At the same time, the Education indicator ranking advanced one position, but was still relatively backward. The Society indicator ranking dropped one position, while the digital divide index ranking remain unchanged. During 2002 and 2007, the Government indicator showed the greatest positive impact on the digital divide index ranking of Hainan, followed by the Economy indicator, but the pulling effect of the Economy indicator was gradually increasing, which meant Hainan was relatively positive in national ICT investment, scientific and educational input and innovation, and economic development. It is noteworthy that Hainan’s Education indicator rankings have always been low and the Technology indicator (which mainly refers to a variety of coverage and hardware subindicators) rankings sharply declined since 2006. 36

b . z h a n g a n d r . d . t ay l o r

case 3: using time distance analysis Typically in research, time is used as a key dimension for analysis and comparison. However, in studies measuring the digital divide, deep information contained in the time dimension has often been overlooked. The TDA approach can put the time dimension into such studies, showing the degree of change and the trends of variables over time, and reflecting the dimensions of the digital divide more clearly while providing the basis for further policy research. The development trend for every province from 2002 to 2007 can easily be obtained by comparing the difference between the digital divide indexes of each province and the national average, as the calculation of time distance is based on interpolation and compound growth rate. For example, assume that after 2007 the growth rate of certain indexes of Sichuan keep the same compound growth rate as between 2002 and 2007. Based on this premise, the strength of input of every index can be calculated as a reference for every province. The concept of time distance provides a new perspective for research on the digital divide. It also can measure the gaps between selected provinces and the national average. Figure 2-6 shows an example of the relationship between growth, efficiency and inequality. Static relative measure and time distance lead to different conclusions. Now let us take a broader view of the situation. The concept of time distance

Per capita income (log scale)

10000

Unit 1 Scenario A 1000

Unit 2

S121(t) S122(t)

R12(t) Scenario B

Unit 1

S121(t) S122(t) 100 1960

1970

1980

1990

R12(t)

2000

Unit 2

2010

2020

2030

2040

Time

figure 2-6.

The relations between growth, efficiency and inequality when based on a dynamic concept of overall degree of disparity. u n d e r s ta n d i n g d i g i ta l g a p s

37

for a given level of an indicator as one of the dimensions of disparity leads to a different conclusion about the degree of disparity in scenario A and in scenario B, as can bee seen in Figure 2-6. In the 4 percent growth rate for scenario A with the 50 percent static relative disparity (the more developed region has a 50 percent higher value of the indicator) the time distance between the two regions is ten years. In scenario B with 1 percent growth rate and the same static relative disparity the time distance between the compared regions is forty years. It is highly unlikely that people would perceive such situations as equal degrees of disparity, even though the static measure of disparity is 50 percent in both cases. Higher growth rates lead to smaller time distances, and thus have an important effect on the overall degree of disparity. This is based on both static and time distance, as both matter. Static measures alone are inadequate. digital divide measurement based on time distance analysis Actually there exists a significant digital divide between the thirty-one provinces in China. The level and the penetration rate of ICT technology among the provinces are totally distinct. Time distance analysis, which is different from the absolute and relative difference analysis, can show us the digital divide from another perspective. Digital divide index system. The same index system and weight distribution will be employed again, but will be analyzed through time distance analysis. The index system and weight distribution were presented in case 1. Time distance comparison of the digital divide index between thirty-one provinces and the national average. According to the indicator system and corresponding data of the digital divide index above, the digital divide index of thirty-one provinces and the national average from 2002 to 2007 were obtained. The procedure is as follows: first, the data from 2002 to 2007 was aggregated, and the national average value of every index in these six years as the standard value was considered. Then, based on the standardized data and the weight of every indicator, the composite digital divide index of every province was calculated, including the indexes of Economy, Education, Government, Technology, and Society. Then one-dimensional and twodimensional time distance comparisons were made with these indexes. One-dimensional time distance comparison. The time distance comparison of the digital divide index between every province and the national average from 2002 to 2007 was calculated. It showed that the digital divide is gradually widening. For example, the digital divide index of Beijing, Shanghai, Tianjin, and so on is leading the national average level by three to six years in 2002 and by four to eight years in 2007. The majority of backward provinces fall behind the national average by two years in 2002, but it becomes 38

b . z h a n g a n d r . d . t ay l o r

two to four years in 2007. Therefore, it can be concluded from the general trend of the digital divide index from 2002 to 2007 that the positive and negative digital divide are all getting deeper. Considering the trend of the time distance between the digital divide index of thirty-one provinces and the national average, it can be seen that three districts have an obvious difference, as shown in Table 2-2. District 1 is a positive area. The aggregative indicator value has kept ahead of the national average level. District 1 is mainly in the eastern part of China, such as Beijing and Shanghai. District 2 is composed of both positive areas and negative areas. The aggregative indicator value is almost equal to the national average, so the digital divide is not so apparent. District 2 is mainly in such provinces as Hainan and Shaanxi. District 3 is negative area. The aggregative indicator is behind the national average level. It should be highly developed in the future. If the current situation continues, it will lead to an increasing digital divide in this district. That will make these provinces continuously lag behind. District 2 is mainly in such provinces as Guizhou and Gansu. Two-dimensional time distance comparison. Two provinces (Zhejiang and Yunnan), which respectively represent the well developed and the poorly developed in terms of the digital divide, were chosen for deeper analysis. The broken line graph of the 2D time distance comparison is shown as Figure 2-7. The developing trend is obvious. The two polygonal lines are distributed in the first quadrant and third quadrant. If the digital divide of a district is less than the national average, the absolute difference is positive, and the time distance is also positive. So the digital development of all the provinces in the first quadrant is better than the national average. It can be seen from the polygonal line that Zhejiang is in the first quadrant, and Zhejiang developed well from 2002 to 2007, with its digital divide index and increasing rate always above the national average. So the development

table 2-2.

.

three districts with obvious differences in their digital divides

Area

Time distance in 2007

Trend of digital divide

Main provinces

District 1

More than 0

Positively widen

Beijing, Shanghai, Tianjin, Guangdong, etc.

District 2

More than –1, less than 1

In the range of plus Hainan, Shanxi, Jilin, or minus Xinjiang, etc.

District 3

Less than 0

Negatively widen

Neimeng, Jiangxi, Hebei, Xizang, etc.

u n d e r s ta n d i n g d i g i ta l g a p s

39

Absolute difference

Yunnan

Zhejiang 2004

2004

2006 2007 2005

−4

2002

2002

−2

2006

2003

2007

0.1

2005

0

2003

0

0.2

2

4

−0.1

Time distance (year)

figure 2-7.

2D polygonal lines show the time distance comparison of observed provinces (Yunnan, Zhejiang), 2002–2007; positive figures mean ahead of the national average.

track of Zhejiang shows a kind of linear increase and the same for the time distance and absolute difference. The absolute difference of the digital divide index of Yunnan is below the national average by 0.05 from 2002 to 2005, and its fluctuation is not quite obvious. However, removing consideration of the slight reduction of the time distance from 2002 to 2003 and from 2004 to 2005, the time distance dimension is increasing rapidly. The digital divide between Yunnan and the national average is growing. Because the digital divide index in 2007 shows negative growth, this leads to the rapid increase of the time distance. In order to understand the specific reasons for the changes in the digital divide, a detailed analysis is presented here only of Zhejiang and Yunnan. Time distance analysis of Zhejiang. The digital divide index consists of five sub-indexes: Technology, Economy, Government, Education, and Society. And these indexes consist separately of several specific actions each. Therefore, the key elements of the digital divide and the reasons which lead to the change of time distance can be shown clearly by comparing the time distance between the five subindexes of Zhejiang and the national average. The digital divide index of Zhejiang led the national average by 1.36 years in 2002, mainly due to the Technology and Economy indexes. The Technology index led the national average by 3.07 years and the Economy index led by two years. This is the main reason that the digital divide index of Zhejiang led the national average. But the performance of its Education index and Government index is poor. The Education index and Government index are respectively 0.67 and 0.55 years behind the national average. Therefore, in 2002 the leading position of Zhejiang mainly depended on its Technology and Economy indexes. At the same time, the Education and Government indexes still should be improved. The digital divide index of Zhejiang in 2007 exceeded the national average in 2002 by more than 3.19 years. It is again due to the Technology and Economy indexes. The Government index fell behind the national average 40

b . z h a n g a n d r . d . t ay l o r

compared with the prior situation. Therefore, although the advantages of Zhejiang are its Technology and Economy indexes, the Government index should be emphasized more, in order to improve its position. Time distance analysis of Yunnan. The digital divide index of Yunnan lagged the national average by 1.17 years in 2002. The root reason is its Education index, which fell behind the national average by 6.02 years. This enormous weakness led to the large gap between the digital divide index of Yunnan and the national average. The Technology, Government, and Society indexes are each behind their own national averages, but they are all above the digital divide index. The only leading index is the Economy index, which performed the best in all indexes in 2002, with a lead of 0.3 year over the national average. Therefore, the Education index mostly enlarged the gap between Yunnan and the national average in 2002. More attention needs to be given to improving it. The digital divide index of Yunnan lagged the national average in 2007 by 3.65 years more than in 2002. And again it is due to the Education index. The whole situation has not been improved since 2002 because of the poor basis of the Education index. It can’t be changed in the short term. The Technology and Economy indexes are both below the national average. It leads to their time distance getting larger in different degrees. The Government and Society indexes have both been improved. Both of them have a little lead over the national average. So the Education index is the key point which should be emphasized to gain further improvement.

case 4: using data envelopment analysis Data envelopment analysis (DEA) is an efficiency evaluation method based on the concept of relative efficiency. As an efficiency evaluation method, DEA is a unified approach to the various evaluation systems used to evaluate the effectiveness of units, based mainly on the input-output indicators evaluation system, which establishes the evaluation model for deriving the efficiency value for each unit. In case 4, DEA is used to study the digital divide based on five core components: Technology, Economy, Government, Education, and Society. In order to better analyze the impact of the various factors on the digital divide, the efficiency of input-output was divided into two levels for comparison: the analysis of the conversion efficiency of technology and of the conversion efficiency of social applications. Then, based on a production-profitability matrix, all the efficiency values are plotted into two dimensions in four quadrants, divided by the two reference points of the national average efficiency in each year. In each quadrant, representative provinces are selected for study, and for each, recommendations are proposed for improvement. u n d e r s ta n d i n g d i g i ta l g a p s

41

application of data envelopment analysis Indicators system’s structure. The indicators system follows the same structure as in the previous three cases. Procedure of analysis: In order to get a better analysis of the impact on the digital divide of the various factors, we divide the efficiency of input-output into two levels to provide a comparison, leading to a further analysis of the conversion efficiency of Technology and Social applications (see Tables 2-3 and 2-4). Based on the input-output models shown in Tables 2 and 3, Economy, Government, and Education are used as indicators of inputs in the DEA model of influence factors; Technology is used as a first-class output indicator to inspect the efficiency of input factors into the related indicators. Social factors are treated as a second output indicator, estimating whether economic policy, education, and development of ICT technology applications in the community are effectively transformed. data collection The data sources for this analysis are the China Statistical Yearbook (2003– 2007), CNNIC Statistical Survey Report on the Development of the Internet (2003–2008) and the China Information Yearbook (2003–2007). Next, we have to convert the Economy, Government, Education, Technical, and Social components into the corresponding indicators of inputs and outputs with integrated values.

table 2-3.

.

first level of input-output analysis— technical efficiency

Input indicators

Output indicator

1 Economic

Technique

2 Government 3 Education

table 2-4.

.

second level of input-output analysis— social efficiency

Input indicators

Output indicator

1 Economic

Society

2 Government 3 Education 4 Technique 42

b . z h a n g a n d r . d . t ay l o r

The geometric mean is used instead of the algebraic average because the geometric mean represents more balanced value judgments, not the development of indexes of inequality. The index is: Iˆt

n

I

i t

i 1

It on behalf of the value of the index T factor; n on behalf of the number of independent indicator composed of each factor; i on behalf of one of the ith indicators. For technology, n=10; for the economy n=4; for the government n=5; for education n=3; for society n=7. evaluation results and analysis The C 2R model is used to establish the corresponding linear programming model for the provinces and municipalities (decision-making unit) to obtain the decision-making units (DMUs), in order to express more clearly the concept of conversion efficiency. Based on the adoption of a DEA efficient defined model, we can see: • When technical efficiency or social efficiency is 100 percent, and the remaining variables or the slack variables of the influencing factors are 0, the result is referred to as technical DEA efficient or social DEA efficient, indicating that when the application of the impact of technology or social factors has reached the maximum input-output, the conversion has produced the best results. • When technical efficiency or social efficiency is 100 percent, but the remaining variables or slack variables of the influencing factors are not 0, the result is referred to as technical DEA weakly efficient or social DEA weakly efficient. However, in this study, as in the model, there is a relatively large number of variables as to the decisionmaking units for provinces, so there are no examples of the DEA weakly efficient case in which the optimal solution of 1 when only the variable value is 1, the value of other variables are 0, and slack variables and the remaining variables are 0. • When technical efficiency or social efficiency are less than 100 percent, it is referred to as technical non-DEA efficient or social non-DEA efficient, indicating that the impact of technology or social factors do not achieve the desired output results, which shows that the applications of technology or society need more development. At that point, the remaining variables and slack variables need to be u n d e r s ta n d i n g d i g i ta l g a p s

43

analyzed to find the root causes of the result being non-DEA efficient. Analysis shows that in recent years, technical efficiency is slightly higher than social efficiency, that is, the input of the influencing factors is better able to transfer into technology, while the application in the transformation of society is less certain. At the same time, the efficiency levels are increasing each year, and social efficiency levels are increasing faster than technical efficiency levels. In fact, for all provinces, the annual values of efficiency levels have increased in varying degrees, as a result of actively taking measures to enhance the level of ICT technology and the corresponding level of social application. For technical efficiency of output, the top provinces are Beijing, Jiangsu, Shanghai, Shandong, and Anhui. However, for efficiency of the social output, the top provinces are: Beijing, Shanghai, Guangdong, and Tibet. Thus, it can be seen that Beijing and Shanghai are leading the country in the technological and social application of conversion efficiency; Guangdong province, although a little behind on technical efficiency, has a high efficiency of social applications. Jiangsu, Shandong, Anhui, and other provinces, although they show a good use of government, education, economic, and other factors that raise the level of technology, do not have the same levels of application in this community. This shows that strengthening the ICT infrastructure should be followed by strengthening the community’s understanding and application of technology, in order to get the full benefit of the infrastructure construction. analysis of the relative position of technical efficiency and social efficiency The relationship between the efficiency scores obtained from both the technical efficiency and social efficiency assessments can be explored by means of a production-profitability matrix as proposed by Boussofiane, Dyson, and Thanassoulis. We use social efficiency as abscissa and technical efficiency as the longitudinal coordinates, and the average point of technical efficiency and social efficiency as the criteria for the classification, to divide the figure into four quadrants. These useful scores can also be plotted in two dimensions. The overall trends show that the criteria for classification gradually move to the right, and the technical and social efficiency of the average is increasing on a year-by-year basis. For all provinces, technical and social efficiency correspondingly increased more quickly. Judging from the details, each quadrant has its own characteristics. Two characteristic provinces from two of the four quadrants are selected and compared. The remaining variable 44

b . z h a n g a n d r . d . t ay l o r

s1*- shows the output of the economic component; s2*- refers to the government component, s3*- to the education component, and s4*- to the technical component. Representative province of the first quadrant: Zhejiang. Zhejiang Province has usually been located in the first quadrant (except for 2002); the quadrant is characterized by relatively high technical and social efficiency values. From the point of view of the actual numbers (except for social efficiency in 2002), all of its technical and social efficiency measures are more than 75 percent, and efficiency at all levels is ranked at the top, that is, the technical and social conversion efficiencies of the applications are relatively good. However, through analysis of the remaining variables, it can be seen that there are still factors constraining the efficiency of technical efficiency and social efficiency. From the point of view of technical efficiency, the remaining variable is only s2*-, showing that the main factor constraining transformation is Government. Government should intensify its policy efforts to invest in the information industry and in technical research and development so as to enhance the rate of technology transfer to achieve the desired results. From the impact factors of social efficiency, although there are some changes in six years, the main variables remaining are s1*- and s4*-, which means the major influencing factors are the Economic and Technical aspects. Although the implementation of technology is improving the infrastructure, its availability did not sufficiently increase the consumer economy, so the application of ICT technology alone is not enough. Available technical facilities are idle in this situation, so that the efficiency of social applications has not achieved the best results. These areas should be targeted for future improvements, so that consumers, through more channels and means of using ICT applications, improve their utilization rate of the infrastructure. Representative province of the third quadrant: Heilongjiang. From the point of view of the actual numbers, during the period 2002 and 2007, Heilongjiang Province, whose application of technical and social efficiency has been lower than the national average, held its technical efficiency basically at the 60 percent level with fluctuations, and an upward trend was not apparent. Social efficiency was even lower, on average only about 45 percent, and except for 2003, when the figure increased a bit more than a basic fluctuation from the average, there was no obvious increase. From the point of view of technical efficiency, the remaining variable is only s2*-, showing that the main factor constraining transformation is Government; from the components of social efficiency, the main variables for the remaining factors are s1*- and s2*-, indicating the major influencing factors are Economic and Government. A major factor impacting technology is the government. Lacking sufficient local technology infrastructure investment results in a low rate of technology transfer, the technology infrastructure is not adequate without the u n d e r s ta n d i n g d i g i ta l g a p s

45

corresponding technology in support, and that, coupled with the awareness that ICT spending is weak means it cannot bring about a high level of social application. The most urgent tasks are first to strengthen technology-related facilities, and second, within the existing technology level, to strengthen economic development and improve people’s consumption level and their level of awareness of new technologies, in order to fundamentally solve the problem of informatization.

the future: challenges and opportunities strengths of digital divide measures Most current digital divide and e-readiness measures are basically descriptive and comparative. They are good for describing and comparing the status between and within economies and regions, and showing changes over time. The value of this is not to be underestimated. It allows us to compare countries and regions, indicate areas of strengths and weakness, and provide timeseries data for interpretation. The use of composite indicators to create ordinally ranked (i.e., ranked in order of priority) sets of economies provides policy makers with relevant directional guidance. Studies of this type have been widely used internationally, and developed by both governmental and private bodies. Indeed, there are many models of such studies, not all of which are compatible. They have been very useful in drawing attention to the nature and size of the digital divide, and helping focus energies on overcoming it. This is very significant and useful. Such studies, however, are not as strong at answering the questions of how and why changes occur; at showing causality; at identifying the relative importance of factors; and at presenting methods to predict the future. They do not always make evident the internal mechanisms of informatization in ways that can be tested and validated. Taking this to the next level requires developing approaches that are both explanatory and predictive, and that will require a different way of thinking about the theory of informatization. Information technology in society is a complex adaptive system. It needs to be seen comprehensively. It is an integrated system of technologies, networks, industries, content, policies, laws, regulations, and social factors. In a narrow sense, it is linked to productivity and growth; in its broad sense, it shapes the foundation of all social relations. All these aspects need to be seen together. Emergent systems like information networks often involve such a high degree of causal complexity that traditional modes of analysis may not be adequate. Thus some shy away from this area simply because it is difficult to operationalize. This type of analysis is not economics, but the application of social science statistical tools to an array of economic, social, cultural, 46

b . z h a n g a n d r . d . t ay l o r

political, and other data to identify relationships and effects. There is much to be gained from rising to this challenge. limitations of digital divide ⁄e-readiness measurements There is no perfect way to measure either the digital divide or e-readiness, due to the limitations of dealing with proxies (indicators). They both involve a complex analysis of inhomogeneous dynamic adaptive systems. Most existing approaches are subject to certain inherent methodological limitations. It is possible that these approaches have reached the limits of their potential. Two general critiques are paramount: conceptual imprecision and issues of measurement. Without clear definitions of concepts, scientific discourse is difficult if not impossible, and any resulting theory will lack clarity and precision. This conceptual imprecision (e.g., information, digital divide) may be responsible, to some extent, for the subjective interpretation, implicit value judgment, or ideological claims, which we often find in the literature on the process of social informatization. According to the OECD, in general composite indicators suffer from the following weaknesses, which reduce their analytic value: • Conceptual imprecision • Studies not comparable across time or place • Choice of indicators and weights subjective • Data/indicators often not comparable • Computation ad hoc • Lack of concurrent validity • Built on different values • Sensitivity to different weighting and aggregation techniques • Problem of subjective indicators (e.g., well-being) • Problem of handling missing data This approach is built on empirical data, and involves sophisticated mathematical procedures, so it is scientific, but it generally does not proceed in the traditional scientific model of hypothesis testing (observation, hypothesis, experiment, theory, and prediction). toward a general model The global process of identifying indicators, establishing accepted definitions of indicators, uniformly collecting data, and storing it and making it accessible in a consistent and timely manner is well underway. At the same time, it is necessary to be mindful that information needs a theoretical u n d e r s ta n d i n g d i g i ta l g a p s

47

framework – that is the difference between description and explanation. And the role of experimentation and hypothesis-testing for suggesting causal relationships between variables is a critical factor that cannot be ignored if theory is to be grounded in data linked to the world we live in and be used as a foundation for policy and economic choices. That said, and with all due recognition of the challenges, difficulties, and limitations of a large-scale and meaningful information metrics program at the national level, useful contributions from these experiences are possible for directing economic and social development in real time. A good example of this is well underway in China. exploration of new models What is called for is the development of a general theory of informatization, of which the analysis of the digital divide and e-readiness are just particular applications. That is, the digital divide is not informatization. Informatization is an approach to understanding and resolving the digital divide. They are each specific applications of the idea of informatization, designed to answer different questions. Informatization is not just about measuring the digital divide or e-readiness, it is a general theory of the transformation of society by information technologies, of which they are both subsets. Thus far, the study of informatization shows a lack of a coherent theoretical approach. China has an implicit experimental approach—“stepping on stones to cross the river” (try, evaluate, proceed or retry)—which may ultimately lead to theory from the “bottom up,” building theory based on experience through a “natural experiment.” However, this kind of inductive approach, while it may allow for some qualified predictions, does not necessarily explain the mechanisms of why or how effects occur (causality), or help us to model informatization in specific circumstances. statistical methodologies We have presented four statistical approaches to understanding the digital divide. They are very useful tools that add to our collective understanding, and continuously move us toward better models. The discussion in this section does not advocate for or against the use of any particular statistical approach to analyzing data. What is important is the ability to test theory-based hypotheses, discern causal relationships, and permit at least probabilistic predictions. To this end, approaches that reduce the level of subjectivity with respect to selection of indicators and the weighting of factors are to be preferred. The goal is to “let the data speak for itself.” This will likely require statistical tools and computer programs capable of analyzing very large amounts of data. And the results may be multidimensional, not just one-dimensional rankings. 48

b . z h a n g a n d r . d . t ay l o r

going forward The study of informatization needs to develop testable hypotheses, and move from an inductive approach to a deductive, experimentalist approach. But evidence of progress (or the opposite) must be measured against initial goals, so first the ultimate goals must be defined. They could be: • Economic prosperity • High quality of life • Harmonious society • Satisfaction/Happiness Each of these possible goals is different, but aspects of them are not mutually exclusive. What is critical is that the chosen goal(s) must be carefully defined, and then specified as measurable/quantifiable components. Based on experience, it is possible to formulate initial assumptions to test. Then going forward, as policies or investments are made in IT-related factors, the process will need to: • Define the goal • Look at past data • State the hypothesis to be tested • Formulate the question related to achieving the goal • Present a prediction as to the outcome • Specify the main components (factors) related to the goal • Quantify the components • Identify and quantify the relevant indicators • Apply appropriate mathematical tools (evaluate relevant methodologies) • Implement • Measure • Evaluate • Revise The use of testable hypotheses is the path to determining if it is possible to develop a generally applicable theory of informatization, which will apply to both developed and developing nations, indicate the relevance of particular factors under specified conditions, and (based on an analysis of complex cross-causalities) permit a probabilistic prediction about the consequences u n d e r s ta n d i n g d i g i ta l g a p s

49

of particular policy choices. The paradox appears to be that the more extensive and detailed the information used to analyze any particular situation, the more the outcomes will relate only to that particular situation. So any general theory will have to be adapted to the specifics of the case at hand, and the goals of the users. However, it is the development of such a general theory that is the key to further future advancement.

50

b . z h a n g a n d r . d . t ay l o r

CHAPTER 3

Broadband Microfoundations The Need for Traffic Data steven bauer, david clark, and william lehr To date, most of the empirical effort to understand broadband service markets has focused on availability and adoption metrics and data. Data of this sort is indeed valuable when the dominant policy questions concern penetration and uptake. However, as broadband availability and penetration saturate, such data will become less informative. The next set of questions, both for service providers and regulators, will center on the continued health of the broadband access market: levels of investment, the competitive landscape, the evolving definition of broadband, the degree of neutrality in consumer access, and the nature of interconnection among providers. Our position is that network traffic data will be central to understanding and answering many of these questions. To answer these sorts of questions it will be helpful to know such things as the distribution of usage across the user population, the characteristics of users that participate during peak periods of network congestion, and the variance in usage and how it differs by type of user. This data will help inform forecasts of capacity/infrastructure investment needs (e.g., how much bandwidth does a subscriber need? How much sharing is feasible at which points in the network?), to understand ISP costs, and to assess network management practices (e.g., traffic engineering). Better traffic data will provide insights into consumer adoption decisions and the evaluation of product offerings (e.g., how important are peak rates versus average data rates?).

51

background The Internet is often compared to the network of highways, streets, and roads that make up the transportation system. Both are vital infrastructures that provide businesses with access to materials and markets, and provide people with access to goods, services, recreation, jobs, and each other. For transportation networks, it is generally recognized that traffic data (i.e., the volume of traffic, congestion information, incident reports, and so on) is as important to understanding the state of the network as is information about where the roads or links actually are. The same is true for the Internet. In transportation networks, traffic data is valuable over both short time scales (e.g., allowing real-time traffic management to reroute commuters around a rush hour accident) and over longer time scales (e.g., for planning maintenance cycles and capacity expansion investments). During periods of congestion,1 traffic data and real-time traffic management via lights, tolls, and special commuter lanes has proved especially important in enabling more efficient utilization of the existing transportation infrastructure. Improving the efficiency of the existing infrastructure delivers benefits in the form of reduced commute times (contributing directly to labor productivity), improved safety, and reduced pollutant emissions through intelligent traffic management policies.2 On the Internet, traffic data is similarly important to network operations. Over short time scales ranging from less than a second to hours or days, traffic data is an input into systems (both automated and human-centered) that make routing decisions (e.g., balancing loads across different network links), identify suspected or actual security or transmission failures, and implement traffic management policies.3 Over longer time scales measuring months or years, traffic data is vital to capacity planning and provisioning, allowing capacity to be efficiently installed in advance of demand, thereby better accommodating future traffic growth without congestion-related disruptions. Thus traffic data is essential to almost all the practical dimensions of network management and to the political, regulatory, and theoretical questions of what constitutes good, acceptable, or socially desirable network management. While traffic conditions on the highway and roadways can be observed externally (via both technical sensors and human observations), information about Internet traffic and the level of congestion of the different autonomous networks that collectively compose the Internet is limited. While individual network operators generally have a good idea about the state of their own networks, outside stakeholders have little visibility into the state of traffic on networks. Networks can be probed and tested by outside observers to derive some measurements, but the scope and confidence of such measurements is 52

s. bauer, d. clark, and w. lehr

limited compared to the accuracy and breadth of information available to network operators. The majority of users have very little visibility or understanding of what is happening to their traffic once it enters a network. Without better visibility, it is not surprising that there are widely diverging opinions about the true state of networks. For example, what are the congestion and utilization levels now and in the predictable future? What are the underlying cost structures for carrying traffic and expanding capacity? Or, what are the effects of different traffic management policies?4 The problem is that this limited visibility by outside stakeholders into the traffic and congestion state of networks makes it hard to have confidence in the regulatory and investment decisions that affect such networks. On the one hand, traffic management policies that are efficient and “fair”5 could be disrupted or private investments in expanding capacity could be deterred. On the other hand, network operators could be exploiting their control to thwart or discourage disruptive new innovations and competitors (either intentionally or accidentally). Furthermore, in contrast to our transportation grids where most of the physical infrastructure (roads, terminals, bridges) are publicly funded and managed, most of the physical infrastructure that composes the Internet is investor-funded and privately managed. We rely chiefly on profit-motivated firms and market competition to direct resources to their best uses for the collective benefit of society and the economy. For markets to work efficiently, they depend on the public availability of relevant market information to allow buyers and sellers to formulate their strategic decision-making. For example, market participants need to understand what they are purchasing or selling, between whom the exchanges are to occur, and the timing and terms for market transactions. Markets generally work best when they are lightly regulated, so ensuring that the appropriate information is produced by the market process presents an interesting challenge for institutional design and incentive compatibility.6 Our thesis is that better visibility by outside stakeholders into the traffic data of networks is required to improve the regulatory processes, investment/market decision-making, and technical research. More generally, better visibility of what traffic looks like in networks will promote understanding and trust between what ultimately has to be a cooperative community of interconnecting and communicating parties. At least some network operators are interested in sharing their internal data in order to facilitate this process (see below). The challenges to making this happen are multidisciplinary: engaging aspects that are technical (how to sample or share the potentially terabyte-sized data sets), analytic (how to compare and combine data generated by different measurement processes), b r o a d b a n d m i c r o f o u n d at i o n s

53

policy-oriented (how to preserve the privacy of individual subscribers), and business strategy related (how to protect competing providers’ business interests). Addressing these problems while still producing data capable of providing useful insights into the important questions noted above is nontrivial and requires a multifaceted and process-oriented approach that is capable of evolving as the Internet evolves. In the following sections we give an overview of what traffic data is generally available in networks, how this data may prove important in answering questions that are relevant to the entire community of stakeholders, and why collecting the data and making it more generally available is challenging. We conclude with a discussion of open research questions and a brief overview of some of the interesting research efforts that have been initiated in recent years.

traffic data The amount of information that could be collected by network operators from their networks is enormous. Individual network elements—which include routers, switches, servers, caches, and subscriber modems—can report hundreds of different statistics. With hundreds of thousands of elements in a network, and millions of subscriber lines, the volume of potential data is enormous. For example, one network operator we spoke with indicated that the total volume of data records could exceed three hundred terabytes of data a year. Collecting and transporting the raw data in real time to the network operations center where it can be processed, analyzed, and managed presents a difficult challenge that incurs significant operational costs. Determining what data to archive and how to compress or summarize the data and manage access presents complex statistical, logistical, and policy challenges. In spite of the costs, network operators do systematically collect real-time traffic data because it is essential for successful network operation. The data is an input into strategic and operational decision-making across virtually all ISP functions. The data informs decisions about the capacity of internal links, routing policies, security policies, and interconnection contracting. It is used for high availability and disaster recovery planning, for financial projections, employee evaluations, technical strategy discussions, and sales and marketing. In larger network operations, there are specialized departments focused on managing the collection and analysis of network traffic data, and the sharing of relevant portions and views of the data across the organization. One of the most important uses for the traffic data, after monitoring the health of the existing network, is capacity planning. This is accomplished by studying the utilization of network links averaged over multiple time intervals. 54

s. bauer, d. clark, and w. lehr

The utilization data of a link is collected from a router using a protocol such as SNMP (Simple Network Management Protocol). Figure 3-1 shows utilization graphs from one of the gigabit Ethernet links connecting our lab to the main MIT campus network. Our lab sends more traffic (top line) than it receives (bottom solid color) because we host a number of popular sites including mirrors of software distributions. One can see in the graph the diurnal variations in traffic displayed. On the top are the average bits per second and on the bottom are the average packets per second. Both are potentially important statistics as a router can be congested because of the volume of data (each packet carrying the maximum amount of data) or the number of packets (each packet could have little data but there could be hundreds of thousands of packets). Congestion in most networks today is more likely related to excess volume than excess packets. For this particular link at MIT, there is no particular evidence of any persistent performance or congestion problems. While we don’t display it here, the long-term trends on this link also don’t suggest congestion or

figure 3-1.

Measurement of bits per second. b r o a d b a n d m i c r o f o u n d at i o n s

55

performance problems in the near future as there isn’t significant growth in the aggregate traffic levels. However, if there were hints of impending traffic congestion, other forms of data would be instrumental in analyzing the causes and planning the course of action, and to understand trends, it would be necessary to have time-series data documenting utilization across time. A network operator at our lab might first look at flow-level details using data such as Netflow records.7 These records provide a way of looking inside the aggregate flow to better understand what combinations of edge sources and destinations (forming a traffic matrix) are actually communicating. Such data is essential to understanding whether peering or upstream connections should be modified. For instance, this data might indicate that our lab sent and received a significant amount of traffic to and from the Harvard campus. Therefore we might be able to reduce the utilization on the loaded link by establishing a separate direct peering connection to Harvard, thereby offloading some traffic to the new link (see Figure 3-2). Another way of reducing link utilization is to constrain the top contributors to traffic on a link. Table 3-1 is a list of top traffic contributors, again derived from Netflow data. At our lab this data is monitored primarily to identify anomalous sources of traffic such as hosts that have been infected and are unwittingly serving as bots or data depots for hackers. But given that it is a shared research network, disputes can arise as to what constitutes acceptable use of network resources. These tools provide a way of objectively identifying “hot spots” and measuring the impact of different experiments, websites, and uses. Another way in which detailed traffic data is employed is to examine what protocols and applications are being used on a network. This data is significant

Internet

MIT CSAIL Lab

Harvard

Establish direct peering link?

figure 3-2.

56

Traffic matrixes derived from Netflow style data is one way of determining where peering links should be established.

s. bauer, d. clark, and w. lehr

table 3-1

.

top contributors to traffic on the csai network for one period in september

Rank Address

Bit/sec In

Bits/sec Out

1

carver.debian.org 128.31.0.50 (1 sample)

60.0 M 1.4 M

2

infmite-state.csail.mit. 48.2 M edu 128.30.24.177 (1 sample)

3

rore.debian.org 128.31.0.49 (1 sample)

4

Pkts/sec Pkts/sec Flows/sec Flows/sec In Out In Out 6.3 k

3.2 k

223.3 m 233.3 m

1.3 M

4.1 k

2.1 k

3.3 m

3.3 m

38.1 M

735.2 k

4.1 k

463.3

29.7

28.9

mosdef.w3.org 128.30.55.83 (4 samples)

35.4 M

573.8 k

3.1 k

1.3 k

10.8 m

10.0 m

5

newsswitch.csail.mit. edu 128.30.2.35 (37 samples)

18.1 M

7.6 M

1.9 k

1.7 k

241.5 m 240.9 m

6

thursday.csail.mit.edu 10.2 M 128.30.100.224 (1 sample)

200.7 k 897.9

463.8

13.3 m

7

30-7-158.wireless. csail.mit.edu 128.30.7.158 (3 samples)

4.6 M

71.2 k

390.7

184.9

250.0 m 248.9 m

8

planetlab3.csail.mit. edu 128.31.1.13 (37 samples)

4.1 M

3.5 M

956.2

957.6

85.9

93.5

9

mdemaine.csail.mit. edu 128.30.48.115 (2 samples)

3.4 M

57.3 k

300.5

109.3

1.4

1.4

10

xyz.csail.mit.edu 128.31.0.28 (19 samples)

3.3 M

3.7 M

504.7

514.5

1.7

2.1

23.3 m

both to identify anomalies (a significant rise or drop in any category might indicate a problem) and to understand and predict future traffic growth. (Figure 3-3 shows a sample of the protocols in use on our lab network.) Particularly as new video-centric applications and services become more b r o a d b a n d m i c r o f o u n d at i o n s

57

figure 3-3.

Top contributors to traffic on CSAILs lab network for one period in September 2009.

popular, monitoring their adoption will be crucial to capacity planning.8 Many of the emerging applications transmit their data over random ports or standard web ports (thereby mixing in with other types of web traffic) so Netflow data records may become less useful for monitoring the adoption of “new” applications over time. Other tools and measurement devices—often referred to as Deep Packet Inspection or DPI—enable more detailed traffic analysis on a per flow or per packet basis. These techniques seek to classify traffic flows by looking at other information both within the packets and other predictable signatures such as the pattern of communication (bytes transmitted during the initial connection handshake, and so on).9 While individual ISPs collect such data on their networks, they have little insight into the detailed traffic patterns on other networks, even ones they may be directly connected to. In spite of the need for such data to monitor the macro-economic health and direction of the broadband marketplace, such data is not readily available publicly. A notable exception is the excellent collaborative research project among ISPs that has been underway in Japan since 2004.10 That project represents the most advanced publicly reported broadband data project undertaken to date—seven large ISPs, carrying roughly 40 percent of Japanese traffic, contributed summary data on traffic characteristics at least twice yearly since 2004. This data offered a 58

s. bauer, d. clark, and w. lehr

compelling picture of the growth and distribution of broadband traffic as experienced in Japan. While there are many different interesting details that emerged from their work, we highlight some here to give concrete examples of how traffic data connects to important macro-economic issues. Unsurprisingly, the transition to broadband has fundamentally changed traffic patterns on the Internet. The effects of this transformation are sometimes obvious, but also sometimes surprising. For many years, the peak usage periods of access networks (which generally serve both residential and commercial customers) were during the business day. However, for at least the last several years, the peak usage hours of many access networks are in the evening, roughly between 9 p.m. and 11 p.m. (see Figure 3-4). This is important for understanding the economics of networks as the previously off-peak residential customers used to more easily “fit” in the pipes that had been provisioned for the peak-using commercial users. Now, however, the usage patterns of the residential customers are often driving the provisioning decisions of network providers.11 This has obvious implications for cost sharing and service pricing. The Minnesota Internet Traffic Studies (MINTS), run by Andrew Odlyzko at the University of Minnesota, has been monitoring traffic growth

figure 3-4.

Residential broadband traffic in May 2005 (top) and May 2008 (bottom) as measured in Japan. Source: Cho, 2008. b r o a d b a n d m i c r o f o u n d at i o n s

59

levels on networks for a number of years.12 The traffic data in his study is derived from publicly available data sources such as peering points and sites, such as universities, that post information about their traffic. Most of his data comes directly from Multi Router Traffic Grapher (MRTG) and Round Robin Database (RRD) graphs (very similar to the previous figures in this chapter). The raw data used to generate the graphs would be even more informative and presumably preferred by most analysts, but most sites do not make it available. The MINTS data shows that the aggregate level of traffic continues to grow at double-digit rates, recently averaging around 50–60 percent compound annual growth rate (CAGR) per year (see Figure 3-5).13 While these growth rates are impressive, they are substantially lower than the rates widely cited in the trade press over the years.14 Kenjiro Cho et al. (2006) used the Japanese ISP data to investigate how the mix of applications on broadband networks is changing. Addressing one of the most significant questions for the near-term traffic growth—the macro-level impact of video—Cho noted that “the current traffic is heavily affected by an eruption of peer-to-peer applications but the crust underneath is also slowly rising with video and other rich media content. The crustal movement is slow at the macro level so that it is unlikely to cause a major quake in the near future.”15 This is a good metaphor as the increasingly ∧ ∧

8

∧ ∧∧ + +

+

4

Growth rate

+ +

+

2 +

+ + ++ +

+

+

++

+

+ + + + + + + + +

1 + +

+

+ + + +

+

+

+ ++ +

+

+ ++ + ++ + + + + + + + + +

+ ++ + ++

+ + +

+

1/2 + +

1/4 1.0e+05

1.0e+06

1.0e+07

1.0e+08

1.0e+09

1.0e+10

1.0e+11

2008 Traffic, bits/s

figure 3-5.

60

2008 traffic growth rates from publicly observed sites in the MINTS traffic study.

s. bauer, d. clark, and w. lehr

popular video traffic does not pose an imminent threat to the stability of the Internet, but the growth in video traffic will be significant, eventually fundamentally reshaping the traffic mix on broadband networks. This will have unmistakable economic impacts on regulatory policy, innovation in new applications and services, competition, and the value chain of network vendors and suppliers. The final question on which existing public data sheds some light is the distribution of traffic among subscribers on a network. This is significant because networks are shared resources where not all traffic demands can necessarily be simultaneously satisfied. So there is a very basic question as to what constitutes fair sharing of a network. Users are sometimes categorized as exhibiting “heavy” versus “regular” or “light” usage patterns. The relationship between the aggregate volumes of traffic a subscriber sends or receives and their contribution to congestion (in terms of causing packets to be dropped) is not always clear. It is possible that a “heavy user” does not disproportionately contribute to either packet dropping congestion or to usage during the aggregate peaks on a network. What is clear though is that there are very large differences in the volume of traffic sent and received by different subscribers. While most users may download less than two gigabytes of traffic in a month, the top users on a system can easily exceed one hundred gigabytes. Figure 3-6 displays the average daily inbound and outbound traffic per user on a fiber network in Japan measured over a week in 2008. Each dot represents one user.

Daily inbound traffic (byte)

1011

(c) Fiber (2008)

1010 109 108 107 106 105 104 104

105

figure 3-6.

106 107 108 109 1010 Daily outbound traffic (byte)

1011

Correlation of daily inbound and outbound traffic volumes per user in one Japanese metropolitan prefecture for a fiber optic network in 2008. Each dot above the dashed line represents users that sent more than 100 megabytes of traffic in a day. Source: Cho, 2008. b r o a d b a n d m i c r o f o u n d at i o n s

61

As peak rates increase, and hence the possibility for sending and receiving ever larger amounts of traffic grows, there exists the potential for an increasing divergence between the volumes of traffic that different segments of the market send and receive.16 This is not problematic in and of itself. A challenge will arise, however, if these very different usage patterns are associated with different underlying cost structures either in terms of the congestion they contribute to or in terms of the variable costs their usage incurs (e.g., transit charges from an upstream network provider).

the importance of traffic data The previous section provides just a sample of the extensive history in the networking community on research detailing the technical behavior of networks. Indeed, the properties of individual links, paths, hosts, and networks have been extensively analyzed. While these measurements have served the purpose for which they were designed, connecting these technical details and data to inform the economic, regulatory, and policy challenges of networking is a relatively new challenge.17 What are missing in most regions of the world are collections of data and measurements that provide a richer picture of the overall state of networks. As demonstrated, this is data that broadband providers routinely collect and analyze in their individual network operations centers, but is rarely understood or shared with the wider community, including other operators. By aggregating views of multiple individual networks, a picture of the issues, opportunities, and problems confronting both individual networks and the collection of networks that comprise the Internet18 can be developed while still protecting the confidentiality of individual network operators and subscribers. In particular we see this data as important to establishing traffic trends, growth, and characterizations at both the aggregate and subscriber level; vital inputs into a data-driven discussion of network management practices; promoting public and industry awareness of the challenges, successes, and opportunities in the broadband marketplace; and assisting in diagnosing and understanding traffic problems and phenomena. Broadband traffic characterization. Data about broadband and network traffic is needed to develop representative aggregate and subscriber traffic models that are used to analyze and forecast market trends and plan network provisioning and management. While aggregate growth statistics indicating the total volume of Internet traffic or subscribership are clearly important and regularly cited in company annual reports, municipal broadband plans, policy debates, research papers, and the popular press, more detailed and less aggregated data are needed to understand the composition of the aggregates, 62

s. bauer, d. clark, and w. lehr

to identify local phenomena, and to discern the drivers and relationships among the subcomponents. Data on top-line growth alone is not adequate to address questions about the changing mix of applications (e.g., peer-topeer versus streaming video), differences in platform technologies (e.g., cable modem versus DSL versus wireless), or changes over time (in response to changing technology, network architecture, or the industry ecosystem). Data to allow the decomposition of aggregate growth are needed for the development of rich future scenarios and to support flexible “what-if” analyses. User traffic data is needed to understand “within” and “across” user traffic distributions (e.g., how do subscriber usage patterns vary across subscribers and across time?). When linked to cost and revenue data, representative traffic data underlies a fuller understanding of broadband economics. Traffic diagnosis. Better traffic data will enable the analysis of significant traffic events. A number of organizations currently produce analyses based upon their view of both public and private data. These include analyses of the effects of the de-peering incidents,19 significant cable cuts,20 routing incidents,21 major media events,22 and security incidents.23 Each of these provides important lessons learned in terms of understanding the actual and potential effects of the incidents and also learning how they might be prevented in the future. More traffic data would provide a richer picture of the effects of these incidents on communications, business, and users. Traffic management. Representative traffic samples and broadband traffic models will prove useful in enhancing simulations and in testing network management approaches, including congestion management strategies. This is particularly important now since access providers have met with opposition, from a mix of stakeholders, to the deployment of network devices that implement a variety of congestion management policies.24 These policies often change the network resource allocations that would result from the distributed actions of host applications and TCP stacks. While the result is certainly different than what would occur without these devices, it is not de facto unfair or welfare reducing. However, the wider Internet community might regard it as unfair or inefficient depending upon the policies that are implemented.25 Promoting public and industry awareness of the challenges, successes, and opportunities in broadband. There is a lack of awareness across the Internet value chain and within the wider community of the challenges posed for infrastructure investment, especially in last-mile networks, from Internet traffic growth. In 2004, as part of the Broadband Working Group in the MIT Communications Futures Program (a collaborative effort with academic and industry partners from across the broadband value chain) we examined what we termed the broadband incentive problem—the challenge of incentivizing ISPs to continue investing in expanded capacity in the face b r o a d b a n d m i c r o f o u n d at i o n s

63

of rising traffic-related costs.26 As household subscribership approaches saturation, the growth in access revenues priced at a flat monthly rate per subscriber line will slow, but aggregate traffic will continue to grow. Reduced investment by ISPs in expanding network capacity poses a threat to innovative, high-bandwidth uses of the Internet. How best to resolve this quandary is a challenge for the entire Internet value chain and will likely require a mix of new investment, new usage and pricing models, and better network management. Evaluating the economic health of the broadband marketplace is also important. There is a growing research literature documenting the economic benefits of the Internet and broadband for employment and productivity.27 More granular data about subscriber usage patterns would help improve these studies and offer insight into how best to promote universal adoption and how best to target public broadband funds.28 Public traffic data also would offer a perspective on where the opportunities in broadband are. In the future, understanding where broadband service is available and which households are subscribers will become less interesting relative to questions about how broadband is being used to support novel applications directed at improving education, health care, business processes, entertainment, and communication. Which regions are leading and lagging in the adoption of these innovations will be of general interest.29

challenges In this section we discuss some of the challenges of collecting an appropriate multi-ISP traffic data set. One of the primary challenges is that the data requirements evolve over time as the measurement infrastructure changes (both in terms of measurement locations and methodology), the questions asked of the data change (requiring more granular or detailed data on a particular topic), and legal and regulatory obligations are modified (changing what can or must be collected). Thus the institutional frameworks put in place to gather data must be flexible and able to accommodate changes. This is nontrivial because forging even temporary agreement on a methodology requires the assent of the technical, legal, and management teams of all participating organizations.30 Technical challenges. All the typical technical challenges associated with data collection arise: missing data, spurious data, missing metadata, and ambiguous fields. Data can and is commonly lost as systems are moved, upgraded, and reconfigured. If a data collection process for a network temporarily fails, it is often impossible to restore past data.31 Varying network measurement methodologies are common over time, across ISPs and measurement equipment providers, and even within a single 64

s. bauer, d. clark, and w. lehr

provider’s network because the provider may have a mix of vendor equipment and legacy systems. The precise location of traffic probes in a network determines what traffic is measured. In the case of analysis boxes (or DPI)32 that identify applications and protocols, the choice of equipment vendor and the rules in effect at any point in time (i.e., what measurement options are set and the current generation of vendor software) have a considerable impact on how traffic is classified (e.g., how much traffic may be classified as “other”). Because traffic classification techniques differ across equipment vendors, one could expect different traffic classification results even for an identical stream of traffic. While we are not aware of any systematic study of the differences, the network operators we have spoken to indicate that such differences are common. The sheer size of the data sets can also present challenges. Depending on the ISP, the data sets may range from small comma-separated data files of less than one megabyte in size to specialized databases that collect hundreds of terabytes of data a year. Large data sets often dictate that sampling procedures be employed otherwise even basic queries can take hours to run. One network operator we spoke with indicated that, in their initial data collection setup, running a database query at the same time as data was being collected was impossible. Analytic challenges. An area of particular interest is how to match traffic characterizations with other types of data in ways that allow analysts to better understand aggregate and per-user behavior while protecting against ex postuser identification (a challenge we discuss below). There is a great deal of information that would be desirable to collect and compare but that, in practice, is challenging to acquire. For instance, to better understand the drivers of user behavior, it would be desirable to understand what other services (telephony, video, premium video, and so on) a subscriber takes, the advertised service characteristics (peak rate, service pricing, and so on) of each subscriber, subscription timing (when was service first initiated, when changed, when terminated), geographic location data about the subscribers, and other types of demographic data.33 At least in some providers’ networks, it is hard to bring together service plan information with usage data. Not only are the databases physically separate, they also reside in separate organizational units within the business. Even the internal analysis teams are stymied at times when they seek to match usage and service description data. It is also challenging to answer some analytic questions for technical reasons. If one wanted to analyze how traffic demands shifted immediately following a capacity upgrade, it is difficult in practice to identify the precise timing for when the upgrade took effect for an individual subscriber. Just because a subscriber has been authorized to utilize higher peak sending and receiving rates, the subscriber may b r o a d b a n d m i c r o f o u n d at i o n s

65

not have rebooted their modem to pick up the new settings and hence would still be running with older and slower rates until they do. Even once data is collected, analysis is complicated because there are no generally accepted “correct” metrics for many of the questions that come up in discussion of traffic data. In a related paper,34 we presented several different definitions for congestion and recounted some of the intellectual history in the evolution of thinking about congestion to suggest the importance of this complexity. Each metric has implications and is important in particular contexts, but can be misleading if not understood properly. As another example, consider the Organization for Economic Cooperation and Development (OECD) report that documents the “fastest advertised broadband speeds” per country in 2008 (see Figure 3-7). What is advertised is technically very different in different countries, rendering cross-country comparisons of even something as seemingly straightforward as peak advertised rates difficult. For example, for similar technologies or services, the rates that are advertised in Japan and the United States are different. For example, the Japanese advertise a maximum peak rate of 160 Mbps that is not achievable in practice.35 Legal challenges. The privacy implications of measuring individual subscriber behaviors must be carefully managed. There are obvious concerns that detailed information may enable socially undesirable forms of discrimination. Thus, there is a growing awareness that Internet traffic data needs to be managed so as to respect and protect subscriber privacy. Today, there are

Japan Finland France Korea Spain Germany Portugal Australia Luxembourg Norway Austria Denmark Canada New Zealand Switzerland The Netherlands Sweden Czech Republic Slovak Republic Belgium Hungary Ireland Poland United Kingdom United States Mexico 0

figure 3-7.

66

20

40

60

80

100

120

140

160

180

OECD data from September 2008 on the fastest advertised broadband speeds using cable (in Mbps).

s. bauer, d. clark, and w. lehr

no clear or universally accepted norms or rules for protecting user data on the Internet. This is an active and important area of ongoing research.36 Even in the absence of consensus on norms or rules,37 ISPs have brand images to protect from perceptions that they may have inadequately protected subscriber privacy or misused subscriber data (regardless of whether any such perceptions are well-founded).38 Thus, addressing concerns about protecting individual subscriber data are very important in generating support for the collection and sharing of appropriate public data. One of the benefits of pooling data from multiple ISPs is that it provides additional options for preserving provider and subscriber confidentiality, while enabling sufficiently rigorous and detailed sampling to obtain statistically accurate traffic characterizations. For academic researchers, many universities, including MIT, require that all affiliated personnel that are engaged in research involving human subjects submit their proposal to an institutional review board (IRB)39 that has the responsibility to confirm that the research is in compliance with federal regulations designed to protect human subjects from harm that may arise as a consequence of the proposed research. Risks to individual privacy are one of the potential harms that the IRB process is intended to address. While the original focus of these rules was on humans engaged in medical research, prompted by several well-known cases of abuse,40 the IRB process has now been extended to all research involving human subjects. This process provides an additional layer of protection to ensure adequate privacy protection. There are a variety of techniques including sampling design and anonymization that may be used to ensure that the data that is collected does not include any personally identifiable information (PII). Business challenges. As noted earlier, the efficiency of markets depends on the availability of adequate information to key stakeholders (buyers and sellers). There is a rich economics literature documenting the importance of private and asymmetric information, and its potential to effect the allocation of resources and profits.41 The fact that better traffic data may make the Internet ecosystem more competitive and efficient means that such data is inherently strategic. Better traffic data may allow an ISP to better plan its investments and target its service offerings to capture market share from other providers. We believe efforts to collect data would be most successful if the data is voluntarily supplied. While the data could be compelled, using strong regulation to collect or force disclosure of the data would introduce regulatory costs (e.g., direct overhead as well as distorting incentives) and rigidities (technology evolves faster than regulations). If successful mechanisms can be crafted to make sharing the data incentive compatible, we believe the result will be to get more granular, timely, and better data publicly available sooner and at a lower total cost to stakeholders.42 b r o a d b a n d m i c r o f o u n d at i o n s

67

There is also a free-rider problem that will need to be addressed. Even if one accepts the value of better public information on traffic, most folks would be happier if they can derive those benefits without having to pay the costs of making the data available. While this will pose a challenge, it is hardly a new challenge or one limited to the problem of Internet traffic data, and so there are a host of well-known approaches for addressing this challenge. We expect that industry associations, consortia, and standardization bodies may play useful roles in figuring out how to resolve these issues.

conclusion In the initial phase of broadband Internet access, the focus of policy-makers and many researchers has been on ensuring universal availability and adoption of broadband service. As broadband subscribership saturates, broadband infrastructure continues to evolve (e.g., toward much higher potential peak rates, toward mobile broadband, and so on), and the applications enabled by broadband become more widely relied upon (e.g., interactive rich multimedia applications), questions about how broadband is being used will be increasingly of interest. To properly address such questions, we will need much better insight into Internet traffic (its growth, statistical characterization, drivers, and so on) over both short (operational) and longer (investment) time frames. We have argued in this chapter that traffic data will be central to monitoring and resolving the inevitable tussles of this next stage of development. Given that such data is not publicly visible, the cooperation of network operators is essential. We are optimistic that the challenges of sharing traffic data can be addressed. The technical community (including academics, operators, vendors, and interested individuals) has a long history of collaborating through institutions such as the IETF,43 NANOG (and its equivalents in other regions),44 and other forums. A similar cooperative capacity can be developed which produces data about traffic on the Internet. While we have a clear understanding of why such traffic data is important now, we also recognize the importance of collecting data in anticipation of future use. In “Looking Over the Fence at Networks: A Neighbor’s View of Networking Research” it was noted that “good data outlives bad theory.”45 Data can be useful to later generations of researchers in ways not yet understood. The report noted the heavy dependence of the scientific community’s knowledge and understanding of climate change on a record of atmospheric carbon dioxide measurements that Charles David Keeling started collecting on Mauna Loa in 1957. An analogous historical data set of traffic data for the Internet might be similarly important for future networking research providing a baseline for evaluating the large-scale impact of both evolutionary and revolutionary changes in the Internet. 68

s. bauer, d. clark, and w. lehr

CHAPTER 4

Adoption Factors of Ubiquitous Broadband sangwon lee and justin s. brown Broadband networks are widely recognized as an indicator of the knowledge economy. Employing secondary data, this chapter examines adoption factors of “ubiquitous broadband” that includes both fixed and mobile technologies. Along with other industry, ICT (Information and Communication Technology) and demographic variables, the results of regression analysis suggest network competition between fixed and mobile broadband, platform competition in fixed broadband markets and multiple standardization policies in mobile markets are all significant factors in ubiquitous broadband deployment. However, some of these ubiquitous broadband deployment factors vary between developed and developing countries. Among other conclusions, the results of this study imply that for the initial fourth generation (4G) mobile markets, in which fixed and mobile broadband networks will converge, governments need to be open to diverse competitive standards instead of government-mandated standards.

background Both fixed and mobile broadband networks are increasingly recognized as indicators of the knowledge economy. Technological innovation and higher bandwidth enable entry into the era of ubiquitous broadband access, which means broadband access anytime, anywhere, by anyone and anything through both fixed and mobile broadband.1 After all, successful diffusion of fixed and mobile broadband is necessary for the provision of advanced IP-based services such as VoIP (Voice over Internet Protocol), IPTV (Internet Protocol Television) mobile television, and 4G applications. 69

Initially broadband was defined as communication technologies that provide high-speed, always-on connections to the Internet for large numbers of residential and small-business subscribers.2 However, this conception of broadband commonly emphasizes fixed broadband services such as DSL and cable modem service. Currently the International Telecommunication Union (ITU) defines broadband as a network offering a combined speed of equal to, or greater than, 256 Kbps in one or both directions, which may include more diverse broadband technologies such as mobile broadband and portable Internet.3 In the near future fixed and mobile broadband technologies will be converged in the Next Generation Networks (NGN) to offer 4G services.4 In spite of this broader definition of broadband, many previous empirical studies on broadband focused only on fixed-broadband technology and did not readily analyze mobile broadband. Also, despite existing research efforts to better understand broadband deployment, previous empirical studies tended to employ limited numbers of independent variables with insufficient numbers of observations. In addition, only a few empirical studies focused on broadband factors directly related to developing countries. Employing secondary data from the ITU, this chapter analyzes the determinants of global ubiquitous (total) broadband deployment. Specifically, this chapter examines whether network competition between fixed and mobile broadband has influenced broadband deployment. It also examines whether different policy types of platform competition in fixed broadband markets and standardization policy in mobile markets lead to different deployment levels of ubiquitous broadband. Finally, we examine whether there is any difference between developed and developing countries in terms of factors contributing to ubiquitous broadband adoption.

global broadband deployment According to Organization for Economic Cooperation and Development (OECD) broadband penetration data (December 2008), Denmark, Netherlands, Norway, and Switzerland were the leading fixed broadband economies among OECD countries.5 With respect to wireless, there exists a wide range of mobile broadband diffusion levels across countries. As of December 2007, Korea, Japan, and Italy were the top three mobile broadband economies among OECD countries in terms of mobile broadband penetration.6 In terms of total broadband penetration rates that include both fixed and mobile broadband subscribers, as of December 2007, Korea, Japan, Italy, Sweden, and Switzerland were the top five broadband economies among OECD countries.7 70

s. lee and j. s. brown

concept of ubiquitous broadband A primary purpose of our study was to examine adoption factors of ubiquitous (total) broadband. The concept of ubiquitous computing was first employed in 1991 by Marc Weiser. Weiser pointed to the “invisibility” of technology through the transformation of everyday items into small computers.8 In the first paradigm of computing, mainframes were shared by many people (one computer for many people).9 Now we are moving from the personal computer era (one computer per person) to proceed to the phase of the ubiquitous computing era (many computers per person).10 With the development of mobile technologies, this notion of ubiquitous computing can be applied to ubiquitous broadband access, which means an individual may use multiple forms of broadband access through a number of different devices. In the ubiquitous broadband environment, broadband access anytime, anywhere, by anyone and anything through both fixed and mobile technologies is possible (see Figure 4-1). For ubiquitous broadband, network competition between fixed and mobile networks may be a driver of broadband deployment. Further, for ubiquitous broadband, platform competition among different fixed broadband technologies and competition between different mobile standards as well as intra-modal competition are possible. These multiple modes of competition may suggest important policy and regulatory implications for nations wishing to expand their ubiquitous broadband infrastructure. In the near future, in the environment of the Next Generation Networks (NGNs) which will be achieved after fixed and mobile networks are converged by fully IP-based integrated systems, the notion of ubiquitous broadband could be replaced by converged ubiquitous (cubiqutous) broadband over a network (see Figure 4-1). In this NGN environment, single or multiple standardization policies are possible within a country. In this context, an examination of adoption factors of ubiquitous broadband is necessary. As technology continues to evolve, it is logical to extend the definition to ubiquitous broadband—the total network offering across various types of platforms—including the simultaneous access of fixed and mobile services.11

platform ⁄ network competition Platform competition might be a key driver of broadband adoption. Platform competition occurs when different technologies compete to provide telecommunication services to users.12 Platform competition in network industries involves competition between technologies that are not only differentiated, but also are competing networks.13 Strong platform competition among different technologies may lead to lower prices, increased feature adoption factors of ubiquitous broadband

71

One Broadband Access for One Person

Multiple Broadband Access for One Person with Multiple Devices

Ubiquitous Broadband (Fixed, Portable Internet, and Mobile Broadband Access) Network (Infrastructure) Competition

Fixed Telephone

Cellular Phone

Mobile Broadband

Data PSTN

Fixed Broadband

PC/ Narrowband

Wi Fi Fiber DSL Cable Modem

1G 2G 2.5G 3G W-CDMA HSDPA CDMA-2000 (Single or Multiple Standardization Policy)

Voice

Portable Internet

(Platform Competition)

Ubiquitous Computing

Video (Mobile TV)

Video (IP TV) (VoIP)

One Computer for One Person

Voice

Data

Converged Ubiquitous Broadband (Cubiquitous Broadband) 4G Mobile Technology Fixed Mobile Convergence (FMC) Next Generation Networks (NGNs) Packet-based IP Network (Single or Multiple Standardizaton Policy)

figure 4-1.

Concept of ubiquitous (total) broadband.

offerings, and more extensive broadband networks.14 Some empirical studies find that platform competition (intermodal competition) positively influences broadband deployment.15 Relating to ubiquitous broadband, which includes both fixed and mobile broadband, platform competition may be related to the competition between fixed and mobile networks. Competition between networks might lead to lower prices, improving the quality of service, increasing the number of customers, and promoting investment and innovation.16

research on broadband adoption factors Policy Factors: Previous studies suggest standardization policy plays a significant role in the success of wireless communication.17 Specifically, multiple 72

s. lee and j. s. brown

standardization policy might be positively related to rapid mobile broadband deployment. Though multiple standardization policy might lead to limited network externalities and economies of scale, multiple standards and different types of services across technologies enable the existence of diverse competing systems which may lead to better and differentiated services in the market.18 Based on previous studies, it is also expected that both platform (network) competition in fixed broadband markets and multiple standardization policy in mobile broadband markets might lead to higher levels of ubiquitous broadband diffusion. The institutional environment might also influence broadband deployment. For example, Andonova suggests political rights and civil liberties are correlated with Internet deployment.19 Despite this research, there is no empirical study that specifically tests the influences of institutional environmental factors such as political and economic freedom on ubiquitous broadband diffusion. Industry Factors: Previous empirical studies find that industry factors like price, speed, and competition influence broadband deployment. Cava-Ferreruela and Alabau-Mun˘oz suggest technological competition and low cost of deploying infrastructures might be drivers for broadband deployment.20 Employing regression analysis, Lee and Brown determined platform competition, broadband speed, and content contribute to global broadband adoption.21 They also contend the impact of platform competition is strong when market share of dominant technology and nondominant technology is similar.22 Recently some studies found low fixed broadband price is correlated with a higher level of broadband diffusion.23 Higher bandwidth also might be correlated with broadband adoption. Growth in demand for higher capacity is a key driver of broadband diffusion.24 Fransman suggests that bandwidth capacity of broadband is a measure of national performance in broadband.25 Growth in demand for higher capacity and telecommunication infrastructure investment from private and public sectors might lead to higher levels of broadband diffusion,26 but thus far empirical research has yet to analyze this contention. Demographic Economic Factors: Previous studies on new media technology adoption suggest higher levels of income, education, urban population share, and population density might be positively correlated with the higher levels of fixed broadband and mobile diffusion. Grosso found income measured by GDP per capita is related to broadband penetration among OECD countries.27 An empirical study suggests high levels of income contributed to high levels of mobile adoption.28 Ahn and Lee, in their study of the determinants of demand for mobile telephone networks, argued that the probability of subscribing to the mobile telephone adoption factors of ubiquitous broadband

73

networks was positively correlated with per capita GDP.29 Using a panel data set of fifty-six countries to investigate the economic factors influencing the growth of mobile phone services, Madden et al. concluded that higher income and a larger user base tend to promote mobile diffusion.30 In a separate empirical study about the same topic, Andonova also found GDP per capita to be a major contributing variable to mobile diffusion.31 Some previous empirical studies on fixed broadband deployment suggest that high levels of education are positively correlated with broadband penetration.32 Through a nationwide U.S. survey, Horrigan demonstrates younger age, higher education and income, and urban living share of population may lead to greater broadband adoption.33 Some empirical studies on fixed broadband contend a high level of population density is related to rapid fixed broadband deployment.34 Through data analysis of approximately one hundred countries, Garcia-Murillo revealed that population density has positive effects on the number of broadband subscribers.35 Trkman et al. found population density and education are influential demographic factors of fixed broadband deployment in EU countries.36 Most of the studies on the telecommunications sector suggest a younger population is correlated with greater levels of new media technology adoption. Some previous studies found age is negatively correlated with broadband adoption in OECD countries.37 Specifically, de Ridder determined the 35–39 and 40–44 year-old groups were positively correlated with fixed broadband adoption.38 Based on this previous OECD study, it is interesting to examine whether age groups such as 35–44 year-olds are correlated with ubiquitous broadband diffusion. Information and Communication Technology (ICT) Factors: Recent studies on fixed broadband deployment demonstrate ICT factors such as PC infrastructure and teledensity have influenced broadband penetration.39 Denni and Gruber found that teledensity has been an influential factor in broadband deployment in the United States.40 More recently, through a factor analysis, Trkman et al. found that communication technology expenditures, household PC access rate, Internet penetration, and fixed phone penetration are factors of fixed broadband deployment in EU countries.41

research questions Even though there is a growing body of comparative scholarship concerning broadband deployment, many existing empirical studies use a limited number of independent variables with an insufficient number of observations. In addition, previous studies focus on fixed broadband and exclude mobile technology, a key to understanding the concept of ubiquitous (total) broadband. 74

s. lee and j. s. brown

Furthermore, only a few empirical studies contain broadband policy implications for developing countries. Prior research also neglects some important independent variables such as network competition between fixed and mobile networks, institutional environment, bandwidth, and telecommunication infrastructure investment. By employing secondary data from the ITU, this empirical study examines determinants of global ubiquitous broadband deployment. Based on the above-cited studies, we propose the following research questions: • Has network competition between fixed and mobile broadband networks influenced the deployment of ubiquitous (total) broadband? • Have policy factors—different types of platform competition in fixed broadband markets and standardization policies in mobile markets and institutional environment (political and economic freedom)— affected the deployment of ubiquitous (total) broadband? • Have industry factors—specifically fixed-broadband price, broadband speed, bandwidth, telecommunication infrastructure investment, and mobile service price—influenced the deployment of ubiquitous (total) broadband? • Have demographic/economic factors—specifically income, education, urban population share, population density, and age— influenced the deployment of ubiquitous (total) broadband? • Have ICT factors—specifically PC penetration, content, Internet usage, and teledensity—influenced the deployment of ubiquitous (total) broadband? • Is there any difference between developed and developing countries in adoption factors of ubiquitous (total) broadband?

t he empirical model To examine determinants of total broadband deployment, we employ a log linear regression model. The linear regression model employs approximately 216 observations of broadband services from ITU member countries. To examine the influences of quantifiable variables on total broadband deployment, we formulate the following log linear regression model. Since the distribution of many variables in this linear regression model is positively skewed, data transformation with logarithm was utilized. Ln Yt (BPR) = β0 + β1(ln Network Competition (or Standardization Policy Type) + adoption factors of ubiquitous broadband

75

β2(ln Speed) + β3(ln Political Freedom) + β4(ln Economic Freedom) + β5(ln Fixed Broadband Price) + β6(ln Mobile price) + β7(ln Bandwidth) + β8(ln Telecommunication Investment) + β9(ln Income) + β10(ln Education) + β11(ln Population Density) + β12(ln Urban Population) + β13(ln Age Group) + β14(ln PC Penetration) + β15(ln Content) + β16(ln Internet Usage) + β17(ln Teledensity) + εt This empirical model for multivariate analysis was a composite model from previous empirical studies about fixed and mobile broadband deployment. In the empirical model, the dependent variable (Yt) is broadband deployment that accounts for both fixed and mobile services. Independent variables are policy factors such as standardization policy type, political freedom, and economic freedom. Three different standardization policy types are examined in the model—policy type I: platform competition for fixed broadband without mobile broadband services; policy type II: platform competition with single standard for mobile broadband services; and policy type III: platform competition with multiple standards for mobile broadband services. The model also incorporates industry factors such as speed, network competition, telecommunication infrastructure investment, fixed broadband price, bandwidth, and mobile service price. In addition, demographic factors such as income, education, population density, urban population, and age, as well as ICT factors such as PC penetration, Internet usage, content, and teledensity are contained in the model. Since some of the standardization policy types might be conceptually and empirically correlated, only one variable (network competition or policy types for standardization policy) was included in an empirical model.

data and measurement Table 4-1 shows the variables, their measures, and the corresponding data sources used to analyze ubiquitous broadband adoption. Broadband services can be deployed through fixed and mobile networks. Therefore we examine adoption factors of ubiquitous (total) broadband deployment (fixed plus mobile). Ubiquitous (total) broadband is obtained by summing the total number of fixed and mobile broadband connections and is measured by the total number of broadband lines per one hundred inhabitants.42 Policy factors. Three different types of categories were used to measure standardization/platform competition policy: platform competition only in fixed broadband markets and no mobile broadband services (policy type I); 76

s. lee and j. s. brown

table 4-1

.

variables, measurement, and data sources for total broadband deployment

Variables

Measurement

Data Sources

Total Broadband Deployment

Total number of broadband lines per 100 inhabitants

ITU (2004–2005)

Policy I

Dummy (1 for with platform ITU (2004–2005) competition in fixed broadband markets and no standardization in mobile broadband markets, 0 for otherwise)

Policy II

Dummy (1 for with platform competition ITU (2004–2005) in fixed broadband markets and single standardization policy in mobile broadband markets, 0 for otherwise)

Policy III

Dummy (1 for with platform competition ITU (2004–2005) in fixed broadband markets and multiple standardization policy in mobile broadband markets, 0 for otherwise)

LLU Economic Freedom

Dummy (1 for with LLU, 0 for no LLU) Index of economic freedom

OECD (2004–2005) Heritage Foundation (2004–2005)

Political Freedom Network Competition

Inverse of the score on civil liberties Dummy (1 for DSL and cable modem for fixed broadband and multiple standards for mobile broadband are available, 0 for otherwise) Broadband monthly charge (USD)

Freedom House (2004–2005) ITU (2004–2005)

Price of Fixed Broadband Service Mobile Service Price Per minute local call (USD) peak charge

ITU (2004–2005)

Income

GDP per capita

ITU (2004–2005)

PC Infrastructure

Estimated PCs per 100 inhabitants

ITU (2004–2005)

Education

UNDP Education Index

UNDP (2004–2005) 2

ITU (2004–2005)

Population Density

Population density (per km )

ITU (2004–2005)

Internet Usage

Internet user per 100 inhabitants

Urban Population

Percentage of urban population

ITU (2004–2005) Euromonitor (2004–2005)

Telecommunication Infrastructure Investment Teledensity

Annual telecommunication investment

Age Content Bandwidth Speed

Main telephone lines per 100 inhabitants Percentage of age between 35–44 Internet hosts per 10000 inhabitants International Internet bandwidth (bits per inhabitant) Broadband speed (Kbps)

ITU (2004–2005) ITU (2004–2005) ITU (2004–2005) World Bnak (2004–2005) ITU (2004–2005) ITU (2004–2005) ITU (2004–2005)

platform competition in fixed broadband markets and single standardization policy in mobile broadband markets (policy type II); and platform competition in fixed broadband markets and multiple standardization policy in mobile broadband markets (policy type III). Political freedom is measured by the inverse of the score on civil liberties.43 For the measurement of economic freedom, the index of economic freedom is employed. Industry factors. Competition between fixed and mobile networks might be an influential factor of ubiquitous broadband deployment. Ubiquitous broadband access means broadband access that is unconstrained by time, place, and location through either fixed or mobile broadband.44 In the empirical model, network competition is measured by a dummy variable (1 for DSL and cable modem is available for fixed broadband and multiple standards for mobile broadband, 0 for otherwise). We adopted the per-minute local call (USD) peak charge to measure the cost of mobile services in each country. Regarding the factor of nonvoice mobile applications, the cost of short message services (SMS) is used as the price proxy for mobile broadband relevant applications. Fixed broadband price is measured by broadband monthly charge (in USD). Telecommunication infrastructure investment is measured by annual telecommunication investment. For the measurement of bandwidth, international Internet bandwidth (bits per inhabitant) is utilized. Demographic/Economic factors. In terms of demographic variables, level of education could be measured by illiteracy rate and average education/degree level. For the measurement of education, we employed the UNDP education index. A share of urban population is used to measure the demographic aspect of population density.45 We measure population density by population per km2. Age is measured by percentage of population between thirty-five to fortyfour years old. For the measurement of income, the GDP per capita is used. ICT factors. To measure the PC infrastructure, estimated PCs per 100 inhabitants are used. Teledensity is measured by main telephone lines per 100 inhabitants. For the proxy measurement of content, Internet hosts per 10,000 inhabitants is employed. Internet usage is measured by Internet users per 100 inhabitants. Most of the secondary data has been collected from the ITU, OECD, World Bank, Heritage Foundation, and Freedom House. We employ the statistical analyses of linear regression analysis to assess the influential factors of ubiquitous broadband deployment.

results of regression analysis of total (ubiquitous) broadband deployment A total of 216 observations were analyzed employing multiple regression analysis. For ubiquitous broadband deployment, we examined both the model 78

s. lee and j. s. brown

with network competition variable and the model with different platform completion-standardization policy type variable.46 Extended and reduced models were identified from the data analysis for both models. Note that the dependent variable and independent variables were transformed using logarithmic functions since data were positively skewed. The results of both models show similar results. Model with network competition variable. In the initial model, seventeen independent variables were included in the multiple regression analysis. Because a multicollinearity problem might occur when independent variables are highly correlated, a correlation analysis was conducted to check potential multicollinearity problems. To assess the strength of correlations, the criterion of .80 Pearson correlations was employed. PC penetration, teledensity, Internet use, and bandwidth were removed from the initial model because of high correlation with other independent variables. The extended regression model was significant at the .01 level. Specifically, network competition, income, and content were statistically significant at the .01 level. Fixed broadband price, speed, and political freedom were statistically significant at the .05 level. Other variables such as mobile price, education, population density, urban population, age (35–44), telecommunication investment, and economic freedom were not statistically significant. R-squared for the extended model was .81. In the reduced model, nonsignificant variables such as mobile price, education, age (35–44), telecommunication investment, and economic freedom were removed. The reduced model was significant at the .01 level. In the reduced model, income and political freedom were statistically significant at the .01 level. Fixed broadband price, speed, content, urban population, and network competition were statistically significant at the .05 level. Also, population density was statistically significant at the .1 level. R-squared for the reduced model was .79. Table 4-3 provides the results of the regression analysis. Model with platform competition-standardization policy variable. The main interest of this model is to analyze the effects of different platform competitionstandardization policy types. We categorized three different types: platform competition only in fixed broadband markets and no mobile broadband services (policy type I); platform competition in fixed broadband markets and single standardization policy in mobile broadband markets (policy type II); and platform competition in fixed broadband markets and multiple standardization policy in mobile broadband markets (policy type III). For the extended model, initially, all nineteen independent variables were included in the multiple regression analysis. A correlation analysis was conducted to check potential multicollinearity issues. Based on the benchmark of .80 Pearson correlations, PC penetration, teledensity, Internet use, and bandwidth were removed from the initial model because of high correlation with adoption factors of ubiquitous broadband

79

other independent variables. Also, for the extended model, other insignificant independent variables such as mobile price, education, age (35–44), and economic freedom were removed from the initial model. The extended model was significant at the .01 level. In particular, policy type III (platform competition in fixed broadband markets and multiple standardization policy in mobile broadband markets) and income were statistically significant at the .01 level. Policy type II (platform competition in fixed broadband markets and single standardization policy in mobile broadband markets), speed, political freedom, and population density were statistically significant at the .05 level. Policy type I (platform competition only in fixed broadband markets and no mobile broadband services) was significant at the .1 level. Other independent variables such as fixed broadband price, content, telecommunication investment, and urban population were not statistically significant. R-squared for the extended model was .80. To check the stability of results in the empirical study, the nonsignificant independent variable telecommunication investment was removed from the reduced model. The reduced model is significant at the .01 level. In the reduced model, policy type III, income, and political freedom were statistically significant at the .01 level. Policy type II, speed, and population density were statistically significant at the .05 level. Policy type I, fixed broadband price, and urban population were significant at the .1 level. Content was not statistically significant. R-squared for the reduced model was .80. A total of 190 observations were available for this model. Table 4-2 provides the results of the reduced models used in the regression analysis.

results of ubiquitous broadband deployment for developed and developing countries A total of 73 observations were analyzed employing multiple regression analysis for developed countries and a total of 120 observations were analyzed for developing countries. R-squared for the final reduced model for developed countries was .77, and R-squared for the final reduced model for developing countries was .56. To examine effects of different platform competitionstandardization policy types, policy types I, II, and III were included in the model. Dependent variables and independent variables were transformed using a logarithmic function since data were positively skewed. Regression analysis, developed countries. In the initial model, all nineteen independent variables were included in the multiple regression analysis. PC penetration was removed from the initial model because of its high correlation with other independent variables. Also, for the extended model, some insignificant variables like teledensity and age were removed from the model. 80

s. lee and j. s. brown

table 4-2

.

results of regressions of total broadband deployment

Variable

Model 1 (with Network Competition Variable)

Model 2 (with Policy I, II, III)

Extended Model

Extended Model

Reduced Model

Reduced Model

Coefficients B

t-stat

Coefficients B

t-stat

Coefficients B

t-stat

Coefficients B

t-stat

Constant

−4.74

−3.50***

−4.11

−7.74***

−4.61

−7.64***

−4.41

−8.29***

Speed

.23

2.16**

.25

2.55**

.22

2.03**

.23

2.36**

Fixed Broadband Price

−.19

−2.11**

−.17

−2.11**

−.12

−1.45

−.13

−1.69*

Income Mobile Price

.67 −.04

5.65*** −.491

.71 −

7.48*** −

.76 −

7.42*** −

.79 −

8.32*** −

Education

−.89

−1.16

−

−

−

−

−

−

Content

.20

3.05***

.12

2.34**

.09

1.62

.07

.17

Political Freedom

−.37

.18**

−.50

3.27***

−.40

−2.42**

−.50

−3.35***

Age (35–44)

.32

.43

−

−

−

−

−

−

Telecom Investment

.02

.03

−

−

.03

1.08

−

−

.06

.08

−

−

−

−

−

− (Continued)

table 4-2

.

(continued)

Variable

Model 1 (with Network Competition Variable)

Model 2 (with Policy I, II, III)

Extended Model

Extended Model

Reduced Model

Reduced Model

Coefficients B

t-stat

Coefficients B

t-stat

Coefficients B

t-stat

Coefficients B

t-stat

Economic Freedom

.08

1.54

.09

1.96*

.11

Population Density

.40

1.50

.49

2.05**

.41

2.24**

12 .

2.48**

1.64

.45

1.96*

Urban Population

−

−

−

−

−

−

−

− − − − −

−

− − − − −

Teledensity

−

−

−

−

−

PC Penetration

−

−

−

−

−

Internet Use

−

−

−

−

−

Bandwidth

.23

2.67***

.18

2.32**

Network Competition

−

−

−

Policy I***

−

−

−

−

.18

1.86*

.16

1.84*

−

.25

2.46**

.21

2.28*

Policy II*** Policy III***

−

−

−

−

.56

4.33**

.53

4.33**

R-Squared

0.81

0.79

0.80

0.80

Number of Observations

169

191

177

191

− − −

* Statistically significant at the 10% level. ** Statistically significant at the 5% level. *** Statistically significant at the 1% level. **** Policy I: Cable modem-DSL platform competition + no standardization policy for mobile broadband/Policy II: Cable modem-DSL platform competition + single standardization policy for mobile broadband/ Policy III: Cable modem-DSL platform competition + multiple standardization policy for mobile broadband

In the reduced model, other insignificant variables such as fixed broadband price, income, mobile price, education, content, telecommunication investment, urban population, and economic freedom were also removed from the model. Table 4-3 provides the ANOVA table of the reduced regression model, which illustrates the model’s significance at the .01 level (P 200 kbps) is glossing over a new and rapidly developing issue.31 Even if there is relatively wide availability of low-grade broadband, there may be substantially greater unevenness in access to high-quality, high data rate services that could come to define a new digital divide. This may be even truer for advanced broadband services that will define new levels of functionality in the near future. The FCC data seem to indicate that, on the one hand, availability of some (at least low) level of broadband services seems to be a rapidly diminishing the determinants of disconnectedness

187

issue for most of the US population. On the other hand, the number of providers, the degree of competition (and presumably, impacts on pricing), the quality of available service, and not the mere presence or absence of any availability at all, are likely to be the real up-and-coming issues in national policy on broadband Internet services.

188

kenneth flamm

CHAPTER 11

European Broadband Spending Implications of Input-Output Analysis and Opportunity Costs ibrahim kholilul rohman and erik bohlin Currently, many advanced countries have proposed broadband as part of the government’s responsibility to combat the digital divide. The idea is that government should increase the accessibility of broadband to people who have not yet been able to access it, especially in rural areas. A prime example has been the so-called Obama Package, a $787 billion stimulus bill signed into law on February 17, 2009, as the American Recovery and Reinvestment Act of 2009. This act included $7.2 billion to support a variety of broadbandrelated programs, including $4.7 billion for the build-out of broadband in unserved and underserved areas. Moreover, in Europe there have been initiatives to fund broadband as part of the European Economic Recovery Package. Decisions have been taken toward this end by the European Council, including the goal of 100 percent broadband coverage in Europe within the period 2010–2013 by spending 1 billion euro to stimulate broadband in rural areas in the European Recovery Plan. While these programs have received considerable public attention, several issues come to the fore. The immediate one is whether government is spending money wisely. More precisely, a recommendation for using government funds should consider the opportunity cost, that is, whether other alternatives for spending the available money will create greater or less benefit to society. This chapter will present a methodology to evaluate the opportunity cost of using government funds for broadband, with particular attention to the European proposal to publicly fund broadband infrastructures, and ask 189

whether such investments would be justified as improving welfare. More generally, the methodology will address whether any investment, private or public, improves welfare. In particular, this study will consider the opportunity costs of broadband policy using the input-output (IO) table methodology, especially the output multipliers derived from the IO table as a basis for the opportunity cost assessment. Applying the IO methodology in this way provides a counterfactual, as-if analysis to the question of investment impacts due to broadband. Therefore, this study is conducted as an exploratory as-if study based on the IO literature without necessarily striving to provide implications for decision makers. Despite IO having strong theoretical background, the IO method has to be treated carefully due to its limitations and simplifications. The chapter is structured as follows. First, a literature review of previous studies on the relationship between broadband and Information and Communication Technology (ICT) as well as the discussion on IO method will be presented. It will be followed by the opportunity cost concept from a macroeconomic point of view as a tool providing a common point of departure on this concept even though this study will not discuss the micro approach in a more detailed way. The next section will present the macro perspective of the opportunity cost using the IO methodology. Following this, an application of the method to selected European countries is conducted. Conclusions and implications for research follow.

literature review This chapter investigates the impact of broadband represented by the ICT sectors, where the definition is derived from Organisation for Economic Cooperation and Development (OECD)1. The analysis makes the assumption that there is a close relationship between the broadband and ICT sectors. This link can be seen from many previous studies which refer to the Internet—where broadband itself serves always-on and faster Internet connections—to exemplify the terminology of ICT for instance in Hollenstein,2 Bozinis3, Franklin et al.,4 Martin,5 Cortés and Navarro,6 and Haller and Siedschlag.7 As other ICT products, broadband is also embedded with the general purpose technology (GPT). The pervasive use of broadband technology in a wide range of sectors, therefore, will generalize productivity gains transferred to the rest of economy. In this view, Majumdar8 identifies a greater impact of broadband to wages as the sector is associated with the need for a higher skill of labor. In a later study, Majumdar et al.9 find a positive relationship between broadband deployment and the carriers’ productivity and suggest to encourage the deployment of broadband technologies. Other studies also 190

i. k. rohman and e. bohlin

support the importance of broadband technology increasing employment and productivity level, for instance in Van Gaasbeck10 and Grimes et al.11 Despite the tight link with ICT, broadband is very distinctive in comparison to other ICT devices since it brought a much wider impact on the economy. Previous studies predominantly identified the impact of ICT on the services sector as shown in Baily and Lawrence,12 Armistead and Meakins,13 and Antonelli,14 denoted by a growing and greater ratio of services to the world’s gross domestic product (GDP). But, recent studies show the impact of broadband moves beyond the service sector, for instance in agriculture (in the study by Ha et al.15 and Kandilov and Renkow16). Moreover, there is also a growing interest to investigate the impact of broadband on small and medium-sized enterprises (SMEs) and rural and local economic development (Lindskog and Johansson,17 Arbore and Ordanini,18 Howick and Whalley,19 and Glass and Stefanova20). Not only does broadband directly affect economic and productivity as GPT, it also contributes a greater impact on socio-economic development, for instance, health and education (Savage et al.,21 Sivakumar and Robertson,22 Lindsay et al.,23 Collins and Wellman,24 Hale et al.,25 and Ward and Prosser26). With regards to supporting public administration, recent studies found that broadband is a major catalyst for the implementation of e-government, for instance, in Italy (Arduini et al.27), and Estonia (Kalja et al.28). In terms of methodology, the IO table has been largely used to estimate economic impact both solely as the measurement of traditional cost-benefit analysis or in relation to the opportunity cost concepts. The calculation of economic multipliers using an IO table is one of the alternative methods to measure cost-benefit analysis, especially in dealing with public policy choices. Fuguitt and Wilcox also argue that by implementing an IO table, one can derive induced income changes by linkages and for consumption expenditure, thus the contribution of a particular sector can be identified.29 Therefore, the evaluation of public policy choices will consider not only the direct effect of a project or program but also the indirect effect induced by interconnection among economic sectors. As an example, the use of an input-output table as the framework to analyze the opportunity cost concept can be found in Hughes et al. In this study they investigated the net effect of a farmers’ market in West Virginia making alternative spending using the same amount of money as the second alternative. The opportunity cost impact in this study is concerned with the question of what would have occurred had the consumer spending at the farmers’ market was compared to the spending at West Virginia grocery stores. This scenario assumes that expenditure made at the farmer markets is the same as what would have been spent at grocery stores. european broadband spending

191

The gross impact on the economy is measured by calculating the job and income creation from the farmers’ market activity, especially when local and regional economies benefit from an enhanced retention of local dollars. The impact is then compared with the alternatives of spending of money at the grocery level. The study reveals that the gross economic impacts of the farmers’ market sales are to a small degree offset by losses in agriculture from reduced grocery store sales of similar products produced in West Virginia, confirming the positive opportunity cost of farmers’ market development in the near future.30 Another study related to the opportunity concept can be found in Cho and Hooker’s study, in which they build a model based on microeconomics analysis to measure the effects of a food safety program. The study calculates an opportunity cost, which is defined as the shadow value of productive resources used to enhance food safety that could alternatively be used to increase revenue through the sale of a larger volume of output. In other words, the study tries to estimate the effect of regulation on firm behavior. That is, loss in efficiency due to regulation that reduces a firm’s choice of behavior which, at the same time, has a potential impact on economic revenue. To operationalize such a method, Cho and Hooker formulated the output directional distance function which was replaced by the use of IO data.31 Hence, when the IO method is used to evaluate the opportunity cost concept, it will be dealing with alternative outcomes, which can be produced by using the same amount of resources as the alternative choice. Nevertheless, as presented earlier, the use of the IO method to measure economic impact needs to be treated carefully. For instance, there is a choice between closed and open IO tables. A previous study by Grady and Muller suggested that the use of a closed IO table usually creates a misleading result since it yields exaggerated estimates of the impact of program expenditures on the economy.32 This is the case because closed IO models do not take into account the macroeconomic feedback that tends to cause the multiplier to decrease over time. In the end, the use of open IO tables that put all final demand categories, including consumption, as exogenous variables is recommended. In addition, the study by Elder and Butcher shows that when the IO is used to measure the benefit of the project or program, it is also essential to look at the cost induced by the program at the same time.33 Hence it will carry out the net effect of a particular project giving the balanced calculation from the cost and benefit side from both direct and indirect effects. Related to studies on broadband, the ubiquitous adoption of broadband and the current generation of technologies will generate $63.6 billion of capital expenditures in the US economy, according to Crandall et al. This will result in a cumulative increase of the GDP of $179.7 billion and sustain 61,000 jobs per year. Moreover, the identification of consumer benefits from 192

i. k. rohman and e. bohlin

universal broadband deployment can be separated into various sources including shopping, entertainment, telecommuting, telephone services, and telemedicine. Thus, the cumulative $63.6 billion residential broadband capital expenditure will result in $179.7 billion between 2003 and 2021, or $9.5 billion annual impact.34 An empirical study conducted by Kelly on the impact of fiber deployment, found that a city in Iowa that invested in a fiber network at an earlier stage was able to attract 140 companies with 4,250 incremental jobs, while a city that invested later attracted only 9 companies.35 Ford and Koutsky found that the impact of broadband on economic growth is reaching 28 percent.36 Lehr et al. found that technology, based on data since 1999, had an impact on employment growth of 1.5 percent.37 In addition, these studies frame the economic impact of broadband as the benefits listed in Table 11-1. As depicted earlier, the strength of the IO method is its ability to capture both the direct and indirect impact of a certain project.38 Katz explains that the first impact relates to the jobs created from the deployment of the infrastructure (e.g., construction), while the second impact is generated by the network externalities from other sectors of the economy.39

table 11-1

.

framework for measuring economic impact of broadband

Impact area

Benefit

Productivity

Labor productivity in ICT intensive and nonintensive sectors Productivity in supply chain and distribution functions

Firm relocation

Relocation of firm for labor pool search Relocation of functions resulting from value chain decomposition Enhancement of the quality of life, which attracts educated labor force

Employment

Enhancement of self-labor force enabled by telecommunications infrastructure Enhancement of the radius of tele-commuting allowing for tapping into additional labor pools Creation of new firms/services requiring additional labor force

Economic growth

Strengthening of industries by lowering transaction costs (trade, finance, etc.) Consumer surplus derived from telecommunications services, saving of transportation time, etc.

source: Crandall, Jackson and Singer (2003).

european broadband spending

193

The study by Katz also synthesizes previous analyses of the calculation of the multiplier effect in the broadband sector, with studies conducted at the regional level (e.g., Strategic Network Group, 2003) and country level (Crandall et al., 2003; Katz et al., 2008; Atkinson et al., 2009; Katz et al., 2009.a; Katz et al., 2009.b and Liebenau, 2009).40 An overview is presented in Table 11-2. Regarding Table 11-2, the Type I effect measures the impact as the ratio of the total of direct and indirect, divided by direct (direct+indirect/direct), while the Type II effect estimates the impact as the ratio of the sum of direct, indirect and induced, divided by direct (direct+indirect+induced/direct).41 It has been found in previous studies that the multiplier of broadband varies between 1.43 to 3.60 depending upon the region investigated and the type of multiplier, confirming the study by Grady and Muller, which suggested that the Type II multiplier exhibits a greater value than the Type I multiplier.42 Katz and Suter assert that the expenditures for providing broadband service in the United States create employment as a result of network externalities. The study estimates that, due to the employment multiplier, providing of broadband services created 32,000 jobs per year for the four years of the project.43 In addition, the Strategic Network Group estimated that the impact of the investment in fiber optic networks in a small city in Ontario can be investigated through the effect of new job creation, expansion of commercial facilities, increased revenue and decreased cost. That study calculates that between June 2001 and April 2003, there was a $2.8 million industry expansion and $140,000 in increased revenue and decreased costs.44 In addition there is a correlation between the use of broadband technology and job growth: 50 percent of businesses using broadband experienced job growth; 27 percent of businesses with dial-up Internet access experienced job growth; 5.6 percent of businesses with no Internet access experienced job growth.

table 11-2

194

.

the multiplier effect of broadband

Author

Location

Type I

Type II

Crandall et al. (2003)

United States

NA

2.17

Strategic network group

A Canadian County

2.03

3.42

Katz et al. (2008)

Switzerland

1.4

NA

Atkinson et al. (2009)

United States

NA

3.6

Katz et al. (2009.a)

United States

1.83

3.43

Libenau et al. (2009)

United Kingdom

NA

2.76

Katz et al. (2009.b)

Germany

1.45

1.93

i. k. rohman and e. bohlin

The above results, which indicate the gradation of the level of Internet technology, were discovered by interviewing business entities after the project’s implementation in the United Kingdom. Liebenau et al. provide a more detailed study in the case of the UK economy. Their study estimates that the impact will be significant in contributing immediate direct and indirect job growth in the UK economy, creating a network effect throughout the economy which will support additional jobs, providing a foundation for longer-term benefits including government cost saving, economy-wide productivity, and improved quality of life. Liebenau et al. summarize the impact in the United Kingdom as creating around 280,500 new jobs following a GBP 5 billion investment in broadband deployment.45 In addition to the calculation of the economic impact of broadband, broadband policy focusing on spreading out the technology needs to be studied. According to the OECD, the information society is a phenomenon that results from the role of ICTs in every aspect of life, especially its economic and social implications. In economic respects, businesses are transforming their supply and demand chains, as well as their internal organization, to fully exploit ICTs. In addition, ICTs have greatly contributed to the process of creative destruction through the birth of new firms with resulting implications for productivity improvement and economic growth. Therefore, ICTs are called general purpose technologies that can be used for a broad range of everyday activities. Thus, it should be taken into account as an alternative view, that amid possible lower multiplier effects, broadband policy should be implemented because broadband and the ICT sectors play important roles as GPTs for the rest of economy.46 In summary, in comparison to previous studies, this study will seek to adopt previous analyses in using IO tables as the basis of calculation, especially comparing countries within the European region. The use of an open IO table is implemented to avoid overestimation of the multiplier effect. In addition, the opportunity cost concept is introduced by taking other alternatives of spending broadband funds on non-ICT sectors by assuming that all sectors have similar probability ratios of success. Furthermore, this study only investigates the benefit side of the program, assuming that the cost structure of the ICT and non-ICT sectors investigated do not change before and after the implementation of the project. Hence, the cost side is assumed to remain the same.

t he io model The flow of transaction on the IO table, which is used as the main method in this study, can be explained in the following system equation (Equation 1). Suppose we have four sectors in the economy. european broadband spending

195

x11 + x12 + x13 + x14 + c1 = x1 x21 + x22 + x23 + x24 + c2 = x2 x31 + x32 + x33 + x34 + c3 = x3 x41 + x42 + x43 + x44 + c4 = x4 From Equation 1, xij denotes the output from sector i which is used by sector j as an intermediate input (or in the other words, it reflects the input from sector i which is used for further production process in sector j ). On the IO quadrant, these values are located in quadrant 1. Moreover, ci refers to total final demand of sector i whereas xi refers to total output of sector i. ci is put in quadrant 2. Introducing the matrix notation, equation 1 can be modified to obtain the following matrix column: ⎛ x1 ⎞ ⎛ c1 ⎞ ⎜ ⎟ ⎜ ⎟ x = ⎜  ⎟;c = ⎜  ⎟ ⎜x ⎟ ⎜c ⎟ ⎝ 4⎠ ⎝ 4⎠ Equation 2 consists of two matrixes that show the other representation of total output and total final demand. Thus, from equation 2, x denotes the column matrix of output and c is the column matrix of the final demand. The following matrixes I and A are the identity matrix and technology matrix respectively.

I

⎛1 ⎜0 ⎜ ⎜0 ⎜ ⎝0

0 1 0 0

0 0 1 0

0⎞ ⎡ a11 … a14 ⎤ 0 ⎟⎟ ; A = ⎢⎢    ⎥⎥ 0⎟ ⎢⎣ a41 a44 ⎥⎦ ⎟ 1⎠

The left-hand side of equation 3 is the identity matrix; a diagonal matrix whose off-diagonals are zero. Furthermore, A is the technology matrix which consists of the ratio of the intermediate demand to the total output, xij/x. Hence, a14, for instance, explains the ratio of output from sector 1, which is further used by sector 4 to produce their output, divided by total output from sector 1. Combining equations 1, 2, and 3, the equilibrium of the equation for demand and supply in equation 1 can be modified as follows: Ax + c = x (I – A) = c x = (I – A)–1c 196

i. k. rohman and e. bohlin

The first row of this equation is the general form of equation 1. Then, from equation 4, the multiplier is defined as the inverse Leontief matrix, (I − A)−1. The multiplier measures the ratio of output changes in the equilibrium as the result of the change in the final demand. Therefore, the output multiplier measures total change throughout the economy from a unit change in final demand. The change in the final demand might arise from private consumption, government expenditure, investment, and export. As the consequence of production linkages, a change of output will be larger than a change in the final demand. For instance, if the final demand of the ICT sector (e.g., additional purchasing of personal computers) increases by ten unit values of money, the output in the economy will grow by more than ten unit values of money or as much as the multiplier coefficient of this sector. (Additional purchasing of personal computers leads to increased packaging services and transportation services, for instance). This study employs a simple multiplier that measures the impact both from direct and indirect impacts, thus the matrix (I − A)−1 consists of all sectors within the economy putting household exogenous.47 Hence, the dimension of the matrix is 59 x 59 in the case of the European IO table. In addition, the IO table used in this method is open IO. Thus, the macroeconomic impact of broadband can be evaluated as shown in Figure 11-1. From Figure 11-1, it is assumed that the budget that is spent for broadband promotion is carried out from a specific pool of funds that actually can be spent for other purposes. Thus, there is a choice for the government in using these funds for public broadband expenditures. The broadband project will be indirectly benefiting the ICT sector, which is identified as the sector utilizing broadband relatively more than other sectors. This explanation describes the “benefit of the broadband package.” On the other hand, the

Benefit Multiplier effect of the sector that benefits from the implementation of Broadband Promotion (ICT sector)

figure 11-1.

Cost Multiplier effect if the budget is spent for any purposes other than broadband (Non-ICT sector)

Cost and benefit framework. european broadband spending

197

(opportunity) cost is the alternative cost of spending. Thus, any alternative sector to invest the money in is considered as the next best alternative. However, it should be noted that there is a weakness in the method: considering current economic development, almost all economic sectors are affected by broadband either directly or indirectly. Therefore, from the calculation of the IO method, benefits are underestimated and costs are overestimated. Moreover, this study employs the IO table deflated based on the GDP deflator for each country to enable the comparison between years of observation. The more thorough estimation on deflating the IO table is explained, for instance, by Celasun in the case of the Turkish structural change of the economy.48 The same method using the sectoral producer price index and import price index can be seen in Zakariah and Ahmad on the Malaysian economy.49 This study only uses the GDP deflator to obtain the constant value of the IO table. A similar method can be found in Akita (1991).50

applying the io model to broadband in the eu Moving concretely into an application of the IO approach to assess impacts of broadband investment on a selected sample of countries to represent the European economy,51 the following steps will be taken: 1. Defining the ICT sector. 2. Matching the ICT sector with the 59-sector input-output table. 3. Descriptive analysis of ICT sectors based on the IO table. 4. Multiplier analysis. 5. Evaluation. These five steps are explained below. Step 1: Defining the ICT sector. The fact that the ICT sector has contributed to the performance of the economy and other sectors has been proven by many empirical studies. Therefore, the need for statistics and analysis to support and inform policy making has grown alongside the rapid emergence of new ways of communicating, processing, and storing information in a common statistical standard which can be used uniformly to evaluate the information society within the OECD countries.52 To give the common understanding of what the ICT sectors are and how the ICT sectors have contributed to the information society, OECD provides details to guide in the measurement of the information society which defines and limits the classification of the ICT sectors. Focusing on the definition, there are two categories which are attributed to the ICT sectors: ICT 198

i. k. rohman and e. bohlin

products and media and content products. The basic principle of identifying the ICT products is adapted from the OECD definition of the ICT sectors: “ICT products must primarily be intended to fulfill or enable the function of information processing and communication by electronic means, including transmission and display.”53 In addition to that, there is also an OECD definition of the content and media products which is used to determine the content and media sectors: “Content corresponds to an organized message intended for human beings published in mass communication media and related media activities. The value of such a product to the consumer does not lie in its tangible qualities but in its information, educational, cultural or entertainment content.”54 Step 2: Matching the ICT sector with the 59-sector IO table. Based on the definition of the OECD regarding ICT products and media and content products, the aggregation of ICT sectors based on the European IO table is described below. Table 11-3 shows the detailed classification of the sectors grouped as the ICT sectors based on the IO 59 table. This transformation process is conducted by matching the definition of each ICT and media and content product with the definition of IO sectors. Table 11-3 shows that there are twelve ICT sectors among the fifty-nine sectors in the European IO table. The classification is also very interesting in

table 11-3

.

classification of ict sectors based on european 59 sectors io table

No.

Sector Number

Sector Name

1

16

Printed matter and recorded media

2

23

Machinery and equipment

3

24

Office machinery and computers

4

25

Electrical machinery and apparatus

5

26

Radio, television and communication equipment and apparatus

6

27

Medical, precision and optical instruments, watches and clocks

7

36

Wholesale trade and commission trade services, except of motor vehicles and motorcycles

8

43

Post and telecommunications services

9

49

Computer and related services

10

50

Research and development services

11

51

Other business services

12

53

Education services

european broadband spending

199

the sense that half the sectors are defined as goods and the rest are service sectors. Here, the economic impact and the contribution of the ICT sectors in this study correspond to these twelve sectors. However, since the aggregation level of the International Standard of Industrial Classification (ISIC) categories are more detailed than the IO categories, some ICT products and media and content products are aggregated in a particular IO sector. Step 3: Descriptive analysis. The descriptive analysis from the IO table is implemented by directly comparing the value of economic sectors from the table in terms of GDP, output, export and intermediate transactions. In this study the comparison is investigated distinguishing the ICT sectors and the rest of the economy. Table 11-4 shows the growth of output between 1995– 2000 and 2000–2005 for a selected sample of European countries. From Table 11-4, it can be inferred that the growth of ICT sectors in the sample of European countries follows a general path of economic growth. While the output growth decreased from 4.89 percent during 1995–2000 to 2.37 percent from 2000–2005, the growth of ICT sectors also dropped from 7.31 percent during the first period to only 2.33 percent in the second subperiod. Thus, from Table 11-4, it can be concluded that, in general, ICT sectors still show a positive output growth for the average European countries. Nevertheless, the growth rate is decreasing compared to the rest of the economy. Figures 11-2 through 11-5 show the contribution of the ICT sector to output, GDP, export and intermediate transaction for each country.55 While Table 11-4 describes that, in general, the growth rate of the ICT sector is lower than the rest of the economy, Figure 11-2 shows that the contribution of ICT sectors to output varies between countries. Except for Austria and Denmark, which reached a higher contribution during the second period, most of the European countries suffered from lower contribution of the ICT sectors (Finland, France, the Netherlands, Norway, Spain, and Sweden), while Germany was stagnant. The contribution of the ICT sectors on average is 25 percent for the whole European economy in 2005, slightly lower compared to 2000 (26 percent). Figure 11-3 shows the contribution to the GDP. Since GDP has a proportional relationship to the output, the contribution of the ICT sectors in terms of GDP is similar to those explained earlier

table 11-4

200

.

growth of output 1995–2000

2000–2005

Economic output

4.89%

2.37%

ICT sectors

7.31%

2.33%

i. k. rohman and e. bohlin

0.35 1995

2000

2005

0.30 0.25 0.20 0.15 0.10 0.05

ag

e

en

er

Th e

N

et

Av

he

Sw

ed

ai n Sp

s rla

m er G

D

nd

y an

e nc Fr a

k m en

Fi nl an

ar

ria st Au

d

0.00

figure 11-2.

Contribution of ICT sectors to economic output (%). Source: Author’s calculations.

0.30 2005

2000

1995 0.25 0.20 0.15 0.10 0.05

Av

er a

ge

en ed Sw

Th e

N

et

Sp ai n

nd s he rla

an y G

er m

nc e Fr a

nd la Fi n

en m D

Au s

tri

ar k

a

0.00

figure 11-3.

Contribution of ICT sectors to the GDP (%). Source: Author’s calculations.

in terms of output. In this case, except for Germany and the Netherlands, all European countries have a smaller contribution by ICT sectors to the GDP. On average, the contribution was 22 percent in 2005 compared to 23 percent in 2000. european broadband spending

201

Figure 11-4 describes the contribution of the ICT sectors to the total export of European countries. The figure also indicates that having reached higher contributions in 1995–2000, the proportion of the ICT sectors declined in most countries except for Finland and Germany. Intermediate transactions measure the output of the economy, which, instead of being consumed in final demand, is used for further production process in the other sectors. Thus this measurement reflects the interrelatedness of ICT sectors to the rest of the economy. Figure 11-5 shows that the contribution of the ICT sectors declined for all countries except for Finland and Germany. In summary, the output of ICT sectors has declined due both to production level (which is indicated by the lower proportion in intermediate transaction) and to consumption level (both from domestic and export). Step 4: Multiplier analysis. The most powerful aspect of the input-output model used is the fact that European countries apply the compatible fiftynine economic sectors; thus they can be compared with each other. As a detailed example of implementing the IO methodology, the Swedish IO tables—generated for the years 1995, 2000 and 2005—will be examined in more detail as an example. The multiplier effect of the Swedish economy is derived by calculating the inverse of the Leontief matrix and summing each column in order to get a multiplier coefficient for a particular sector.56 The result in 2000 is shown in Figure 11-6.

0.50 1995

2000

2005

0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05

ag e Av

er

en Sw

ed

n Sp ai

s

Th

e

N

et

G

er

he r

m

la

nd

an y

an ce Fr

d an nl Fi

ar k en m D

Au

st

ria

0.00

figure 11-4.

202

Contribution of ICT sectors to the GDP (%). Source: Author’s calculations.

i. k. rohman and e. bohlin

0.50 1995

0.45

2000

2005

0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05

ag

e

en

er Av

Sw

Sp

rla

ed

ai n

s nd

y et Th e

N

D

G

he

er

Fr a

m

nc

an

e

d Fi nl an

en

Au

m

st

ar

k

ria

0.00

figure 11-5.

Contribution of ICT sectors to intermediate transaction (%). Source: Author’s calculations.

4 3.5 3 Multiplier

2.5 2 1.5 1 0.5

po rt ra ns

Ai

rt

itu re rn

pr o od

Fu

du ct

e cl

l

eh i Fo

ru de C

M

pe t

ot or v

C oa

ro le um

et al s

n

m c

Ba si

W

at er t

ra n

sp or ta tio

er y hi n ac

al m ric

ec t El

R ef

in ed

pe

tro

le um

0

Sector

figure 11-6.

Multiplier for Swedish economy, 2000.

Referring to the OECD definition of the ICT sector, the multiplier effect of the ICT sector for Sweden would be as shown in Table 11-5. As suggested by Table 11-5, an average spending of 1 SEK in the ICT sector will generate accumulated economic output of about 1.89 SEK. It then becomes important to compare it with the impact of the expenditure for the non-ICT sector. Applying the same analysis, the average multiplier for the european broadband spending

203

table 11-5

.

ict and non-ict sector multiplier for sweden based on io table 2005

Year

ICT

Non-ICT

2005

1.89

2.15

non-ICT sector is 2.15. The implication is that investments in the non-ICT sector will create a greater multiplier effect. Hence, the opportunity cost of investing in ICT is higher than investing in non-ICT for Sweden. As the ICT sector is used as an indicator of the broadband sector, the initial results suggest that Sweden would be better off by investing in the nonbroadband sector. Applying the same analysis, Table 11-6 presents the comparison between two groups (ICT and non-ICT) measured by the average multiplier coefficients for all selected European countries investigated in this study. As has been explained previously, the output multiplier measures a change in output as one unit value change in the final demand. From Table 11-6, 1 euro of spending in the ICT sector’s final demand enabled growth of economic output by as much as 1.53 euro in 1995. Table 11-6 also indicates that in general the output multiplier of the ICT sector is smaller than the nonICT sectors. In other words, based on the multiplier analysis, it demonstrates that the ability of the ICT sectors to contribute to the economy from final demand effect is lower than that of non-ICT sectors. To have a better understanding of the variation between countries, Figures 11-7 through 11-9 compare the multiplier for both ICT and non-ICT sectors from 1995–2000. Figure 11-7 shows that except for Sweden, the output multiplier of the ICT sectors was smaller than the non-ICT sectors in 1995. It indicates that the contribution of output generated for every euro spent in ICT sectors product is less than the average sectors. The results for the year 2000 are shown in Figure 11-8.

table 11.6 Year

.

multiplier effect

ICT

Non-ICT

1995

1.53

1.58

2000

1.57

1.62

2005

1.57

1.61

source: Author’s calculation.

204

i. k. rohman and e. bohlin

1.8 ICT

Non-ICT

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2

s nd

en Th e

N

et

he

Sw

rla

ed

ai n Sp

m G

er

Fr a

nc

an

y

e

d Fi nl an

D

en

m

ar

k

lg iu m Be

Au

st

ria

0

figure 11-7.

Output multiplier in the European economy (1995).

1.8 ICT

Non-ICT

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2

nd s

Th e

N

et h

er la

en

ai n

Sw ed

Sp

or w ay N

Ita ly

er m an y G

Fr an ce

nl an d Fi

ar k

iu lg

en m D

Be

Au

st

ria

m

0

figure 11-8.

Output multiplier in the European economy (2000).

As in the 1995 analysis, the performance of the ICT sectors is generally lower in terms of output of multiplier in 2000. Nevertheless, unlike those results in 1995, Sweden has a lower multiplier effect in the ICT sectors. In the other Scandinavian countries, Norway has a higher multiplier effect in the ICT sectors. In addition, there is a significant growth of the ICT multiplier, european broadband spending

205

even though the numbers are still lower than the non-ICT sectors. Figure 11-9 presents the comparison between the two categories in 2005. The last investigation of the multiplier effect demonstrates that the ICT sectors contributed even less in 2005 in comparison to 1995–2000. Figure 11-9 shows that, except France, none of the European countries has ICT sectors with a greater contribution than non-ICT sectors. The range of multiplier is 1.3–1.7, where the ICT sectors have an approximately 15 percent lower multiplier compared to non-ICT sectors. Step 5: Evaluation. In conclusion, the output of the ICT sectors has declined due both to the production level (which is indicated by the lower proportion in intermediate transaction) and to the consumption level (both from domestic and export). The growth rate of the ICT sectors was greater in the European economy in 1995–2000 but decreased in 2000–2005 compared to the rest of the economic sectors. The contribution of the output of the ICT sectors to the total output generally increased in 1995–2000 and decreased in 2000–2005 except for Austria. The ratio to the total GDP generally increased in 1995–2000 and decreased in 2000–2005 except for Germany. The ratio for export also gives the same sign—that it increased in the first period and decreased in the second period except for Finland. The last ratio is related to the interrelatedness to other sectors through the production side. It leads to the conclusion that except for Finland and Germany, which increased in both periods, the other countries have been experiencing a lower proportion in the second period. In all countries observed in this

1.8 ICT

Non-ICT

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2

nd s

n

rla

ed e Th

e

N

et

he

Sw

ai n Sp

ay w

Ita ly

or N

G

er m

an

y

ce an Fr

k

nl an d Fi

ar m en D

Au st ria

0

figure 11-9. 206

Output multiplier in the European economy (2005).

i. k. rohman and e. bohlin

study, the ICT sectors contributed lower multiplier effects. On average, these ranged from 1.5 to 1.6, in 1995–2005, while in the non-ICT sectors, the multiplier ranged from 1.6 to 1.7. This conclusion brings two important messages: • If the evaluation of broadband deployment is conducted strictly based on the analysis of sector performance, undoubtedly, the opportunity cost of spending on broadband deployment is large in the sense that choosing to spend the funding on other sectors is believed to achieve a greater economic impact. • But, as other studies have shown—for instance Rosenberg,57 Milgrom et al.,58 Bresnahan and Trajtenberg,59 and OECD60—the technology attached to the ICT sector has the potential for pervasive use in a wide range of sectors and by its technological dynamism enables generalized productivity gains to be transferred to the rest of economy. Accordingly, the role of technology has become more important as the catalyst in the process of innovation and thus broadband policy is an important agenda to be realized.

conclusion The following conclusions can be drawn from the above analysis: • The implications of a policy, the goal of which is to promote broadband in rural areas in Europe, have to be treated carefully. Does this policy need special funds competing with funding for other purposes? Or is there some free money that has zero opportunity cost? • In dealing with the possible opportunity cost, this study applies a cost-benefit analysis based on the value of the economic multiplier, using the 59-sector IO table of European countries. The sample of the study covers eleven selected European countries that have a similar technological level to represent the European economy. • For the basic analysis, it is presumed that any broadband policy will have significant economic impact from the production point of view. Hence, promoting broadband will directly affect the sectors that utilize broadband as the main component of their operation. These sectors are then called the ICT sector. This study adopts the OECD definition related to the classification of the ICT sector based on the International Standard of Industrial Classification (ISIC) • From the descriptive analysis, in most countries, the ICT sectors contributed a greater proportion during the first half of the european broadband spending

207

observation period (1995–2000) but a lower proportion to the output, GDP, and intermediate input for the rest of the economy during the second part (2000–2005). Moreover the multiplier analysis leads to the conclusion that ICT sectors contributed a lower multiplier during the second part of the observation. • Two implications of this study suggest that if the evaluation of broadband deployment is conducted based on ex post analysis of sector performance, the opportunity cost of spending on broadband deployment is large in the sense that choices to spend the funding for other sectors would create a greater economic impact. However, considering that the technology attached to broadband enables further use by the other sectors (non-ICT sectors), as a general purpose technology, the policy to promote broadband is an important agenda. • Further research after this study should elaborate the IO method in defining the employment multiplier and wage multiplier. Thus, forecasting the effect of employment generation for the whole economy given an additional job available from the ICT sector would be of interest. The same interpretation is applied for the wage multiplier, which describes additional wages (or income) for the whole economy, given an additional one euro in income from the ICT sector. The employment data is a significant challenge however. Further research can also be applied to carry out the primary survey data to measure additional welfare from consumer or the individual point of view in a more precise way. The main weaknesses of this study are the weaknesses of the IO model itself, which is based on static analysis, constant price, and a homogeneous product. In the worst case, the method aggregates many dissimilar products into one group. In addition, comparing one country to another is apparently less realistic given the fact that many countries have different economic platforms or even exchange rates as the basis for calculating prices. Therefore, this study is basically an exploratory study based on the IO table as an alternative approach to address broadband investment. Future research can be implemented by utilizing the ongoing project of interregional input-output (IRIO) table of European countries. Using this table, the study can investigate the impact of the project not only by looking at a single country impact but also at the interconnection between all European countries, which is more sensible in producing a policy recommendation. Another enrichment study can be done by using the computable general equilibrium (CGE) model, which can investigate the change in production, household and factor price simultaneously. 208

i. k. rohman and e. bohlin

CHAPTER 12

Using Data for Policy Development Designing a Universal Service Fund for Tanzania heather e. hudson Many developing countries have established universal service funds to subsidize the extension of a telecommunications infrastructure to areas that are unattractive to private investment because of high costs or low revenue potential. There are many challenges in implementing these funds, particularly where there is little experience in estimating demand for new services such as Internet and broadband, and limited time and resources to collect field data. This chapter examines how data, primarily from existing sources, is being used to implement a universal service fund for Tanzania. It is presented as an example of relying primarily on existing data from multiple sources and from local researchers that could be used in other developing regions where time or resources to conduct dedicated field studies may be limited. The government of Tanzania (population about forty million, of which 80 percent live in rural areas) has approved the creation of a Universal Communications Access Fund (UCAF) to extend communications services to unserved and underserved areas of Tanzania, primarily in rural regions. Initial primary emphasis is to be on the extension of voice telephony to rural areas without connectivity and provision of public access to the Internet in rural communities. The author was a member of a team providing advice to the Ministry of Communications, Science and Technology on implementation of the UCAF, including definition of services, identification of unserved and underserved areas, prioritization of access requirements, and design of procedures to allocate funds.1 Preliminary studies of availability of telephony through mobile phones and fixed lines and projections of demand had been conducted, but there 209

was a delay of four years between the initial studies and the final authorization of the UCAF. In that time period, mobile penetration had increased dramatically through licensing of several competitive providers, but there appear to be remaining gaps in coverage in some rural areas. Pricing is also a deterrent to use among low-income populations. The government has indicated strong interest in including support for Internet service, primarily through public access, which could include post offices, telecenters (publicly accessible locations with computers and Internet access), Internet cafes, and other entrepreneurial outlets, and possibly schools. Key questions to be addressed included: What priorities should be used to provide total voice coverage in Tanzania, in terms of criteria such as population densities, geographic areas, or estimated demand? Could innovative technological solutions and business models to extend Internet access be fostered through financial incentives from the UCAF? Could broadband services be extended with support from the fund without waiting for the completion of 100 percent coverage for voice services? What services were likely to be provided, and what regions would gain access, through operator investment without subsidies? Since funding and time for field research were limited, the availability and interpretation of existing data were important in understanding the extent of the gap in availability and affordability, and for identifying viable models of shared access to Internet services in most rural areas of the country. Where field research appeared necessary, involvement of local researchers from universities, development organizations, and local consulting firms was recommended to conduct the studies.

tanzania and its telecommunications sector Located in the east part of Africa, Tanzania has a land area of about 342,000 square miles (about twice the size of California) and a population of 41 million of whom approximately 45 percent are under the age of 16.2 Classified as a low income or least developed country, Tanzania has a per capita income of only about $400,3 although the amount in terms of purchasing power parity (PPP) is estimated at $1200. About 70 percent of the population is rural. After independence in 1964 (with the merger of Tanganyika and Zanzibar), President Julius Nyerere implemented a socialist agenda, which increased access to primary school education and basic health care, but left a stagnating economy. Following liberalization of the economy in the past decade, the economy has grown more than 7 percent annually since 2002, but with an average inflation rate of 5.2 percent.4 Liberalization has been accompanied by increased entrepreneurial activity in the informal sector, more foreign aid, expansion of mining and exports of raw materials, and increases in commer210

h e at h e r e . h u d s o n

cial (as opposed to subsistence) agriculture. Though still primarily agricultural, Tanzania’s economy also benefits from recent expansion of mobile communications, as mobile and other communication services now account for about 20 percent of its economic growth.5 Tanzania’s teledensity in 2008 was estimated at 31.6 percent by the regulator, the Tanzania Communications Regulatory Authority (TCRA). The telecommunications market was approximately $1 billion for 2008, of which more than 90 percent came from mobile services, based on a mobile average revenue per user (ARPU) of $6 to $7 and about 11 million subscribers.6 More than 99 percent of customers use prepaid services. The number of fixed lines actually decreased from 174,000 in 2000 to 124,000 in 2008, while mobile subscribers grew more than a hundredfold (see Figure 12-1). Mobile coverage is estimated by the operators at 70 percent of the population and 50 percent of the land area. Most of the operators use GSM technology, as is common throughout much of Africa. However, the fixed carrier TTCL operates a network using CDMA technology, and one mobile carrier offers both a limited CDMA service and a GSM service provided as a mobile virtual network operator (MVNO) using another GSM network. The GSM networks have been upgraded to the general packet radio service (GPRS)

35%

14,000 Fixed lines (000s) Mobile lines (000s)

30%

Penetration (%)

10,000

25%

8,000

20%

6,000

15%

4,000

10%

2,000

5%

Subscribers/100

Subscribers (000)

12,000

0%

0 Year 2000–2008

figure 12-1.

Tanzania subscriber growth, fixed and mobile. Source: TCRA data and company reports. u s i n g d at a f o r p o l i c y d e v e l o p m e n t

211

technology (also referred to as 2.5G) across the networks, while third-generation (3G) networks have been launched in urban areas by two operators. Users may simply switch SIM cards in the same handset to access operators using the same technology. To switch between operators using different technologies, they would need a second handset (see below). The mobile market is among the most competitive in Africa, with six major operators that compete on coverage or price (see Figure 12-2).7 Competition has become especially aggressive since 2007. The three largest operators have each deployed more than one thousand base stations and cover most of the same territory. Mobile users are likely to have several SIM cards to take advantage of variations in coverage and price. For example, one carrier may provide coverage in a rural area where family members live while another may provide cheaper on-network calls. Some consumers may have multiple SIM phones that allow them to switch between operators, also allowing them to take advantage of short-term promotions. Fixed service coverage is very limited, and the number of fixed service lines has actually decreased since 2002. The incumbent fixed line operator, TTCL, originally a government-owned postal, telegraph, and telephone (PTT) service provider, was partially privatized in 2001, with the government retaining a 65 percent share (the remaining 35 percent are now owned by Celtel, an African mobile operator purchased by Kuwait-based Zain in 2005, which, in 60% Vodacom

Zain

Tigo

Zantel mobile

TTCL mobile

All fixed

50%

Market share

40%

30%

20%

10%

0% 2005

2006

2007

2008

Year

figure 12-2.

212

Telecom sector growth, 2005–2008. Source: TCRA data. Note: A new sixth operator, Benson, had a market share of 0.02% in 2008.

h e at h e r e . h u d s o n

turn, was bought by Bharti Airtel of India in 2010). TTCL retains a de facto monopoly on domestic fixed services, although there has been competition in domestic and fixed international services since 2005. The continuing fixed service monopoly, coupled with remaining significant government ownership and several different management structures, as well as the high investment required to build out fixed lines, appear to have stunted the fixed network. However, the government has contracted with a Chinese vendor to build a fiber backbone ring to link all district centres (comparable to county seats) around the country and to connect local fiber networks. This fiber network is to be managed by TTCL. International connectivity was provided exclusively by satellite until late July 2009, when the Seacom submarine cable linking East African countries went into operation.8 An additional submarine cable, EASSy, went into service in 2010.9 As a result, international quality of service (QOS) is expected to improve, and bandwidth prices to drop. For example, the price of a 1 Mbps satellite leased line via satellite in 2009 was around $1,500 per month, compared to about $200 where competing undersea fiber cables were available. The introduction of competition in international gateways in Tanzania should also result in lower prices.10 Concerning Internet access, the TCRA estimated some 1.8 million Internet users in Tanzania by the end of 2008, including some fixed-line subscribers and customers of Internet cafes, as well as 600,000 mobile GPRS users and more than 1 million students who access the Internet for schoolrelated purposes. Computers in schools remain scarce, so that most students would use telecenters and Internet cafes. Factors that contribute to very limited Internet use include the low penetration of personal computers (less than 1 percent), lack of fixed digital subscriber lines (DSL), and the high price of broadband connectivity where it does exist (only in major centers). In addition, there is very limited rural connectivity, and lower levels of education and income in rural areas. Limited Internet content in local languages also appears to discourage Internet usage. Fixed operator TTCL offers some DSL services in Dar es Salaam (the largest city) and large towns. Competitive Internet service providers (ISP) offer dial-up and fixed broadband using CDMA-based facilities or very small aperture terminals (VSATs). In 2009 broadband connectivity cost about TZS 25,000 (about $19) for 250 MB per month at 512 Kbps, ranging to TZS 80,000 (or about $60) for 1 GB where available. DSL modems cost TZS 125,000 (about $95) and service activation an additional TZ 36,000 (about $27). Clearly such prices were beyond the reach of most Tanzanians. At least two mobile operators were offering Internet access on upgraded HSPA networks to phones and some Internet cafes; two providers were starting to offer 3G service, initially in major cities. However, prices remained u s i n g d at a f o r p o l i c y d e v e l o p m e n t

213

beyond reach of all but the wealthiest individuals or small businesses. In 2009 a prepaid 3G package was priced at TZS 90,000 (about $68) per month for a 3 MB connection.

the universal communication access fund The Tanzania Universal Communications Service Access Act of 2006 established broad objectives for communications and authorized establishment of a Universal Communications Access Fund (UCAF). The act states that telecommunications service providers will contribute up to 1.5 percent of their gross revenues to subsidize access to “communications services” in underserved areas. Among specifications for the fund are: • Communication services include those for “the provision of postal, electronic communications [including broadcasting] or content services.” Thus the fund may be used to subsidize broadcasting services such as local or community radio, postal services, and content. • The fund is to increase availability of communication services in “rural and underserved urban areas.” • The act specifically seeks to promote socio-economic development. However, there are no specified limitations on technologies to be used or carriers/operators eligible for support and there are no specific definitions of what constitutes unserved and underserved areas.11 The broad definition of communication services reflects a unified regulatory framework adopted by the regulator, which includes telecommunications and broadcasting. In addition to telephony, communication services can include both broadcasting and Internet services, and broadband connectivity to provide such services. While universal service funds generally prioritize coverage for regions or populations without access to voice telephony (traditionally provided over fixed lines and public pay phones),12 many countries are now extending definitions to include mobile telephony and Internet access through community facilities such as telecenters, libraries, Internet cafes, or other public facilities. Developing country examples in addition to Tanzania include Ghana, India, Indonesia, Mongolia, Morocco, Mozambique, Nicaragua, Nigeria, Paraguay, South Africa, Venezuela, and Vietnam.13 Organizationally, the fund falls under the jurisdiction of the Ministry of Communication, Science and Technology. It is to be managed by a board whose eight members are to be appointed by the minister to include three from industry and one from a consumer organization, one from the regulator, and the other three from the government. This structure is not typical of universal service funds in other countries, some 60 percent of which are 214

h e at h e r e . h u d s o n

administered through the national regulator, and an additional 19 percent through an independent agency.14 Disadvantages of being under the ministry include the potential for bureaucratic delays and the possibility of political influence in selection of projects eligible for support. For example, the fixed line carrier is still 65 percent government owned, the post office is a government agency, and content providers that may be perceived as competing with the national broadcaster also appear eligible for support. Thus the UCAF differs from some other universal service funds in that it includes the postal service and broadcasting, has a specific goal of promoting socioeconomic development, and is to be administered by a board essentially controlled by the communications ministry.

the potential of ict s for development in tanzania In planning and setting priorities for the universal service fund, it is important to recognize the potential role of ICTs in Tanzania, particularly for its rural, social, and economic development. Information is critical to the social and economic activities that comprise the development process. If information is critical to development, then ICTs, as a means of sharing information, are not simply a connection between people, but a link in the chain of the development process itself. In general, the ability to access and share information can contribute to the development process by improving: • Efficiency, or the ratio of output to cost (for example, to get price information, to coordinate delivery of agricultural products to market; to obtain information on weather and soil content to improve agricultural yields); • Effectiveness, or the quality of products and services (such as improving health care through telemedicine); • Equity, or the distribution of development benefits throughout society (such as to rural and remote areas, to minorities and disabled populations); and • Reach, or the ability to contact new customers or clients (e.g., craftspeople reaching global markets on the Internet; educators reaching students at work or at home).15 Ugandan farmers and fishermen can now get the daily prices of coffee and lake fish on their mobile phones for the price of a single text message (SMS). African examples of the contribution of telecommunications to effectiveness include the activities of Schoolnet Africa that provide ICTs and content for u s i n g d at a f o r p o l i c y d e v e l o p m e n t

215

schools such as online “virtual labs” and reference materials that can enhance education where schools lack lab equipment and libraries.16 In telehealth and telemedicine, information about the patient transmitted electronically can help distant specialists to diagnose conditions and recommend treatment. The African Medical and Relief Foundation (AMREF) has used two-way radios in Tanzania and Kenya for several decades to provide medical advice and coordinate its flying doctor service.17 Tanzania has recognized the importance of ICTs for its development through its Vision 2025 strategic plan which envisages the nation with five key attributes: high-quality livelihood; peace, stability, and unity; good governance; a well-educated and learning society; and a strong and competitive economy capable of producing sustainable growth and shared profits. Tanzania’s National ICT policy, adopted in March 2003, includes several goals, among them development of a nationwide e-education system, the teaching of ICT at all levels of education and training, and the use of ICT to improve the quality of delivery of education.18 Extending Internet access is critical to achieving these goals. Several initiatives are already underway. Many government ministries have plans to extend e-services. There are already a number of projects providing access through community telecenters and at some schools and vocational training colleges.19 Some rural entrepreneurial Internet access points are being established; others have been associated with radio stations and community centers. Tanzania’s Ministry of Education and Vocational Training (MoEVT) has recognized the role of ICTs in extending basic education to all Tanzanians and providing skills that will be needed to improve productivity and competitiveness, as well as to deliver government and other services. All regional and district offices have been provided with computers and printers (although the Ministry does not state whether all locations have connectivity). A project to introduce ICTs in all government teachers’ colleges has been implemented,20 and the Institute of Adult Education (IAE), which offers correspondence courses using print materials, is promoting the introduction and integration of ICT and e-learning in its regional centers across the country. Tanzania is also committed to achieving the Millennium Development Goals (MDGs) for poverty reduction. The MDGs were enshrined in the Millennium Declaration adopted by 189 countries at the UN Millennium Summit in September 2000 to attain 8 specific goals by the year 2015: • Eradication of extreme poverty and hunger; • Achievement of universal primary education; • Promotion of gender equality and empowerment of women; • Reduction of child mortality; 216

h e at h e r e . h u d s o n

• Improvement of maternal health; • Combatting HIV/AIDS, malaria, and other diseases; • Ensuring environmental sustainability; and • Development of global partnerships for the attainment of a more peaceful, just, and prosperous world. Specific targets have been identified for each goal. The MDG goals and targets have been developed into an MDG goals/ICT indicator matrix.21

implementing the ucaf Research questions. Many questions need to be addressed in implementing Tanzania’s universal service fund, for example: General questions: • What priorities should be adopted to extend voice services in rural areas? For example, starting with largest unserved communities or with the most remote? Extending service from existing coverage areas? Targeting certain geographic or ethnic areas? • What kind of voice service is required in rural areas? For example, pay phones (public call offices) in addition to household or individual services typically by mobile phone? • How important is affordability (as opposed to availability) in influencing actual usage by rural residents? • Should projects that are integrated into rural economic and social development initiatives be given priority? • What priority should be given to improving access in underserved (as opposed to unserved) areas (e.g., areas with poor QOS because of hills or other barriers, areas with low signal strength, areas where there is an insufficient number of available circuits, and so on)? • What models of community or institutional access could be used to extend Internet services, such as telecenters, commercial Internet cafes, post offices, libraries? Should connectivity for schools be a priority? • Should voice networks be required to be upgradeable to provide broadband Internet services? • How can the UCAF provide incentives for innovative technological solutions? For example, should it subsidize satellite phones for remote areas difficult to serve with terrestrial technologies? Should existing u s i n g d at a f o r p o l i c y d e v e l o p m e n t

217

microwave networks be used to provide backhaul for competitive Internet providers? • How can the other services included in the Act, such as postal services, broadcasting, and content, be included in the funding process? What criteria should be used, and what priority should they get? Voice services: • Voice access: How many Tanzanians now have access to voice services through owning mobile phones, through shared use of mobile phones, or through access to pay phones? • Access demographics: What demographic characteristics distinguish people with voice access from those without (e.g., location, income, education level)? • Mobile coverage: What parts of the country have no mobile coverage, and how many people live there? • Affordability: To what extent is price, or affordability, restricting access (as opposed to availability or signal coverage)? • Cultural factors: What cultural factors may influence access and use (e.g., large and extended families, sharing within families, communities of interest—tribal, ethnic, work location)? • Mobile investment: What factors influence the investment decisions of mobile operators to continue to extend coverage (e.g., cost of financing, estimated demand, topography, lack of electricity, and so on)? Internet and broadband: • Demographics of Internet users: What are the characteristics of current Internet users (e.g., age, location, education, income)? • Internet access: How and where are people accessing the Internet? For example, where are community access facilities located, and what is their usage? How many small and medium enterprises (SMEs) and nongovernmental organizations (NGOs) have Internet connections? How many people are accessing Internet applications on mobile phones? • Computer access: Why is the number of personal computers so limited? If price is the major barrier, how could prices be reduced? • Internet use: What factors influence Internet use (e.g., availability, price, skills required, language required, and so on)? 218

h e at h e r e . h u d s o n

• Mobile Internet: To what extent are subscribers living where 2.5G or 3G service is available using it for e-services or for Internet access? • Current models of Internet access: What is the experience with community access to date (telecenters, radio stations, schools, Internet cafes, and so on)? • Broadband demand: Will cheaper broadband prices significantly increase demand? • Broadband investment: Would increased demand result in more broadband investment by operators? Sources of data. Some demographic and technical data was available from government and industry. However, there was neither time nor budget to carry out extensive field studies to get an understanding of how mobile phones are used and shared, and what factors inhibit Internet usage. Preliminary field research included interviews carried out in Dar es Salaam with operators, ISPs, government officials, and Tanzanian researchers. Most of the information had to be gleaned from secondary sources, including data collected by the Tanzanian regulator and census, and several studies of rural telecommunications access and demand in African countries including Tanzania, as well as ICT case studies in Tanzania. These sources include: • Household budget expenditure data including money spent on ICTs collected by the National Bureau of Statistics, but disaggregated only to urban versus rural level; • Other recent national and rural government surveys that contain rural socio-economic data (census data, although detailed, dated from 2002 before mobile competitive mobile services were introduced); • Data on coverage and number of subscribers submitted by operators to the regulator regularly as a requirement of their licenses, but not disaggregated by location or combined into a single data base or mapping system; • Field studies recently conducted or in process on rural use or impact of ICTs that could provide some insights into rural demand and use, although not carried out in the projected highest-priority regions; and • Pilot projects to establish community access through telecenters and entrepreneurial access points and Internet cafes that could provide anecdotal or possibly case study evidence of usage, costs, and sustainability. There are several potential problems with relying on these studies to gain a better understanding of rural communications access and usage. Because of u s i n g d at a f o r p o l i c y d e v e l o p m e n t

219

the difficulties in reaching rural populations (poor roads or no roads, scattered villages, and so on) researchers tend to concentrate more on urban and periurban areas, or to overgeneralize from small rural samples or case studies. Where significant resources have been put into rural data collection, such as for the national census and HIV/AIDS and malaria monitoring, available data may not be sufficiently disaggregated to be very useful.

insights from available data What is known about access? The following section discusses findings from existing research and field case studies that could help to understand rural access and demand, and to assess the likelihood of the mobile industry extending rural coverage without subsidies. Most users are urban. Most mobile phone owners live in urban areas. The National Bureau of Statistics (NBS) survey in 2007 found that almost twothirds of households in Dar es Salaam had mobile phones and 42.5 percent had phones in other urban areas, but only 13.9 percent of the households surveyed in rural areas owned mobile phones.22 Yet many more rural households have phones than electricity, which is available in only about 2 percent of rural households, compared to 40 percent in urban areas.23 (Rural entrepreneurs offer charging services for phone owners using diesel generators, vehicle batteries, and solar panels.) Sharing mobile phones and minutes is common. On a per capita basis, the NBS statistic of only 13.9 percent of households with a mobile phone would indicate a rural teledensity of less than three, assuming at least six members per household. But there is convincing evidence that family members share their mobile phones. ICT Africa’s survey of Tanzanian mobile phone owners found that 46.2 percent indicated they did not share their phones, while 44.6 percent said they shared with family members, 25.4 percent with friends, and 15.8 percent with neighbors (multiple responses permitted).24 Using these estimates, access to a phone where there is coverage could be widespread among family members and fairly extensive for friends and neighbors of a mobile phone owner. This conclusion was collaborated by a study on ICTs and poverty in Tanzania that found that mobile phones were widely shared among poor populations.25 Airtime is also apparently widely shared, through transfer of minutes from one phone account to another. The ICT Africa survey found that of all Tanzanians surveyed (20.5 percent of whom were estimated to have mobile phones), 61.7 percent said they had received minutes as a favor or for cash or barter.26 Rural mobile phone owners are older. While Tanzania has a very young population with about half under the age of sixteen, many mobile phone owners tend to be from older cohorts. A survey of four hundred mobile 220

h e at h e r e . h u d s o n

owners in twenty-two villages in Iringa province found that 49 percent were aged thirty-six or older (of course, this means that half were under thirty-six years).27 This finding is not surprising in that it is likely that only adults with disposable income from rural jobs, cash crops, or remittances from relatives working in cities could afford mobile phones. Demand for mobile phone service is high. ICT Africa’s survey indicates that on average, mobile phone users spend 28.9 percent of their disposable income on mobile phone use, while the poorest 75 percent of the population with phones spend about 41 percent of their disposable incomes on phone use.28 Clearly, Tanzanians view access to communications as extremely important, likely displacing other discretionary purchases, particularly in rural areas where incomes are generally very low. These findings may also help explain the fact noted earlier that communication services now account for about 20 percent of Tanzania’s economic growth.29 Cost is a significant barrier to mobile phone ownership and usage. While demand for communication is high, cost is a constraint in usage among those with mobile phones, and a barrier to those without. In the ICT Africa survey, when asked what prevented them from making more calls, 72.3 percent cited the cost of calls, and 46.7 percent cited the cost of calls “very high.” Among those without phones, demand for mobile phones and mobile phone use is high, with about 20 percent of the population interested in purchasing a phone or SIM card. However, the average amount they were willing to pay to use a phone was very low, only $2.61. Next to Ethiopia, this was the lowest amount found among the seventeen sub-Saharan African countries covered in the survey. To purchase a phone, respondents on average were willing to pay $10.89, and among rural users $9.74, although they realized that phones would cost the equivalent of about $17. The study concluded that availability of $10 handsets could result in 2.26 million additional subscribers.30 Subscriber overcounting. It is likely that there are fewer individual subscribers than most estimates indicate because of the prevalence of users having multiple SIM cards to deal with variations in coverage areas (e.g., to make sure they can reach family members in rural areas), to use cheaper on-network calling rates among contacts on the same network, and to take advantage of shortterm promotions. Research ICT Africa estimated 1.16 SIMS per mobile user in Tanzania in 2008 based on their survey of mobile phone owners and nonowners in seventeen countries.31 Yet operators and SIM resellers interviewed in Dar es Salaam estimated most mobile subscribers had two to four SIM cards. Indeed, one mobile executive interviewed in April 2009 stated: “People walk around Dar with their pockets loaded with SIMs.”32 However, multiple SIM use is likely to be less common in rural areas because of more limited network coverage and fewer convenient local SIM resellers. u s i n g d at a f o r p o l i c y d e v e l o p m e n t

221

Also, operators may have incentives to inflate their subscriber figures in order to attract others to their network—using the rationale articulated in Metcalfe’s Law (as popularized by Internet pioneer Robert Metcalfe) that being able to connect to more people offers significantly greater value.33 Further, data are not available on the amount of churn in subscribership (i.e., changing carriers as opposed to simply using different SIM cards). Thus, aggregating numbers of SIM cards from all operators is likely to significantly overestimate the number of individual subscribers. These findings provide both a “good news” and “bad news” view of rural access. Clearly, telephony is highly valued by the rural population, but most find they cannot afford to purchase phones or to use phones much. However, those with phones tend to share their phones and their minutes within families, and may also be willing to sell or barter minutes with others in the community, so that where there is coverage, actual access may be much higher than reflected in numbers of handsets or SIM cards. It is not clear whether sharing of phones and selling and bartering of minutes have effectively replaced the function of a community public payphone. While it would be useful to have more data from rural areas of Tanzania with and without mobile voice coverage, it seems clear that if coverage were provided in currently unserved areas, residents would benefit from individual and shared use of mobile phones.

the importance of subsidies to expanding rural access In 2008 Tanzania’s mobile growth was 56.4 percent, while the mobile cumulative annual growth rate (CAGR) from 2002 to 2007 was 68.8 percent.34 Extrapolating the 2008 growth rate to the existing estimated mobile density of 31.6 percent would result in more than 100 percent mobile density by the end of 2011, a highly unlikely scenario. In fact growth rates have generally declined in sub-Saharan Africa including Tanzania in the past few years. Figure 12-3 shows growth rates from 2002 to 2007 declining compared to growth rates in 1999 to 2004 in several sub-Saharan countries (the exception being Zambia). Although growth rates may decline, given heavy urban penetration, we might expect that most network expansion will be in rural areas, since about ten million people or 25 percent of the population live in rural areas without mobile coverage or fixed line access. Yet Tanzanian mobile operators at the time of planning implementation of the fund were apparently much more concerned with capturing additional urban customers than extending into unserved areas. Interviews with several mobile operators in April 2009 indicated that they were not likely to be extending networks much beyond their 222

h e at h e r e . h u d s o n

100.0 1999–2004 CAGR 2002–2007 CAGR 2007 Penetration

160.0 140.0

80.0 70.0

120.0 CAGR

90.0

60.0 100.0 50.0 80.0

40.0

60.0

30.0

40.0

20.0

20.0

10.0

Mobile subs/100

180.0

0.0

0.0 w

a an

ts

Bo

a

n bo

a

G

figure 12-3.

bi

am

G

Ke

a ny

a

a

i ib

am

c fri

h

N

S

t ou

il

A

az

Sw

a nd

a

i an

d an

nz

Ta

ga

U

a

bi

m

Za

Mobile growth and penetration in SubSaharan Africa. Source: Based on ITU data, www.itu.int/ITU-D/icteye/.

current coverage, but rather upgrading their existing networks to attract more business and high income users, and increasing urban capacity to compete with new entrants. Competition among operators should be expected to drive continued expansion. As noted above, there are six mobile competitors (plus two new licensees), of which three have significant rural coverage. Developing countries with mobile competition (even a duopoly) have experienced mobile growth rates significantly higher than those that remain monopolies.35 The three largest operators in Tanzania do compete on coverage; ICT Africa’s survey found that for 62 percent of Tanzanians with mobile phones, coverage was the primary reason for selecting a carrier. Yet the operators appear to prefer to compete on services and features, which were cited as important factors in about 28 percent of survey responses. Of course, anticipated average revenue per user (ARPU) is likely to be much higher in urban areas, as many rural residents live at subsistence levels, although they may also generate significant revenue from incoming calls paid for by urban relatives. The global recession also apparently squeezed mobile operators as well as contributing to a revenue shortfall in the Tanzanian economy. One mobile executive stated that his company had effectively stopped extending rural coverage; another executive stated that the two largest mobile operators had halted network expansion.36 Thus, it would appear that subsidies are required to extend coverage to unserved rural areas. Yet the available research on access indicates that there u s i n g d at a f o r p o l i c y d e v e l o p m e n t

223

is significant pent-up demand among those without mobile phones, that those with phones are willing to spend a very high percentage of their disposable income on mobile use, and that reducing the price of handsets could result in more than one million new subscribers, who in turn might provide access to many million more relatives and neighbors. Simply reducing the price of handsets, therefore, where wireless coverage exists, could significantly increase rural access. Could operators be persuaded to lower handset prices, given the potential increase in usage revenues? Or would a handset discount subsidized by the fund (similar to the Linkup model in the United States) be more appropriate?

options for increasing internet access Access models. As discussed earlier, the Tanzania Universal Communication Service Access Act of 2006 and its specifications for the UCAF, use the term communication services (rather than telecommunications or telephony), which can be interpreted to include Internet services. The Ministry of Communications, Science and Technology and the manager of the UCAF believe that extending Internet access should be a major means for the fund to contribute to Tanzania’s development.37 Tanzania is also committed to using ICTs to help achieve the Millennium Development Goals, and several ministries and government organizations have plans to provide services electronically in rural areas. Thus, while voice communications remains the first priority, the UCAF is expected to fund Internet projects also as a high priority, without waiting until all unserved areas receive voice coverage. There are a variety of alternatives in terms of access models and technologies that could be adopted to extend Internet access. The access models include: • A single nationwide model of public access (e.g., at post offices); • A variety of public access models (telecenters, cybercafes, other shops, NGOs, etc.); • Schools and libraries; • Other institutions, such as government offices, community centers, and banks; and • Personal Internet access using wireless phones, PDAs, laptops, or netbooks. Also, there are a variety of technologies that could be used to reach these access locations including 3G or next-generation mobile networks; fixed wireless (e.g., via WiFi, WiMax) from existing backbone—fixed or mobile; 224

h e at h e r e . h u d s o n

DSL in towns where TTCL has fixed lines; and very small aperture terminals (VSATs). It should be noted that these models are not mutually exclusive. Different access models may accommodate users with varying needs (e.g., literacy levels, ability to pay, and familiarity with technology). Further, there is considerable experience in sub-Saharan Africa (and other developing regions) in implementing Internet projects, in terms of organization, training, content, and models for sustainability.38 Additionally, there are already a number of private and public sector plans or initiatives for Internet expansion or adoption. The UCAF could coordinate its Internet/broadband initiatives with other government priorities, such as Internet connectivity for all schools (or middle and secondary schools), for postsecondary education (including universities, teachers’ colleges, technical schools, and nursing schools), or for the achievement of targets adopted for the Millennium Development Goals (MDGs) program. A pilot project approach. The UCAF will not have sufficient funds to address all Internet needs, and is likely to address completing voice service coverage as its first priority. Therefore, in the first phase (i.e., the first three years), several pilot projects could be supported that would take a variety of approaches to providing Internet access in different regions of the country. Also, by the end of the three-year period, more of the domestic fiber backbones should be in place, and impact of the availability of international submarine cables on demand should be better understood. Advances in wireless technologies may also indicate which middle-mile or last-mile technologies are likely to dominate. Bidders for pilot project support in rural areas could include operators (mobile or fixed) that have announced plans for extensive national broadband coverage. They could use the pilot project support to demonstrate how their networks would be deployed in rural areas, and the projects would provide an opportunity to assess the suitability and sustainability of their strategies in rural areas. The pilot project approach would also provide an opportunity for smaller operators and ISPs to obtain seed money to demonstrate their approaches, also subject to the evaluation criteria. The following are three examples of projects that could yield valuable experience for planning national Internet and broadband access. A key component of this pilot phase would be an independent evaluation to collect utilization data, to document lessons learned from the projects’ implementation and operation, identify individual and collective benefits, and assess potential sustainability. Tanzanian researchers could be contracted to carry out these evaluation studies, which would be used in determining future strategies for investment in Internet access. u s i n g d at a f o r p o l i c y d e v e l o p m e n t

225

Other forms of support. The following are examples of possible pilot projects that could be funded during the start-up phase of the UCAF. Since the UCAF mandate is broad, other strategies could be tried to increase user access, provide more affordable equipment, and develop relevant content. For example: Individual user-based subsidies. The UCAF could experiment with providing some subsidies for individuals who would otherwise be unable or unlikely to use the Internet. For example, vouchers for a certain amount of time online could be subsidized for distribution to target populations, such as lowincome users, first-time users, rural users, or students. This approach was used successfully in a United States Agency for International Development (USAID) telecenter project in Bulgaria.39 Discounts on equipment. The UCAF could also reduce the cost of computers, cited by several people interviewed as a hindrance to computer and Internet use in Tanzania. Netbooks, and other basic personal computers are now available for a few hundred dollars in industrialized countries, but typically cost much more in the developing world. The UCAF might be able to provide funds to schools, telecenters, and libraries, among other entities, to arrange bulk purchases or offset taxes and duties. Other government procurement initiatives with vendors or donors for schools and community access should be coordinated with the UCAF so that participating communities and institutions receive high priority for connectivity projects. Support for training. The UCAF could experiment with supporting training in computer and Internet skills because ability to use computers or other devices and to find and use online materials are critical for Internet use. Training could be done at telecenters, or by NGOs, community groups, or professional trainers or volunteers. Support for content or applications. Another aspect of UCAF’s Internet access program could involve the issuing of grants for the development of online content and/or applications that could increase the utilization and effectiveness of online resources. While such content support is unusual for universal service funds, the UCAF mandate is broad enough to include content. The experience of Uganda, one of the few countries to support content creation and training, could be used as a model.

conclusion While available data are limited, several conclusions can be drawn from the above analysis that can inform UCAF implementation. Concerning voice services, the number of mobile subscribers is likely to be considerably lower than most estimates because of the common use of multiple SIM cards. However, the actual number of Tanzanians with access to voice communications is 226

h e at h e r e . h u d s o n

significantly greater than the number of mobile subscribers because of the common practice of sharing phones and airtime. Factors affecting rural access to voice services include not only signal coverage but also the price of mobile handsets and calls. Without subsidies, mobile carriers are not likely to extend rural coverage, but to upgrade their urban networks and services to attract more urban customers and generate higher ARPUs. Concerning the Internet, usage is extremely low, and occurs largely through Internet cafes and telecenters, primarily in urban areas. Few rural schools, government offices, or businesses have computers. Prices for Internet devices, whether computers or 3G phones and personal digital assistants (PDAs), are likely to remain beyond the reach of the vast majority of Tanzanians. Pilot projects to increase Internet access in rural and underserved areas could be designed to encourage innovation in terms of technologies and implementation strategies, and participation by small and new operators and ISPs. Evaluation of the projects could provide valuable data on demand and barriers to usage, and contribute to a better understanding of the role of ICTs in Tanzanian development. Several of the strategies recommended for implementation of Tanzania’s UCAF could be relevant for other developing regions. For example, data from the national census and other surveys can be a useful source, particularly if data can be disaggregated according to relevant demographic and geographic criteria. Studies and experience from neighboring countries and others with comparable conditions (in this case, Uganda, and more generally sub-Saharan Africa) can be useful to estimate demand and to understand potential barriers to adoption. A pilot project approach with support for evaluation can provide valuable experience at minimum cost before national rollout plans are implemented. Finally, involving local researchers can be advantageous in identifying relevant existing studies, planning and carrying out field research, and building local capacity in ICT planning and evaluation.

u s i n g d at a f o r p o l i c y d e v e l o p m e n t

227

This page intentionally left blank

NOTES

introduction: numbers that matter Richard D. Taylor and Amit M. Schejter

1. Mike Peters, “Obama Views Broadband as Key to Economic Recovery,” Broadbandinfo.com (December 16, 2008), http://www.broadbandinfo.com/ news-archives/2008/obama-views-broadband-as-key-to-economic-recovery.html. 2. D. Bell, The Coming of Post-Industrial Society: A Venture in Social Forecasting (New York: Basic Books, 1976).

1. beyond broadband access: what do we need to measure, and how do we measure it? Catherine A. Middleton

1. European Commission, “EU Guidelines on Public Funding of Broadband Networks in Member States—FAQs,” http://www.egovmonitor.com/node/28328. 2. Organisation for Economic Co-operation and Development, Broadband Growth and Policies in OECD Countries (Paris: OECD, 2008); OECD Directorate for Science Technology and Industry, OECD Policy Guidance on Convergence and Next Generation Networks (Seoul: OECD Ministerial Meeting on the Future of the Internet Economy, 2008). 3. It is acknowledged that there are broader goals for broadband investment, goals that involve engagement with government and industry as well as individuals. The overall impact of broadband investment on an economy is not measured just by considering individual uptake and use. While the focus of this chapter is on the relationship between individual broadband adoption and 229

societal benefits, this is just one piece of a much larger puzzle about how benefits are realized from investment in broadband technologies. 4. Paschal Preston and Anthony Cawley, “Broadband Development in the European Union to 2012—A Virtuous Circle Scenario,” Futures 40 (2008): 813. 5. Statistics Canada, “Canadian Internet Use Survey, 2009,” http://www .statcan.gc.ca/daily-quotidien/100510/dq100510a-eng.htm. 6. Canadian Radio-television and Telecommunications Commission, Communications Monitoring Report (Ottawa: Canadian Radio-television and Telecommunications Commission, 2009). 7. Australian Bureau of Statistics, “Internet Activity, Australia, December 2009—Internet Subscribers by Access Connection, for ISPs with More Than 1,000 Active Subscribers,” http://tinyurl.com/abs-internet-activity. 8. Government of Australia, “21st Century Broadband,” http://www.dbcde .gov.au/__data/assets/word_doc/0017/112454/National_Broadband_Network_ policy_brochure.doc. 9. These data are available online at http://epp.eurostat.ec.europa.eu/portal/ page/portal/information_society/data/main_tables. Numbers reported here are for the EU15 countries. These countries report higher broadband penetration and higher usage rates than the EU25 or EU27 countries. 10. Council of the European Union, Council Resolution on the Implementation of the eEurope 2005 Action Plan (Brussels: European Union, 2005), 16. 11. Organisation for Economic Co-operation and Development, “Broadband Penetration, Historical Broadband Penetration Rates (December 2009),” http:// www.oecd.org/dataoecd/22/12/39574779.xls. 12. Organisation for Economic Co-operation and Development, “Percentage of Fibre Connections in Total Broadband among Countries Reporting Fibre Subscribers, December 2009,” http://www.oecd.org/dataoecd/21/58/ 39574845.xls. 13. OECD Directorate for Science Technology and Industry, Developments in Fibre Technologies and Investment (Paris: OECD Working Party on Communication Infrastructures and Services Policy, 2008). 14. Federal Communications Commission, Broadband Performance—OBI Technical Paper No. 4 (Washington, DC: Federal Communications Commission, 2010). 15. Epitiro, Australia Internet Performance Index—Summary Findings 2008 (Q4) (Sydney: Epitiro, 2009). 16. Ofcom, UK Fixed Broadband Speeds, November/December 2010—The Performance of Fixed-Line Broadband Delivered to UK Residential Consumers (London: Ofcom, 2011). 17. Caps are very common in Australia, Belgium, Canada, Iceland, and New Zealand. For data on monthly download caps and pricing for additional usage, see OECD, “Average Monthly Bit/Data Cap Size and Price Per Additional Mb, by Country, October 2009,” http://www.oecd.org/dataoecd/22/46/39575020.xls. 230

n ot e s to pag e s 9 – 1 2

18. A discussion of network neutrality is beyond the scope of this chapter. See the collection by Thomas M. Lenard and Randolph J. May, Net Neutrality or Net Neutering: Should Broadband Internet Services Be Regulated (New York: Springer, 2006) and the International Journal of Communication’s Net Neutrality issue (2007) for discussion of the key issues. 19. For a discussion of “user-generated content,” see OECD Directorate for Science Technology and Industry, Participative Web: User-Created Content (Paris: OECD, 2007). 20. ADSL refers to asymmetrical DSL (digital subscriber line) service, for instance. 21. Broadband Stakeholder Group, Predicting UK Future Residential Bandwidth Requirements (Cambridge: Broadband Stakeholder Group, 2006); California Broadband Task Force, The State of Connectivity—Building Innovation through Broadband (State of California, 2008); Communications Workers of America, Speed Matters: Affordable High-Speed Internet for All (Washington, DC: Communications Workers of America, 2007); Deloitte, National Broadband Network—A User’s Perspective (Sydney: Deloitte, 2009); Stephen Ezell et al., The Need for Speed: The Importance of Next-Generation Broadband Networks (Washington, DC: The Information Technology and Innovation Foundation, 2009). 22. Broadband Stakeholder Group, “Pipe Dreams?: Prospects for Next Generation Broadband Deployment in the UK,” http://www.broadbanduk.org/ content/view/236/7/; OECD Directorate for Science Technology and Industry, Developments in Fibre Technologies and Investment (Paris: OECD Working Party on Communication Infrastructures and Services Policy, 2008). 23. For a broader discussion of the characteristics required for broadband networks to meet the needs of the public, see Amelia Bryne Potter and Andrew Clement, “A Desiderata for Broadband Networks in the Public Interest,” (paper presented at the 35th Research Conference on Communication, Information and Internet Policy, Arlington, VA, 2007). 24. Over time, it may make more sense to develop individual broadband profiles, reflecting the increased uptake of mobile broadband, and the proliferation of access devices, both of which allow for more personalized individual broadband services. But in the short term, it is suggested that an individual’s primary broadband access point is the home, and as such the household likely provides the appropriate location for assessing broadband capacity. This assumption should be revisited periodically. 25. Sharon E. Gillett et al., Measuring Broadband’s Economic Impact: Final Report (Washington, DC: US Department of Commerce, Economic Development Administration, 2006), 3. 26. Brief reviews that cover this territory are provided by Middleton and colleagues to frame analysis of Canadian Internet usage data: Catherine A. Middleton and Jordan Leith, “Intensity of Internet Use in Canada: Exploring

n ot e s to pag e s 1 2 – 1 5

231

Canadians’ Engagement with the Internet” (paper presented at the Statistics Canada Socio-economic Conference, Ottawa, 2007); Catherine A. Middleton and Jonathan Ellison, Understanding Internet Usage among Broadband Households: A Study of Household Internet Use Survey Data (Ottawa: Statistics Canada—Science, Innovation and Electronic Information Division Working Papers, 2008); Catherine A. Middleton and Jordan Leith, “An Analysis of Canadians’ Scope of Internet Usage” (paper presented at the Statistics Canada Socio-economic Conference, Ottawa, 2008); Catherine A. Middleton, Ben Veenhof, and Jordan Leith, Intensity of Internet Use in Canada: Understanding Different Types of Users (Ottawa: Statistics Canada—Business Special Surveys and Technology Statistics Division Working Papers, 2010). 27. See for example Darin D Barney, The Network Society (Cambridge: Polity Press, 2004); Daniel Bell, The Coming of Post-Industrial Society: A Venture in Social Forecasting (New York: Basic Books, 1973); Manuel Castells, The Rise of the Network Society (Oxford; Malden, Mass.: Blackwell Publishers, 2000); Alistair S. Duff, David Craig, and David A. McNeill, “A Note on the Origins of the ‘Information Society,’ ” Journal of Information Science 22 (1996); Yoneji Masuda, The Information Society (Washington, DC: World Future Society, 1981); and Frank Webster, “Making Sense of the Information Age,” Information, Communication & Society 8 (2005). This is by no means an exhaustive list of research regarding the information society. The articles noted here are examples of this work, but there are many other important works beyond the ones noted here. 28. Manuel Castells, The Informational City: Information Technology, Economic Restructuring, and the Urban-Regional Process (Oxford and Cambridge, MA: B. Blackwell, 1989); William H. Dutton, Jay G. Blumler, and Kenneth L. Kraemer, eds., Wired Cities: Shaping the Future of Communications, (Communications Library) (Boston: G. K. Hall, 1987); Harold Sackman and Barry W Boehm, eds., Planning Community Information Utilities (Montvale, NJ: AFIPS Press,1972); Harold Sackman and Norman H. Nie, eds., The Information Utility and Social Choice (Montvale, NJ: AFIPS Press, 1970). 29. T. Randolph Beard, George S. Ford, and Lawrence J. Spiwak, “The Broadband Adoption Index: Improving Measurements and Comparisons of Broadband Deployment and Adoption,” (Phoenix Center Policy Paper Number 36), (Washington, DC: 2009); Kenneth Flamm et al., “Measuring Broadband: Improving Communications Policymaking through Better Data Collection,” Pew Internet Project, http://www.pewinternet.org/Reports/2007/MeasuringBroadband/01-Executive-Summary.aspx. Payam Hanafizadeh, Mohammad Reza Hanafizadeh, and Mohsen Khodabakhshi, “Extracting Core ICT Indicators Using Entropy Method,” The Information Society 25 (2009); Philip N. Howard et al., “Sizing up Information Societies: Toward a Better Metric for the Cultures of ICT Adoption,” The Information Society 25 (2009); Michel J. Menou and Richard D. Taylor, “A ‘Grand Challenge’: Measuring Information Societies,” The Information Society 22 (2006). 232

n ot e s to pag e 1 5

30. William H. Dutton et al., “The Politics of Information and Communication Policy: The Information Superhighway,” in Information and Communication Technologies: Visions and Realities, ed. William H. Dutton and Malcolm Peltu (Oxford: Oxford University Press, 1996); Milton Mueller and Becky Lentz, “Revitalizing Communication and Information Policy Research,” The Information Society 20 (2004); Audrey N. Selian, “The World Summit on the Information Society and Civil Society Participation,” The Information Society 20 (2004). 31. Examples include the European Commission, the Information Highway Advisory Council, the Information Society Technologies Advisory Group, the National Information Infrastructure Advisory Council, the World Bank, the World Summit on the Information Society, and UNCTAD. UNCTAD Secretariat, Information Economy Report 2006—The Development Perspective (New York and Geneva: United Nations Conference on Trade and Development, 2006); UNCTAD Secretariat, Information Economy Report 2007–2008—Science and Technology for Development: The New Paradigm of ICT; Commission of the European Communities, Digital Divide Forum Report: Broadband Access and Public Support in Under-Served Areas (Brussels: 2005); Commission of the European Communities, Europe’s Digital Competitiveness Report — Volume 1: i2010—Annual Information Society Report 2009. Benchmarking i2010: Trends and Main Achievements (Brussels: 2009), Information Society Technologies Advisory Group, “Shaping Europe’s Future through ICTs,” ftp://ftp.cordis.europa .eu/pub/ist/docs/istag-shaping-europe-future-ict-march-2006-en.pdf; National Information Infrastructure Advisory Council, Common Ground: Fundamental Principles for the National Information Infrastructure: First Report of the National Information Infrastructure Advisory Council (Washington, DC: National Information Infrastructure Advisory Council, 1995); World Bank, Information and Communications for Development 2009: Extending Reach and Increasing Impact (Washington, DC: World Bank, 2009); International Telecommunication Union, World Information Society Report (Geneva: International Telecommunication Union, 2006); International Telecommunication Union, World Information Society Report 2007—Beyond WSIS (Geneva: International Telecommunication Union, 2007). 32. European Commission, Europe and the Global Information Society: Recommendations of the High-Level Group on the Information Society to the Corfu European Council (Bangemann Group) (Brussels: European Commission, 1994), 10. 33. International Telecommunication Union, World Information Society Report. 34. European Commission, “Knowledge Society,” http://tinyurl.com/ ec-knowledge. 35. Kristin A. Van Gaasbeck, “A Rising Tide: Measuring the Economic Effects of Broadband Use across California,” Social Science Journal 45 (2008). n ot e s to pag e 1 5

233

36. Martin Fornefeld, Gilles Delaunay, and Dieter Elixmann, The Impact of Broadband on Growth and Productivity (Düsseldorf: MICUS Management Consulting GmbH, 2008). 37. Shane Greenstein and Ryan C. McDevitt, The Broadband Bonus: Accounting for Broadband Internet’s Impact on U.S. GDP (Chicago: Kellogg School of Management and Department of Economics, Northwestern University, 2009). They also note that their estimates are much more conservative than those produced earlier in the decade by Crandall and colleagues, for example, Robert Crandall, William Lehr, and Robert Litan, The Effects of Broadband Deployment on Output and Employment: A Cross-Sectional Analysis of U.S. Data (Washington, DC: Brookings Institution, 2007); Robert W Crandall and Charles L Jackson, The $500 Billion Opportunity: The Potential Economic Benefit of Widespread Diffusion of Broadband Internet Access (Washington, DC: Brookings Institution, 2001). Crandall later described the figures in the 2007 report as “grossly overestimated” and not relevant for predicting the impacts of proposed stimulus investment “Broadband Job Creation Figures Overestimated,” http://www.redorbit.com/ news/technology/1631755/broadband_job_creation_figures_overestimated/. 38. To be precise, it is actually the Internet connectivity that broadband enables that provides information (society) access, rather than broadband per se. Today, there are few information society type services available without Internet access, but as governments (for example) become more involved in developing next generation networks and services, it may be possible to engage in the information society without using the Internet. This may be an arcane observation, but it becomes relevant as governments try to increase the reach of the information society to people who are not interested in what the Internet offers. 39. An example that does explore the benefits of broadband at the community level is a recent report on rural broadband produced by the US Department of Agriculture’s Economic Research Service. Peter Stenberg et al., Broadband Internet’s Value for Rural America: Economic Research Report Number 78 (Washington, DC: United States Department of Agriculture, 2009). 40. For instance, Deloitte offers an extensive overview of applications and services that will be available to individuals using Australia’s National Broadband Network. Deloitte, National Broadband Network—A User’s Perspective. Ezell et al. also offer detailed descriptions of applications that will be possible with higher capacity broadband access. Ezell et al., The Need for Speed: The Importance of Next-Generation Broadband Networks. 41. Jed Kolko, Does Broadband Boost Local Economic Development? (San Francisco: Public Policy Institute of California, 2010), 2. 42. Ibid. 43. World Bank, Information and Communications for Development 2009: Extending Reach and Increasing Impact, 5. 44. These are questions used by the Canadian Internet Use Survey. Statistics Canada, “Canadian Internet Use Survey, 2009.” 234

n ot e s to pag e s 1 5 – 1 7

45. Allen’s study of the experience of connectivity for Australian Internet users is consistent with this objective. Matthew Allen, “The Experience of Connectivity—Results from a Survey of Australian Internet Users,” Information, Communication & Society 13 (2010). 46. International Telecommunication Union, Measuring the Information Society—The ICT Development Index (Geneva: International Telecommunication Union, 2009). 47. Soumitra Dutta and Irene Mia, Global Information Technology Report 2009–2010: ICT for Sustainability (Geneva: World Economic Forum and INSEAD, 2010). 48. Leonard Waverman and Kalyan Dasgupta, “How to Maximize the Economic Impact of Mobile Communications: The Four Waves,” in Global Information Technology Report 2008–2009, ed. Soumitra Dutta and Irene Mia (Geneva: World Economic Forum and INSEAD, 2009), 56. 49. Pepper et al., “From Mobility to Ubiquity: Ensuring the Power and Promise of Internet Connectivity . . . for Anyone, Anywhere, Anytime.” 50. Leonard Waverman, Kalyan Dasgupta, and Nicholas Brooks, Connectivity Scorecard 2009 (2009); Leonard Waverman and Kalyan Dasgupta, Connectivity Scorecard 2010 (2010). 51. Pepper et al., “From Mobility to Ubiquity: Ensuring the Power and Promise of Internet Connectivity . . . for Anyone, Anywhere, Anytime,” 39. 52. For examples of various critiques regarding these data, see George S. Ford, Thomas M. Koutsky, and Lawrence J. Spiwak, The Broadband Performance Index: A Policy-Relevant Method of Comparing Broadband Adoption among Countries (Phoenix Center Policy Paper Number 29) (Washington, DC: 2007); Market Clarity, Broadband Wars: The OECD’s International Broadband Arms Race (Sydney: 2007); and Suzanne Blackwell, “Canada’s Broadband Price Per Mbps—GIGO,” http://www.giganomics.ca/observations-old/2009/8/20/ canadas-broadband-price-per-mbps-gigo.html. 53. Nielsen, The Global Online Media Landscape—Identifying Opportunities in a Challenging Market (New York: 2009). 54. Statistics Canada, “Canadian Internet Use Survey, 2009.” 55. Jed Kolko, “How Broadband Changes Online and Offline Behaviors,” Information Economics and Policy 22 (2010), 152. 56. Michael Gurstein, “Effective Use: A Community Informatics Strategy Beyond the Digital Divide,” First Monday 8 (2003); Michael Gurstein, “Effective Use and the Community Informatics Sector: Some Thoughts on Canada’s Approach to Community Technology/Community Access,” in Seeking Convergence in Policy and Practice, ed. Marita Moll and Leslie Regan Shade (Ottawa: Canadian Centre for Policy Alternatives, 2004). 57. See for example Eszter Hargittai and Steven Shafer, “Differences in Actual and Perceived Online Skills: The Role of Gender,” Social Science Quarterly 87 (2006); Eszter Hargittai and Amanda Hinnant, “Digital Inequality: Differences

n ot e s to pag e s 1 7 – 2 1

235

in Young Adults’ Use of the Internet,” Communication Research 35 (2008); Eszter Hargittai and Gina Walejko, “The Participation Divide: Content Creation and Sharing in the Digital Age,” Information, Communication & Society 11 (2008). 58. Middleton and Leith, “Intensity of Internet Use in Canada: Exploring Canadians’ Engagement with the Internet”; Middleton and Ellison, “Understanding Internet Usage among Broadband Households: A Study of Household Internet Use Survey Data”; Middleton and Leith, “An Analysis of Canadians’ Scope of Internet Usage”; Middleton, Veenhof, and Leith, Intensity of Internet Use in Canada: Understanding Different Types of Users.

2. understanding digital gaps: a quartet of empirical methodologies Bin Zhang and Richard D. Taylor

1. This chapter is a condensation of a much longer paper of the same title presented at the Beyond Broadband Access workshop, which includes extensive tables showing the underlying data and calculations. That paper is available on request from the authors.

3. broadband microfoundations: the need for traffic data Steven Bauer, David Clark, and William Lehr

1. For a discussion of congestion in the Internet, see our companion paper: S. Bauer, D. Clark, and W. Lehr, “The Evolution of Internet Congestion,” prepared for the 37th Research Conference on Communication, Information and Internet Policy, Arlington, VA, September 2009. 2. D. Jonsson, S. Berglund, P. Almström, and S. Algers, “The Usefulness of Transport Models in Swedish Planning Practice,” Transport Reviews: A Transnational Transdisciplinary Journal 31 (2011): 251–265. 3. For an introduction to network management, see M. Subramanian, Network Management: Principles and Practice (Reading, MA: Addison Wesley, 1999). 4. Over 10,000 comments were filed in the FCC’s proceeding (07-52) regarding Comcast’s management of peer-to-peer traffic (see http://fjallfoss.fcc.gov/ prod/ecfs/comsrch_v2.cgi). 5. We put “fair” in quotes because there is an on-going debate in the technical community as to what constitutes fair allocations of network traffic. See for example B. Briscoe, “Flow Rate Fairness: Dismantling a Religion,” SIGCOMM Computer Communication Review 37 (2007): 63–74; S. Floyd and M. Allman, “RFC 5290: Comments on the Usefulness of Simple Best-Effort Traffic,” Network Working Group, Internet Engineering Task Force, July 2008, http:// www.faqs.org/rfcs/rfc5290.html.

236

n ot e s to pag e s 2 1 – 5 3

6. For example, since public funds are used to build roads and bridges, and there is a strong interest in maintaining transparency in government processes, the need to collect and publish relevant information and statistics is well established and (relatively) straightforward. In contrast, where private investment is involved, ensuring adequate public disclosure of relevant information is more complex. 7. Netflow is the common name for this record type, but it has been standardized now as Internet Protocol Flow Information eXport (see http://en.wikipedia .org/wiki/IPFIX). 8. Different types of traffic have different profiles in terms of upstream/downstream bit rates, tolerance for delay, jitter, or bit error losses, and amenability to being multicast. Knowing the mix of applications facilitates planning to ensure an appropriate quality of service for the applications that are expected. 9. These techniques are imperfect because the traffic signatures change as new applications are introduced and because they are based on sampling techniques that are subject to stochastic measurement error. 10. K. Cho, K. Fukuda, H. Esaki, and A. Kato, “Observing Slow Crustal Movement in Residential User Traffic,” in Proceedings of the 2008 ACM CoNEXT Conference, Madrid, Spain, December 2008; K. Cho, K. Fukuda, H. Esaki, and A. Kato, “The Impact and Implications of the Growth in Residential User-to-User Traffic,” in Proceedings of the 2006 Conference on Applications, Technologies, Architectures, and Protocols for Computer Communications, September 2006. 11. Cho et al. (2008). 12. See http://www.dtc.umn.edu/mints/home.php. 13. See Minnesota Internet Traffic Studies (MINTS) at http://www.dtc.umn .edu/mints/home.php, for data on traffic growth rates. 14. For example, see “Net Traffic Doubling Every Six Months,” a report from August 2001, http://www.theregister.co.uk/2001/08/17/net_traffic_doubling_ every_six/. 15. Cho, et al. (2008). 16. In addition to distinct differences in the usage patterns of different types of users, there may be different numbers of each type; and they may be distributed differently across a network in ways that may be related to what they are doing (e.g., different on-net/off-net patterns) with resulting implications for aggregate traffic flows. 17. This is not to say there are not a number of researchers addressing this challenge. For example, in addition to the work by Odlyzko/MINTS and Cho/ Japanese ISPs, see the work of Caida at http://www.caida.org/publications/ papers/2009/aims_report/aims_report.xml#topten. We are aware of a number of other projects and suspect there are many more we are unaware of being undertaken at universities and in industry labs (e.g., Cable Labs, AT&T Labs) across the United States and abroad.

n ot e s to pag e s 5 3 – 6 2

237

18. Most subscriber traffic is a mix of on-net (i.e., traffic that originates and terminates on the access ISPs network) and off-net (i.e., traffic that either originates or terminates on another ISP’s network). Because ISPs generally lack detailed insight into traffic conditions on the networks of other ISPs, better pooled traffic data may allow ISPs a more complete understanding of the factors driving local phenomena on their own networks (e.g., separating local from general trends) as well as conditions in the wider Internet. 19. “Internet Captivity and the De-peering Menace,” http://www.renesys .com/tech/presentations/pdf/nanog-45-Internet-Peering.pdf. 20. “Deja Vu All Over Again: Cables Cut in the Mediterranean,” http:// www.renesys.com/blog/2008/12/deja-vu-all-over-again-cables.shtml#more. 21. “The Day the YouTube Died: What Happened and What We Might Do About It,” http://www.renesys.com/tech/presentations/pdf/nanog43-hijack.pdf. 22. Akami’s “Net Usage Index for News” enables users to monitor global news consumption 24 x 7, seeing in real time the impact of current events on online media consumption. http://www.akamai.com/html/technology/nui/news/ index.html 23. Conficker/Conflicker/Downadup worm as seen from the UCSD Network Telescope http://www.caida.org/research/security/ms08-067/conficker.xml. 24. For example, regulatory authorities have initiated proceedings to examine ISP traffic management practices (e.g., for Canada see http://www.crtc.gc.ca/ENG/ archive/2008/pt2008-19.htm, November 2008; and in the United States see http:// hraunfoss.fcc.gov/edocs_public/attachmatch/DA-08-92A1.pdf, January 2008). 25. For further discussion of why these issues are contentious, see Bauer, Clark, and Lehr (2009). 26. Broadband Working Group, “Broadband Incentive Problem,” a white paper by the MIT Communications Futures Program, September 2005, http:// cfp.mit.edu/publications/CFP_Papers/Incentive_Whitepaper_09-28-05.pdf. 27. See for example H. Varian, R. Litan, A. Elder, and J. Shutter, “The Net Impact Study: The Projected Economic Benefits of the Internet in the United States, United Kingdom, France, and Germany,” research report, funding support from Cisco Systems is acknowledged, January 2002; W. Lehr, C. Osorio, S. Gillett, and M. Sirbu, “Measuring Broadband’s Economic Impact,” paper prepared at the Telecommunications Policy Research Conference, Arlington, VA, September 2005; S. Greenstein and R. McDevitt, “The Broadband Bonus: Accounting for Broadband Internet’s Impact on U.S. GDP,” white paper, Technology Policy Institute, Washington, D.C., January 2009, http://www .techpolicyinstitute.org/files/greenstein-broadband bonus1.pdf; M. Dutz, J. Orszag, and R. Willig, “The Substantial Consumer Benefits of Broadband Connectivity for U.S. Households,” Compass Lexecon, A Study Commissioned by the Internet Innovation Institute, July 2009. 28. For example, the American Recovery and Reinvestment Act of 2009 (Pub. L. 111–5, 123 Stat. 115, 2009) has targeted US$7.2 billion in public funding for the 238

n ot e s to pag e s 6 2 – 6 4

promotion of broadband. Most of the data that is available regarding usage patterns is highly aggregated over geographic regions, time periods, users, and applications in ways that obscures many potentially interesting details. 29. See http://www.connectivityscorecard.org/images/uploads/media/ TheConnectivityReport2009.pdf. 30. In Cho’s Japanese study, the challenges were described as mainly political not technical. See Cho et al. (2008); Cho et al. (2006). 31. To economize on data storage costs, raw data is summarized in real time. 32. DPI refers to Deep Packet Inspection. Equipment and software are available from a variety of vendors that allows detailed inspection of the contents of packets, not just the header-information which must be examined in order to forward the packets toward their destination. Looking inside the packets for additional information about what the application is or other traffic identifier information is referred to as DPI. 33. This is information that is not available in the Japanese studies described above. 34. Bauer, Clark, and Lehr (2009). 35. See Figure 3-8, noting in particular Japan, which offers 160 Mbps broadband. This particular service is offered at 6,000 yen ($60, according to a 2009 New York Times blog, http://bits.blogs.nytimes.com/2009/04/03/the-cost-tooffer-the-worlds-fastest-broadband-20-per-home/) and is a DOCSIS 3.0 service offered over a cable network. What is interesting is that if four channels are bonded together the maximum synchronization speed of a subscriber’s modem and the CMTS is around 170 Mbps but the maximum usable speed is 152 Mbps (i.e., less than the advertised speed). 36. For a sample, see J. Camp, Trust and Risk in Internet Commerce (Cambridge, MA: MIT Press, 2001), Wik-Consult/Rand Europe/CLIP/CRID/GLOCOM, Comparisons of Privacy and Trust Policies in the Area of Electronic Communications, Final Report, report prepared for the European Commission, July 2007, http:// ec.europa.eu/information_society/policy/ecomm/doc/library/ext_studies/privacy_ trust_policies/final_report_20_07_07_pdf.pdf; P. Ohm, “The Rise and Fall of Invasive ISP Surveillance,” University of Illinois Law Review 2009, 1417. Links to current policy debates and further research are available at Electronic Privacy Information Center (http://epic.org/); privacy.org (http://privacy.org/); and Electronic Frontier Foundation (http://www.eff.org/issues/privacy). 37. ISPs do face regulatory rules on the disclosure of subscriber data, but these rules vary by context and are not comprehensive. 38. In today’s Internet blogosphere rumors of bad behavior can be damaging. 39. The IRB is a review board established by MIT and with similar boards at other universities with representation from across the community. 40. These included the Milgram Obedience to Authority experiments of 1988 and the Public Health Service Syphilis Study (1932–1972), more commonly referred to as the “Tuskegee Syphilis” experiments.

n ot e s to pag e s 6 4 – 6 7

239

41. See for example J. Tirole, The Theory of Industrial Organization (Cambridge, MA: MIT Press, 1988). 42. Government-mandated data collection of detailed data comes with strong nondisclosure obligations. 43. Internet Engineering Task Force, http://www.ietf.org/. 44. The North American Operators Group, http://www.merit.edu/nanog. 45. CSTB, Looking Over the Fence at Networks: A Neighbor’s View of Networking Research (National Academy Press, 2001).

4. adoption factors of ubiquitous broadband Sangwon Lee and Justin S. Brown

1. International Telecommunication Union, Digital.life (Geneva: ITU, 2006). 2. Robert W. Crandall, “Broadband Communications” in The Handbook of Telecommunications Vol. 2.: Technology Evolution and the Internet, ed. Sumit Kumar Majumdar, Martin Cave, and Ingo Vogelsang (Amsterdam: Elsevier, 2005), 156–190; Martin Fransman, ed., Global Broadband Battles; Why the U.S. and Europe Lag While Asia Leads (Stanford, CA: Stanford University Press, 2006). 1–58. 3. International Telecommunication Union, The Internet of Things (Geneva: ITU, 2005); International Telecommunication Union, Digital.life. 4. Organization for Economic Cooperation and Development, Mobile Multiple Play (Paris: OECD, 2007). 5. Organization for Economic Cooperation and Development, Communication Outlook 2009 (Paris: OECD, 2009). 6. Ibid. 7. Ibid. 8. Marc Weiser, “The Computer for the 21st Century,” Scientific American (September 1991): 94–100. 9. International Telecommunication Union, Digital.life. 10. Ibid. 11. Ibid. 12. Jeffrey Church and Neil Gandal, “Platform Competition in Telecommunications,” in The Handbook of Telecommunications Vol. 2.: Technology Evolution and the Internet, ed. Sumit Kumar Majumdar, Martin Cave, and Ingo Vogelsang (Amsterdam: Elsevier, 2005), 119–154. 13. Ibid. 14. International Telecommunication Union, Promoting Broadband (Geneva: ITU, 2003). 15. David E. Burnstein and Debra J. Aron, “Broadband Adoption in the United States: An Empirical Analysis,” in Down to the Wire: Studies in the Diffusion and Regulation of Telecommunications Technologies, ed. Allan L. Shampine 240

n ot e s to pag e s 6 7 – 7 2

(Hauppauge, NY: Nova Science Publishers, 2003), 119–38; Mario Denni and Harald Gruber, “The Diffusion of Broadband Telecommunications: The Role of Competition,” paper presented at International Telecommunication Conference, Pontevedra, 2005; Walter Distaso, Paolo Lupi, and Fabio M. Maneti, “Platform Competition and Broadband Uptake: Theory and Empirical Evidence from the European Union,” Information Economics and Policy 18, no. 1 (2006): 87–106; Sangwon Lee and Justin S. Brown, “Examining Broadband Adoption Factors: An Empirical Analysis between Countries,” Info 10 (2008): 25–39. 16. International Telecommunication Union, Promoting Broadband. 17. Joshua S. Gans, Stephen P. King, and Julian Wright, “Wireless Communications,” in The Handbook of Telecommunications Vol. 2.: Technology Evolution and the Internet, ed. Sumit Kumar Majumdar, Martin Cave, and Ingo Vogelsang (Amsterdam: Elsevier, 2005), 243–287. 18. Harold Gruber and Frank Verboven, “The Evolution of Markets under Entry and Standards Regulation: The Case of Global Mobile Telecommunications,” International Journal of Industrial Organization 19 (2001): 1189– 1212; Church and Gandal, “Platform Competition in Telecommunications,” 119–154. 19. Veneta Andonova, “Mobile Phones: the Internet and the Institutional Environment,” Telecommunications Policy 30 (2006): 29–45. 20. Inmaculada Cava-Ferreruela and Antonio Alabau- Munoz, ˘ “Broadband Policy Assessment: A Cross-National Empirical Analysis,” Telecommunications Policy 30 (2006): 445–463. 21. Lee and Brown, “Examining Broadband Adoption Factors,” 25–39. 22. Ibid. 23. John de Ridder, “Catching-up in Broadband: What will It Take?” paper presented at the 35th Research Conference on Communication, Information and Internet Policy, Arlington, VA, 2007; Robert D. Atkinson, Daniel K. Correa, and Julie A. Hedlund, Explaining International Broadband Leadership (Washington, DC: The Information Technology and Innovation Foundation, 2008). 24. International Telecommunication Union, Digital.life. 25. Fransman, Broadband Battles, 1–58. 26. International Telecommunication Union, Promoting Broadband; International Telecommunication Union, Digital.life. 27. Marcelo Grosso, “Determinants of Broadband Penetration in OECD Nations,” paper presented at the Australian Communications Policy and Research Forum, 2006. 28. Gary Madden, Grant Coble-Neal, and Brian Dalzell, “A Dynamic Model of Mobile Telephony Subscription Incorporating a Network Effect,” Telecommunications Policy 28 (2004): 133–144. 29. Hyungtaik Ahn and Myenong-Ho Lee, “An Econometric Analysis of the Demand for Access to Mobile Telephone Networks,” Information Economics and Policy 11 (1999): 297–305. n ot e s to pag e s 7 2 – 7 4

241

30. Madden, Coble-Neal, and Dalzell, 133–144. 31. Andonova, 29–45. 32. Anindya Chaudhuri, Kenneth Flamm, and John Horrigan, “An Analysis of the Determinants of Internet Access,” Telecommunications Policy 29 (2005): 731–55; Peter Trkman, Borka Blazic, and Tomas Turk, “Factors of Broadband Development and the Design of a Strategic Policy Framework,” Telecommunications Policy 32 (2008): 101–15. 33. John Horrigan and Aron Smith, Home Broadband Adoption 2007 (Washington, DC: Pew Research Center, 2007). 34. Martha A. Garcia-Murillo, “International Broadband Deployment: The Impact of Unbundling,” Communications & Strategies 57 (2005): 83–108; Lee and Brown, “Examining Broadband Adoption Factors,” 25–39. 35. Garcia-Murillo, 83–108. 36. Trkman, Blazic, and Turk, 101–115. 37. de Ridder; Atkinson, Correa, and Hedlund. 38. de Ridder. 39. Trkman, Blazic, and Turk, 101–115.; Lee and Brown, “Examining Broadband Adoption Factors,” 25–39. 40. Denni and Gruber. 41. Trkman, Blazic and Turk, 101–115. 42. International Telecommunication Union, Digital.life. 43. Andonova, 29–45. 44. International Telecommunication Union, Digital.life. 45. Harold Gruber, “Competition and Innovation: The Diffusion of Mobile Telecommunications in Central and Eastern Europe,” Information Economics and Policy 13 (2001): 19–34; Jukka Liikanen, Paul Stoneman, and Otto Toivanen, “Intergenerational Effects in the Diffusion of New Technology: The Case of Mobile Phones,” International Journal of Industrial Organization 22 (2004): 1137–1154. 46. A goal of this empirical model was to examine effective policy variables, which might influence a higher level of ubiquitous broadband diffusion. Both network competition variable and platform competition-standardization policy type variable could be included in a single model. However, network competition variable and some of the platform competition-standardization policy type were correlated with each other. Therefore we examine two different models: the model with the network competition variable and the model with the platform competition-standardization policy variable. 47. Sangwon Lee and Justin Brown, “A Cross-country Analysis of Fixed Broadband Deployment: Examination of Adoption Factors and Network Effect,” paper presented at the 2009 Association for Education in Journalism and Mass Communication Annual Convention, Boston, MA, 2009. 48. Eli M. Noam, “The Next Frontier for Openness: Wireless Communications,” in Competition for the Mobile Internet, ed. Eli M. Noam and Dan Steinbock (Norwell, MA: Kluwer Academic Publishers, 2003), 21–37. 242

n ot e s to pag e s 7 4 – 8 6

49. Church and Gandal, 119–154. 50. Sangwon Lee, Sylvia M. Chan-Olmsted, and Heejung Kim, “The Deployment of Third-Generation Mobile Services: A Multinational Analysis of Contributing Factors,” paper presented at the 2007 Association for Education in Journalism and Mass Communication Annual Convention, Washington DC, 2007. 51. Lee and Brown,” Examining Broadband Adoption Factors,” 25–39. 52. de Ridder, “Catching-up in broadband: What will it take?” Paper presented at the 35th Research Conference on Communication, Information and Internet Policy, Arlington, VA, 2007.

5. data and modeling challenges in international comparisons Johannes M. Bauer and Sungjoong Kim

1. See, for example, Berkman Center, Next Generation Connectivity: A Review of Broadband Internet Transitions and Policy from Around the World (Cambridge, MA: Berkman Center for Internet and Society, Harvard University, 2010). 2. International Telecommunication Union, World Information Society Report 2007 (Geneva: International Telecommunication Union, 2007). 3. Economist Intelligence Unit, E-Readiness Rankings 2009 (London: Economist Intelligence Unit, 2009). In 2010, the name was changed to Digital Economy Rankings but so far only one year of data is available. 4. World Economic Forum, The Networked Readiness Index (Davos, Switzerland: World Economic Forum, published annually). 5. L. Waverman and K. Dasgupta. Connectivity Scorecard 2011 (London: LECG and NokiaSiemens, 2009), http://www.connectivityscorecard.org/ images/uploads/media/TheConnectivityReport2011.pdf. 6. J. M. Bauer, “Learning from Each Other: Promises and Pitfalls of Benchmarking in Communications Policy,” Info, The Journal of Policy, Regulation and Strategy for Telecommunications, Information and Media 12 (2010): 8–20. 7. E. Ostrom, Understanding Institutional Diversity (Princeton, NJ, and Oxford: Princeton University Press, 2005); D. Rodrik, One Economics Many Recipes: Globalization, Institutions, and Economic Growth (Princeton, NJ: Princeton University Press, 2007). 8. For an earlier discussion, see P. Xavier, “Price Discount Schemes and InterCountry Comparisons of Telecommunication Tariffs,” Telecommunications Policy 22 (1998): 289–314. 9. B. A. Cherry and J. M. Bauer, “Institutional Arrangements and Rate Rebalancing: Evidence from the United States and Europe,” Information Economics and Policy 14 (2002): 495–517. 10. A. Sen, Development as Freedom (New York: Knopf, 1999). n ot e s to pag e s 8 6 – 9 3

243

11. J. M. Bauer, S. Kim, and T. Koh, “International Comparisons of Mobile Voice and Data Prices: Coping with Heterogeneity and Differentiation,” unpublished manuscript, East Lansing, Michigan: Michigan State University, 2010. 12. J. Triplett, Handbook on Hedonic Indexes and Quality Adjustments in Price Indexes: Special Application to Information Technology Products (Paris: OECD, 2006). 13. For a discussion of these issues in fixed broadband service, see S. Greenstein and R. McDevitt, “Evidence of a Modest Price Decline in US Broadband Services,” Working Paper #0102, Northwestern University: The Center for the Study of Industrial Organization. 14. One simple way is to assume that outgoing and incoming calls are equally distributed. In a calling-party-pays (CPP) environment, one hundred minutes of outgoing and one hundred minutes of incoming calls cost a user the equivalent of the one hundred outgoing calls. In a mobile-party-pays (MPP) environment, the subscriber has to pay for two hundred minutes as both outgoing and incoming calls are counted. To correct for this difference, one hundred CPP minutes must be equated to two hundred MPP minutes in determining LEF. 15. Purchasing power parity is an economic technique employed in order to determine the relative values of two or more number of currencies. It is necessary because the amount of goods that a currency can purchase differs substantially depending on various factors including availability of goods, demand for the goods, and inflation. 16. World Bank, World Development Indicators, http://data.worldbank.org/ data-catalog/world-development-indicators. 17. Organization for Economic Cooperation and Development, Communications Outlook 2009 (Paris: OECD, 2009), 276–277. 18. CTIA, “The United States and World Wireless Markets,” Written ex parte communication, RM-11361; GN Docket No. 09–51; WC Docket No. 07–52. Washington, DC: CTIA—The Wireless Association, 2009. 19. Bauer, “Learning from Each Other.” 20. H. Sawhney, “Dynamics of Infrastructure Development: The Role of Metaphors, Political Will and Sunk Investment,” Media, Culture and Society 23 (2001): 33–51. 21. M. Fransman, ed., Global Broadband Battles: Why the U.S. and Europe Lag While Asia Leads (Palo Alto, CA: Stanford University Press, 2006). 22. C. C. Ragin, Fuzzy-set Social Science (Chicago, IL: University of Chicago Press, 2000). 23. Ibid. 24. See, for example, J. M. Bauer, J. Kim, and S. S. Wildman, “Regulation and the Diffusion of Broadband in the OECD,” paper presented at the University of Florida–London Business School joint conference, 2005. 25. J. D. Sterman, Business Dynamics: Systems Thinking and Modeling for a Complex World (Boston, MA: McGraw-Hill, 2000). 244

n ot e s to pag e s 9 5 – 1 0 1

26. K. DeMaagd and J. M. Bauer, “Modeling the Dynamic Interactions of Agents in the Provision of Network Infrastructure,” Information Systems Frontiers, DOI 10.1007/s10796-010-9244-2, 2010.

6. data, policy, and democracy Jorge Reina Schement

1. A. Hamilton, “Objections to the Proposed Constitution from Extent of Territory Answered,” federalist No. 14, November 30, 1787, in A. Hamilton et al., Federalist Papers (New York, Bantam Books, 1982). 2. M. Luo, “Frustration and Despair as Job Search Drags On,” New York Times, July 17, 2010, http://www.nytimes.com/2010/07/18/us/18unemployed .html?scp=1&sq=cable%20bill%20unemployed%20&st=cse. 3. In this chapter, I concentrate on access to broadband, rather than all Internet access, which also includes dial-up, on the premise that full functionality in access requires broadband. 4. J. R. Schement, “Broadband, Internet, and Universal Service: Challenges to the social contract of the twenty-first century,” in . . . And Communications for All: A Policy Agenda for a New Administration, ed. A. Schejter (Lanham, MD: Lexington Books, 2009), 3–27. 5. F. Bar and A. M. Riis, “Tapping User-Driven Innovation: A New Rationale for Universal Service,” The Information Society 16 (2000): 99–108; K. Jayakar and H. Sawhney, “Universal Access in the Information Economy: Tracking Policy Innovations Abroad,” Benton Foundation Universal Service Report, 2007, http://benton.org/node/5598. 6. Organization for Economic Cooperation and Development, “OECD Broadband Portal,” June 2010, http://www.oecd.org/document/54/0,3343, en_2649_34225_38690102_1_1_1_1,00.html. 7. Federal Communications Commission, Universal Service, undated. 8. J. R. Schement and M. Tate, Rural America in the Digital Age (Columbia, MO: Rural Policy Research Institute, 2003); L. Kvasny et al., “Communities, Learning and Democracy in the Digital Age,” The Journal of Community Informatics 2 (2006), online. 9. Ibid. 10. J. Horrigan, “Home Broadband Adoption,” Pew Internet American Life Project, 2009, http://www.pewinternet.org/~/media/Files/Reports/2009/ Home-Broadband-Adoption-2009.pdf. 11. J. R. Schement, “Telephone Penetration at the Margins: An Analysis of the Period Following the Breakup of AT&T, 1984–1994,” in Progress in Communication Sciences, Vol. XV: Advances in Telecommunications, ed. H. Sawhney and G. A. Barnett (Stamford, CT: Ablex, 1998), 187–215. 12. Government Printing Office, Series H 878–893, R 93–105, R 1–12, Historical Statistics of the United States, Colonial Times to 1970 (Washington, DC: GPO, 1975). n ot e s to pag e s 1 0 1 – 8

245

13. J. Horrigan, “Closing the Broadband Divide,” Pew Internet & American Life Project, August 2007, http://www.pewinternet.org/Reports/2007/ Closing-the-Broadband-Divide/Why-it-will-Be-Hard-to-Close-the-BroadbandDivide.aspx?r=1. 14. S. Strover, G. Chapman, and J. Waters, “Beyond Community Networking and CTC’s: Access Development and Public Policy,” Telecommunications Policy 28 (2004): 465–485. 15. I contributed portions of this chapter on my sister’s ranch north of Ridgway, Colorado. At the end of a long and winding dirt road, the ranch lies nestled against the San Juan Mountains beyond the reach of wireline broadband service. My efforts to connect to the Internet via satellite uplink proved slow and frustrating. Downloading a web page required my willingness to wait up to fifteen minutes for a complete screen; email was equally slow and erratic, and atmospheric fluctuations were noticeable. Dial-up service is available, but offers no advantage. Consequently, business at the ranch is conducted almost exclusively by voice over the telephone. 16. Open Cape, “Creating Regional Broadband Opportunities,” undated, http://www.opencape.com/the-project. 17. Connect Ohio, “Rural Community Building,” undated, http:// ruralcommunitybuilding.fb.org/2010/06/07/grassroots-efforts-bring-broadband-to-ohiorural-areas/. 18. United States Department of Agriculture, Connecting Rural America, USDA Broadband Initiative Program (Washington, DC: USDA, 2010). 19. Bar and Riis. 20. Jayakar and Sawhney. 21. Federal Communications Commission, National Broadband Plan, April 2010, http://www.broadband.gov/plan/executive-summary/, xi. 22. Ibid. 23. Ibid., 23. 24. Schement, 2009. 25. Bar and Riis, 106. 26. Organization for Economic Cooperation and Development, June 2010.

7. “rulers of thousands, hundreds, fifties, and tens”: does democracy count? Amit M. Schejter

Note: The title of this chapter is taken from Exodus 18:25. 1. A. Broache and D. McCullagh, “Technology Voters’ Guide: Barack Obama,” CNET, January 2, 2008, http://tinyurl.com/broache-tech-voter-guide. 2. The White House, “Promoting Innovation and Competitiveness: President Bush’s Technology Agenda,” June 24, 2004, http://georgewbush-whitehouse .archives.gov/infocus/technology/.

246

n ot e s to pag e s 1 0 9 – 1 3

3. Ibid. 4. R. Frieden, “Lessons from Broadband Development in Canada, Japan, Korea and the United States,” Telecommunications Policy 29 (2005): 595–613. 5. A. Schejter,“’From all my teachers I have grown wise, and from my students more than anyone else’: What Lessons Can the US Learn from Broadband Policies in Europe?” International Communication Gazette 71 (2009): 429–445. 6. S. Meinrath, “2008 OECD Broadband Statistics. How’s the US Doing? Survey Says—Stagnation and Decline,” May 20, 2008, http://www.saschameinrath .com/2008/may/20/2008_oecd_broadband_statistics_hows_us_doing_survey_ says_stagnation_and_decline. 7. A. Schejter, “International Benchmarks: The Crisis in U.S. Communications Policy Through a Comparative Lens,” in . . . and Communications for All: A Policy Agenda for a New Administration, ed. A. Schejter (Lanham, MD: Lexington Books, 2009), 57. 8. N. Pace, “China Surpasses U.S. in Internet Use,” Forbes, April 3, 2006, http://www.forbes.com/2006/03/31/china-internet-usage-cx_nwp_0403china .html. 9. Ibid. 10. S. Jones, CyberSociety: Computer-Mediated Communication and Community (Thousand Oaks, CA, London, and New Delhi: Sage Publications, 1995), 2. 11. M. Castells, The Rise of the Network Society, 2nd ed. (Chichester, UK: Wiley-Blackwell, 2010); J. Van Dijk, The Network Society, 2nd ed. (London, Thousand Oaks, CA, and New Delhi: Sage Publications, 2006). 12. J. Martin, The Wired Society: A Challenge for Tomorrow (Englewood Cliffs, NJ: Prentice Hall, 1978). 13. E. Katz, “Communications Research since Lazarsfeld,” Public Opinion Quarterly 51 (1987): S25–S45. 14. A. Schejter and M. Yemini, “ ‘Justice, and only justice, you shall pursue:’ Network Neutrality, the First Amendment and John Rawls’ Theory of Justice,” Michigan Telecommunications and Technology Law Review 14 (1992): 137–174. 15. M. Castells, “Communication, Power and Counter-power in the Network Society,” International Journal of Communication 1 (2007): 239. 16. Ibid. 17. Tricycle, “Mitchell Kapor on Dharma, Democracy, and the Information Superhighway,” 1994, http://www.kapor.com/writing/Tricycle-interview.htm. 18. K. Raman, “Cyberdemocracy and the Public Sphere: Chat Rooms as Agoras of the Modern World?” in Philosophical Perspectives on Globalization, ed. P. Cam, R. Ibana, and P. Duc (Korean National Commission for UNESCO, 2006), 77. 19. Y. Breindl, “Critique of the Democratic Potentialities of the Internet: A Review of Current Theory and Practice,” tripleC 8 (2010): 43–59. n ot e s to pag e s 1 1 3 – 1 7

247

20. P. Aufderheide, “Cable Television and the Public Interest,” Journal of Communication 42 (1992): 52–65. Aufderheide’s statement, though, referred to cable television. 21. Schejter and Yemini. 22. F. Siebert, T. Peterson, and W. Schramm, Four Theories of The Press: The Authoritarian, Libertarian, Social Responsibility, and Soviet Communist Concepts of What the Press Should Be and Do (Urbana, IL: University of Illinois Press, 1956). 23. J. Curran and J. Seaton, Power Without Responsibility: The Press, Broadcasting and New Media in Britain, 6th ed. (London: Routledge, 2003). 24. A. Schejter and S. Lee, “The evolution of cable regulatory policies and their impact: A comparison of South Korea and Israel,” Journal of Media Economics 20 (2007): 1–28. 25. Ibid. 26. P. Napoli, “The Marketplace of Ideas Metaphor in Communications Regulation,” Journal of Communication 49 (1999): 151–169. 27. J. Blumler, The Role of Public Policy in the New Television Marketplace (Washington, DC: Benton Foundation, 1989). 28. R. Entman and S. Wildman, “Reconciling Economic and Non-Economic Perspectives on Media Policy: Transcending the ‘Marketplace of Ideas,’ ” Journal of Communication 42 (1992): 5–19. 29. A. Schejter, “Jacob’s Voice, Esau’s Hands: Transparency as a First Amendment Right in an Age of Deceit and Impersonation,” Hofstra Law Review 35 (2007): 1489–1518. 30. See, for example, W. Hirczy, “The Impact of Mandatory Voting Laws on Turnout: A Quasi-Experimental Approach,” Electoral Studies 13 (1994): 64–76; L. Hill, “Compulsory Voting: Residual Problems and Potential Solutions,” Australian Journal of Political Science 37 (2002): 437–455. 31. R. Alvarez, T. Hall and M. Llewellyn, “On American Voter Confidence,” University of Arkansas at Little Rock Law Review 29 (2007): 651–668. 32. Schejter and Yemini, 170. 33. J. Balkin, “Digital Speech and Democratic Culture: A Theory of Freedom of Expression for the Information Society,” New York University Law Review 79 (2004): 1–55. 34. N. Garnham, “Universal Service,” in Telecom Reform: Principles, Policies and Regulatory Practices, ed. W. Melody (Lyngby, Denmark: Technical University of Denmark, 2001), 199. 35. Ibid. 36. K. Hacker, “Missing Links in the Evolution of Electronic Democratization,” Media, Culture & Society 18 (1996): 213–232; K. Brant, M. Huizenga, and R. van Meerten, “The New Canals of Amsterdam: An Exercise in Local Electronic Democracy,” Media, Culture & Society 18 (1996): 233–247. 37. Schejter and Yemini. 248

n ot e s to pag e s 1 1 7 – 2 5

38. J. Downey, “Surveillance from Below: The Internet and the Intifada,” in Ideologies of the Internet, ed. K. Sarikakis and D. Thussu (Cresskill, NJ: Hampton Press, 2006), 158. 39. G. Giacomello, National Governments and the Control of the Internet: A Digital Challenge (Oxon, UK: Routledge, 2005). 40. S. Vegh, “Profits Over Principles: The Commercialization of the Democratic Potentials of the Internet,” in Ideologies of the Internet, ed. K. Sarikakis and D. Thussu (Cresskill, NJ: Hampton Press, 2006), 62. 41. A. Friedland, “Electronic Democracy and the New Citizenship,” Media, Culture & Society 18 (1996): 185–212. 42. On file with the author. 43. Napoli. 44. The Constitution of The Kingdom of Bhutan, Article 9, http://karari.org/ it/node/28502. 45. Bhutan in itself has only adopted democracy in recent years. 46. J. Stiglitz, A. Sen, and J.-P. Fitoussi. Report by the Commission on the Measurement of Economic Performance and Social Progress, 2009, 14, http://www .stiglitz-sen-fitoussi.fr/documents/rapport_anglais.pdf. 47. Ibid., 50.

8. p h d heal thyself: in search of evidence-based research for evidence-based policy Eli Noam

1. Ben Bagdikian, The New Media Monopoly (Boston, MA: Beacon Press, 2004). 2. Lawrence Lessig, Free Culture: How Big Media Uses Technology and the Law to Lock Down Culture and Control Creativity (New York: Penguin, 2004). 3. Robert W. McChesney, The Problem of the Media: U.S. Communication Politics in the 21st Century (New York: Monthly Review Press, 2004). 4. Adam D. Thierer, Media Myths: Making Sense of the Debate over Media Ownership (Washington DC: Progress & Freedom Foundation, 2005). 5. Benjamin M. Compaine and Douglas Gomery, Who Owns the Media? Competition and Concentration in the Mass Media Industry, 3rd ed. (New Jersey: Erlbaum, 2000). 6. Edwin C. Baker, Media Concentration and Democracy: Why Ownership Matters (New York: Cambridge University Press, 2007). 7. Eli M. Noam, Media Ownership and Concentration in America (New York: Oxford University Press, 2009). 8. Antitrust enforcement guidelines classify markets: HHI < 1,000 Unconcentrated Market 1,000 < HHI, Moderately Concentrated Market 1,800 < HHI, Highly Concentrated Market n ot e s to pag e s 1 2 5 – 3 6

249

9. case studies in results-driven decision making at the fcc Rob Frieden

1. For example, the Communications Act requires the FCC to reduce market entry barriers for entrepreneurs and other small businesses in the provision and ownership of telecommunications services and information services that serve “the public interest, convenience, and necessity.” United States Code, 47 §257(c)(1)(2008). 2. “[A regulatory] agency must examine the relevant data and articulate a satisfactory explanation for its action including a rational connection between the facts found and the choice made.” Motor Vehicle Manufacturers Association v. State Farm Mutual Automobile Insurance Company, 463 U.S. 29, 43, 103 S.Ct. 2856 (1983). Courts will set aside agency decisions found to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” Administrative Procedure Act, 5 U.S.C. § 706(2)(A). 3. American Radio League, Inc. v. F.C.C., 524 F.3d 227 (D.C. Cir. 2008) (reversing the FCC based on the commission’s dismissal of empirical data submitted at its invitation without reason or analysis). 4. Rob Frieden, Winning The Silicon Sweepstakes: Can The United States Compete In Global Telecommunications? (New Haven, CT: Yale University Press, 2010), 16. 5. Federal Communications Commission, Existing Shareholders of Citadel Broadcasting Corp. and of The Walt Disney Co., etc. for Consent to Transfers of Control, Memorandum Opinion and Order and Notice of Apparent Liability, 22 F.C.C.R. 7083 (2007). 6. Federal Communications Commission, Appropriate Framework for Broadband Access to the Internet over Wireline Facilities, Report and Order and Notice of Proposed Rulemaking, 20 F.C.C.R. 14853, 14901 (2005), affirmed Time Warner Telecom, Inc. v. F.C.C., 507 F.3d 205 (3d Cir. 2007) (forbearing, on the FCC’s own motion, from applying tariffing requirements to providers of wireline broadband Internet access service that offer the underlying transmission component of broadband Internet access service as a telecommunications service). Rob Frieden, “Lies, Damn Lies and Statistics: Developing a Clearer Assessment of Market Penetration and Broadband Competition in the United States,” Virginia Journal of Law and Technology 14 (2009): 100. 7. Federal Communications Commission, Applications of Cellco Partnership D/B/A Verizon Wireless and Atlantis Holdings LLC, for Consent to Transfer Control of Licenses, Authorizations, and Spectrum Manager and De Facto Transfer Leasing Arrangements, Memorandum Opinion and Order and Declaratory Ruling, 23 F.C.C.R.17444 (2008) (conditionally approving Verizon Wireless acquisition of Alltel wireless assets resulting in a 90 percent market share held by four firms). 8. Federal Communications Commission, 2006 Quadrennial Regulatory Review—Review of the Commission’s Broadcast Ownership Rules and other Rules 250

n ot e s to pag e 1 4 3

Adopted Pursuant to Section 202 of the Telecommunications Act of 1996, Report and Order and Order on Reconsideration, 23 F.C.C.R. 2010, 2018–19 (2007). “For the better half of the existence of federal ownership regulations, which date back to the 1940s, the Commission offered and the courts required little evidence of the connection between ownership and viewpoint diversity.” Daniel E. Ho and Kevin M. Quinn, “Viewpoint Diversity and Media Consolidation: An Empirical Study,” Stanford Law Review 61 (2009): 789. 9. Federal Communications Commission, AT&T Inc. and BellSouth Corporation Application for Transfer of Control, Memorandum Opinion and Order, 22 F.C.C.R. 5662 (2007); SBC Communications, Inc. and AT&T Corp. Applications for Approval of Transfer of Control, Memorandum Opinion and Order, 20 F.C.C.R. 18290 (2005); Verizon Communications, Inc. and MCI, Inc. Applications for Approval of Transfer of Control, Memorandum Opinion and Order, 20 F.C.C.R. 18433 (2005); Applications of AT&T Wireless Services, Inc. and Cingular Wireless Corporation, Memorandum Opinion and Order, 19 F.C.C.R. 21522 (2004); General Motors Corporation and Hughes Electronics Corporation, Transferors, and the News Corporation Limited, Transferee, Memorandum Opinion and Order, 19 F.C.C.R. 473 (2004). 10. John Kneuer, former assistant secretary for communications and information and administrator at the Commerce Department’s National Telecommunications and Information Administration claimed in 2008 that the United States “has the most effective multiplatform broadband in the world.” “True or False: U.S.’s Broadband Penetration is Lower than Even Estonia’s; Answer: True,” Newsweek, July 9, 2007, 58. 11. American Family Association, Inc. v. FCC, 365 F.3d 1156, 1166 (D.C. Cir. 2004) (noting “necessarily wide latitude to make policy based on predictive judgments deriving from [the commission’s] general expertise”). 12. “Ancillary jurisdiction may be employed, in the Commission’s discretion, when Title I of the Act gives the Commission subject matter jurisdiction over the service to be regulated and the assertion of jurisdiction is ‘reasonably ancillary to the effective performance of [its] various responsibilities.’ ” Federal Communications Commission, IP-Enabled Services, WC Docket No. 04-36, E911 Requirements for IP-Enabled Service Providers, WC Docket No. 05–196, First Report and Order and Notice of Proposed Rulemaking, 20 F.C.C.R. 10245, 10261 (2005) citing United States v. Southwestern Cable Co., 392 U.S. 157, 177–78 (1968); United States v. Midwest Video Corp., 406 U.S. 649, 667–68 (1972); FCC v. Midwest Video Corp., 440 U.S. 689, 700 (1979). Rob Frieden, “What Do Pizza Delivery and Information Services Have in Common? Lessons From Recent Judicial and Regulatory Struggles with Convergence,” Rutgers Computer and Technology Law Journal 32 (2006): 276. 13. Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004); Pacific Bell Telephone Co. v. Linkline Communications, Inc., 555 U.S. 544 (2009). n ot e s to pag e s 1 4 3 – 4 4

251

14. Center for Public Integrity v. FCC, 505 F.Supp.2d 106 (D.D.C. 2007), reconsideration denied, 515 F.Supp.2d 167 (D.D.C. 2007). 15. United States, Office of Management and Budget, Final Information Quality Bulletin for Peer Review, Federal Register 70 Fed. Reg. 2664 (January 14, 2005). 16. The FCC appears to interpret its peer review obligation as limited to matters that involve technical or scientific determinations. “We note that if the Commission determines to rely on a scientific or technical study (or studies) as a basis for its decision-making in this proceeding, such study (or studies) may need to meet any applicable peer review requirements set forth in the Peer Review Bulletin issued by the Office of Management and Budget (OMB).” Federal Communications Commission, Effects of Communications Towers on Migratory Birds, Notice of Proposed Rulemaking, 21 F.C.C.R. 13241, n. 105 (2006). 17. “The Commission shall determine whether any new technology or service proposed in a petition or application is in the public interest within one year after such petition or application is filed. If the Commission initiates its own proceeding for a new technology or service, such proceeding shall be completed within 12 months after it is initiated.” 47 U.S.C. §157(b). “The Commissioner may prescribe such rules and regulations as may be necessary in the public interest to carry out the provisions of this Act.” 47 U.S.C. §201(b). 18. Federal Communications Commission, Commissioner Robert M. McDowell Announces Staff Change, Public Notice (rel. September 18, 2009) (announcing appointment of Christine Kurth as Policy Director and Wireline Counsel). 19. 47 C.F.R. §§ 0.457, 0.459, 1.7001(d), 43.11(c); Federal Communications Commission, Examination of the Current Policy Concerning the Treatment of Confidential Information Submitted to the Commission, Report and Order, 13 F.C.C.R. 24816 (1998). 20. Federal Communications Commission, Applications of Cellco Partnership d/b/a Verizon Wireless and Atlantis Holdings LLC For Consent to Transfer Control of Licenses, Authorizations, and Spectrum Manager and De Facto Transfer Leasing Arrangements and Petition for Declaratory Ruling that the Transaction is Consistent with Section 310(b)(4) of the Communications Act, WT Docket No. 08-95, Protective Order, 23 F.C.C.R. 11154 (WTB 2008) (agreeing to treat as confidential data filed to support acquisition of a competitor). 21. The Telecommunications Act of 1996 requires the FCC to forbear from any statutory provision or regulation if the Commission determines that: (1) enforcement of the regulation is not necessary to ensure that charges and practices are just and reasonable, and are not unjustly or unreasonably discriminatory; (2) enforcement of the regulation is not necessary to protect consumers; and (3) forbearance is consistent with the public interest. 47 U.S.C. § 160(a) (2008). In making such determinations, the FCC must also consider “whether forbearance from enforcing the provision or regulation will promote competitive market conditions.” 47 U.S.C. § 160(b). 252

n ot e s to pag e s 1 4 4 – 4 5

22. “The record of competition compiled in this proceeding and, significantly, the other market-opening regulations that we leave in place today, support our finding that supply elasticity in this market is high for all mass market services. Cox’s extensive facilities build-out in the Omaha MSA, and growing success in luring Qwest’s mass market customers, indicates that . . . [ample facilitates-based competition exists] for both switched access and broadband Internet access services.” Federal Communications Commission, Petition of Qwest Corporation for Forbearance Pursuant to 47 U.S.C. § 160(c) in the Omaha Metropolitan Statistical Area, WC Docket No. 04-223, Memorandum Opinion and Order, 20 F.C.C.R. 19415, 19432–33 (2005), affirmed, Qwest Corp. v. F.C.C., 482 F.3d 471 (D.C. Cir. 2007). 23. Ibid., 19416. 24. “Even Cox, which is the competitive LEC with the most extensive facilitiesbased coverage in Qwest’s territory in the Omaha MSA, depends on Qwest for interconnection, collocation, and reasonable notice of changes in Qwest’s network in order to exchange telecommunications traffic in the Omaha MSA.” Ibid., 19457. 25. Federal Communications Commission, Petition of Verizon Telephone Companies for Forbearance Pursuant to 47 U.S.C. § 160(c) in the Boston, New York, Philadelphia, Pittsburgh, Providence and Virginia Beach Metropolitan Statistical Areas, Inc., WC Docket No. 06-172, Memorandum Opinion and Order, 22 F.C.C.R.21293 (2007), remanded, Verizon Telephone Companies v. F.C.C., 570 F.3d 294, 47 Communications Reg. (P&F) 1487 (D.C. Cir. Jun 19, 2009) (No. 08-1012); Petition of Qwest Corporation for Forbearance Pursuant to 47 U.S.C. § 160(C) in the Phoenix, Arizona Metropolitan Statistical Area, WC Docket No. 09-135, Memorandum Opinion and Order, 25 F.C.C.R. 8622 (2010). 26. The FCC acknowledges this in a report and order establishing more specific criteria for evaluating forbearance petitions: “We acknowledge that we have not previously required petitioners to specify in the petition how the requested relief meets each of the three forbearance criteria, and that a requirement to do so will burden applicants to the extent that they must develop their supporting arguments in advance of filing. We do not, however, consider this an unreasonable expectation, and we find that the benefit to both commenters and the Commission of clarity and precision outweighs the burden on the petitioner of explaining how forbearance from each regulation or statutory provision meets each prong.” Federal Communications Commission, Petition to Establish Procedural Requirements to Govern Proceedings for Forbearance Under Section 10 of the Communications Act of 1934, as Amended, Report and Order, 24 F.C.C.R 9543, 9551 (2009) [hereafter cited as Forbearance Criteria Order]. 27. In 2009 the FCC belatedly specified the documentation petitioners must submit: “A petition for forbearance must include in the petition the facts, information, data, and arguments on which the petitioner intends to rely to make the n ot e s to pag e 1 4 6

253

prima facie case for forbearance. Specifically, the prima facie case must show in detail how each of the statutory criteria are met with regard to each statutory provision or rule from which forbearance is sought.” Forbearance Criteria Order, 9553. 28. Verizon Communications, Inc. v. Law Office of Curtis V. Trinko, LLP, 540 U.S. 398 (2004). 29. Pacific Bell Telephone Co., v. Linkline Communications, Inc., 555 U.S. 438 (2009). 30. Ibid., 1123. 31. Sean M. Carroll, “Main Dish with a Side of Voluntary Commitments: Dish Network-DirecTV Revisited,” Administrative Law Review 61 (2009): 661. 32. “Based on the record before us, we conclude that the proposed transfer of control would violate our rule against one licensee controlling both SDARS licenses. We also conclude that, absent Applicants’ voluntary commitments and other conditions discussed below, the proposed transaction would increase the likelihood of harms to competition and diversity. Applicants, however, have proposed significant voluntary commitments regarding steps the merged company would take to mitigate harms and achieve public interest benefits. We find that absent those voluntary commitments and other conditions, the harms of the transaction would outweigh the potential public interest benefits. On balance, however, we find that with Applicants’ voluntary commitments and other conditions, the potential public interest benefits outweigh the harms.” Federal Communications Commission, Applications for Consent to the Transfer of Control of Licenses XM Satellite Radio Holdings Inc., Transferor to Sirius Satellite Radio Inc., Transferee, Memorandum Opinion and Order and Report and Order, 23 F.C.C.R. 12348, 12352 (2008). See also Federal Communications Commission, Applications of Comcast Corporation, General Electric Company and NBC Universal, Inc. for Consent to Assign Licenses and Transfer Control of Licensees, MB Docket No. 10–56, Memorandum Opinion and Order, 26 F.C.C.R. 4238 (2011). 33. “We find that competitive harm is unlikely in any market, primarily because multiple other service providers in these markets would be an effective competitive constraint on the behavior of the merged entity. We also conclude that the transaction will result in major public interest benefits by facilitating the provision of a nationwide WiMAX-based network that will lead to increased competition, greater consumer choice, and new services.” Federal Communications Commission, Sprint Nextel Corporation and Clearwire Corporation Applications for Consent to Transfer Control of Licenses, Leases, and Authorizations, Memorandum Opinion and Order, 23 F.C.C.R. 17570, 17572 (2008) (approving merger of two major wireless carriers, Sprint-Nextel and Clearwire Corp.). 34. “This lack of present competition between these two incumbent LECs is hardly surprising—both carriers largely serve rural local exchanges and the adjacent exchanges are almost all small and rural.” Federal Communications 254

n ot e s to pag e s 1 4 6 – 4 8

Commission, Applications Filed for the Transfer of Control of Embarq Corporation to Centurytel, Inc. Memorandum Opinion and Order, 24 F.C.C.R. 8741, 8750 (2009). 35. Federal Communications Commission, Constellation, LLC, Carlyle Panamsat I, LLC, Carlyle Panamsat II, LLC, Pep Pas, LLC, And Peop Pas, LLC, Transferors, and Intelsat Holdings, Ltd., Transferee, Consolidated Application For Authority to Transfer Control of Panamsat Licensee Corp. and Panamsat H-2 Licensee Corp., IB Docket No. 05-290, Memorandum Opinion and Order, 21 F.C.C.R.7368 (2006). 36. Federal Communications Commission, Applications for Consent to the Transfer of Control of Licenses XM Satellite Radio Holdings Inc., Transferor, to Sirius Satellite Radio Inc., Transferee, Memorandum Opinion and Order and Report and Order, 23 F.C.C.R. 12348 (2008). 37. Federal Communications Commission, AT&T Inc. and BellSouth Corporation, Application for Transfer of Control, 22 F.C.C.R. 5662 (2007), on partial reconsideration, 22 F.C.C.R. 6285 (2007). 38. Federal Communications Commission, Applications Filed for the Transfer of Control of Embarq Corporation to Centurytel, Inc., Memorandum Opinion and Order, 24 F.C.C.R. 8741, 8756 (2009). 39. Federal Communications Commission, Application of Echostar Communications Corporation, (a Nevada Corporation), General Motors Corporation, and Hughes Electronics Corporation (Delaware Corporations) (Transferors) and Echostar Communications Corporation (a Delaware Corporation) (Transferee), Hearing Designation Order, 17 F.C.C.R. 20559, 20631 (2002) (designating a hearing to resolve issues pertaining to the public interest merits in the merger of two major direct broadcast satellite firms). 40. The FCC has experienced several judicial reversals of the commission’s attempt to relax broadcast and MVPD ownership rules. In Fox Television Stations, Inc. v. F.C.C., 280 F.3d 1027, modified on rehearing, 293 F.3d 537 (D.C. Cir. 2002) the D.C. Circuit Court of Appeals remanded the FCC’s retention of the then congressionally established 35 percent national television ownership rule. In Sinclair Broadcasting Group, Inc. v. FCC, 284 F.3d 148 (D.C. Cir. 2002) the court remanded the commission’s 1999 revision of its local television multiple ownership rule. 41. 373 F.3d 372, 382 (3d Cir. 2004), cert. denied, 545 U.S. 1123 (2004). The Third Circuit Court of Appeals rejected the FCC’s decision to count media outlets with no differentiation on market penetration and significance. 42. “Most importantly, the Commission has not sufficiently justified its particular chosen numerical limits for local television ownership, local radio ownership, and cross-ownership of media within local markets. Accordingly, we partially remand the Order for the Commission’s additional justification or modification” (ibid., 382). 43. Ibid., 395. n ot e s to pag e s 1 4 8 – 5 1

255

44. Federal Communications Commission, 2000 Biennial Regulatory Review Spectrum Aggregation Limits for Commercial Mobile Radio Services, Report and Order, 16 F.C.C.R. 22668, 22682 (2001). The FCC rejected as a significant barrier to market entry the need to acquire spectrum, in light of the commission’s view that resale opportunities would suffice. “Nonetheless, there are factors that moderate concern regarding the spectrum access barrier to entry. In particular, the need for direct access to spectrum is not absolute because carriers can compete in the provision of CMRS without direct access to spectrum through resale, or a mobile virtual network operator (MVNO) arrangement.” Ibid., 16 F.C.C.R. 22690. 45. Federal Communications Commission, Annual Report and Analysis of Competitive Market Conditions with Respect to Mobile Wireless, Including Commercial Mobile Services, 14th Report, 25 F.C.C.R. 11407 (2010). 46. “Using the various data sources and metrics discussed above, we have met our statutory requirement to analyze the competitive market conditions with respect to commercial mobile services, and conclude that the CMRS marketplace is effectively competitive.” Federal Communications Commission, Implementation of Section 6002(B) of the Omnibus Budget Reconciliation Act of 1993, Annual Report and Analysis of Competitive Market Conditions With Respect to Commercial Mobile Services, Twelfth Report, 23 F.C.C.R. 2241, 2354 (2008). 47. Federal Communications Commission, Service Rules for the 698–746, 747– 762 and 777–792 MHz Bands, Second Report and Order, 22 F.C.C.R.15289, 15361(2007), on reconsideration, 22 F.C.C.R.17,935, partially modified, 23 F.C.C.R.5319 (2008), 24 F.C.C.R. 4782 (2009). Wireless carriers remain subject to conventional common carrier regulation of their telecommunications services, a status the FCC has generally ignored except for the matter of compulsory interconnection to provide subscribers with access to other carriers when “roaming” outside their home territory. Federal Communications Commission, Reexamination of Roaming Obligations of Commercial Mobile Radio Service Providers, Report and Order and Further Notice of Proposed Rulemaking, 22 F.C.C.R. 15817 (2007) (specifying that cellular operators must provide their subscribers automatic access to other carriers for making and receiving telephone calls when traveling outside subscribers’ home service regions). 48. “According to an analysis by The Associated Press, the two telecom companies bid more than $16 billion, constituting the vast majority of the overall $19.6 billion that was bid in the FCC auction. With Verizon Wireless and AT&T dominating the auction so completely, hopes that the auction would allow for the creation of a new nationwide wireless service provider were dashed.” W. David Gardner, “Verizon, AT&T Big Winners in 700 MHz Auction,” InformationWeek, March 20, 2008, http://www.informationweek.com/news/206905000. 49. Illinois Public Telecommunications Association v. FCC, 117 F.3d 555 (1997) (court rejects FCC’s determination that local and toll free calls from pay 256

n ot e s to pag e s 1 5 1 – 5 2

telephone have similar costs, because the record compiled by the FCC showed significantly different costs). 50. “We must defer to the Commission’s expert judgment in the absence of record evidence indicating that the Commission’s assumption is a clear error of judgment, or a showing that the empirical assumption is facially implausible or inconsistent.” American Family Association, Inc. v. FCC, 365 F.3d 1156, 1165 (D.C. Cir.2004) (FCC’s method for assigning noncommercial educational broadcast licenses among competing applicants deemed valid). 51. FCC v. National Citizens Committee for Broadcasting, 436 U.S. 775, 814 (1978) quoting FPC v. Transcontinental Gas Pipe Line Corp., 365 U.S. 1, 29; Industrial Union Dept., AFL-CIO v. Hodgson, 499 F.2d 467, 474–475 (1974). 52. 545 U.S. 967 (2005). 53. Chevron U.S.A. v. Natural Resources Defense Council, Inc. 467 U.S. 837 (1984). 54. “If a statute is ambiguous, and if the implementing agency’s construction is reasonable, Chevron requires a federal court to accept the agency’s construction of the statute, even if the agency’s reading differs from what the court believes is the best statutory interpretation.” Brand X, 545 U.S. at 980, citing Chevron at 843–844 and n. 11. 55. Ibid., 982. 56. Ibid., 991. 57. Ibid., 992. 58. “The important fact, however, is that the Commission has chosen to achieve this [result] through an implausible reading of the statue, and thus exceeded the authority given it by Congress.” Brand X, Scalia Dissenting Opinion, ibid., 1005. 59. Scalia Dissenting Opinion, Brand X, ibid., 1009. 60. Rob Frieden, “The FCC’s Name Game: How Shifting Regulatory Classifications Affect Competition,” Berkeley Technology Law Journal 19 (2004): 1275. 61. FCC. v. Fox Television Stations, Inc., 129 S.Ct. 1800, 1813 (2009), on remand, 613 F.3d 317 (2d Cir. 2010). 62. 524 F.3d 227 (D.C. Cir. 2008). 63. American Radio Relay League, Inc. v. FCC, 524 F.3d 227 (D.C. Cir. 2008). 64. Ibid., 239. 65. 258 F.3d 1191 (10th Cir. 2001). 66. Ibid., 1205. 67. Federal Communications Commission, Wireless Telecommunications Bureau Seeks Comment on Petition for Rulemaking of Rural Telecommunications Group, Inc. to Impose a Spectrum Aggregation Limit on all Commercial Terrestrial Wireless Spectrum Below 2.3 GHz, Public Notice, DA 08–2279 (rel. October 10, 2008), http://wireless.fcc.gov/index.htm?job=headlines&y=2008#10. 68. For example Verizon Wireless bid $9,363,160,000 of the net bidding total amounting to total $18,957,582,150. AT&T bid $6,636,658,000. See Federal

n ot e s to pag e s 1 5 2 – 5 6

257

Communications Commission, Auction 73, 700 MHz Band, available at http:// wireless.fcc.gov/auctions/default.htm?job=auction_summary&id=73. 69. Federal Communications Commission, Harvard’s Berkman Center to Conduct Independent Review of Broadband Studies to Assist FCC, Public Notice (rel. July 14, 2009), http://hraunfoss.fcc.gov/edocs_public/attachmatch/ DOC-291986A1.pdf. 70. See Federal Communications Commission, Broadband.gov website, http://www.broadband.gov/. 71. The FCC has acknowledged “that broadband is not yet being deployed ‘to all Americans’ in a reasonable and timely fashion.” Inquiry Concerning the Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fashion, and Possible Steps to Accelerate Such Deployment Pursuant to Section 706 of the Telecommunications Act of 1996, as Amended by the Broadband Data Improvement Act, GN Docket No. 11-121, Eighth Broadband Progress Report, FCC 12-90, ¶1, 27 F.C.C.R. 10342 (rel. Aug. 21, 2012); http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-1290A1.doc. 72. Annual Report and Analysis of Competitive Market Conditions with Respect to Mobile Wireless, Including Commercial Mobile Services, Fifteenth Report, FCC 11-103, 26 F.C.C.R. 9664 (rel. June 27, 2011); http://www.fcc.gov/ reports/mobile-wireless-competition-report-15th-annual.

10. the determinants of disconnectedness: understanding us broadband unavailability Kenneth Flamm

1. Influential studies suggesting links between IT deployment and aggregate productivity growth include S. Oliner and D. Sichel, “The Resurgence of Growth in the Late 1990s: Is Information Technology the Story?” Journal of Economic Perspectives 14 (2000), 3-22; D. Jorgenson, “Information Technology and the U.S. Economy,” American Economic Review 91 (2001), 1-32; U.S. President, Council of Economic Advisors, “The Making of the New Economy,” in Economic Report of the President (2001), 25–38. A more skeptical view can be found in R. Gordon, “Does the ‘New Economy’ Measure Up to the Great Inventions of the Past?” Journal of Economic Perspectives 14 (2000): 49–74. 2. Federal Communications Commission, Industry Analysis and Technology Division, Wireline Competition Bureau, High-Speed Services for Internet Access: Status as of December 31, 2005, July 2006, 4. 3. Federal Communications Commission, Fifth Report, Inquiry Concerning the Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fashion, and Possible Steps to Accelerate Such Deployment Pursuant to Section 706 of the Telecommunications Act of 1996, FCC 08–88, June 2008, 18–19. 258

n ot e s to pag e s 1 5 6 – 5 9

4. Federal Communications Commission (2008), 34. 5. Federal Communications Commission, Industry Analysis and Technology Division, Wireline Competition Bureau, High-Speed Services for Internet Access: Status as of June 30, 2008, July 2009, Table 15. 6. John B. Horrigan, Home Broadband Adoption 2006, Pew Internet and American Life Project, May 2006; John B. Horrigan, Home Broadband Adoption 2009, Pew Internet and American Life Project, June 2009. 7. Seventeen percent of dial-up Internet users in Pew’s April 2008 survey said that broadband service was unavailable where they live (Horrigan 2009). 8. Ken Belson, “Rural Areas Left in Slow Lane of High-Speed Data Highway,” New York Times, September 28, 2006, A1. 9. Such voluntarily reported lines accounted for less than .05 percent of high-speed lines in recent submissions. Federal Communications Commission, Industry Analysis and Technology Division, Wireline Competition Bureau, High-Speed Services for Internet Access: Status as of December 31, 2003, June 2004, 2. 10. Federal Communications Commission, Industry Analysis and Technology Division, Wireline Competition Bureau, High-Speed Services for Internet Access: Status as of June 30, 2006, December 2006, Table 7. I thank Scott Wallsten for first bringing this point to my attention. 11. My source for the “typical” numbers here is an ESRI PowerPoint presentation. 12. This has the practical effect of making most if not all prior research using this (pre-June 2005) data, which typically assumed that “geographic” zip codes not disclosed in the FCC list of zip codes with broadband service constitute the “no service” zip codes, deeply flawed. Examples of research making this erroneous assumption include J. E. Prieger, “The Supply Side of the Digital Divide: Is There Equal Availability in the Broadband Internet Access Market?” Economic Inquiry 41 (2003); K. Flamm, “The Role of Economics, Demographics, and State Policy in Broadband Availability,” presented at the PURC/London Business School Conference on “The Future of Broadband: Wired and Wireless, 2005,” University of Florida, Gainesville, 2005. My initial, incorrect, and sometimes conflicting misunderstandings of FCC procedures were derived from a teleconference with FCC staff in January 2005, and a follow-up discussion in September 2006. 13. Federal Communications Commission (July 2009), 4n8. 14. Tony H. Grubesic, “Zip Codes and Spatial Analysis: Problems and Prospects,” Socio-Economic Planning Sciences 42 (2008): 136–137. 15. ZCTA-based Census data are approximations corresponding to actual zip codes. Their construction is explained at http://www.census.gov/geo/ZCTA/ zcta_brch_prnt.pdf, and http://www.census.gov/geo/ZCTA/zcta.html. I have discarded “artificial” ZCTAs (unclassified areas, or areas consisting of bodies of water), which do not have a corresponding “real” zip code in the analysis that follows.

n ot e s to pag e s 1 5 9 – 6 2

259

The census data correspond to the estimates in the Census SF-3 (long form) database, and were taken from the “Gazeteer” ZCTA file available at http:// www.census.gov/geo/www/gazetteer/places2k.html, and from the version of the Census SF-3 database as extracted and made accessible at the University of Missouri’s Missouri Census Data Center through http://mcdc2.missouri.edu/ cgi-bin/uexplore?/pub/data/sf32000x. I have omitted 1,152 additional alphanumeric “HH” and “XX” codes defined by the US census to cover primarily water and unpopulated areas from this count. 16. These are available in various annual revisions of IRS Publication 3496, downloadable from the IRS website. 17. There is also a possible matching issue with the population and income data that the FCC has used in its six-month reports on broadband availability. The FCC in its public reports on high-speed service combines these provider numbers by zip code with 2000 census population estimates by zip code, to construct an estimate of the percentage of the year 2000 US population living in US zip codes with at least one service provider. The 2000 population estimates (and ancillary year 2000 data on zip code area and median household income by zip) apparently come from yet another proprietary data source, Mapinfo’s Demographic Power Pack, Current Year Update (2000). It is unclear that there is any necessary relationship between Mapinfo’s 2000 zip code boundaries— even if the zip code numbers were to match up perfectly, which is most unlikely, given the constant morphing of the FCC zip codes in use—and the current TeleAtlas/GDT definitions. There is no discussion of this issue in any FCC publication that I have been able to locate. 18. A final point to make is that as part of a program intended to encourage the development of e-filing, the IRS publishes information on precise numbers of approved e-Filing service vendors, and operating e-filing companies, by zip code. This information on number of vendors by zip code is either not deemed sensitive or business confidential by the IRS, or the objective of encouraging entry and growth in e-filing of tax returns was thought to outweigh any such concerns. The FCC, by way of contrast, has refused to reveal analogous data (number of vendors by zip code) for high-speed service providers on grounds of business information sensitivity (one to three providers have always been lumped into a single bin). Clearly, there does not seem to be some immutable principle obstructing disclosure in one case, but not the other, and the objective of encouraging broadband deployment would seem to be at least as worthy as— indeed, complementary to—the goal of encouraging e-filing. 19. Until recently, my neighborhood was beyond the three-mile range for a DSL connection from a DSL-capable central office telephone switch, and DSLbased service was unavailable. The hilly neighborhood I live in also may lack the clear, unobstructed view of the southern horizon needed for higher cost satellite service. The broadband service I purchased is actually an Earthlink-branded service (my ISP appears to be Mindspring.com when my IP is identified on the Internet). 260

n ot e s to pag e s 1 6 2 – 6 4

Earthlink has agreements with Time Warner to market this service in a number of US markets. FCC instructions on its Form 477 website seemed to require Time Warner, and not Earthlink, to report my broadband connection: “An ISP that obtains a broadband connection or service from an unaffiliated entity, which it incorporates into its own high-speed Internet-access service, is exempt from reporting broadband connections in Part I of Form 477. Instead, the obligation to report the broadband connection rests with the underlying facilities-based provider.” See http://www.fcc.gov/broadband/broadband_data_faq .html, question 10. 20. Belson. 21. Government Accountability Office, “Broadband Deployment is Extensive Throughout the United States, but It Is Difficult to Assess the Extent of Deployment Gaps in Rural Areas,” May 2006, 18. Adjustments the GAO (which had access to the proprietary data in the FCC database) made included removing satellite-only service providers, business-only service providers, DSL-based service providers to residences more than 2.5 miles from a central office, and not double-counting cable service operators serving disjoint groups of households in a zip code. 22. Ibid., 17. 23. The Pew data actually use survey weights to estimate the number of adults in households with broadband connections, so it would also seem possible that the number of adults in households might vary systematically between areas primarily served by cable and areas primarily served by DSL, perhaps explaining a small part of the discrepancy. 24. Possible explanations include respondent error in the Pew and GAO/SRI surveys (people thinking that their telephone line connections are broadband when they are not), sample bias (homes without a conventional landline telephone, who are therefore excluded from many telephone-based sampling frames, would be more likely to use cable for a broadband connection), and the (unlikely?) possibility that a significant number of residences may have multiple broadband connections. 25. FCC (July 2006), Table 3. 26. Government Accountability Office, 10. 27. Note that cable providers provide broadband services to office premises passed by their cable infrastructure in many instances. 28. This analysis is based on Table 14 in FCC’s Form 477 reports, for June 2005 and June 2008. The numbers shown are the share of households where ILECs offered telephone service where xDSL (including both symmetric and asymmetric variants) was also offered, and the share of households with cable TV service that also had access to cable modem service. 29. Having no zero provider zip codes, once it occurs, is not necessarily a permanent state. I note that Kentucky went from having over 3 percent of its zip codes with no providers in December 2006, to no zip codes with zero providers n ot e s to pag e s 1 6 4 – 7 1

261

in June 2007, to over 3 percent of its zip codes once again with zero providers in December 2007! 30. To undertake this task, I made use of the Missouri Census Data Center’s MABLE/Geocorr2K geographic correspondence engine, available at http:// mcdc2.missouri.edu/websas/geocorr2k.html. 31. See for example, K. Flamm, A. Friedlander, J. B. Horrigan, and W. Lehr, “Measuring Broadband: Improving Communications Policymaking through Better Data Collection,” 2007, http://www.pewinternet.org/PPF/r/227/report_ display.asp.

11. european broadband spending: implications of input-output analysis and opportunity costs Ibrahim Kholilul Rohman and Erik Bohlin

1. Organisation for Economic Co-operation and Development (OECD), Information Economy Product Definition Based on the Central Product Classification (Geneve: OECD, 2009). 2. H. Hollenstein, “Determinants of the Adoption of Information and Communication Technologies (ICT) an Empirical Analysis Based on FirmLevel Data for the Swiss Business Sector,” Structural Change and Economic Dynamics 15, no. 3 (2004): 315–342. 3. A. I. Bozinis, “International Economic Relations and Information Communication Technologies (ICT) Use: Economic Globalization via Economic Digitalization,” American Journal of Applied Sciences 4, no. 4 (2007): 188–191. 4. M. Franklin, P. Stam, and T. Clayton, “ICT Impact Assessment by Linking Data,” Economic and Labour Market Review 3, no. 10 (2009): 18–27. 5. L. Martin, “The Effects of ICT Use on Employee’s Motivations: An Empirical Evaluation,” Economics Bulletin 31, no. 2 (2011): 1592–1605. 6. E. A. Cortés and J. A. Navarro, “Do ICT Influence Economic Growth and Human Development in European Union Countries?,” International Advances in Economic Research 17, no. 1 (2011): 28–44. 7. S. A. Haller and I. Siedschlag, “Determinants of ICT Adoption: Evidence from Firm-Level Data,” Applied Economics 43, no. 26 (2011): 3775–3788. 8. S. K. Majumdar, “Broadband Adoption, Jobs and Wages in the US Telecommunications Industry,” Telecommunications Policy 32, no. 9–10 (2008): 587–599. 9. S. K. Majumdar, O. Carare, and H. Chang, “Broadband Adoption and Firm Productivity: Evaluating the Benefits of General Purpose Technology,” Industrial and Corporate Change 19, no. 3 (2010): 641–674. 10. K. A. Van Gaasbeck, “A Rising Tide: Measuring the Economic Effects of Broadband Use across California,” Social Science Journal 45, no. 4 (2008): 691–699. 262

n ot e s to pag e s 1 7 1 – 9 1

11. A. Grimes, C. Ren, and P. Stevens, “The Need for Speed: Impacts of Internet Connectivity on Firm Productivity,” Journal of Productivity Analysis 37, no. 2 (2012): 187–201. 12. M. N. Baily and R. Z. Lawrence, “Do We Have a New E-Conomy?,” American Economic Review 91, no. 2 (2001): 308–312. 13. C. Armistead and M. Meakins, “A Framework for Practising Knowledge Management,” Long Range Planning 35, no. 1 (2002): 49–71. 14. C. Antonelli, “The Digital Divide: Understanding the Economics of New Information and Communication Technology in the Global Economy,” Information Economics and Policy 15, no. 2 (2003): 173–199. 15. L. Ha, R. N. Okigbo, and P. Igboaka, “Knowledge Creation and Dissemination in Sub-Saharan Africa,” Management Decision 46, no. 3–4 (2008): 392–405. 16. I. T. Kandilov and M. Renkow, “Infrastructure Investment and Rural Economic Development: An Evaluation of USDA’s Broadband Loan Program,” Growth and Change 41, no. 2 (2010): 165–191. 17. H. Lindskog and M. Johansson, “Broadband: A Municipal Information Platform: Swedish Experience,” International Journal of Technology Management 31, no. 1–2 (2005): 47–63. 18. A. Arbore and A. Ordanini, “Broadband Divide among SMEs—the Role of Size, Location and Outsourcing Strategies,” International Small Business Journal 24, no. 1 (2006): 83–99. 19. S. Howick and J. Whalley, “Understanding the Drivers of Broadband Adoption: The Case of Rural and Remote Scotland,” Journal of the Operational Research Society 59, no. 10 (2008): 1299–1311. 20. V. Glass and S. K. Stefanova, “An Empirical Study of Broadband Diffusion in Rural America,” Journal of Regulatory Economics 38, no. 1 (2010): 70–85. 21. S. Savage, G. Madden, and M. Simpson, “Broadband Delivery of Educational Services: A Study of Subscription Intentions in Australian Provincial Centers,” Journal of Media Economics 10, no. 1 (1997): 3–15. 22. S. C. Sivakumar and W. Robertson, “Developing an Integrated Web Engine for Online Internetworking Education: A Case Study,” Internet ResearchElectronic Networking Applications and Policy 14, no. 2 (2004): 175–192. 23. S. Lindsay, P. Bellaby, S. Smith, and R. Baker, “Enabling Healthy Choices: Is ICT the Highway to Health Improvement?,” Health 12, no. 3 (2008): 313–331. 24. J. L. Collins and B. Wellman, “Small Town in the Internet Society: Chapleau Is No Longer an Island,” American Behavioral Scientist 53, no. 9 (2010): 1344–1366. 25. T. M. Hale, S. R. Cotten, P. Drentea, and M. Goldner, “Rural-Urban Differences in General and Health-Related Internet Use,” American Behavioral Scientist 53, no. 9 (2010): 1304–1325. n ot e s to pag e 1 9 1

263

26. A. Ward and B. T. Prosser, “Reflections on Cyberspace as the New “Wired World of Education,” Educational Technology & Society 14, no. 1 (2011): 169–178. 27. D. Arduini, A. Zanfei, M. Denni, and G. Giungato, “The Provision of E-Government Services: The Case of Italy,” in Proceedings of the 11th European Conference on eGovernment ed. M. Klun, M. Decman, and T. Jukic (Reading, UK: Academic Publishing, 2011), 53–64. 28. A. Kalja, J. Pold, T. Robal, and U. Vallner, “Modernization of the E-Government in Estonia,” in 2011 Proceedings of Picmet 11: Technology Management in the Energy-Smart World, ed. by D. F. Kocaoglu, T. R. Anderson and T. U. Daim (N.p.: IEEE, 2011). 29. D. Fuguitt and S. J. Wilcox, Cost-Benefit Analysis for Public Sector Decision Makers (Westport: Greenwood Publishing Group, 1999). 30. W. Hughes, C. Brown, C. Brown, S. Miller, and T. McConnell, “Evaluating the Economic Impact of Farmers’ Markets Using an Opportunity Cost Framework,” Journal of Agricultural and Applied Economics 40, no. 1 (2008): 253–265. 31. B.-H. Cho and N. H. Hooker, “The Opportunity Cost of Food Safety Regulation: An Output Directional Distance Function Approach,” Working Paper AEDE-WP-0038–04, Ohio State University, 2004. 32. P. Grady and R. A. Muller, “On the Use and Misuse of Input-Output Based Impact Analysis in Evaluation,” Canadian Journal of Program Evaluation 3, no. 2 (1988): 49–61. 33. E. E. Elder and W. R. Butcher, “Including the Economic Impact of Cost Paying in Regional Input-Output Analysis,” Western Journal of Agricultural Economics 14, no. 1 (1989): 78–84. 34. R. Crandall, W. Lehr, and R. Litan, “The Effect of Broadband Deployment on Output and Employment: A Cross Sectional Analysis of US Data,” Issues in Economics Policy no. 6 (July 2007): 1–34. 35. D. A. Kelly, Study of the Economic and Community Benefits of Cedar Falls (Iowa Association of Municipal Utilities, 2004). 36. G. S. Ford and T. M. Koutsky, “Broadband and Economic Development: A Municipal Case Study from Florida,” Review of Urban and Regional Development Studies 17, no. 3 (2005): 216–229. 37. W. Lehr, C. Osorio, S. Gillet, and M. Sirbu, “Measuring Broadband Economic Impact.” The 33rd Research Conference on Communications, Information and Internet Policy. Arlington, VA, 2006. 38. R. Katz, “Estimating Broadband Demand and its Economic Impact in Latin America,” Proceedings of the 3rd ACORN-REDECOM Conference, Mexico City, 2009. 39. Ibid. 40. Ibid. 41. Yan gives details that the induced effect is due to the impact from additional income. As a result of direct and indirect effects, the level of the household 264

n ot e s to pag e s 1 9 1 – 9 4

income throughout the economy will increase. Some portions of this increased income will be spent on final goods and services which then represent the size of the induced effect. Thus, when measuring the effect, the IO includes households in the production system (closed IO table). Consequently, the dimension of matrix increases by an additional one column (consumption) and one row (wages). C.-S. Yan, Introduction to Input-Output Economics (New York: Holt, Reinhart and Winston, 1968). 42. Grady and Muller. 43. R. Katz and S. Suter, “Estimating the Economic Impact of the Broadband Stimulus Plan,” Working Paper, Columbia Institute for Tele-Information, 2009. 44. Strategic Network Group, “Economic Impact Study of the South Dundas Township Fiber Network,” Report for the UK Department of Trade and Industry, Ontario, Canada, 2003. 45. J. Liebenau, R. Atkinson, P. Karrberg, D. Castro, and S. Ezell, The UK’s Digital Road to Recovery (LSE Enterprise LTD and the Information Technology and Innovation Foundation, 2009). 46. Organisation for Economic Co-operation and Development, Information Economy Product Definition Based on the Central Product Classification (Geneve: OECD, 2009). 47. R. E. Miller and D. Blair, Input-Output Analysis: Foundations and Extensions (Cambridge: Cambridge University Press, 2009). 48. M. Celasun, “Sources of Industrial Growth and Structural Change: The Case of Turkey,” Working Papers 614, The World Bank, 1984. 49. A. R. Zakariah and E. E. Ahmad, “Sources of Industrial Growth Using the Factor Decomposition Approach: Malaysia, 1978–1987,” Developing Economies 37, no. 2 (1999): 162–196. 50. T. Akita, “Industrial Structure and the Source of Industrial Growth in Indonesia: An I-O Analysis Between 1971 and 1985,” Asian Economic Journal 5, no. 2 (1991): 139–158. 51. Measuring the impact of ICT sectors in these countries is relevant in the sense that they are identified as having long histories of R&D activities and technology transfer. B. Eichengreen. The European Economy Since 1945: Coordinated Capitalism and Beyond (Princeton: Princeton University Press, 2008), 26, Table 2.6. The complete sets of the IO data (1995, 2000, 2005) in this study can be found in Austria, Denmark, Finland, France, Germany, Netherlands, Spain, and Sweden. The missing data are Belgium (2005), Italy (1995), Norway (1995) where the data in 2000 represents 2001. Since the comparison requires at least two time periods to investigate the structural change in the economy, the United Kingdom is then dropped from the country study. The unavailability of the UK data has been confirmed (personal communication with a UK statistical officer in January 13, 2010). The confirmation informs that other priorities stopped the construction of IO tables in the United Kingdom, but work is now in process to produce tables for 2005. n ot e s to pag e s 1 9 4 – 9 8

265

52. Organisation for Economic Co-operation and Development, Guide to Measuring the Information Society (Geneve: OECD, 2008). 53. Ibid. 54. Ibid. 55. Output refers to the sum of intermediate demand and final demand (GDP), which is equal to the sum of intermediate and primary inputs. By definition, output is greater than the GDP. 56. Data is based on the Swedish statistical agency; see hhtp://www.scb.se. 57. N. Rosenberg, Inside the Black Box: Technology and Economics (Cambridge: Cambridge University Press, 2004). 58. P. Milgrom, Y. Quaian, and J. Robert, “Complementarities, Momentum and the Evolution of Modern Manufacturing,” American Economic Review, 81, no. 2 (1991): 84–88. 59. T. F. Bresnahan and M. Trajtenberg, “General Purpose Technologies: Engines of Growth?,” Journal of Econometrics 65, no. 1 (1995): 83–108. 60. Organisation for Economic Co-operation and Development, “Good Practice Paper on ICTs for Economic Growth and Poverty Reduction,” DAC Journal 6, no. 3 (2005): 27–95.

12. using data for policy development: designing a universal service fund for tanzania Heather Hudson

1. Research for this paper was carried out in the process of gathering information to advise the government of Tanzania under a contract to Kalba International funded by the World Bank. Konrad (Kas) Kalba and Heather Hudson collaborated in conducting field interviews, reviewing secondary data, and drafting initial reports. No confidential information is included in this chapter. The views expressed are those of the author. 2. A recent study estimates that 44 percent of the population is under fifteen; see Population Reference Bureau, “2008 World Population Data Sheet,” http:// www.prb.org/Publications/Datasheets/2008/2008wpds.aspx. 3. The World Bank estimates the average Tanzanian per capita GDP at $400 (without PPP adjustment. See World Bank,”Tanzania at a glance,” September 24, 2008. 4. Robert J. Utz, ed., Sustaining and Sharing Economic Growth in Tanzania (Washington, DC: World Bank, 2008). 5. Interview with director of the National Bureau of Statistics, April 2009. 6. Estimates by Konrad Kalba, May 2009. 7. The major mobile operators are Vodacom (subsidiary of Vodafone), Zain owned by MTC of Kuwait which purchased Celtel International (sold in 2010 to Bharti Airtel), Zantel (also the fixed carrier in Zanzibar), owned by Etisalat, TTCL Mobile, and Tigo (owned by Milicom).

266

n ot e s to pag e s 1 9 8 – 2 1 2

8. See http://www.seacom.mu. 9. The Eastern Africa Submarine Cable System, http://www.eassy.org. 10. Three carriers were operating international gateways in competition with TTCL in August 2009, according to the UCAF manager. Personal communication, August 2009. 11. United Republic of Tanzania, the Universal Communication Service Access Act, 2006. 12. InfoDev, “Universal Access and Service.” Module 4 of ICT Regulation Toolkit, December 2008, http://www.ictregulationtoolkit.org. 13. Heather E. Hudson,”Defining Universal Service Funds: Are They Accelerators or Anachronisms?” Intermedia 38 (2010). 14. Ibid. 15. Heather E. Hudson, From Rural Village to Global Village: Telecommunications for Development in the Information Age (New York: Routledge, 2006). 16. See http://www.schoolnetafrica.org. 17. See http://www.amref.org. 18. Tanzania ICT Policy, 2003, quoted in “Information and Communication Technology (ICT) for Basic Education,” http://www.moe.go.tz. 19. Information and Communication Technology (ICT) Policy for Basic Education, “ICT for Improved Education,” August, 2007, http://www.moe.go.tz. 20. Project “ICT Implementation in Teachers’ Colleges,” http://www.teachers .or.tz. 21. See http://www.un.org/millenniumgoals/. 22. National Bureau of Statistics (Tanzania) and Macro International, Tanzania 2007–08 HIV/AIDS and Malaria Indicator Survey Key Findings (Calverton, MD: NBS and Macro International, 2009). 23. Smaller households in emerging markets are statistically correlated with higher mobile penetration levels. See Kas Kalba, “The Adoption of Mobile Phones in Emerging Markets,” International Journal of Communication, January 2009. 24. Alison Gillwald, “Towards Evidence Based ICT Policy: Access & Usage in 17 African Countries,” Research ICT Africa, presented at the Telecommunications Policy Research Conference, Arlington, VA, September 2008. 25. Interview with Ophelia Mascarenhas, researcher for a study of ICTs and poverty in Tanzania funded by the International Development Research Centre (ICRC), April 2009. 26. Christoph Stork, “Mobile Access & Use in East Africa” (n.d.), http:// researchICTafrica.net. 27. Hosea Mpogole, Hidaya Usanga, and Matti Tedre, “Mobile Phones and Poverty Alleviation: A Survey Study in Rural Tanzania” (Iringa, Tanzania: Tumaini University, n.d.). 28. Stork. 29. Utz. n ot e s to pag e s 2 1 3 – 2 1

267

30. Gillwald. 31. Ibid. 32. Personal interview, April 2009. 33. Metcalfe’s Law states that the value of the network is n(n–1), or almost the square of the number of users. 34. ITU data at http://www.itu.int/ITU-D/icteye/. 35. Hudson, 2006. 36. Personal interviews, Dar es Salaam, April 2009. 37. Ibid. 38. See for example, Florence E. Etta, and Sheila Parvyn-Wamahiu, eds. Information and Bibliography Communication Technologies for Development in Africa, vol. 2; and Tina James, ed. Information and Communication Technologies for Development in Africa, vol. 3 (Networking Institutions of Learning— SchoolNet) (Ottawa: International Development Research Centre, 2004). 39. “Bulgaria Projects: LearnLink,” http://www.learnlink.aed.org/Projects/ bulgaria.htm.

268

n ot e s to pag e s 2 2 1 – 2 6

BIBLIOGRAPHY

introduction Bell, Daniel. The Coming of Post-Industrial Society: A Venture in Social Forecasting. New York: Basic Books, 1976. Peters, Mike. “Obama Views Broadband as Key to Economic Recovery,” Broadbandinfo.com, December 16, 2008. http://www.broadbandinfo.com/ news-archives/2008/obama-views-broadband-as-key-to-economic-recovery .html.

chapter 1 Allen, Matthew. “The Experience of Connectivity—Results from a Survey of Australian Internet Users.” Information, Communication & Society 13 (2010): 350–374. Australian Bureau of Statistics. “Internet Activity, Australia, December 2009— Internet Subscribers by Access Connection, for ISPs with More Than 1,000 Active Subscribers.” http://tinyurl.com/abs-internet-activity. Barney, Darin D. The Network Society. Cambridge: Polity Press, 2004. Beard, T. Randolph, George S. Ford, and Lawrence J. Spiwak. The Broadband Adoption Index: Improving Measurements and Comparisons of Broadband Deployment and Adoption (Phoenix Center Policy Paper Number 36). Washington, DC: Phoenix Center for Advanced Legal and Economic Public Policy Studies, 2009. Bell, Daniel. The Coming of Post-Industrial Society: A Venture in Social Forecasting. New York: Basic Books, 1973. 269

Blackwell, Suzanne. “Canada’s Broadband Price Per Mbps—GIGO.” http:// www.giganomics.ca/observations-old/2009/8/20/canadas-broadband-priceper-mbps-gigo.html. ——. “Broadband Job Creation Figures Overestimated.” http://www.redorbit.com/ news/technology/1631755/broadband_job_creation_figures_overestimated/. Broadband Stakeholder Group. “Pipe Dreams? Prospects for Next Generation Broadband Deployment in the UK.” www.broadbanduk.org/content/ view/236/7/. ——. “Predicting UK Future Residential Bandwidth Requirements.” Cambridge: Broadband Stakeholder Group, 2006. Bryne Potter, Amelia, and Andrew Clement. “A Desiderata for Broadband Networks in the Public Interest.” Paper presented at the 35th Research Conference on Communication, Information and Internet Policy, Arlington, VA, 2007. California Broadband Task Force. “The State of Connectivity—Building Innovation through Broadband.” In Final Report of the California Broadband Task Force: State of California, 2008. Canadian Radio-television and Telecommunications Commission. Communications Monitoring Report. Ottawa: Canadian Radio-Television and Telecommunications Commission, 2009. Castells, Manuel. The Informational City: Information Technology, Economic Restructuring, and the Urban-Regional Process. Oxford and Cambridge, MA: B. Blackwell, 1989. ——. The Rise of the Network Society. 2nd ed., Information Age, vol. 1. Malden, MA: Blackwell Publishers, 2000. Commission of the European Communities. Digital Divide Forum Report: Broadband Access and Public Support in under-Served Areas. Brussels, 2005. ——. Europe’s Digital Competitiveness Report, Volume 1: i2010—Annual Information Society Report 2009 Benchmarking i2010: Trends and Main Achievements. Brussels, 2009. Communications Workers of America. Speed Matters: Affordable High Speed Internet for All. Washington, DC: Communications Workers of America, 2007. Council of the European Union. Council Resolution on the Implementation of the eEurope 2005 Action Plan. Brussels: European Union, 2005. Crandall, Robert, William Lehr, and Robert Litan. The Effects of Broadband Deployment on Output and Employment: A Cross-Sectional Analysis of U.S. Data. Washington, DC: Brookings Institution, 2007. Crandall, Robert W., and Charles L. Jackson. The $500 Billion Opportunity: The Potential Economic Benefit of Widespread Diffusion of Broadband Internet Access. Washington: Brookings Institution, 2001. Deloitte. National Broadband Network—A User’s Perspective. Sydney: Deloitte, 2009. 270

bibliography

Duff, Alistair S., David Craig, and David A. McNeill. “A Note on the Origins of the ‘Information Society.’ ” Journal of Information Science 22 (1996): 117–122. Dutta, Soumitra, and Irene Mia. Global Information Technology Report 2009– 2010: ICT for Sustainability. Geneva: World Economic Forum and INSEAD, 2010. Dutton, William H., Jay G. Blumler, Nicholas Garnham, Robin Mansell, James Cornford, and Malcolm Peltu. “The Politics of Information and Communication Policy: The Information Superhighway.” In Information and Communication Technologies: Visions and Realities, edited by William H. Dutton and Malcolm Peltu, 387–405. Oxford: Oxford University Press, 1996. Dutton, William H., Jay G. Blumler, and Kenneth L. Kraemer, eds. Wired Cities: Shaping the Future of Communications, Communications Library. Boston: G. K. Hall, 1987. Epitiro. Australia Internet Performance Index—Summary Findings 2008 (Q4). Sydney: Epitiro, 2009. European Commission. “EU Guidelines on Public Funding of Broadband Networks in Member States—FAQs.” http://www.egovmonitor.com/ node/28328. ——. Europe and the Global Information Society: Recommendations of the HighLevel Group on the Information Society to the Corfu European Council (Bangemann Group). Brussels: European Commission, 1994. ——. “Knowledge Society.” http://tinyurl.com/ec-knowledge. ——. “Telecoms: Commission Endorses Amended Version of Austrian Broadband Access Market Definition.” http://tinyurl.com/telecomscommission. Ezell, Stephen, Robert Atkinson, Daniel Castro, and George Ou. The Need for Speed: The Importance of Next-Generation Broadband Networks. Washington, DC: The Information Technology and Innovation Foundation, 2009. Federal Communications Commission. Broadband Performance—OBI Technical Paper No. 4. Washington, DC: Federal Communications Commission, 2010. Flamm, Kenneth, Amy Friedlander, John Horrigan, and William Lehr. “Measuring Broadband: Improving Communications Policymaking through Better Data Collection.” Pew Internet Project. http://www.pewinternet.org/ Reports/2007/Measuring-Broadband/01-Executive-Summary.aspx. Ford, George S., Thomas M. Koutsky, and Lawrence J. Spiwak. The Broadband Performance Index: A Policy-Relevant Method of Comparing Broadband Adoption among Countries (Phoenix Center Policy Paper Number 29). Washington, DC: Phoenix Center for Advanced Legal and Economic Public Policy Studies, 2007. Fornefeld, Martin, Gilles Delaunay, and Dieter Elixmann. The Impact of Broadband on Growth and Productivity. Düsseldorf: MICUS Management Consulting GmbH, 2008. bibliography

271

Gillett, Sharon E., William H. Lehr, Carlos A. Osorio, and Marvin A. Sirbu. Measuring Broadband’s Economic Impact: Final Report. Washington, DC: US Department of Commerce, Economic Development Administration, 2006. Government of Australia. “21st Century Broadband.” http://www.dbcde.gov .au/__data/assets/word_doc/0017/112454/National_Broadband_Network_ policy_brochure.doc. Greenstein, Shane, and Ryan C. McDevitt. The Broadband Bonus: Accounting for Broadband Internet’s Impact on U.S. GDP. Kellogg School of Management and Department of Economics, Northwestern University, 2009. Gurstein, Michael. “Effective Use and the Community Informatics Sector: Some Thoughts on Canada’s Approach to Community Technology/Community Access.” In Seeking Convergence in Policy and Practice, edited by Marita Moll and Leslie Regan Shade, 223–244. Ottawa: Canadian Centre for Policy Alternatives, 2004. ——. “Effective Use: A Community Informatics Strategy Beyond the Digital Divide.” First Monday 8 (2003). Hanafizadeh, Payam, Mohammad Reza Hanafizadeh, and Mohsen Khodabakhshi. “Extracting Core ICT Indicators Using Entropy Method.” Information Society 25 (2009): 236–247. Hargittai, Eszter, and Amanda Hinnant. “Digital Inequality: Differences in Young Adults’ Use of the Internet.” Communication Research 35 (2008): 602–621. Hargittai, Eszter, and Steven Shafer. “Differences in Actual and Perceived Online Skills: The Role of Gender.” Social Science Quarterly 87 (2006): 432–448. Hargittai, Eszter, and Gina Walejko. “The Participation Divide: Content Creation and Sharing in the Digital Age.” Information, Communication & Society 11 (2008): 239–256. Horrigan, John. Home Broadband Adoption. Washington, DC: Pew Internet & American Life Project, 2009. Howard, Philip N., Ken Anderson, Laura Busch, and Dawn Nafus. “Sizing up Information Societies: Toward a Better Metric for the Cultures of ICT Adoption.” Information Society 25 (2009): 208–219. Information Society Technologies Advisory Group. “Shaping Europe’s Future through ICTs.” ftp://ftp.cordis.europa.eu/pub/ist/docs/istag-shaping-europefuture-ict-march-2006-en.pdf. International Telecommunication Union. Measuring the Information Society— The ICT Development Index. Geneva: International Telecommunication Union, 2009. ——. World Information Society Report. Geneva: International Telecommunication Union, 2006. ——. World Information Society Report 2007—Beyond WSIS. Geneva: International Telecommunication Union, 2007. ——. World Telecommunication/ICT Indicators Database. Geneva: International Telecommunication Union, 2008. 272

bibliography

Kolko, Jed. Does Broadband Boost Local Economic Development? San Francisco: Public Policy Institute of California, 2010. ——. “How Broadband Changes Online and Offline Behaviors.” Information Economics and Policy 22 (2010): 144–152. Lenard, Thomas M., and Randolph J. May. Net Neutrality or Net Neutering: Should Broadband Internet Services Be Regulated. New York: Springer, 2006. Market Clarity. Broadband Wars: The OECD’s International Broadband Arms Race. Sydney, 2007. Masuda, Yoneji. The Information Society. Washington, DC: World Future Society, 1981. Menou, Michel J., and Richard D. Taylor. “A ‘Grand Challenge’: Measuring Information Societies.” Information Society 22 (2006): 261–267. Middleton, Catherine A., and Jonathan Ellison. Understanding Internet Usage among Broadband Households: A Study of Household Internet Use Survey Data. Ottawa: Statistics Canada—Science, Innovation and Electronic Information Division Working Papers, 2008. Middleton, Catherine A., and Jordan Leith. “An Analysis of Canadians’ Scope of Internet Usage.” Paper presented at the Statistics Canada Socio-economic Conference, Ottawa, 2008. ——. “Intensity of Internet Use in Canada: Exploring Canadians’ Engagement with the Internet.” Paper presented at the Statistics Canada Socio-economic Conference, Ottawa, 2007. Middleton, Catherine A., Ben Veenhof, and Jordan Leith. Intensity of Internet Use in Canada: Understanding Different Types of Users. Ottawa: Statistics Canada—Business Special Surveys and Technology Statistics Division Working Papers, 2010. Mueller, Milton, and Becky Lentz. “Revitalizing Communication and Information Policy Research.” Information Society 20 (2004): 155–157. National Information Infrastructure Advisory Council. Common Ground: Fundamental Principles for the National Information Infrastructure: First Report of the National Information Infrastructure Advisory Council. Washington, DC: National Information Infrastructure Advisory Council, 1995. Nielsen. The Global Online Media Landscape—Identifying Opportunities in a Challenging Market. New York: The Nielsen Company, 2009. OECD. “Average Monthly Bit/Data Cap Size and Price Per Additional Mb, by Country, October 2009.” OECD. http://www.oecd.org/dataoecd/22/46/ 39575020.xls. ——. “Percentage of Fibre Connections in Total Broadband among Countries Reporting Fibre Subscribers, December 2009.” http://www.oecd.org/dataoecd/ 21/58/39574845.xls. OECD Directorate for Science Technology and Industry. Developments in Fibre Technologies and Investment. Paris: OECD Working Party on Communication Infrastructures and Services Policy, 2008. bibliography

273

——. OECD Policy Guidance on Convergence and Next Generation Networks. Seoul: OECD Ministerial Meeting on the Future of the Internet Economy, 2008. ——. Participative Web: User-Created Content. Paris: OECD, 2007. Ofcom. UK Fixed Broadband Speeds, November/December 2010—The Performance of Fixed-Line Broadband Delivered to UK Residential Consumers. London: Ofcom, 2011. Organisation for Economic Co-operation and Development. Broadband Growth and Policies in OECD Countries. Paris: OECD, 2008. ——. “Broadband Penetration, Historical Broadband Penetration Rates (December 2009).” http://www.oecd.org/dataoecd/22/12/39574779.xls. Otsuka, Yasuhiro. Mobile Broadband: Pricing and Services. Paris: OECD Directorate for Science, Technology and Industry—Committee for Information, Computer and Communications Policy, 2009. Pepper, Robert, Enrique J. Rueda-Sabater, Brian C. Boeggeman, and John Garrity. “From Mobility to Ubiquity: Ensuring the Power and Promise of Internet Connectivity . . . for Anyone, Anywhere, Anytime.” In Global Information Technology Report 2008–2009, edited by Soumitra Dutta and Irene Mia, 37–52. Geneva: World Economic Forum and INSEAD, 2009. Preston, Paschal, and Anthony Cawley. “Broadband Development in the European Union to 2012—A Virtuous Circle Scenario.” Futures 40 (2008): 812–821. Sackman, Harold, and Barry W. Boehm, eds. Planning Community Information Utilities. Montvale, NJ: AFIPS Press, 1972. Sackman, Harold, and Norman H. Nie, eds. The Information Utility and Social Choice. Montvale, NJ: AFIPS Press, 1970. Selian, Audrey N. “The World Summit on the Information Society and Civil Society Participation.” Information Society 20 (2004): 201–215. Statistics Canada. “Canadian Internet Use Survey, 2009.” http://www.statcan .gc.ca/daily-quotidien/100510/dq100510a-eng.htm. Stenberg, Peter, Mitch Morehart, Stephen Vogel, John Cromartie, Vince Breneman, and Dennis Brown. Broadband Internet’s Value for Rural America: Economic Research Report Number 78. Washington, DC: United States Department of Agriculture, 2009. UNCTAD Secretariat. Information Economy Report 2006—The Development Perspective. New York and Geneva: United Nations Conference on Trade and Development, 2006. ——. Information Economy Report 2007–2008—Science and Technology for Development: The New Paradigm of ICT. New York and Geneva: United Nations Conference on Trade and Development, 2007. Van Gaasbeck, Kristin A. “A Rising Tide: Measuring the Economic Effects of Broadband Use across California.” Social Science Journal 45 (2008): 691–699. 274

bibliography

Waverman, Leonard, and Kalyan Dasgupta. Connectivity Scorecard 2010. LECG, 2010. ——. “How to Maximize the Economic Impact of Mobile Communications: The Four Waves.” In Global Information Technology Report 2008–2009, edited by Soumitra Dutta and Irene Mia, 53–61. Geneva: World Economic Forum and INSEAD, 2009. Waverman, Leonard, Kalyan Dasgupta, and Nicholas Brooks. Connectivity Scorecard 2009. LECG, 2009. Webster, Frank. “Making Sense of the Information Age.” Information, Communication & Society 8 (2005): 439–458. World Bank. Information and Communications for Development 2009: Extending Reach and Increasing Impact. Washington, DC: World Bank, 2009.

chapter 2 Babe, R. E. “The Place of Information in Economics.” In Information and Communication in Economics, edited by R. E. Babe. Boston, Dordrecht & London: Kluwer Academic, 1994. Barzilai-Nahon, K. “Gaps and Bits: Conceptualizing Measurements for Digital Divide/s.” Information Society 22 (2006): 269–278. Bridges.org. “E-Readiness Assessment Tools Comparison,” February 28, 2005. http://www.bridges.org/files/active/0/ereadiness_tools_bridges_10Mar05 .pdf. Constantin, C., D. Grigorovici, K. Jayakar, R. Taylor, and J. Schement. “Infometrics: A Structural Equation Modeling Approach to Information Indicators and E-Learning Measurement, Part II.” Presented at the 32nd Telecommunications Policy Research Conference, Arlington, VA, 2004. ——. “Infometrics: A Structural Equation Modeling Approach to Information Indicators and ‘E-Readiness’ Measurement.” In Governing Communications Networks: Connecting Societies and Markets with IT, edited by B. Preissl and J. Mueller. Elsevier, 2006. Constantin, C., E. Park, Y. S. Cho, P. Liao, and R. Taylor. “Measuring and Comparing Changes in the ‘Digital Divide’ Through Time: A Structural Equation Modeling Approach.” Presented at the 16th Biennial Conference of the International Telecommunications Society, Beijing, 2006. Du, Dong, and Pang Qinghua. Comprehensive Evaluation of Modern Methods and Case Selection. Beijing: Tsinghua University Press, 2005. Everitt, B., S. Landau, and M. Leese. Clusteranalysis. London: Arnold, 2001. Giokas, D. I., and G. C. Pentzaropoulos. “Efficiency Ranking of the OECD Member States in the Area of Telecommunications: A Composite AHP/DEA Study.” Telecommunications Policy 32 (2008): 672–685. Giovannini, E. “Statistics and Politics in a ‘Knowledge Society.’ ” Social Indicators Research 86 (2008): 177–200. bibliography

275

Granger, C. and Y. Jeon. “A Time-Distance Criterion for Evaluating Forecasting Models.” International Journal of Forecasting 19 (2003): 9–215. Grigorovici, D., J. R. Schement, and R. Taylor. “Weighing the Intangible: Towards a Theory-Based Framework for Information Society Indices.” In Global Economy and Digital Society, edited by E. Bohlin, S. Levin, N. Sung, and C. Yoon. Amsterdam: Elsevier Science, 2004. Han, S. “Comparison of Undimensionalization in SPSS Cluster Analysis.” Science Mosaic 3 (2008): 229–231. Hao, Y., and J. Fan, eds. Methods and Application of Systems Engineering. Beijing: Science Press, 2007. Hou, J., and W. Yin. Regional Economic Analysis. Shanghai: Commercial Press, 2004. International Telecommunications Union. “Global Event on Measuring the Information Society, Final Report.” Partnership on Measuring ICT for Development. ITU: Geneva, 2008. http://www.itu.int/ITU-D/ict/ conferences/geneva08/Global_Event_final_report.pdf. ——. “Measuring the Information Society 2007: ICT Opportunity Index and World Telecommunication/ICT Indicators,” 2007. http://www.itu.int/ ITU-D/ict/publications/ict-oi/2007/index.html. International Telecommunications Union and Orbicom. “From the Digital Divide to Digital Opportunities Measuring Infostates for Development,” 2005. http://www.orbicom.ca/media/projets/ddi2005/index_ict_opp.pdf. International Telecommunications Union and UNCTAD. The Digital Opportunity Index, World Information Society Report 2007: Beyond WSIS. Geneva: ITU, 2007. Jin, J., and C. Xiong. “Digital Divide in Terms of National Informatization Quotient: The Perspective of Mainland China.” Presented at International Conference on The Digital Divide: Technology and Politics in the Information Age, Hong Kong, 2002. Jingxin, H., and Y. Weihong. Regional Economic Analysis Methods. Commercial Press, 2004. Li, Meijuan, and Chen Guohong. “Data Envelopment Analysis (DEA) of the Research and Application.” China Engineering Science 5 (2003): 88–94. Li, Yizhi, and Xu Xuanhua. The Number of Business Decision-Making Method. Beijing: Economic Science Press, 2003. Lin, Qining. Decision Analysis. Beijing: BUPT Press, 2003. Mutula, S., and van Brakel, P. “An Evaluation of E-Readiness Assessment Tools with Respect to Information Access: Towards an Integrated Information Rich Tool.” International Journal of Information Management 26 (2006): 212–223. Organisation for Economic Co-operation and Development. “Communications Outlook 2007.” OECD: Paris, 2007. http://213.253.134.43/oecd/pdfs/ browseit/9307021E.PDF. 276

bibliography

——. “Composite Indicators of Country Performance: A Critical Assessment.” STI Working Paper 2003/16 (Michael Freudenberg, JT00153477), 2003. ——. “Draft Guide to Information Society Measurement.” Working Party on Indicators for the Information Society in the Committee for Information, Computer and Communications Policy of Directorate for Science, Technology and Industry, 2005. http://www.escwa.un.org/wsis/meetings/ 7-0june/0B8Guide_IS_measurementsOECD05.pdf. ——. “The Future of the Internet Economy.” Policy Brief, OECD, June 2008. http://www.oecd.org/dataoecd/20/41/40789235.pdf. ——. “Guide to Measuring the Information Society 2007.” OECD: Paris, 2007. www.oecd.org/sti/measuring-infoeconomy/guide/. Partnership on Measuring ICT for Development. “Core ICT Indicators.” 2005. http://new.unctad.org/upload/docs/Core%20ICT%20Indicators_Eng.pdf. ——. “The Global Information Society: A Statistical View.” 2008. http://www .ifap.ru/library/book303.pdf. Pruulmann-Vengerfeldt, P. “Exploring Social Theory as a Framework for Social and Cultural Measurements of the Information Society.” Information Society 22 (2006): 303. Qin, Shoukang. Principle and Application of Comprehensive Evaluation. Beijing: Publishing House of Electronics Industry, 2002. Sanguthevar, Rajasekaran. “Efficient Parallel Hierarchical-Clustering Algorithms.” IEEE Transactions on Parallel and Distributed Systems, 6 (2005). Sciadas, G. “The Interdependence of Information Society-Measurements and Policies.” Presented at 2008 Global Event on Measuring the Information Society, Geneva, 2008. http://tinyurl.com/sciadas-interdependence. ——, ed. “Monitoring the Digital Divide . . . and Beyond.” Montreal: Orbicom, 2003. http://www.orbicom.ca/projects/ddi2002/2003_dd_pdf_en.pdf. Sheng, Zhaohan, Lu Gang, and Jiang Yiqing. DEA Theory, Method and Application. Beijing: Science Press, 1996. Sicherl, P. “Different Statistical Measures Provide Different Perspectives on Digital Divide.” Presented at VI Conference of the European Sociological Association, Murcia, 2003. ——. “Integrating Comparisons across Time and Space: Methodology and Application to Disparities within Yugoslavia.” Journal of Public Policy 12 (1992): 377–403. ——. “The Intertemporal Aspect of Well-Being and Societal Progress.” Presented at OECD-JRC workshop, 2006. ——. Methods of Measuring Disparity Between Men and Women. Santo Domingo: United Nations International Research and Training Institute for the Advancement of Women, 1989. ——. Slovenija zdaj (Slovenia Now). Ljubljana, Gospodarski vestnik, 1990. ——. “Time Distance—A Missing Perspective in Comparative Analysis.” e-WISDOM 2a/2004. bibliography

277

——. “Time Distance as an Additional Measure of Discrepancy Between Actual and Estimated Values in Time Series Models.” Presented at the International Symposium on Economic Modeling, Washington, World Bank, 1994. ——. Time Distance as a Dynamic Measure of Disparities in Social and Economic Development. Kyklos XXVI, Fasc. 3, 1973. ——. “Time Distance Measure in Economic Modeling.” Presented at the International Symposium on Economic Modeling, University of London, London, 1997. Taylor, R. “The Nature and Measurement of Information: Two Grand Challenges.” Presented at 16th Biennial Conference of the International Telecommunications Society, Beijing, 2006. Teltscher, S. “Partnership Core List of indicators.” Presented at Global Event on Measuring the Information Society, Geneva, 2008. http://new.unctad.org/ upload/Global%20Event%202008/Session3_Teltscher_Partnership.ppt. United Nations. Measuring ICT: The Global Status of ICT Indicators— Partnership on Measuring ICT for Development. United Nations Information and Communication Technologies Task Force, 2005. http://tinyurl.com/ un-measuring-ict. ——. “Information Economy Report 2007–2008: Science and Technology for Development—The New Paradigm of ICT.” Prepared by the UNCTAD Secretariat. Geneva: United Nations, 2007. http://www.unctad.org/en/docs/ sdteecb20071_en.pdf. UNCTAD. Manual for the Production of Statistics on the Information Economy. Geneva: UNCTAD, 2007. http://measuring-ict.unctad.org. UNESCO Institute for Statistics, Montreal. Measuring and Monitoring the Information and Knowledge Societies—Statistical Challenge. Montreal: UNESCO Publications for the World Summit for the Information Society, 2003. http://tinyurl.com/unesco-statistical-challenge. Vehovar, V., P. Sicherl, T. Husing, and V. Dolnicar. (2006). “Methodological Challenges of Digital Divide Measurements.” Information Society 22 (2006): 279–290. Wei, Quanling. Relatively Efficient Evaluation of the DEA Method. Beijing: Chinese People’s University Press, 1988. World Economic Forum. Global Information Technology Report 2007–2008. Geneva: World Economic Forum, 2008. http://tinyurl.com/wef-report. Wu, Yu Hua, Liu Xihua, and Guo Junpeng. Economic Management Methods. Beijing: Economic Science Press, 2008. Yong, H., and F. Junhui. Systems Engineering Methods and Applications. Beijing: Science Press, 2007. Xue, W., and J. Liu. “The Relationship System Analysis of the Main Influence on the Digital Divide.” Theory and Practice of Systems Engineering 5 (2008): 85–91. ——. “An analysis of relative structure on main influence factors of digital divide.” Systems Engineering-Theory & Practice 5 (2008). 278

bibliography

Zhang, B., and R. Taylor. “Modeling IT for Growth: How China’s Informatization Index System Informs Its Development Policy.” Presented at PTC05, Broadband and Content: From Wires to Wireless conference, Honolulu, 2005. Zhang, X., and R. Wang. “Multi-Dimensional Perspective of National Informatization Evaluation System.” Modern Information, August 2007. Zhihang, T., and Y. Baoan. “Investigation and Application of Hierarchical Clustering in Customer Relationship Management.” Computer Engineering and Applications 13 (2007): 220–235.

chapter 3 Bauer, S., D. Clark, and W. Lehr. “The Evolution of Internet Congestion.” Paper prepared at Telecommunications Policy Research Conference, Arlington, VA, September 2009. Briscoe, B. “Flow Rate Fairness: Dismantling a Religion.” SIGCOMM Computer Communication Review 37 (2007): 63–74. Broadband Working Group. “Broadband Incentive Problem.” White paper by the MIT Communications Futures Program, September 2005. http://cfp.mit .edu/publications/CFP_Papers/Incentive_Whitepaper_09-28-05.pdf. Camp, J. Trust and Risk in Internet Commerce. Cambridge, MA: MIT Press, 2001. Cho, K., K. Fukuda, H. Esaki, and A. Kato. “Observing Slow Crustal Movement in Residential User Traffic.” In Proceedings of the 2008 ACM CoNEXT Conference, Madrid, Spain, December 2008. ——. “The Impact and Implications of the Growth in Residential User-to-User Traffic.” In Proceedings of the 2006 Conference on Applications, Technologies, Architectures, and Protocols for Computer Communications, September 2006. CSTB. Looking Over the Fence at Networks: A Neighbor’s View of Networking Research. National Academy Press, 2001. Dutz, M., J. Orszag, and R. Willig. “The Substantial Consumer Benefits of Broadband Connectivity for U.S. Households.” Compass Lexecon, A Study Commissioned by the Internet Innovation Institute, July 2009. Faratin, P., D. Clark, P. Gilmore, A. Berger, and W. Lehr. “Complexity of Internet Interconnections: Technology, Incentives and Implications for Policy.” Paper prepared at Telecommunications Policy Research Conference, Arlington, VA, September 2007. Floyd, S., and M. Allman. “RFC 5290: Comments on the Usefulness of Simple Best-Effort Traffic.” Network Working Group, Internet Engineering Task Force, July 2008. http://www.faqs.org/rfcs/rfc5290.html. Fukuda, K., K. Cho, and H. Esaki, “The Impact of Residential Broadband Traffic on Japanese ISP Backbones.” SIGCOMM Computer Communication Review 35 (2005): 15–22. bibliography

279

Greenstein, S. and R. McDevitt. “The Broadband Bonus: Accounting for Broadband Internet’s Impact on U.S. GDP.” White paper, Technology Policy Institute, Washington, D.C., January 2009. http://tinyurl.com/ greenstein-broadband. Jacobson, V., and M. J. Karels. “Congestion Avoidance and Control.” ACM Computer Communication Review; Proceedings of the Sigcomm ’88 Symposium in Stanford, CA, August 1988, vol. 18, 314–329. Jonsson, D., S. Berglund, P. Almström, and S. Algers. “The Usefulness of Transport Models in Swedish Planning Practice.” Transport Reviews: A Transnational Transdisciplinary Journal 31 (2011): 251–265. Lehr, W., C. Osorio, S. Gillett, and M. Sirbu. “Measuring Broadband’s Economic Impact.” Paper prepared at Telecommunications Policy Research Conference, Arlington, VA, September 2005. Lehr, W., J. Peha, and S. Wilkie. Special Section on Network Neutrality: International Journal of Communication 1 (2007). http://ijoc.org/ojs/index .php/ijoc/issue/view/1. Leiner, B. M., V. G. Cerf, D. D. Clark, R. E. Kahn, L. Kleinrock, D. C. Lynch, J. Postel, L. G. Roberts, and S. Wolff. “A Brief History of the Internet.” Communications of the ACM 40 (1997). Licklider, J. “Topics for Discussion at the Forthcoming Meeting, Memorandum for: Members and Affiliates of the Intergalactic Computer Network.” April 1963. Washington, DC, Advanced Research Projects Agency. Lyon, M. Where Wizards Stay Up Late. New York: Simon & Schuster, 2004. Mathis, M. “Rethinking TCP Friendly”, March 2009. http://staff.psc.edu/ mathis/papers/TSVAREA73.pdf. Ohm, P. “The Rise and Fall of Invasive ISP Surveillance.” University of Illinois Law Review 2009 (2009): 1417–1496. Peterson, L. L., and B. Davie. Computer Networks: A Systems Approach. 4th ed. New York: Morgan Kaufmann, 2007. Popescu, Alin. “Deja Vu All Over Again: Cables Cut in the Mediterranean.” http://www.renesys.com/blog/2008/12/deja-vu-all-over-again-cables .shtml#more. Renesys Corp. “The Day the YouTube Died: What Happened and What We Might Do About It.” http://www.renesys.com/tech/presentations/pdf/ nanog43-hijack.pdf. ——. “Internet Captivity and the De-peering Menace.” http://www.renesys .com/tech/presentations/pdf/nanog-45-Internet-Peering.pdf. Stallings, W. Operating Systems: Internals and Design Principles. 6th ed. Boston: Prentice Hall, 2008. Subramanian, M. Network Management: Principles and Practice. Reading, MA: Addison Wesley, 1999. Tirole, J. The Theory of Industrial Organization. Cambridge, MA: The MIT Press, 1988. 280

bibliography

Varian, H., R. Litan, A. Elder, and J. Shutter. “The Net Impact Study: The Projected Economic Benefits of the Internet in the United States, United Kingdom, France, and Germany.” Research report, funding support from Cisco Systems is acknowledged, January 2002. Wik-Consult/Rand Europe/CLIP/CRID/GLOCOM. Comparisons of Privacy and Trust Policies in the Area of Electronic Communications, Final Report, report prepared for the European Commission, July 2007. http://ec.europa .eu/information_society/policy/ecomm/doc/library/ext_studies/privacy_trust_ policies/final_report_20_07_07_pdf.pdf.

chapter 4 Ahn, Hyungtaik, and Myenong-Ho Lee. “An Econometric Analysis of the Demand for Access to Mobile Telephone Networks.” Information Economics and Policy 11 (1999): 297–305. Andonova, Veneta. “Mobile Phones: the Internet and the Institutional Environment.” Telecommunications Policy 30 (2006): 29–45. Atkinson, Robert D., Daniel K. Correa, and Julie A. Hedlund. Explaining International Broadband Leadership. Washington, DC: The Information Technology and Innovation Foundation, 2008. Burnstein, David E., and Debra J. Aron. “Broadband Adoption in the United States: An Empirical Analysis.” In Down to the Wire: Studies in the Diffusion and Regulation of Telecommunications Technologies, edited by Allan L. Shampine, 119–138. NY: Nova Science Publishers, 2003. Cava-Ferreruela, Inmaculada, and Antonio Alabau- Munˇoz. “Broadband Policy Assessment: A Cross-National Empirical Analysis.” Telecommunications Policy 30 (2006): 445–463. Chaudhuri, Anindya, Kenneth Flamm, and John Horrigan. “An Analysis of the Determinants of Internet Access.” Telecommunications Policy 29 (2005): 731–755. Church, Jeffrey, and Neil Gandal. “Platform Competition in Telecommunications.” In The Handbook of Telecommunications Vol. 2.: Technology Evolution and the Internet, edited by Sumit Kumar Majumdar, Martin Cave, and Ingo Vogelsang, 119–154. Amsterdam: Elsevier, 2005. Crandall, Robert W. “Broadband Communications.” In The Handbook of Telecommunications Vol. 2.: Technology Evolution and the Internet, edited by Sumit Kumar Majumdar, Martin Cave, and Ingo Vogelsang, 156–190. Amsterdam: Elsevier, 2005. de Ridder, John. “Catching-up in Broadband: What will It Take?” Paper presented at the 35th Research Conference on Communication, Information and Internet Policy, Arlington, VA, 2007. Denni, Mario, and Harald Gruber. “The Diffusion of Broadband Telecommunications: The Role of Competition.” Paper presented at International Telecommunication Conference, Pontevedra, 2005. bibliography

281

Distaso, Walter, Paolo Lupi, and Fabio M. Maneti. “Platform Competition and Broadband Uptake: Theory and Empirical Evidence from the European Union.” Information Economics and Policy 18, no. 1 (2006): 87–106. Fransman, Martin, ed. Global Broadband Battles; Why the U.S. and Europe Lag While Asia Leads. Stanford: Stanford University Press, 2006. Gans, Joshua S., Stephen P. King, and Julian Wright. “Wireless Communications.” In The Handbook of Telecommunications Vol. 2: Technology Evolution and the Internet, edited by Sumit Kumar Majumdar, Martin Cave, and Ingo Vogelsang, 243–287. Amsterdam: Elsevier, 2005. Garcia-Murillo, Martha A. “International Broadband Deployment: The Impact of Unbundling.” Communications & Strategies 57 (2005): 83–108. Grosso, Marcelo. “Determinants of Broadband Penetration in OECD Nations.” Paper presented at the Australian Communications Policy and Research Forum, 2006. Gruber, Harold, “Competition and Innovation: The Diffusion of Mobile Telecommunications in Central and Eastern Europe.” Information Economics and Policy 13 (2001): 19–34. Gruber, Harold, and Frank Verboven. “The Evolution of Markets under Entry and Standards Regulation: The Case of Global Mobile Telecommunications.” International Journal of Industrial Organization 19 (2001): 1189–1212. Horrigan, John, and Aron Smith. Home Broadband Adoption 2007. Washington, DC: Pew Research Center, 2007. International Telecommunication Union. Digital.life. Geneva: ITU, 2006. ——. The Internet of Things. Geneva: ITU, 2005. ——. Promoting Broadband. Geneva: ITU, 2003. Lee, Sangwon, and Justin S. Brown. “A Cross-country Analysis of Fixed Broadband Deployment: Examination of Adoption Factors and Network Effect.” Paper presented at the 2009 Association for Education in Journalism and Mass Communication Annual Convention, Boston, MA, 2009. ——. “Examining Broadband Adoption Factors: An Empirical Analysis between Countries.” Info 10 (2008): 25–39. Lee, Sangwon, Sylvia M. Chan-Olmsted, and Heejung Kim. “The Deployment of Third-Generation Mobile Services: A Multinational Analysis of Contributing Factors.” Paper presented at the 2007 Association for Education in Journalism and Mass Communication Annual Convention, Washington DC, 2007. Liikanen, Jukka, Paul Stoneman, and Otto Toivanen. “Intergenerational Effects in the Diffusion of New Technology: The Case of Mobile Phones.” International Journal of Industrial Organization 22 (2004): 1137–1154. Madden, Gary, Grant Coble-Neal, and Brian Dalzell. “A Dynamic Model of Mobile Telephony Subscription Incorporating a Network Effect.” Telecommunications Policy 28 (2004): 133–144. Noam, Eli M. “The Next Frontier for Openness: Wireless Communications.” In Competition for the Mobile Internet, edited by Eli M. Noam and Dan Steinbock, 21–37. Norwell, MA: Kluwer Academic Publishers, 2003. 282

bibliography

Organisation for Economic Co-operation and Development. Communication Outlook 2009. Paris: OECD, 2009. Organisation for Economic Co-operation and Development. Mobile Multiple Play. Paris: OECD, 2007). Trkman, Peter, Borka Blazic, and Tomas Turk. “Factors of Broadband Development and the Design of a Strategic Policy Framework.” Telecommunications Policy 32 (2008): 101–115. Weiser, Marc. “The Computer for the 21st Century.” Scientific American (September 1991): 94–100.

chapter 5 Bauer, J. M. “Learning from Each Other: Promises and Pitfalls of Benchmarking in Communications Policy.” Info: The Journal of Policy, Regulation and Strategy for Telecommunications, Information and Media 12 (2010): 8–20. Bauer, J. M., J. Kim, and S. S. Wildman. “Regulation and the Diffusion of Broadband in the OECD.” Paper presented at the University of FloridaLondon Business School joint conference, 2005. Bauer, J. M., S. Kim, and T. Koh. “International Comparisons of Mobile Voice and Data Prices: Coping with Heterogeneity and Differentiation.” Unpublished manuscript, East Lansing, Michigan: Michigan State University, 2010. Berkman Center. Next Generation Connectivity: A Review of Broadband Internet Transitions and Policy from Around the World. Cambridge, MA: Berkman Center for Internet and Society, Harvard University, 2010. Cherry, B. A., and J. M. Bauer. “Institutional Arrangements and Rate Rebalancing: Evidence from the United States and Europe.” Information Economics and Policy 14 (2002): 495–517. CTIA. “The United States and World Wireless Markets.” Written ex parte communication, RM-11361; GN Docket No. 09-51; WC Docket No. 07-52. Washington, DC: CTIA—The Wireless Association, 2009. DeMaagd, K., and J. M. Bauer. “Modeling the Dynamic Interactions of Agents in the Provision of Network Infrastructure.” Information Systems Frontiers, DOI 10.1007/s10796-010-9244-2, 2010. Economist Intelligence Unit. E-Readiness Rankings 2009. London: Economist Intelligence Unit, 2009. Fransman, M., ed. Global Broadband Battles: Why the U.S. and Europe Lag While Asia Leads. Stanford, CA: Stanford University Press, 2006. Greenstein, S., and R. McDevitt. “Evidence of a Modest Price Decline in US Broadband Services.” Working Paper #0102, Northwestern University: The Center for the Study of Industrial Organization. International Telecommunication Union. World Information Society Report 2007. Geneva: International Telecommunication Union, 2007. Organisation for Economic Co-operation and Development. Communications Outlook 2009. Paris: OECD, 2009. bibliography

283

Ostrom, E. Understanding Institutional Diversity. Princeton and Oxford: Princeton University Press, 2005. Ragin, C. C. Fuzzy-set Social Science. Chicago, IL: University of Chicago Press, 2000. Rodrik, D. One Economics Many Recipes: Globalization, Institutions, and Economic Growth. Princeton, NJ: Princeton University Press, 2007. Sawhney, H. “Dynamics of Infrastructure Development: The Role of Metaphors, Political Will and Sunk Investment.” Media, Culture and Society 23 (2001): 33–51. Sen, A. Development as Freedom. New York: Knopf, 1999. Sterman, J. D. Business Dynamics: Systems Thinking and Modeling for a Complex World. Boston, MA: McGraw-Hill, 2000. Triplett, J. Handbook on Hedonic Indexes and Quality Adjustments in Price Indexes: Special Application to Information Technology Products. Paris: OECD, 2006. Waverman, L., and K. Dasgupta. Connectivity Scorecard 2009. London: LECG and NokiaSiemens, 2009. http://www.connectivityscorecard.org/images/ uploads/media/TheConnectivityReport2009.pdf. World Bank. World Development Indicators. http://data.worldbank.org/ data-catalog/world-development-indicators. World Economic Forum. The Networked Readiness Index. Davos, Switzerland: World Economic Forum, 2009. Xavier, P. “Price Discount Schemes and Inter-Country Comparisons of Telecommunication Tariffs.” Telecommunications Policy 22 (1998): 289–314.

chapter 6 Bar, F., and A. M. Riis. “Tapping User-Driven Innovation: A New Rationale for Universal Service.” Information Society 16 (2000): 99–108. Connect Ohio. “Rural Community Building.” n.d. http://ruralcommunitybuilding .fb.org/2010/06/07/grassroots-efforts-bring-broadband-to-ohio-rural-areas/. Federal Communications Commission. National Broadband Plan. April 2010. http://www.broadband.gov/plan/executive-summary/. ——.Universal Service, n.d. Government Printing Office. Series H 878-893, R 93-105, R 1-12, Historical Statistics of the United States, Colonial Times to 1970. Washington DC: GPO, 1975. Hamilton, A. “Objections to the Proposed Constitution from Extent of Territory Answered.” FEDERALIST No. 14, November 30, 1787, in A. Hamilton et al., Federalist Papers. New York, Bantam Books, 1982. Horrigan, J. “Closing the Broadband Divide.” Pew Internet and American Life Project, August 2007. http://tinyurl.com/horrigan-closing-divide. ——. “Home Broadband Adoption.” Pew Internet and American Life Project, 2009. http://www.pewinternet.org/~/media/Files/Reports/2009/HomeBroadband-Adoption-2009.pdf. 284

bibliography

Jayakar, K., and H. Sawhney. “Universal Access in the Information Economy: Tracking Policy Innovations Abroad.” Benton Foundation Universal Service Report, 2007. http://benton.org/node/5598. Kvasny, L., et al. “Communities, Learning and Democracy in the Digital Age.” Journal of Community Informatics 2, no. 2 (2006). http://ci-journal.net/index .php/ciej/article/view/341/247. Luo, M. “Frustration and Despair as Job Search Drags On.” New York Times, July 17, 2010. http://tinyurl.com/luo-job-search. Open Cape. “Creating Regional Broadband Opportunities.” n.d. http://www .opencape.com/the-project. Organization for Economic Co-operation and Development. “OECD Broadband Portal.” June 2010. http://tinyurl.com/oecd-broadband-portal. Schement, J. R. “Broadband, Internet, and Universal Service: Challenges to the social contract of the twenty-first century.” In . . . And Communications for All: A Policy Agenda for a New Administration, edited by A. Schejter, 3–27. Lanham, MD: Lexington Books, 2009. ——. “Telephone Penetration at the Margins: An Analysis of the Period Following the Breakup of AT&T, 1984–1994.” In Progress in Communication Sciences, Vol. XV: Advances in Telecommunications, edited by H. Sawhney and G. A. Barnett, 187–215. Stamford, CT: Ablex, 1998. Schement, J. R., and M. Tate. Rural America in the Digital Age. Columbia, MO: Rural Policy Research Institute, 2003. Strover, S., G. Chapman, and J. Waters. “Beyond Community Networking and CTC’s: Access Development and Public Policy.” Telecommunications Policy 28 (2004): 465–485. United States Department of Agriculture. Connecting Rural America, USDA Broadband Initiative Program. Washington, DC: USDA, 2010.

chapter 7 Alvarez, R., T. Hall, and M. Llewellyn. “On American Voter Confidence.” University of Arkansas at Little Rock Law Review 29 (2007): 651–668. Aufderheide, P. “Cable Television and the Public Interest.” Journal of Communication 42 (1992): 52–65. Balkin, J. “Digital Speech and Democratic Culture: A Theory of Freedom of Expression for the Information Society.” New York University Law Review 79 (2004): 1–55. Blumler, J. The Role of Public Policy in the New Television Marketplace. Washington, DC: Benton Foundation, 1989. Brant, K., M. Huizenga, and R. van Meerten. “The New Canals of Amsterdam: An Exercise in Local Electronic Democracy.” Media, Culture & Society 18 (1996): 233–247. Breindl, Y. “Critique of the Democratic Potentialities of the Internet: A Review of Current Theory and Practice.” tripleC 8 (2010): 43–59. bibliography

285

Broache, A., and D. McCullagh. “Technology Voters’ Guide: Barack Obama.” CNET, January 2, 2008. http://tinyurl.com/broache-tech-voter-guide. Castells, M. “Communication, Power and Counter-power in the Network Society.” International Journal of Communication 1 (2007): 238–266. ––––––. The Rise of the Network Society. 2nd ed. Chichester, UK: WileyBlackwell, 2010. Curran, J., and J. Seaton. Power Without Responsibility: The Press, Broadcasting and New Media in Britain. 6th ed. London: Routledge, 2003. Downey, J. “Surveillance from Below: The Internet and the Intifada.” In Ideologies of the Internet, edited by K. Sarikakis and D. Thussu, 147–162. Cresskill, NJ: Hampton Press, 2006. Entman, R., and S. Wildman. “Reconciling Economic and Non-Economic Perspectives on Media Policy: Transcending the ‘Marketplace of Ideas.’ ” Journal of Communication 42 (1992): 5–19. Frieden, R. “Lessons from Broadband Development in Canada, Japan, Korea and the United States.” Telecommunications Policy 29, no. 8 (2005): 595–613. Friedland, A. “Electronic Democracy and the New Citizenship.” Media, Culture & Society 18 (1996): 185–212. Garnham, N. “Universal Service.” In Telecom Reform: Principles, Policies and Regulatory Practices, edited by W. Melody, 207–212. Denmark: Technical University of Denmark, 2001. Giacomello, G. National Governments and the Control of the Internet: A Digital Challenge. Oxon, UK: Routledge, 2005. Greenberg, B., and M. Salwen. “Mass Communication Theory and Research: Concepts and Models.” In An Integrated Approach to Communication Theory and Research, edited by M. Salwen and D. Stacks, 63–78. Mahwah, NJ: Lawrence Erlbaum, 1996. Hacker, K. “Missing Links in the Evolution of Electronic Democratization.” Media, Culture & Society 18 (1996): 213–232. Hill, L. “Compulsory Voting: Residual Problems and Potential Solutions.” Australian Journal of Political Science 37 (2002): 437–455. Hirczy, W. “The Impact of Mandatory Voting Laws on Turnout: A QuasiExperimental Approach.” Electoral Studies 13 (1994): 64–76. Jones, S. CyberSociety: Computer-Mediated Communication and Community. Thousand Oaks, CA, London, and New Delhi: Sage Publications, 1995. Katz, E. “Communications Research since Lazarsfeld.” Public Opinion Quarterly 51 (1987): S25–S45. Martin, J. The Wired Society: A Challenge for Tomorrow. Englewood Cliffs, NJ: Prentice Hall, 1978. Meinrath, S. “2008 OECD Broadband Statistics. How’s the US Doing? Survey Says – Stagnation and Decline.” May 20, 2008. http://www.saschameinrath .com/2008/may/20/2008_oecd_broadband_statistics_hows_us_doing_survey_ says_stagnation_and_decline. 286

bibliography

Napoli, P. “The Marketplace of Ideas Metaphor in Communications Regulation.” Journal of Communication 49 (1999): 151–169. Pace, N. “China Surpasses U.S. in Internet Use.” Forbes, April 3, 2006. http:// www.forbes.com/2006/03/31/china-internet-usage-cx_nwp_0403china.html. Raman, K. “Cyberdemocracy and the Public Sphere—Chat Rooms as Agoras of the Modern World?” In Philosophical Perspectives on Globalization, edited by P. Cam, R. Ibana, and P. Duc, 77. Korean National Commission for UNESCO, 2006. Schejter, A. “Jacob’s Voice, Esau’s Hands: Transparency as a First Amendment Right in an Age of Deceit and Impersonation.” Hofstra Law Review 35, no. 2, (2007): 1489–1518. ––––––. “From all my teachers I have grown wise, and from my students more than anyone else”: What Lessons Can the US Learn from Broadband Policies in Europe? International Communication Gazette 71, no. 5 (2009): 429–445. ––––––. “International Benchmarks: The Crisis in U.S. Communications Policy Through a Comparative Lens.” In . . . And Communications for All: A Policy Agenda for a New Administration, edited by A. Schejter, 57. Lanham, MD: Lexington Books, 2009. Schejter, A., and S. Lee. “The Evolution of Cable Regulatory Policies and Their Impact: A Comparison of South Korea and Israel.” Journal of Media Economics 20, no. 1 (2007): 1–28. Schejter, A., and M. Yemini. “ ‘Justice, and only justice, you shall pursue:’ Network Neutrality, the First Amendment and John Rawls’ Theory of Justice.” Michigan Telecommunications and Technology Law Review 14, no. 1 (1992): 137–174. Shannon, C., and W. Weaver. The Mathematical Model of Communications. Urbana: University of Illinois Press, 1949. Siebert, F., T. Peterson, and W. Schramm. Four Theories of The Press: The Authoritarian, Libertarian, Social Responsibility, and Soviet Communist Concepts of What the Press Should Be and Do. Urbana, IL: University of Illinois Press, 1956. Stiglitz, J., A. Sen, and J.-P. Fitoussi. Report by the Commission on the Measurement of Economic Performance and Social Progress, 2009. http://www.stiglitz-senfitoussi.fr/documents/rapport_anglais.pdf. Tricycle. “Mitchell Kapor on Dharma, Democracy, and the Information Superhighway.” 1994. http://www.kapor.com/writing/Tricycle-interview .htm. Van Dijk, J. The Network Society. 2nd ed. London, Thousand Oaks, CA, and New Delhi: Sage Publications, 2006. Vegh, S. “Profits Over Principles: The Commercialization of the Democratic Potentials of the Internet.” In Ideologies of the Internet, edited by K. Sarikakis and D. Thussu, 62. Cresskill, NJ: Hampton Press, 2006. bibliography

287

The White House. “Promoting Innovation and Competitiveness: President Bush’s Technology Agenda.” June 24, 2004. http://georgewbush-whitehouse .archives.gov/infocus/technology/.

chapter 8 Baker, Edwin C. Media Concentration and Democracy: Why Ownership Matters. New York: Cambridge University Press, 2007. Bagdikian, Ben. The New Media Monopoly. Boston: Beacon Press, 2004. Compaine, Benjamin M., and Douglas Gomery. Who Owns the Media? Competition and Concentration in the Mass Media Industry. 3rd ed. New Jersey: Erlbaum, 2000. Lessig, Lawrence. Free Culture: How Big Media Uses Technology and the Law to Lock Down Culture and Control Creativity. New York: Penguin. 2004. McChesney, Robert W. The Problem of the Media: U.S. Communication Politics in the 21st Century. New York: Monthly Review Press, 2004. Noam, Eli M. Media Ownership and Concentration in America. New York: Oxford University Press, 2009. Thierer, Adam D. Media Myths: Making Sense of the Debate over Media Ownership. Washington DC: Progress & Freedom Foundation, 2005.

chapter 9 Carroll, Sean M. “Main Dish with a Side of Voluntary Commitments: Dish Network-DirecTV Revisited.” Administrative Law Review 61 (2009). Federal Communications Commission. 2000 Biennial Regulatory Review Spectrum Aggregation Limits for Commercial Mobile Radio Services. Report and Order. 16 F.C.C.R. 22668 (2001). ——. 2006 Quadrennial Regulatory Review—Review of the Commission’s Broadcast Ownership Rules and other Rules Adopted Pursuant to Section 202 of the Telecommunications Act of 1996. Report and Order and Order on Reconsideration. 23 F.C.C.R. 2010 (2007). ——. Annual Report and Analysis of Competitive Market Conditions With Respect to Mobile Wireless, Including Commercial Mobile Services. 14th Report. 25 F.C.C.R. 11407 (2010). ——. Applications Filed for the Transfer of Control of Embarq Corporation to Centurytel, Inc. Memorandum Opinion and Order. 24 F.C.C.R. 8741 (2009). ——. Applications of Comcast Corporation, General Electric Company and NBC Universal, Inc. for Consent to Assign Licenses and Transfer Control of Licensees. MB Docket No. 10-56, Memorandum Opinion and Order. 26 F.C.C.R. 4238 (2011). ——. Application of Echostar Communications Corporation, (a Nevada Corporation), General Motors Corporation, and Hughes Electronics Corporation (Delaware Corporations) (Transferors) and Echostar Communications 288

bibliography

Corporation (a Delaware Corporation)(Transferee). Hearing Designation Order. 17 F.C.C.R. 20559 (2002). ——. Applications for Consent to the Transfer of Control of Licenses XM Satellite Radio Holdings Inc., Transferor to Sirius Satellite Radio Inc., Transferee. Memorandum Opinion and Order and Report and Order. 23 F.C.C.R. 12348 (2008). ——. Applications of AT&T Wireless Services, Inc. and Cingular Wireless Corporation. Memorandum Opinion and Order. 19 F.C.C.R. 21522 (2004). ——. Applications of Cellco Partnership D/B/A Verizon Wireless and Atlantis Holdings LLC, for Consent to Transfer Control of Licenses, Authorizations, and Spectrum Manager and De Facto Transfer Leasing Arrangements. Memorandum Opinion and Order and Declaratory Ruling. 23 F.C.C.R.17444 (2008). ——. Applications of Cellco Partnership d/b/a Verizon Wireless and Atlantis Holdings LLC For Consent to Transfer Control of Licenses, Authorizations, and Spectrum Manager and De Facto Transfer Leasing Arrangements and Petition for Declaratory Ruling that the Transaction is Consistent with Section 310(b)(4) of the Communications Act. WT Docket No. 08-95, Protective Order. 23 F.C.C.R. 11154 (WTB 2008). ——. Appropriate Framework for Broadband Access to the Internet over Wireline Facilities. Report and Order and Notice of Proposed Rulemaking. 20 F.C.C.R. 14853 (2005). ——. AT&T Inc. and BellSouth Corporation Application for Transfer of Control. Memorandum Opinion and Order. 22 F.C.C.R. 5662 (2007). ——. Commissioner Robert M. McDowell Announces Staff Change. Public Notice (released September 18, 2009). ——. Constellation, LLC, Carlyle Panamsat I, LLC, Carlyle Panamsat II, LLC, Pep Pas, LLC, And Peop Pas, Llc, Transferors, and Intelsat Holdings, Ltd., Transferee, Consolidated Application For Authority to Transfer Control of Panamsat Licensee Corp. and Panamsat H-2 Licensee Corp. IB Docket No. 05-290, Memorandum Opinion and Order. 21 F.C.C.R.7368 (2006). ——. Effects of Communications Towers on Migratory Birds. Notice of Proposed Rulemaking. 21 F.C.C.R. 13241 (2006). ——. Examination of the Current Policy Concerning the Treatment of Confidential Information Submitted to the Commission. Report and Order. 13 F.C.C.R. 24816 (1998). ——. Existing Shareholders of Citadel Broadcasting Corp. and of The Walt Disney Co., etc. for Consent to Transfers of Control. Memorandum Opinion and Order and Notice of Apparent Liability. 22 F.C.C.R. 7083 (2007). ——. General Motors Corporation and Hughes Electronics Corporation, Transferors, and the News Corporation Limited, Transferee. Memorandum Opinion and Order. 19 F.C.C.R. 473 (2004). ——. Harvard’s Berkman Center to Conduct Independent Review of Broadband Studies to Assist FCC. Public Notice (released July 14, 2009). bibliography

289

——. Implementation of Section 6002(B) of the Omnibus Budget Reconciliation Act of 1993, Annual Report and Analysis of Competitive Market Conditions With Respect to Commercial Mobile Services. Twelfth Report. 23 F.C.C.R. 2241 (2008). ——. IP-Enabled Services, WC Docket No. 04-36, E911 Requirements for IP-Enabled Service Providers, WC Docket No. 05-196. First Report and Order and Notice of Proposed Rulemaking, 20 F.C.C.R. 10245 (2005). ——. Petition of Qwest Corporation for Forbearance Pursuant to 47 U.S.C. § 160(c) in the Omaha Metropolitan Statistical Area. WC Docket No. 04-223, Memorandum Opinion and Order. 20 F.C.C.R. 19415 (2005). ——. Petition of Qwest Corporation for Forbearance Pursuant to 47 U.S.C. § 160(C) in the Phoenix, Arizona Metropolitan Statistical Area. WC Docket No. 09-135, Memorandum Opinion and Order. 25 F.C.C.R. 8622 (2010). ——. Petition of Verizon Telephone Companies for Forbearance Pursuant to 47 U.S.C. § 160(c) in the Boston, New York, Philadelphia, Pittsburgh, Providence and Virginia Beach Metropolitan Statistical Areas, Inc. WC Docket No. 06-172, Memorandum Opinion and Order. 22 F.C.C.R.21293 (2007). ——. Petition to Establish Procedural Requirements to Govern Proceedings for Forbearance Under Section 10 of the Communications Act of 1934, as Amended. Report and Order. 24 F.C.C.R 9543 (2009). ——. Reexamination of Roaming Obligations of Commercial Mobile Radio Service Providers. Report and Order and Further Notice of Proposed Rulemaking. 22 F.C.C.R. 15817 (2007). ——. SBC Communications, Inc. and AT&T Corp. Applications for Approval of Transfer of Control. Memorandum Opinion and Order. 20 F.C.C.R. 18290 (2005). ——. Service Rules for the 698–746, 747–762 and 777–792 MHz Bands. Second Report and Order. 22 F.C.C.R.15289 (2007). ——. Sprint Nextel Corporation and Clearwire Corporation Applications for Consent to Transfer Control of Licenses, Leases, and Authorizations. Memorandum Opinion and Order. 23 F.C.C.R. 17570 (2008). ——. Verizon Communications, Inc. and MCI, Inc. Applications for Approval of Transfer of Control. Memorandum Opinion and Order. 20 F.C.C.R. 18433 (2005). ——. Wireless Telecommunications Bureau Seeks Comment on Petition for Rulemaking of Rural Telecommunications Group, Inc. to Impose a Spectrum Aggregation Limit on all Commercial Terrestrial Wireless Spectrum Below 2.3 GHz. Public Notice, DA 08–2279 (rel. October 10, 2008). Frieden, Rob. “The FCC’s Name Game: How Shifting Regulatory Classifications Affect Competition.” Berkeley Technology Law Journal 19 (2004): 1275– 1314. ——. “Lies, Damn Lies and Statistics: Developing a Clearer Assessment of Market Penetration and Broadband Competition in the United States.” Virginia Journal of Law and Technology 14 (2009): 100–125. 290

bibliography

——. “What Do Pizza Delivery and Information Services Have in Common? Lessons From Recent Judicial and Regulatory Struggles with Convergence.” Rutgers Computer and Technology Law Journal 32 (2006). ——. Winning The Silicon Sweepstakes: Can The United States Compete In Global Telecommunications? New Haven, CT: Yale University Press, 2010. Gardner, W. David. “Verizon, AT&T Big Winners in 700 MHz Auction.” InformationWeek, March 20, 2008. http://www.informationweek.com/ news/206905000. Ho, Daniel E., and Kevin M. Quinn. “Viewpoint Diversity and Media Consolidation: An Empirical Study.” Stanford Law Review 61 (2009): 781–868. “True or False: U.S.’s Broadband Penetration is Lower than Even Estonia’s; Answer: True.” Newsweek, July 9, 2007, 58. United States, Office of Management and Budget. Final Information Quality Bulletin for Peer Review. Federal Register 70 Fed. Reg. 2664. January 14, 2005.

chapter 10 Belson, Ken. “Rural Areas Left in Slow Lane of High-Speed Data Highway.” New York Times, September 28, 2006, A1. Federal Communications Commission. Fifth Report, Inquiry Concerning the Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fashion, and Possible Steps to Accelerate Such Deployment Pursuant to Section 706 of the Telecommunications Act of 1996, FCC 08–88. June 2008. Federal Communications Commission, Industry Analysis and Technology Division, Wireline Competition Bureau. High-Speed Services for Internet Access: Status as of December 31, 2003. June 2004. ——. High-Speed Services for Internet Access: Status as of December 31, 2005. July 2006. ——. High-Speed Services for Internet Access: Status as of June 30, 2006. December 2006. ——. High-Speed Services for Internet Access: Status as of June 30, 2008, July 2009. Flamm, K. “The Role of Economics, Demographics, and State Policy in Broadband Availability.” Presented at the PURC/London Business School Conference on “The Future of Broadband: Wired and Wireless, 2005.” University of Florida, Gainesville, 2005. Flamm, K., A. Friedlander, J. B. Horrigan, and W. Lehr. “Measuring Broadband: Improving Communications Policymaking through Better Data Collection.” 2007. http://www.pewinternet.org/PPF/r/227/report_display.asp. Government Accountability Office. “Broadband Deployment is Extensive Throughout the United States, but It Is Difficult to Assess the Extent of Deployment Gaps in Rural Areas.” May 2006. bibliography

291

Gordon, R. “Does the ‘New Economy’ Measure Up to the Great Inventions of the Past?” Journal of Economic Perspectives 14 (2000): 49–74. Grubesic, Tony H. “Zip Codes and Spatial Analysis: Problems and Prospects.” Socio-Economic Planning Sciences 42 (2008): 129–149. Horrigan, John B. Home Broadband Adoption 2006. Pew Internet and American Life Project, May 2006. ——. Home Broadband Adoption 2009. Pew Internet and American Life Project, June 2009. Jorgenson, D. “Information Technology and the U.S. Economy.” American Economic Review 91 (2001): 1–32. Oliner, S., and D. Sichel. “The Resurgence of Growth in the Late 1990s: Is Information Technology the Story?” Journal of Economic Perspectives 14 (2000): 3–22. Prieger, J. E. “The Supply Side of the Digital Divide: Is There Equal Availability in the Broadband Internet Access Market?” Economic Inquiry 41 (2003): 346–363. US President, Council of Economic Advisors. “The Making of the New Economy.” Economic Report of the President. 2001.

chapter 1 1 Akita, T. “Industrial Structure and the Source of Industrial Growth in Indonesia: An IO Analysis Between 1971 and 1985.” Asian Economic Journal 5, no. 2 (1991): 139–158. Antonelli, C. “The Digital Divide: Understanding the Economics of New Information and Communication Technology in the Global Economy.” Information Economics and Policy 15, no. 2 (2003): 173–199. Arbore, A., and A. Ordanini. “Broadband Divide among SMEs— The Role of Size, Location, and Outsourcing Strategies.” International Small Business Journal 24, no. 1 (February 2006): 83–99. Arduini, D., A. Zanfei, M. Denni, and G. Giungato. “The Provision of eGovernment Services: The Case of Italy.” In Proceedings of the 11th European Conference on eGovernment, edited by M. Klun, M. Decman and T. Jukic, 53–64. Reading, UK: Academic Publishing, 2011. Armistead, C., and M. Meakins. “A Framework for Practising Knowledge Management.” Long Range Planning 35, no. 1 (February 2002): 49–71. Baily, M. N., and R. Z. Lawrence. “Do We Have a New E-Conomy?” American Economic Review 91, no. 2 (May 2001): 308–312. Bozinis, A. I. “International Economic Relations and Information Communication Technologies (ICT) Use: Economic Globalization via Economic Digitalization.” American Journal of Applied Sciences 4, no. 4 (2007): 188–191. Bresnahan, T. F., and M. Trajtenberg. “General Purpose Technologies: Engines of Growth?” Journal of Econometrics 65, no. 1 (1995): 83–108. 292

bibliography

Celasun, M. Sources of Industrial Growth and Structural Change: The Case of Turkey. Working Papers 614, The World Bank, 1984. Cho, B.-H., and N. H. Hooker. “The Opportunity Cost of Food Safety Regulation: An Output Directional Distance Function Approach.” Working Paper AEDE-WP-0038-04, Ohio State University, 2004. Collins, J. L., and B. Wellman. “Small Town in the Internet Society: Chapleau Is No Longer an Island.” American Behavioral Scientist 53, no. 9 (May 2010): 1344–1366. Cortés, E. A., and J. A. Navarro. “Do ICT Influence Economic Growth and Human Development in European Union Countries?” International Advances in Economic Research 17, no. 1 (2011): 28–44. Crandall, R. W., C. L. Jackson, and H. J. Singer. The Effect of Ubiquitous Broadband Adoption on Invetsment, Jobs and US Economy. Washington, DC: Criterion Economics, LLC, 2003. Crandall, R., W. Lehr, and R. Litan. “The Effect of Broadband Deployment on Output and Employment: A Cross Sectional Analysis of US Data.” Issues in Economics Policy, no. 6 (July 2007): 1–34. Eichengreen, B. The European Economy Since 1945: Coordinated Capitalism and Beyond. Princeton: Princeton University Press, 2008. Elder, E. E., and W. R. Butcher. “Including the Economic Impact of Cost Paying in Regional Input-Output Analysis.” Western Journal of Agricultural Economics 14, no. 1 (1989): 78–84. Ford, G. S., and T. M. Koutsky. “Broadband and Economic Development: A Municipal Case Study from Florida.” Review of Urban and Regional Development Studies 17, no. 3 (2005): 216–229. Franklin, M., P. Stam, and T. Clayton. “ICT Impact Assessment by Linking Data.” Economic and Labour Market Review 3, no. 10 (2009): 18–27. Fuguitt, D., and S. J. Wilcox. Cost-Benefit Analysis for Public Sector Decision Makers. Westport: Greenwood Publishing Group, 1999. Glass, V., and S. K. Stefanova. “An Empirical Study of Broadband Diffusion in Rural America.” Journal of Regulatory Economics 38, no. 1 (August 2010): 70–85. Grady, P., and R. A. Muller. “On the Use and Misuse of Input-Output Based Impact Analysis in Evaluation.” Canadian Journal of Program Evaluation 3, no. 2 (1988): 49–61. Grimes, A., C. Ren, and P. Stevens. “The Need for Speed: Impacts of Internet Connectivity on Firm Productivity.” Journal of Productivity Analysis 37, no. 2 (April 2012): 187–201. Ha, L., R. N. Okigbo, and P. Igboaka. “Knowledge Creation and Dissemination in Sub-Saharan Africa.” Management Decision 46, no. 3–4 (2008): 392–405. Hale, T. M., S. R. Cotten, P. Drentea, and M. Goldner. “Rural-Urban Differences in General and Health-Related Internet Use.” American Behavioral Scientist 53, no. 9 (May 2010): 1304–1325. bibliography

293

Haller, S. A., and I. Siedschlag. “Determinants of ICT Adoption: Evidence from Firm-Level Data.” Applied Economics 43, no. 26 (2011): 3775–3788. Hollenstein, H. “Determinants of the Adoption of Information and Communication Technologies (ICT) an Empirical Analysis Based on FirmLevel Data for the Swiss Business Sector.” Structural Change and Economic Dynamics 15, no. 3 (2004): 315–342. Howick, S., and J. Whalley. “Understanding the Drivers of Broadband Adoption: The Case of Rural and Remote Scotland.” Journal of the Operational Research Society 59, no. 10 (October 2008): 1299–1311. Hughes, W., C. Brown, C. Brown, S. Miller, and T. McConnell. “Evaluating the Economic Impact of Farmers’ Markets Using an Opportunity Cost Framework.” Journal of Agricultural and Applied Economics 40, no. 1 (2008): 253–265. Kalja, A., J. Pold, T. Robal, and U. Vallner. “Modernization of the E-Government in Estonia.” In 2011 Proceedings of Picmet 11: Technology Management in the Energy-Smart World, edited by D. F. Kocaoglu, T. R. Anderson and T. U. Daim. N.p.: IEEE, 2011. Kandilov, I. T., and M. Renkow. “Infrastructure Investment and Rural Economic Development: An Evaluation of USDA’s Broadband Loan Program.” Growth and Change 41, no. 2 (June 2010): 165–91. Katz, R. “Estimating Broadband Demand and its Economic Impact in Latin America.” Proceedings of the 3rd ACORN-REDECOM Conference. Mexico City, 2009. Katz, R., and S. Suter. Estimating the Economic Impact of the Broadband Stimulus Plan. Working Paper, Columbia Institute for Tele-Information, 2009. Katz, R., S. Waterlaus, P. Zenhäusern, and S. Suter. “The Impact of Broadband on Jobs and the German Economy.” Working Paper, Columbia Institute for Tele-Information 2009. Kelly, D. A. Study of the Economic and Community Benefits of Cedar Falls. Iowa Association of Municipal Utilities, 2004. Lehr, W., C. Osorio, S. Gillet, and M. Sirbu. “Measuring Broadband Economic Impact.” Paper presented at the 33rd Research Conference on Communications, Information and Internet Policy. Arlington, VA, 2006. Liebenau, J., R. Atkinson, P. Karrberg, D. Castro, and S. Ezell. The UK’s Digital Road to Recovery. London: LSE Enterprise LTD and Information Technology and Innovation Foundation, 2009. Lindsay, S., P. Bellaby, S. Smith, and R. Baker. “Enabling Healthy Choices: Is ICT the Highway to Health Improvement?” Health 12, no. 3 (July 2008): 313–331. Lindskog, H., and M. Johansson. “Broadband: A Municipal Information Platform: Swedish Experience.” International Journal of Technology Management 31, no. 1–2 (2005): 47–63. 294

bibliography

Majumdar, S. K. “Broadband Adoption, Jobs and Wages in the US Telecommunications Industry.” Telecommunications Policy 32, no. 9–10 (October– November 2008): 587–599. Majumdar, S. K., O. Carare, and H. Chang. “Broadband Adoption and Firm Productivity: Evaluating the Benefits of General Purpose Technology.” Industrial and Corporate Change 19, no. 3 (Jun 2010): 641–74. Martin, L. “The Effects of ICT Use on Employee’s Motivations: An Empirical Evaluation.” Economics Bulletin 31, no. 2 (2011): 1592–1605. Milgrom, P., Y. Quaian, and J. Robert. “Complementarities, Momentum and the Evolution of Modern Manufacturing.” American Economic Review 81, no. 2 (1991): 84–88. Miller, R. E., and D. Blair. Input-Output Analysis: Foundations and Extentesions. Cambridge: Cambridge University Press, 2009. Organisation for Economic Co-operation and Development (OECD). “Good Practice Paper on ICTs for Economic Growth and Poverty Reduction.” DAC Journal 6, no. 3 (2005): 27–95. Organisation for Economic Co-operation and Development (OECD). Guide to Measuring the Information Society. Geneva: OECD, 2008. Organisation for Economic Co-operation and Development (OECD). Information Economy Product Definition Based on the Central Product Classification. Geneve: OECD, 2009. Rosenberg, N. Inside the Black Box: Technology and Economics. Cambridge: Cambridge University Press, 2004. Savage, S., G. Madden, and M. Simpson. “Broadband Delivery of Educational Services: A Study of Subscription Intentions in Australian Provincial Centers.” Journal of Media Economics 10, no. 1 (1997): 3–15. Sivakumar, S. C., and W. Robertson. “Developing an Integrated Web Engine for Online Internetworking Education: A Case Study.” Internet ResearchElectronic Networking Applications and Policy 14, no. 2 (2004): 175–192. Strategic Network Group. “Economic Impact Study of the South Dundas Township Fiber Network.” Report for the UK Department of Trade and Industry, Ontario, Canada, 2003. Van Gaasbeck, K. A. “A Rising Tide: Measuring the Economic Effects of Broadband Use across California.” Social Science Journal 45, no. 4 (December 2008): 691–699. Ward, A., and B. T. Prosser. “Reflections on Cyberspace as the New ‘Wired World of Education.’ ” Educational Technology & Society 14, no. 1 (January 2011): 169–178. Yan, C.-S. Introduction to Input-Output Economics. New York: Holt, Reinhart and Winston, 1968. Zakariah, A. R., and E. E. Ahmad. “Sources of Industrial Growth Using the Factor Decomposition Approach: Malaysia, 1978–1987.” Development Economics 37, no. 2 (1999): 162–196. bibliography

295

chapter 12 Academy for Educational Development. “Bulgaria Projects: LearnLink.” http:// www.learnlink.aed.org/Projects/bulgaria.htm. Basu, Indrajit. “Macedonia Transformed Through Broadband.” Digital Communities, November 9, 2007. http://www.govtech.com/dc/articles/174983. Etta, Florence E., and Parvyn-Wamahiu, Sheila, eds. Information and Communication Technologies for Development in Africa. Vol. 2. Ottawa: International Development Research Centre, 2004. Gillwald, Alison. “Towards Evidence Based ICT Policy: Access & Usage in 17 African Countries.” Research ICT Africa, presented at the Telecommunications Policy Research Conference, Arlington, VA, September 2008. Hudson, Heather E. “Defining Universal Service Funds: Are They Accelerators or Anachronisms?” Intermedia 38 (2010). ——. From Rural Village to Global Village: Telecommunications for Development in the Information Age. New York: Routledge, 2006. ——. “The Future of the E-Rate: U.S. Universal Service Fund Support for Public Access and Social Services.” In . . . And Communications for All: An Agenda for a New Administration, edited by Amit Schejter, 239–259. Lanham, MD: Lexington Books, 2009. InfoDev. “Universal Access and Service.” Module 4 of ICT Regulation Toolkit, December 2008. http://www.ictregulationtoolkit.org. James, Tina, ed. Information and Communication Technologies for Development in Africa. Vol. 3: Networking Institutions of Learning—SchoolNet. Ottawa: International Development Research Centre, 2004. Kalba, Kas. “The Adoption of Mobile Phones in Emerging Markets.” International Journal of Communication 2 (2008): 631–661. Mpogole, Hosea, Hidaya Usanga, and Matti Tedre. “Mobile Phones and Poverty Alleviation: A Survey Study in Rural Tanzania.” Iringa, Tanzania: Tumaini University. National Bureau of Statistics (Tanzania) and Macro International. Tanzania 2007–08 HIV/AIDS and Malaria Indicator Survey Key Findings. Calverton, MD: NBS and Macro International, 2009. Population Reference Bureau. 2008 World Population Data Sheet. http://www .prb.org/Publications/Datasheets/2008/2008wpds.aspx. Stork, Christoph “Mobile Access & Use in East Africa.” N.d. http://www .researchICTafrica.net. Tanzania Ministry of Education. “Information and Communication Technology (ICT) for Basic Education.” N.d. http://www.moe.go.tz. United Republic of Tanzania. The Universal Communication Service Access Act, 2006. Utz, Robert J., ed., Sustaining and Sharing economic Growth in Tanzania, Washington, DC: World Bank, 2008. World Bank. “Tanzania at a Glance.” September 24, 2008. 296

bibliography

CONTRIBUTORS

johannes m. bauer is professor in the Department of Telecommunication, Information Studies, and Media at Michigan State University where he is also the director of Special Programs of the Quello Center for Telecommunication Management and Law at Michigan State University. Dr. Bauer joined Michigan State University in 1990, after receiving his doctorate in economics from Vienna University of Economics and Business Administration, Vienna, Austria. Dr. Bauer taught and researched as a visiting professor at the Delft University of Technology, the Netherlands (academic year 2000–2001), the University of International Business and Economics in Beijing, China (May 2002), and the University of Konstanz, Germany (summer 2010). His research covers a wide range of issues related to the evolution of communications and information industries, in particular the design and effect of public policies towards next-generation networks and services, Internet governance, mobile communications, and innovation. He also researches business strategies, often from a comparative and international perspective. Research findings are published in leading journals in the field, including Telecommunications Policy, Information Economics and Policy, the International Journal of Communication, Telematics and Informatics, Info, and Electronic Markets. Dr. Bauer is a frequent speaker at international conferences and has served as an advisor for public and private sector organizations in North and South America, Europe, and Asia. steven bauer is a research affiliate at the Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory. He received his PhD in computer science at MIT working in the Advanced Network 297

Architecture Group with David Clark. Bauer’s research focuses on the architectures and economics of Internet-scale networks. Bauer is the recipient of the Department of Defense National Science and Engineering Fellowship and the National Science Foundation Fellowship. He is also a Harry S. Truman Scholar and Barry Goldwater Scholar. erik bohlin is currently head and professor in Technology Assessment at the Division of Technology and Society, Department of Technology Management and Economics at Chalmers University of Technology. He has published in a number of areas relating to the information society policy, strategy, and management. He is chair of the International Telecommunications Society and chief editor of Telecommunications Policy. He obtained his graduate degree in Business Administration and Economics at the Stockholm School of Economics (1987) and his PhD at Chalmers University of Technology (1995). justin brown is assistant professor in the School of Mass Communications at University of South Florida teaching courses in telecommunications, law and research methods. His research focuses on telecommunications law and policy issues including broadband deployment, freedom of speech and new media. Dr. Brown has made numerous research presentations at conferences organized by the International Communications Association (ICA), Association for Education in Journalism and Mass Communication (AEJMC), Broadcast Education Association (BEA), and the Telecommunication Policy Research Conference (TPRC). His research is represented in such publications as Cardozo Arts & Entertainment Law Journal, Communication Research, Communication Law & Policy, Federal Communications Law Journal, Cornell Journal of Law & Public Policy, IDEA: Intellectual Property Law Review, and Info: The Journal of Policy, Regulation and Strategy for Telecommunications, Information and Media. He also worked as a research assistant at Penn State’s Institute for Information Policy and the Pennsylvania Center for the First Amendment and currently serves on the editorial review board of Journalism and Mass Communication Educator. He earned his BS (Journalism) from the University of Oregon, and both his MA (Telecommunication Studies) and PhD (Mass Communication) from Penn State University, and previously taught at the University of Florida and Winthrop University. david clark is senior research scientist at the Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory, where he has worked since receiving his PhD there in 1973. Since the mid-1970s Dr. Clark has been leading the development of the Internet; from 1981 to 1989 he acted as chief protocol architect in this development, and chaired the Internet Activities Board. More recent activities include extensions to the Internet to support real-time traffic, pricing, and related economic issues, and policy issues surrounding the Internet, such as broadband local loop deployment. His current research looks at redefining the architectural underpinnings of the Internet, and the relation of technology and architecture to economic, 298

contributors

societal, and policy considerations. Dr. Clark is past chairman of the Computer Science and Telecommunications Board of the National Academies, and has contributed to a number of studies on the societal and policy impacts of computer communications. He is codirector of the MIT Communications Futures Program, a project for industry collaboration and coordination along the communications value chain. kenneth flamm, who joined the Lyndon B. Johnson School of Public Affairs at the University of Texas at Austin in fall 1998, is a 1973 honors graduate of Stanford University, and received a PhD in Economics from Massachusetts Institute of Technology in 1979. From 1993 to 1995, Flamm served as principal deputy assistant secretary of defense for economic security, and special assistant to the deputy secretary of defense for dual use technology policy. He was awarded the department’s Distinguished Public Service Medal in 1995 by Defense Secretary William J. Perry. Prior to his service at the Department of Defense, he spent eleven years as a senior fellow in the Foreign Policy Studies Program at Brookings. Flamm has been professor of economics at the Instituto Tecnológico Autónomo de México in Mexico City, the University of Massachusetts, and George Washington University. He has also been an adviser to the director general of income policy in the Mexican Ministry of Finance and a consultant to the Organization for Economic Cooperation and Development, the World Bank, the National Academy of Sciences, the Latin American Economic System, the US Department of Defense, the US Department of Justice, the US Agency for International Development, and the Office of Technology Assessment of the US Congress. Dr. Flamm’s publications include Mismanaged Trade? Strategic Policy and the Semiconductor Industry (1996), Changing the Rules: Technological Change, International Competition, and Regulation in Communications (ed., with Robert Crandell, 1989), Creating the Computer (1988), and Targeting the Computer (1987). He is currently working on studies of the economics technological innovation in the pharmaceutical and semiconductor industries, and broadband policy. Flamm, an expert on international trade and the high-technology industry, teaches classes in the economics of innovation, international trade, and advanced research methods. rob frieden serves as pioneers chair and professor of Telecommunications and Law at Pennsylvania State University. He also provides legal, management, and market forecasting consultancy services. Professor Frieden has written several books, published over sixty articles in academic journals, and appears in a variety of media outlets. He updates a major communications treatise: All About Cable and Broadband (Law Journal Press, 2008). Before accepting an academic appointment, Professor Frieden provided a broad range of business development, strategic planning, policy analysis, and regulatory functions for the IRIDIUM mobile satellite venture. Professor Frieden has held senior policy making positions in international telecommunications at the US contributors

299

Federal Communications Commission and the National Telecommunications and Information Administration. In the private sector, he practiced law in Washington, DC, and served as assistant general counsel at PTAT System, Inc. where he handled corporate, transactional, and regulatory issues for the nation’s first private undersea fiber optic cable company. Professor Frieden holds a BA, with distinction, from the University of Pennsylvania (1977) and a JD from the University of Virginia (1980). heather e. hudson is director of the Institute of Social and Economic Research and Professor of Public Policy at the University of Alaska Anchorage. Previously, she was professor of Communications Technology Management and founding director of the Telecommunications Management and Policy Program at the University of San Francisco. Her work focuses on both domestic and international topics concerning applications of information and communications technologies for socio-economic development, regulation, and policy including universal service and other strategies to extend affordable access to new technologies and services. Dr. Hudson has planned or evaluated communication projects in Alaska, northern Canada, and more than fifty developing countries and emerging economies. She has also consulted for the private sector, government agencies, consumer and indigenous organizations, and international organizations. She is the author of numerous articles and several books including From Rural Village to Global Village: Telecommunications for Developing in the Information Age; Global Connections: International Telecommunications Infrastructure and Policy; Communication Satellites: Their Development and Impact; and When Telephones Reach the Village, and coauthor of Electronic Byways: State Policies for Rural Development through Telecommunications, and Rural America in the Information Age. Dr. Hudson held a Fulbright Policy Research Chair at Carleton University in Canada to conduct a comparative study of US and Canadian broadband policies. She was a Sloan Foundation Industry Fellow at Columbia University’s Institute for Tele-Information and was also awarded a Fulbright Distinguished Lectureship for the Asia/Pacific. She has also been an honorary research fellow at the University of Hong Kong, and senior fellow at CIRCIT in Australia and at the East-West Center in Hawaii. Her research has been funded by inter alia the Sloan Foundation, the Ford Foundation, the Benton Foundation, the International Development Research Centre, the Aspen Institute, the World Bank, and the Telecommunications Education Trust. Dr. Hudson has been a board member of Farm Radio International, Women in Telecommunications (WiT) the Pacific Telecommunications Council, the Telecommunications Policy Research Conference, and the International Council for Computer Communications. She has served on the editorial boards of Telecommunications Policy, Information Technologies and International Development, and the Journal of Community Informatics. 300

contributors

She has been a member of advisory committees for the National Research Council, the Federal Communications Commission, the Department of Commerce, and the Office of Technology Assessment, and was a special advisor to the International Commission on Worldwide Telecommunications Development (the Maitland Commission). sungjoong kim received BA and MA degrees from the Department of Communication, Seoul National University, South Korea. He received his doctoral degree in Information and Media studies at Michigan State University. He has worked as a research fellow in the Department of Interactive Science, SungKyunKwan University, South Korea. He is currently working as a BK21 research fellow in the Department of Communication, Seoul National University, and has been lecturing at several universities. His research interests include diffusion of new media, communication and information policy, and media economics. sangwon lee is assistant professor at Kyung Hee University. He conducts research in telecommunication policy and media management and economics. His work has been published in journals such as Information Economics and Policy; Journalism and Mass Communication Quarterly; International Journal on Media Management; and International Journal of Mobile Marketing. His dissertation on broadband deployment was awarded a NET Institute research grant, which funds scientific research projects related to the telecommunications industries. His research in new media technology adoption has received “top paper” awards at national conferences including Association of Education in Journalism and Mass Communications (2007, 2009 and 2011), Telecommunications Policy Research Conference (2007), and Pacific Telecommunications Council (2008). Lee worked for the International Telecommunication Union in Geneva, Switzerland as a consultant within their regulatory reform unit. Lee received his bachelor’s degree in public administration at Yonsei University and earned a master’s degree in telecommunications at the George Washington University in Washington, DC. He earned his PhD in mass communications at the University of Florida. william lehr is research associate the Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory. Dr. Lehr is an economist whose specialty is the regulatory and industrial economics of the Internet infrastructure industries. He participates in the MIT Communications Futures Program, where his current research focuses on the evolution of wireless networking, broadband Internet access, and spectrum policy. Prior to coming to MIT in 1996, Dr. Lehr was assistant professor in the Graduate School of Business at Columbia University. In addition to his academic teaching and research, Dr. Lehr provides business strategy and litigation consulting services to public and private sector clients in the United States and abroad. Dr. Lehr holds a PhD in Economics from Stanford (1992), an MBA contributors

301

from the Wharton Graduate School (1985), and MSE (1984), BS (1979) and BA (1979) degrees from the University of Pennsylvania. catherine middleton holds a Canada Research Chair at Ryerson University’s Ted Rogers School of Management in Toronto, Canada. Her research focuses on the development and usage of communication technologies in the information society, and takes a critical perspective on the deployment of broadband and mobile technologies. She explores claims about the benefits of broadband technology adoption, showing that benefits can be overstated, infrastructures are not always developed to meet user needs, and that many citizens are not yet ready to participate in a knowledge-based society. In particular, she has conducted extensive work with Canadian Internet usage data, categorizing differences among Internet users, and questioning assumptions that most Internet users are highly engaged with the medium. Her work also considers the shortcomings of market-based approaches to developing communications infrastructure, highlighting important policy questions that have not been well addressed in the Canadian context. Dr. Middleton’s research has been supported by Canada’s Social Sciences and Humanities Research Council, Infrastructure Canada and the Networks of Centres of Excellence program. Current projects include an ongoing analysis of the Australian National Broadband Network (focusing on understanding the benefits of building an open-access wholesale network), and a study of the development of next generation fixed and wireless broadband infrastructures in Canada. She is the New Media theme leader for the GRAND Networks of Centres of Excellence project, and is a member of the board of CANARIE, Canada’s advanced research and innovation network. She is a member of the editorial boards of the Journal of Community Informatics, the Journal of Information Policy, New Media & Society, and the Telecommunications Journal of Australia. eli noam has been professor of Economics and Finance at the Columbia Business School since 1976. He served for three years as a commissioner for public services of New York State, and was appointed by the White House to the president’s IT Advisory Committee. He is director of the Columbia Institute for Tele-Information, a research center focusing on management and policy issues in telecommunications, Internet, and electronic mass media. Noam has published twenty-eight books and over four hundred articles in economics journals, law reviews, and interdisciplinary journals, and is a regular columnist for the Financial Times online edition. Noam has been a member of advisory boards for the federal government’s telecommunications network, of the IRS computer system, of the National Computer Systems Laboratory, the National Commission on the Status of Women in Computing, the governor’s task force on new media, of the Intek Corporation, and of many public interest organizations. He served on advisory boards for the governments of Ireland and Sweden. He received the degrees of BA, MA, PhD (Economics) and JD from Harvard University, and 302

contributors

honorary doctorates from the University of Munich (2006) and the University of Marseilles (2008). ibrahim kholilul rohman is a PhD candidate in the division of Technology and Society at Chalmers University of Technology. He obtained his bachelor’s degree in economics from the University of Indonesia in 2002. He then worked with the Institute for Economic and Social Research, faculty of Economics, University of Indonesia as a researcher. In 2006 he obtained a master’s degree in monetary economics from the same university. In 2008 Rohman received a scholarship from the Ministry of Communication and Information Technology, Government of Indonesia to pursue his PhD. His research interests include demand and impact analysis of ICT sectors, pricing and competition. amit m. schejter is associate professor of Communication Studies at BenGurion University of the Negev and associate professor of Communications and codirector of the Institute for Information Policy at Pennsylvania State University. His research, teaching, and service integrate a comprehensive approach to communication policy and its application to the everyday challenges created by the unequal distribution of resources and the silencing of the public’s voice. His studies have been widely published in both communication and law journals, cited in congressional and Knesset hearings, and have dealt with the challenges raised by the introduction of radio, television, cable, the Internet, mobile phones, and digitization in Israel, the United States, Korea, the European Union, and across wide international comparative settings. His background includes a decade of holding senior executive positions in the telecommunications industry in Israel, among them general counsel for Israeli public broadcasting and vice president of Israel’s largest mobile operator. In addition, he served on and chaired a variety of public committees, counseled media and telecommunication entities in Israel and the Palestinian Authority, and was Mundus Scholar at the universities of Amsterdam and Hamburg. His books include The Wonder Phone in the Land of Miracles: Mobile Telephony in Israel (coauthored with Akiba Cohen and Dafna Lemish, 2008), Muting Israeli Democracy: How Media and Cultural Policies Undermine Freedom of Expression (2009), and . . . And Communications for All: A Policy Agenda for a New Administration (2009). jorge reina schement is dean of the School of Communication and Information at Rutgers University. He is also professor in the Bloustein School of Public Policy, and in the Department of Latino-Hispanic Caribbean Studies. He received his PhD from the Institute for Communication Research at Stanford University, and MS from the School of Commerce at the University of Illinois. He is author of over 250 papers and articles His research contributed to a Supreme Court decision in Metro Broadcasting, Inc. v. FCC. In 1994 he directed the FCC’s Information Policy Project and conducted the original research that led to recognition of the digital divide. In 2008 he contributors

303

advised the FCC Transition Team for the Obama administration. He introduced the idea of universal service as an evolving concept, a view adopted in the Telecommunications Act of 1996. The movement to integrate community museums, libraries, and public broadcasting as partners in public service began in a project he codirected. He conducted the first study of the impact of minority ownership in broadcasting, and authored the telecommunications policy agenda for the Congressional Hispanic Caucus. He cofounded the Institute for Information Policy at Penn State University. Schement has served on advisory boards for several organizations such as the National Academy of Sciences, National Research Council, National Science Foundation, National Endowment for the Humanities, and the president’s Council of Advisors on Science and Technology. In 2007 Schement was listed in the Hispanic Business “100 Most Influential Hispanics.” richard d. taylor holds the Palmer Chair at Pennsylvania State University, where he is professor of Telecommunications Studies, Affiliate Professor of Information Sciences and Technology, and codirector of the Penn State Institute for Information Policy. Prior to joining the faculty at Penn State in 1989, he was vice president and corporate counsel at Warner Cable Communications, where he had overall responsibility for the Law Department of the nation’s second largest cable television operator. Subsequent to his arrival at Penn State, he was appointed to the board of Primestar Partners Ltd. as one of two independent directors. He is cofounder of the Institute for Information Policy at Penn State, which undertakes sponsored research and self-funded programs on the social implications of information technology, with an emphasis on the potential of information technologies for improving democratic discourse, social responsibility, and quality of life. At Penn State, his projects have received funding from Verizon, IBM, Microsoft, and the Ford Foundation, among others. Dr. Taylor is active nationally and internationally. He was a member of the Obama campaign’s Technology/Media/ Telecommunications Advisory Group, and is a former member of the boards of the Telecommunications Policy Research Conference and of the Pacific Telecommunications Council (PTC). He was co-chair of the PTC’s annual conference in 2009 and 2010. He is a member of the American Bar Association, the New York State Bar Association, and the Federal Communications Bar Association. He holds a doctorate in Mass Communications from Columbia University and a law degree from New York University School of Law. dr. bin zhang is professor in the School of Economics and Management at the Beijing University of Posts and Telecommunications (BUPT), Beijing, China, and has been a visiting scholar at Pennsylvania State University. She and Dr. Richard Taylor have collaborated on many papers and comparative studies involving China and the United States. She has written widely on telecommunications developments in China, and is considered one of the leading telecom policy scholars in China. 304

contributors

INDEX

access. See broadband access/availability; dial-up access; Internet access acquisitions and mergers: FCC approval of, 148–50 advanced service lines, 160 African Medical and Relief Foundation (AMRF), 216 American Radio Relay League, Inc. v. FCC, 154–55 American Recovery and Reinvestment Act of 2009, 110, 113, 189 analogies and metaphors: as tools for international comparisons, 97–98 analytic hierarchy process/static and dynamic analysis, 24–32 antitrust remedies, 146–48 Associated Press, 135 AT&T, 135, 149–50 Austin (Texas), 164–65 Australia, 11 average revenue per user, 93 Bagdikian, Ben, 133 Baker, Edwin, 135 barriers to entry, 139–40, 141 Bell, Daniel, 5 BellSouth, 149–50 Bhutan, 126–27

broadband: access (see broadband access/ availability); “broadband digital divide,” 106; democratic/civic participation and, 123–26, 128; economic impact of (see broadband economic impact); incentive problem, 63–64; information society participation and, 9–10, 14–18; measuring individual users’ use, 17–21, 18–21; penetration (see broadband adoption/penetration); public policy considerations (see broadband policy); relationship to ICT, 190–91; social and democratic importance of, 116–17; varying speeds and network characteristics, 10–14. See also Internet broadband access/availability: cable versus DSL, 165, 168–69, 170; versus dial-up, 160; evolution of US residential availability, 168–69; FCC data collection, 159, 160–65; impact of quality-of-service issues, 187–88; overview of conflicting indicators of, 158–60; in Tanzania, 213–14; understanding areas of unavailability in US (see broadband “disconnectedness”); US residential availability maps, 165–68. See also Internet access

305

broadband adoption/penetration: approaches to challenges of increasing, 109–10; in Australia, 11; in Canada, 10–11; demographics of nonusers, 108–9; in Europe, 11; global data, 70, 104–5; “intensive” stage of, 19; in Japan, 11; lagging US performance, 104–5, 114–15; platform/network competition as a driver of, 71–72, 83, 86; research on adoption factors, 72–74; trends in US household adoption, 107–9; as a US policy concern, 111–12, 113–15. See also ubiquitous broadband broadband “disconnectedness”: areas that are “really well connected,” 181, 183, 185; areas with fewer than 7 service providers (level 2), 178–79, 180–81, 182–83; areas with fewer than 12 service providers (level 3), 179, 181, 184–85, 186–87; areas with no service providers (level 0), 171; areas with 1 to 4 service providers (level 1), 171–76; defining levels of, 169, 171, 176–78 broadband economic impact: considerations for European spending policy, 207–8; information society benefits and, 14–18; literature review of, 190–95; macroeconomic impact, 197–98; of spending in Europe, 198–207 broadband policy: adoption/penetration as a US policy concern, 111–12, 113–15; economic impact considerations in Europe, 207–8; need for “democratic capacity” measures, 122–26, 128; need to address quality-of-service issues, 187–88; universal service goals in US, 105–6, 107, 111 broadcast TV, 134, 136 Bush, George W., 113, 115 cable modem service: availability versus DSL, 165, 168–69, 170; court case concerning FCC classification of, 152–54; platform competition with DSL, 86 calling-party-pays (CPP), 92, 244n14 Canada, 10–11 Canadian Internet Use Survey, 10 capacity planning: use of traffic data in, 54–58

306

index

Cape Cod, 110 case studies: as tools for international comparisons, 98 census tracts: use in broadband data collection, 161, 165–68 China: informatization studies in, 23–24, 48; measuring the digital divide in (see digital divide measurement) Cingular Wireless, 149 civic participation. See democratic/civic participation closed IO models, 192 Closing the Broadband Divide (Horrigan), 109 clustering analysis, 32–36 Coming of Post-Industrial Society, The (Bell), 5 Communications Act of 1934, 113–14 communications sector: alternative measures of health/growth, 122–26; concentration in (see media concentration) communications services, international comparisons: application of LEF to mobile voice services, 95–97; challenges of service and price differentiation, 90–92; metrics for price comparisons, 92–95; tools for interpreting comparisons, 97–102; uses and challenges of, 88–90 Compaine, Ben, 134–35 computable general equilibrium (CGE), 208 Connectivity Scorecard, 89 ConnectKentucky, 165, 174 Connect Ohio, 110 converged ubiquitous (cubiquitous) broadband, 71 cost and benefit framework, 197–98 courts: appellate courts on FCC’s lack of empiricism and peer review, 152–55; deference to the FCC, 143–44. See also US Supreme Court Cuba, 119 cubiquitous broadband, 71 cybersociety, 116 data envelopment analysis (DEA), 41–46, 98–99 Deep Packet Inspection (DPI), 58, 239n32

“democratic capacity” measures: need for inclusion in broadband policy, 128; other noneconomic measures as the basis for, 126–27; proposed for assessing communications systems, 122–26 democratic/civic participation: broadband and, 123–26, 128 democratization: Internet as a force for, 117–22, 125 development: noneconomic parameters of, 126–27 dial-up access, 104, 160 digital convergence, 139, 141 digital divide measurement: challenges and opportunities, 46–50; data envelopment analysis, 41–46; hierarchical clustering analysis, 32–36; static and dynamic analysis/analytic hierarchy process, 24–32; time distance analysis, 37–41 Digital Opportunities Index, 89 digital subscriber lines. See DSL “disconnectedness.” See broadband “disconnectedness” distributive justice, 124 download caps, 12 download speeds: broadband definitions and, 11–12; as a factor in broadband deployment, 86–87; NBP goals for, 114 DSL (digital subscriber lines): availability versus cable, 165, 168–69, 170; court case concerning Pacific Bell’s pricing of, 147–48; platform competition with cable, 86 “duty to deal,” 147–48 Dynamap (software), 162 dynamic network externalities, 104, 110 econometrics, 99–100 economic development: information society benefits of broadband, 14–18; literature review of broadband’s impact on, 190–95; role of ICTs in Tanzania, 215–17 economic impact. See broadband economic impact economic output, 266n55 economies of scale, 139–40, 141 Edison Trust, 136 “effective access,” 25 e-learning, 16

empiricism: appellate courts on FCC’s lack of, 152–55 employment, 193, 194 entry barriers. See barriers to entry e-readiness. See digital divide measurement E-Readiness index, 89 Europe: absence of universal service principle, 124; broadband penetration, 11; broadband spending policy considerations, 207–8; calculating the economic impact of broadband spending, 198–207; mobile pricing and service bundling, 91, 92 European Economic Recovery Package, 189 evidence-based policy, 131–32 facilities-based competition, 145, 146 FCC. See Federal Communications Commission Federal Communications Commission (FCC): appellate courts on lack of empiricism and peer review, 152–55; approval of mergers and acquisitions, 148–50; broadband data collection, 159, 160–65; need for reform, 155–58; as a partisan and politicized agency, 144–45; regulatory forbearance and, 145–46, 252n21; results-driven decision making, 143–44; ruling on wireless carrier spectrum caps, 151–52, 156 fiber networks, 11 fixed broadband: as an element of ubiquitous broadband, 69–70; broadband adoption factors and, 72–74, 83, 86 fixed effect models, 100 flat pricing, 91 Florida, 193 forbearance. See regulatory forbearance France, 127 Franklin, Benjamin, 135 free-rider problem, 68 “geo” zip codes, 161 Gomery, Doug, 134–35 government control: of the Internet, 117–22, 125 gross national happiness (GNH), 126–27

index

307

Hainan province (China): digital divide analysis, 35–36 hard-core disconnectedness, 171 Heilongjiang province (China): digital divide analysis, 45–46 Herfindahl-Hirschman Index (HHI), 135, 249n8 hierarchical clustering analysis, 32–36 high-speed Internet. See broadband; Internet horizontal integration, 150–51 Horning, Michael, 125 Horrigan, John, 109 household broadband: profiles, 13–14; trends in adoption, 107–9 IBM, 136 ICT. See information and communications technology ICT Africa surveys, 220, 221 ideology, 131 incumbent local exchange carriers (ILECs), 169 indecency, 154 individual broadband profiles, 231n24 individual user-based subsidies, 226 information activism, 132–33 information and communications technology (ICT): “effective access” concept, 25; factors related to broadband adoption, 74; indicators, 18–20; limitations of current metrics, 1–2; measuring levels of deployment (see digital divide measurement); measuring sector growth in Europe, 198–207; relationship to broadband, 190–91; role in economic development in Tanzania, 215–17; role in information society, 195 information sector concentration: dynamics of, 139–41; empirical evidence for, 136–38. See also media concentration information society: basic definitions, 15; indicators, 18–20; politics of, 132–33; role of broadband in, 9–10, 14–18; role of ICTs in, 195 informatization: building an informatization level index, 25–27; challenges and opportunities in measurement, 46–50; defined,

308

index

23–24. See also digital divide measurement input-output (IO) analysis: of broadband in existing studies, 192–95; of broadband investment in Europe, 198–207; described, 190, 195–98; future directions for research, 208; of non-broadband examples, 191–92 “integrator” firms, 142 Intelsat, 149 interactive web applications, 125–26 intermediate transactions, 202 intermodal competition. See platform/ network competition international comparisons of communications services. See communications services, international comparisons International Telecommunication Union ICT Development Index, 19 Internet: access (see Internet access); connection speed issues, 12; in Cuba, 119; democratization and government control, 117–22, 125; growth in China, 29; market concentration, 137; measuring individual users’ use, 17, 18–21; nonadoption in US, 108; social and democratic importance of, 116–17; traffic issues (see traffic data); ultra-religious opposition to in Israel, 118–19. See also broadband Internet access: approaches to challenges of increasing, 109–10; broadband versus dial-up, 160; “4 C’s” of meaningful access, 106–7; economic and societal value created by, 110–11; Tanzania and, 213–14, 216, 224–26; trends in US household broadband adoption, 107–9; as a US policy concern, 111–12; in US tradition of communications access, 104–6. See also broadband access/ availability Internet service providers (ISPs): collection of traffic data, 58–62 (see also traffic data); court case against Pacific Bell, 147–48; traffic management practices, 12 interregional input-output (IRIO) table, 208 ISPs. See Internet service providers

Israel: media penetration in, 120, 121; ultra-religious opposition to the Internet, 118–19 Japan: broadband data project, 58–59; broadband download speeds, 11, 66; user traffic volumes, 61 job growth, 194–95 Jones, Steve, 116 Kapor, Mitchell, 117 Katz, Elihu, 116 Kentucky, 165, 174, 261n29 Kneuer, John, 251n10 knowledge economy. See information society layered model of broadband, 125 LEF. See lowest expenditure frontier Lessig, Lawrence, 134 lowest expenditure frontier (LEF): application to mobile voice services, 95–97; description and advantages of, 93–95 low-grade broadband, 187 Madison, James, 103 Marconi Company, 135 “marketplace of ideas,” 121–23 Massachusetts, 167, 176–78, 179 mass media: digital convergence and, 141; market concentration, 137, 138 McChesney, Robert, 134 measured pricing, 91 media: democratization and government control, 117–22; “marketplace of ideas” metaphor and, 121–23; penetration in Israel, 120, 121 media concentration: debate over, 133–35; dynamics of, 139–41; empirical evidence for, 136–38; future shape of, 142; historical and global context of, 135–36 “media monopoly,” 133–34 mergers and acquisitions: FCC approval of, 148–50 metaphors and analogies: as tools for international comparisons, 97–98 Metcalfe’s Law, 222, 268n33 Millennium Development Goals, 216–17

Minnesota Internet Traffic Studies (MINTS), 59–60 mobile broadband: as an element of ubiquitous broadband, 69–70, 71; broadband adoption factors and, 72–74, 83, 86; growth in, 11. See also ubiquitous broadband mobile communications: application of LEF to international pricing comparisons, 95–97; global penetration, 114; pricing and service components in, 91–92; in Tanzania, 211–12, 220–24 mobile-party-pays (MPP), 92, 244n14 multipart pricing schemes, 91–92 multiple standardization policy, 72–73, 86 multiplier effect: in broadband sectors worldwide, 194; calculated in IO model, 197; of ICT in Europe, 202–6 Multi Router Traffic Grapher, 60 Munsey’s Magazine, 135 National Broadband Plan (NBP), 111, 113, 114, 128 National Cable & Telecommunications Association v. Brand X Internet Services, 152–54 Netflow, 56, 237n7 Networked Readiness, 89 network/platform competition, 71–72, 83, 86 network quality, 12–13 network traffic data. See traffic data new economy. See information society New York Times, 103, 165 Next Generation Networks, 71 nonparametric statistical analyses, 98–99 Nyerere, Julius, 210 Obama, Barack, 113, 115 Obama Package, 189 Odlyzko, Andrew, 59 Ohio, 110 Open Cape, 110 open IO models, 192 opportunity costs: defined, 189; of European broadband spending, 207, 208; examples of IO analysis of, 191–92

index

309

Pacific Bell Telephone Co., 147–48 PanAmSat, 149 parametric statistical approaches, 99–100 Pearl Valley Cheese Company, 110 peer review: appellate courts on FCC’s lack of, 152–55; FCC’s narrow view of, 252n16 perfect competition, 123 Pew Internet and American Life Project, 104, 107–8, 160, 165 “phoneless,” 108–9 platform/network competition, 71–72, 83, 86 “point” zip codes, 161 policy of inclusion, 124 political freedom, 86 population density, 87 postpaid service, 91 predatory pricing, 147 prepaid service, 91 price basket, 93 price squeeze, 147 privacy, 66–67 production-profitability matrix, 44 Prometheus Radio Project v. FCC, 150 purchasing power parity (PPP), 244n15 qualitative comparative analysis (QCA), 101 Qwest, 145 Qwest Corp. v. FCC, 155 random effects models, 100 RCA, 136 real-time traffic data, 54 reasonable interchangeability test, 123 regulatory forbearance, 145–46, 252n21 representative price, 92–93 residential broadband. See household broadband results-driven decision making, 143–44 Round Robin Database, 60 rural communities: available data on teledensity in Tanzania, 220–22; broadband expansion in US, 109–10; role of subsidies for expanding telephony in Tanzania, 222–24 Sadler, Terri, 103–4 Sarkozy, Nicholas, 127 satellite service providers, 149 scale economies. See economies of scale 310

index

Scalia, Antonin, 153, 154 Schoolnet Africa, 215–16 Sen, Amartya, 127 set theory, 101 similarity measures, 33 simulation models, 101–2 Sirius, 149 SNMP (Simple Network Management Protocol), 55 social contract: Internet access and, 104–6 social efficiency, 43–46 South Carolina, 165–67 standardization policy, 72–73, 86 static and dynamic analysis/analytic hierarchy process, 24–32 Stiglitz, Joseph, 127 Stiglitz-Sen Commission, 127 subscriber privacy, 66–67 subsidies: for increasing rural telephony in Tanzania, 222–24; individual user-based subsidies, 226 Sweden, 202–4 TANA Inc./GDT Inc., 162 Tanzania: available data on rural teledensity, 220–22; expanding communications access (see Universal Communications Access Fund); Internet access, 213–14, 216, 224–26; role of ICTs in economic development, 215–17; role of subsidies for expanding rural telephony, 222–24; telecommunications sector, 210–14 Tanzania Universal Communications Service Access Act of 2006, 214 technical efficiency, 43–46 technological revolution, 133 TeleAtlas, 162 telecenters, 210 Telecommunications Act of 1996: regulatory forbearance, 145; universal service policy goals, 105–6, 107 telecommunications sector: market concentration in US, 137; in Tanzania, 210–14 telehealth and telemedicine, 216 telephone penetration: demographics of the “phoneless,” 108–9 Texas, 164–65 Thierer, Adam, 134 time distance analysis, 37–41

time series analysis, 101 total broadband. See ubiquitous broadband traffic data: background on, 51–54; challenges of collecting multi-ISP data sets, 64–68; existing data collection by individual network operators, 54–58; existing multi-network data sets and uses, 58–62; importance of, 62–64 traffic management: congestion management, 63; importance of traffic data to, 52–53; “throttling” and “shaping,” 12 traffic volume and distribution, 61 transportation networks, 52 TTCL (telecommunications company), 211, 212–13 ubiquitous broadband: analysis of key determinants, 75–83, 84–85; background on, 69–70; concept of, 71, 72; determinants identified as most significant, 83, 86–87; platform/ network competition as a driver of, 71–72, 83, 86; research on broadband adoption factors, 72–74. See also broadband adoption/penetration ubiquitous computing, 71 United Kingdom, 195 universal broadband. See ubiquitous broadband Universal Communications Access Fund (UCAF): characteristics of, 214–15; creation of, 209–10; implementation questions and data sources, 217–20; options for increasing Internet access, 224–26 universal service funds: characteristics of, 214–15; defined, 209 universal service policies: for correcting distortions in access, 124–25; universal service goals in US, 105–6, 107, 111 upload speeds: asymmetries with download speeds, 12; NBP goals for, 114

urbanization, 87 user surveys, 20–21 US Supreme Court: elimination of antitrust remedy, 144, 146–48; on the FCC’s lack of empiricism and peer review, 152–54 Utsumi, Yoshio, 15 Vail, Theodore, 124 Verizon, 146 Vermont, 165 vertical integration, 150–51 video applications, 13, 60 Web 2.0, 125 Weiser, Marc, 71 Western Union, 135 West Virginia, 191–92 wired broadband. See fixed broadband wireless broadband. See mobile broadband wireless carrier spectrum caps, 151–52, 156 wireless communications. See mobile communications World Economic Forum/INSEAD Networked Readiness Index, 19 World Information Society Report, 15 XM, 149 Yunnan province (China): digital divide analysis, 41 “zero service provider” zip codes, 161 Zhejiang province (China): digital divide analysis, 40–41, 45 zip codes: versus census tracts for broadband data collection, 161, 165–68; measurement issues with FCC broadband data collection, 161–65; used to interpret levels of broadband access (see broadband “disconnectedness”); used to understand residential broadband access, 168–69, 170 zip code tabulation areas (ZCTAs), 162

index

311

DONALD MCGANNON COMMUNICATION RESEARCH CENTER’S E V E R E T T C . PA R K E R B O O K S E R I E S

ph ilip m. napoli and minna aslama, eds. Communications Research in Action: Scholar-Activist Collaborations for a Democratic Public Sphere

kari karppinen Rethinking Media Pluralism richard d. taylor and amit m. schejter, eds. Beyond Broadband Access: Developing Data-Based Information Policy Strategies