Beyond Broadband Access Developing Databased Information Policy Strategies Richard D Taylor Amit M Schejter

Beyond Broadband Access Developing Databased
Information Policy Strategies Richard D Taylor
Amit M Schejter download
https://ebookbell.com/product/beyond-broadband-access-developing-
databased-information-policy-strategies-richard-d-taylor-amit-m-
schejter-51900050
Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Beyond Broadband Access Developing Databased Information Policy
Strategies 1st Edition Richard D Taylor Amit M Schejter
https://ebookbell.com/product/beyond-broadband-access-developing-
databased-information-policy-strategies-1st-edition-richard-d-taylor-
amit-m-schejter-51384080
Broadband Connectivity In 5g And Beyond Next Generation Networks
Simranjit Singh
https://ebookbell.com/product/broadband-connectivity-in-5g-and-beyond-
next-generation-networks-simranjit-singh-44975416
Broadband Wireless Mobile 3g And Beyond 1st Edition Willie W Lu
https://ebookbell.com/product/broadband-wireless-mobile-3g-and-
beyond-1st-edition-willie-w-lu-2023694
Mobile Broadband Evolution Towards 5g Rei12 Rei13 And Beyond 4g
Americas
https://ebookbell.com/product/mobile-broadband-evolution-
towards-5g-rei12-rei13-and-beyond-4g-americas-50195190

Deploying Ipv6 In 3gpp Networks Evolving Mobile Broadband From 2g To
Lte And Beyond Jouni Korhonen
https://ebookbell.com/product/deploying-ipv6-in-3gpp-networks-
evolving-mobile-broadband-from-2g-to-lte-and-beyond-jouni-
korhonen-4302818
Awakening Book 02 Beyond
https://ebookbell.com/product/awakening-book-02-beyond-216293636
Awakening Book 03 Beyond
https://ebookbell.com/product/awakening-book-03-beyond-216293650
Awakening Book 01 Beyond
https://ebookbell.com/product/awakening-book-01-beyond-216296374
Beyond The Astral Metaphysical Short Stories William Buhlman Susan
Buhlman
https://ebookbell.com/product/beyond-the-astral-metaphysical-short-
stories-william-buhlman-susan-buhlman-44908450

Beyond Broadband Access

DONALD MCGANNON COMMUNICATION RESEARCH CENTER’S
EVERETT C. PARKER BOOK SERIES
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 deﬁ ned notion of communications policy, in that
it considers not only scholarship addressing speciﬁ c 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
fordham university press . new york . 2013

Copyright © 2013 Fordham University Press
Al l 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.
Fo rdham 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.
Fo rdham 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.
Li brary of Congress Cataloging-in-Publication Data is
available from the publisher.
Pr inted in the United States of America
151413 54321
Fi rst edition

CONTENTS
Ac knowledgments vii
In troduction: Numbers That Matter
Ri chard D. Taylor and Amit M. Schejter
1
PART I: theory
1. Be yond Broadband Access: What Do We Need to Measure
and How Do We Measure It?
Ca therine Middleton
9
2. Un derstanding Digital Gaps: A Quartet of Empirical
Methodologies
Bi n Zhang and R ichard D. Taylor
23
3. Br oadband Microfoundations: The Need for Trafﬁ c Data
St even Bauer, David Clark, and W illiam Lehr
51
4. Ad option Factors of Ubiquitous Broadband
Sa ngwon Lee and J ustin S. Brown
69
5. Da ta and Modeling Challenges in International Comparisons
Jo hannes M. Bauer and S ungjoong Kim
88
6. Da ta, Policy, and Democracy
Jo rge Reina S chement
103

contentsvi
7 . “R ulers of thousands, hundreds, ﬁ fties, and tens”: Does
Democracy Count?
Am it M. S chejter
113
PART II: the use and abuse of data in
information policy making
8. Ph D Heal Thyself: In Search of Evidence-based Research
for Evidence-based Policy
El i Noam
131
9. Ca se Studies in Results-Driven Decision Making at the FCC
Ro b Frieden
143
10 . Th e Determinants of Disconnectedness: Understanding US
Broadband Unavailability
Ke nneth Flamm
158
11 . Eu ropean Broadband Spending: Im plications of Input-Output
Analysis and Opportunity Costs
Ib rahim Kholilul Rohman and E rik Bohlin
189
12 . Us ing Data for Policy Development: Designing a Universal
Service Fund for Tanzania
He ather E. Hudson
209
No tes 229
Bi bliography 269
Li st of Co ntributors 297
In dex 305

vii
Th is book is based on selected edited papers that were ﬁ rst 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 S chement, 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 organiza-
tions for helping make the workshop the success that it was.
Th e 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
ACKNOWLEDGMENTS

acknowledgmentsviii
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 assis-
tance 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.
La st but not least, we have beneﬁ tted from the stimulating, and some-
times 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 ﬁ nancial ofﬁ cers—
Jane Agnelly, Annette Rice, and Dorie Glunt—in keeping our budgets
organized and in-line.

1
In recent years the information policy discourse has been inundated by
numbers and metrics, which purportedly describe the “information society”
and reﬂ ect 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 ﬁ fteenth 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 lim-
ited capabilities. Previously, the mere calculation of penetration levels may
have provided sufﬁ cient 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 ﬁ nd 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
INTRODUCTION: NUMBERS THAT MATTER
ri chard d. taylor and amit m. schejter

r. d. taylor and a. m. schejter2
involve a high degree of subjectivity; it is important to adopt statistical
methods that reduce this aspect. However, method must ﬁ rst be guided by
theory, and in this ﬁ eld, theory is remarkably lacking. Important policy deci-
sions are being made worldwide about information services to promote
innovation, knowledge development, social e quity, 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 ﬁ rst step in developing a strategy for building a comprehensive,
data-based approach to understanding policy consequences, and for improv-
ing policy outcomes through the utilization of meaningful empirical analy-
ses, 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 para-
digm, 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 ﬁ ll 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 meas-
uring information, and with issues that should be considered when design-
ing 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

in troduction 3
indicators, they fail to offer sufﬁ ciently 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 beneﬁ t 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
trafﬁ c 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
Chapter4, Sangwon Lee and Justin Brown demonstrate how the results of
regression analysis suggest that network competition between ﬁ xed and
mobile broadband, platform competition in ﬁ xed broadband markets, and
the multiple standardization policy in mobile markets, are all signiﬁ cant fac-
tors of ubiquitous broadband deployment. While the theoretical reasoning
for quantiﬁ cation of connectivity raises methodological challenges, the social
justiﬁ cation 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 country-
and region-speciﬁ c, 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

r. d. taylor and a. m. schejter4
Schement argues that the roots of information policy can be found in
the implied social contract between governments and the governed and
identiﬁ es a new social contract relevant for the information age. Amit
S chejter, 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 cre-
ated, after reaching consensus on the deﬁ nition 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 ques-
tions whether there is an interest in Congress and the US Federal
Communications Commission (FCC) in serious data analysis. He identiﬁ es
a speciﬁ c structural impediment in the rule-making process, and recom-
mends 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 identiﬁ es several instances in which the
FCC could have used empirical research and peer review to achieve an accu-
rate 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 systemati-
cally model the determinants of “disconnectedness” and assesses what eco-
nomic and demographic factors can be identiﬁ ed as important, and statistically
signiﬁ cant, in reducing or increasing “disconnectedness.”
The ﬁ nal two Chapters discuss cases outside the Americas. In Chapter 11,
Ibrahim Kholilul Rohman and Erik Bohlin address a methodology to evalu-
ate the opportunity cost of using government funds for broadband in the
context of a future European proposal to publicly fund broadband infra-
structures; and in Chapter 12, Heather Hudson examines how data, primar-
ily 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 ﬁ eld
studies may be limited.
. . .
The wide range of issues discussed in part I—from trafﬁ c data to democracy
measures and from statistical analyses to considerations in comparative

in troduction 5
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 theo-
retical approaches will lead to its professionalization, and result in the exclu-
sion 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 conver-
sation 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 ground-
breaking work The Coming of Post-Industrial Society,
2
Daniel Bell noted that
among the distinguishing characteristics of what he perceived to be a revolu-
tionary change in the structure of society were the centrality of theoretical
knowledge, the capability of planning, and the rise of “intellectual technol-
ogy.” Indeed, it would be an irony of our times if at the culmination of these
dramatic changes, the planning for full utilization of their beneﬁ ts for the
sake of all members of society were to be left in the hands of chance.

This page intentionally left blank

9
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 encour-
age 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 build-
ing broadband infrastructure.
2
In instances where the private sector cannot
establish a business case for broadband deployment, governments are com-
mitting public funds to extend the reach of broadband networks, justiﬁ ed 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 beneﬁ ts of a knowledge-based economy and
society.
3
However, Preston, and Cawley observe that broadband develop-
ment is often driven by “supply-side, technology-focused policies” that do
not explicitly consider the needs of users.
4
Deployment of broadband infra-
structure 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 develop-
ment of a knowledge-based society and the ability to deliver the desired
outcomes. This chapter explores this gap.

catherine middleton10
A central motivation for investing in broadband is that providing citizens
with access to broadband connectivity will allow them to engage in the informa-
tion 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 beneﬁ ts of broadband access may not be equal for all broad-
band users; and 3) there is uncertainty as to exactly how to recognize the
broadband-enabled beneﬁ ts 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 sophis-
ticated and wider-ranging data collection in order to fully assess the extent to
which broadband networks actually do enable citizens to become partici-
pants 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 availabil-
ity 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 beneﬁ ts 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 deﬁ nition 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 tel-
ephone connections are asked if their connection is a high-speed connection.
No deﬁ nition 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

beyond broadband access 11
The term broadbandis not applied to these data by Statistics Canada,
but data reported by the CRTC (the Canadian telecommunications regula-
tor) 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, deﬁ ned 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
above256 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 data
9
show that 64 percent of households have a broadband
connection, deﬁ ned as “connectable to an exchange that has been converted
to support xDSL-technology, to a cable network upgraded for Internet trafﬁ c,
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 ﬁ fteen European countries not quite two-thirds of house-
holds have broadband connectivity. After years of promoting the beneﬁ ts 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
ﬁ ber networks that can provide download speeds of 100 Mbps or more, and
several other OECD countries have extensive ﬁ ber 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 ﬁ xed line service.
Examples could be provided from many other countries, but the data
above are sufﬁ cient to demonstrate that there are many deﬁ nitions of broad-
band, and that there are vast differences in the speeds of broadband connections

catherine middleton12
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
conﬁ ned to a speciﬁ c country.
There are three other pertinent issues regarding the speed of Internet
connections that impact the way that citizens experience broadband net-
works. The ﬁ rst is that in a few countries, Internet service providers (ISPs)
impose strict limits on the volume of data that can be downloaded in a ﬁ xed
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, ﬁ nancial or technical penalties (e.g., reduced download speeds) are
imposed on the subscriber. As such, it is important to understand the condi-
tions 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 trafﬁ c. 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 trafﬁ c shaping practices among ISPs, it is useful to under-
stand 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 sym-
metrical 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 trafﬁ c management practices, and
whether the connection supports symmetrical or near-symmetrical uploads
it is difﬁ cult to make assumptions as to the capacity of an individual’s broad-
band 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 con-
sider the speciﬁ c beneﬁ ts that can arise with the adoption and use of faster,

beyond broadband access 13
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 trafﬁ c shaping does not ensure that they will be
able to beneﬁ t 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 beneﬁ ts of higher capacity networks is
the ability to support video applications. These networks will support mul-
tiple high-deﬁ nition 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 inﬂ uence over what
can be done on a particular broadband network. If broadband networks are
to deliver beneﬁ ts to users by enabling certain functionality, then it is essen-
tial 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 varia-
tions in broadband speeds, access restrictions and network quality, there is a
strong case to be made for ensuring that the attributes of broadband net-
works 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.
Table1-1 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 proﬁ le.
24
This proﬁ le would allow for immediate determination
of whether a household has the capacity to engage in speciﬁ c activities, and
identiﬁ cation of gaps to be remedied. The answers to some questions are
clear cut (e.g., advertised speed), others are more subjective (e.g., deﬁ nitions
of affordability, which is relevant in understanding whether a citizen can

catherine middleton14
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 proﬁ les is only a ﬁ rst step in the
exploration of how citizens can beneﬁ t from broadband infrastructure. The
proﬁ le provides information on the current and future capacity of a house-
hold’s broadband connectivity, but capacity is a measure of potential beneﬁ t,
not of realized beneﬁ t. As Gillett et al. observed, in order to beneﬁ t 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 meas-
ures, 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
table 1-1 . sample components of a household
broadband profile
Iss ue: Broadba nd
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 speciﬁ c types
of trafﬁ c? Does such prioritization meet user and application
provider needs?
Service
provider
Do service providers’ policies (e.g., with respect to trafﬁ c shaping
or download caps) negatively impact the user’s experience?
Is the connection affordable?

beyond broadband access 15
communication technologies in enabling a knowledge-based society or digital
economy.
26
There are, for instance, streams of research focusing on concep-
tualizing 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, deﬁ ne,
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 deﬁ nitions
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 2006World 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 any-
where, anytime and to anyone.”
33
The European Commission’s Knowledge
Society Web Portal (now defunct, but accessible through archive.org)
deﬁ ned 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 intan-
gible, 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 beneﬁ ts. 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 speciﬁ c insight. Assertions that broadband enables employment,
education, and healthcare are plentiful, but speciﬁ c studies that consider
how individual usage of broadband
38
actually results in increased engage-
ment with the information society are less common,
39
even when the possi-
bilities for such engagement are clearly articulated.
40
Recent work by Kolko
suggests that beneﬁ ts to households as a result of increased broadband

catherine middleton16
expansion are “ambiguous,” and concludes that “the local economic
development beneﬁ ts 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 difﬁ cult 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 speciﬁ cally considers the relationship
between investments in broadband infrastructure and tangible beneﬁ ts 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 beneﬁ ts, it will be necessary to develop
a range of assessment tools. There are two key requirements for these tools.
The ﬁ rst is that they enable recognition of broadband-enabled information
society beneﬁ ts; 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 beneﬁ t.
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 Web-
based groups that facilitate economic integration and drive development.
Blogs (Web logs, or online diaries), wikis (Web sites where users can contrib-
ute and edit content), video sharing sites, and the like allow new, decentral-
ized, and dynamic approaches to capturing and disseminating information
that allows individuals to become better prepared for the knowledge econ-
omy.”
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-deﬁ nition 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 ﬁ eld trips. These two descriptions of e-learning demonstrate the com-
plexity inherent in trying to understand how broadband enables e-learning,

beyond broadband access 17
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 activi-
ties like email, banking, searching for work, education, watching TV, and
searching for information
44
cannot be used to assess the extent of informa-
tion society beneﬁ ts gained by these types of Internet usage. Such data do
allow for a very basic description of the nature of individuals’ online activi-
ties, but can really only be used to understand the relative popularity of
applications, not to understand how or whether people beneﬁ t from using
them. As the e-learning example suggests, developing a detailed question-
naire about the beneﬁ ts 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 beneﬁ ts in their
own ways, rather than by selecting from a stated list of anticipated beneﬁ ts.
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, beneﬁ ts, and challenges
faced in engaging with the information society) could offer a less resource-
intensive means of gathering richer data, while still allowing individuals to
share their experiences in ways that are meaningful to them. Table 1-2 sum-
marizes the objectives for data collection regarding use of, and beneﬁ ts 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
Beneﬁ ting from Broadband
Information Required
Network use Do citizens access services that enable them to
participate in the information society?
Do citizens accrue beneﬁ ts from the use of such
services? How? What skills are needed to ensure
maximum beneﬁ t?

catherine middleton18
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
speciﬁ c activities are undertaken (this helps to articulate the extent of bene-
ﬁ ts 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
superﬁ cial). An additional problem, especially within various international
ICT indexes that purport to assess use, is that many measures of use are actu-
ally 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 ﬁ rst 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 estab-
lished 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 poten-
tial 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 ﬁ rst step toward participation in the information society, but it may not
produce immediate beneﬁ ts. As such, efforts to understand how having
access to a broadband connection can enable the beneﬁ ts 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’ read-
iness 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 informa-
tion society is a complicated task, and much progress has been made by the

beyond broadband access 19
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-proﬁ le 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 pro-
gress within countries over time. But they offer minimal value in assessing
the extent to which individuals in particular countries are using their broad-
band connections to engage in the information society and to achieve per-
sonal beneﬁ ts. 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 commu-
nications and telecommunications generally has not, to date, looked in
sufﬁ cient detail at factors beyond penetration.”
48
They, as well as Pepper
et al., propose stage models of adoption that do take into account real meas-
ures 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
beneﬁ ts through increased usage of ICTs. Furthering this approach,
Waverman and Dasgupta’s Connectivity Scorecard recognizes the impor-
tance 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 approx-
imately50 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 popula-
tion 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.

catherine middleton20
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 tech-
nologies, 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 speciﬁ c information
society beneﬁ ts. However, there are two ways that the relationship can be
explored. The ﬁ rst 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 con-
tinues to grow steadily around the globe, meaning that the number of users
is increasing, and the potential for more citizens to engage with the informa-
tion society through broadband connections is growing. Outside the indexes,
OECD broadband adoption statistics are widely used, and widely criti-
cized.
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 nonproﬁ t research consortia (e.g., Pew Internet Project, World Internet
Project). Additional data and commentary are available from various consulting
ﬁ rms 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 analy-
sis. 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 indi-
viduals participate in an information society (e.g., by accessing educational
content, government services, healthcare and so on). Although longitudinal
tracking of changes in speciﬁ c 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

beyond broadband access 21
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 ﬁ ndings offer a
tangible example of how broadband usage data can be linked to achievement
of information society beneﬁ ts by individuals, and reinforce the observation
that broadband connectivity alone does not ensure beneﬁ cial outcomes. He
also makes the point that assessing the beneﬁ ts 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 indi-
cate that their ability to beneﬁ t from the use of various Internet and broad-
band 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 beneﬁ ts that individuals
can realize by adopting and using broadband networks, the following infor-
mation 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 broad-
band connectivity is used; assessment of users’ skills and capacity to use their
broadband connectivity, to allow for full beneﬁ ts of connectivity to be real-
ized. 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 beneﬁ ts 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 identiﬁ es research questions and data
required to assess the direct impact of broadband development and use. Any
research that considers the beneﬁ ts of investing in broadband connectivity
must begin with a description of broadband services actually available to users.

catherine middleton22
Applications and services that can offer beneﬁ ts must be identiﬁ ed, 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 reﬂ ect what citizens do with their broad-
band 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 tangi-
ble 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 assess-
ment of the impact of this investment, explicitly funding research, and col-
laborating 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 beneﬁ ts to
individuals and to society as a whole.
As broadband networks are considered an essential infrastructure in the
twenty-ﬁ rst century, improved understanding of how they support individu-
als in their everyday activities will allow for more effective use of this infra-
structure. Research on the speciﬁ c 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 avail-
able to them. Billions of dollars will be invested in broadband infrastructure
development. This chapter offers some speciﬁ c 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.

23
CHAPTER 2
Understanding Digital Gaps
A Quartet of Empirical Methodologies
bin zhang and richard d. taylor
Since the 1960s, the ﬁ eld 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 reﬂ ects 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-calleddigital divide (or e-readiness). Approaches to this matter have grown
more sophisticated over time, and now use complex statistical methodolo-
gies 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 prin-
ciples 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 ﬁ nding coherence in the ﬁ eld 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 ﬁ eld, 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 ﬁ eld, 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

b. zhang and r. d. taylor24
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 ﬁ eld will be of greatest beneﬁ t 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 sev-
eral methods may ﬁ t 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 effec-
tive 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

understanding digital gaps 25
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 ﬁ ve core
factors affecting digital divides in China: technology, economy, government,
education, and society.
Case1 focuses on the measurement of the digital divide: this refers to the
effective access gap between regions in information and communication tech-
nology (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 reﬂ ect the
main factors affecting the digital divide. Indicators measuring effective access
are given higher weights. From this point of view, therefore, AHP has greater
ﬂ exibility, 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 dimen-
sionless 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
Theinformatization level (IL) developed in this case is a composite index.
Speciﬁ cally, there are twenty-nine indicators composing the entire frame-
work—seven indicators for Society, ten for Technology, ﬁ ve 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 infor-
matization level framework were determined. In accordance with the results

b. zhang and r. d. taylor26
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 signiﬁ cant inﬂ uence
upon the regions’ informatization levels. This is evidenced by the
expenditure of communication per capita and rate of literacy over
Technology
Capacity of the
communication
network
Coverage of
information
Communication
network building
condition
Consume ability of
inhabitants
Investment for ICT of
the country
Labor resource
Cultural quality of citizens
Application level of ICT
The Internet local
contents
Contribution of
communication industry
GDP per capita
Disposable income per capita
Expenditure of communication per capita
Expenditure of communication per capita
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
Percentage of employees in the information industry as
part of gross employment
Percentage of on-campus college students as part of total
population
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
Scientific and
educational
investment & innovation
Authorized patents per capita
Scientific and technical investment per capita
Educational funds investment per capita
Length of long-distance optical cable route per square
kilometer
Length of residential trunk optical cable route
Length of optical cable route for access network
Percentage coverage of broadcast (population)
Percentage coverage of television (population)
Percentage of administrative villages with fixed
telephone lines
Capacity of mobile switching exchanges per 10,000 people
Penetration of main (fixed) telephone lines
Number of public telephones per 1,000 people
Capacity of local telephone exchanges per 10,000 people
Economy
Informatization level
Government
Education
Society
f igure 2-1. Hierarchical structure of the informatization level.

understanding digital gaps 27
ﬁ fteen 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):
′
′+
xx=x
x
ijx
ijx
j
ji ji+x
i ji
/
(,x=
j )f l 1′=x
ix
jilet
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 coefﬁ cients 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 varia-
tion, the greater will be the inﬂ uence on the comprehensive analysis. This
kind of undimensionalization tries to preserve the variation information
through the coefﬁ cients of variation of the original variables, not the stand-
ard 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 normali-
zation 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 reﬂ ects that those regions had lower awareness

b. zhang and r. d. taylor28
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 ﬁ nd that Beijing
and Shanghai had much higher index values than all the other
regions, with ﬁ gures 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: Table2-1 shows the results when
applying a descriptive statistical analysis to these numbers.
Table2-1 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 ﬁ rst 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 ﬁ fth 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 classiﬁ ed and applied frequency statistics, so that the following
additional conclusions could be reached.
table 2-1. . 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 111111
Case Number31 31 31 31 31 31

understanding digital gaps 29
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 ﬁ fth 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 com-
puted 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 ﬁ g ure 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 ﬂ uctu-
ates drastically, but it stopped at 8 in 2007, which followed a drop compared
to10 in 2002; the ﬁ g ure of the 0.76–0.875 group climbed with less dramatic
ﬂ uctuation, ending up with the same level in 2007 as it did ﬁ ve 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 cha rts 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.
0.40
2002 2003 2004 2005 2006 2007
0.45
0.50
0.55
0.60
Standard
figure 2-2. Standard deviation of the index values, 2002–2007.

b. zhang and r. d. taylor30
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 equ alization 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 speciﬁ c 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 ﬁ xed):
DigitalDividedd
Indexveealuevv
i
n
n
()DDDD
()i
=
−
=
∑ 1
1
DD stands for Digital Divide; Index value means the index value of
informatization level in the No. i region; n equ als 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.
fi gure 2-3. Frequency statistics for three groups, 2002–2007.
0
2002 2003 2004 2005 2006 2007
2
4
6
8
10
12
0.625–0.75 0.75–0.875 0.875–1

understanding digital gaps 31
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 Figu re 2-5.
To better understand the cause of the narrowing trend, it is necessary to
analyze the changes from the ﬁ ve 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
0.200
2002 2003 2004 2005 2006 2007
0.300
0.400
D
figure 2-4. Changing trend of the digital divide in China, 2002–2007.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
2002 2003 2004 2005 2006 2007
Technology Economics
Government Education
figure 2-5. Digital divide trends comparing ﬁ ve indexed components, 2002–
2007.

b. zhang and r. d. taylor32
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 chieﬂ y represents ICT and educational
investments from the government. Its values ﬂ uctuated 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
Case2 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 under-
standing of the digital divide in China. Case 2 also takes into consideration
the ﬁ ve 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 clus-
tering, case 2 identiﬁ es twelve different types of factors which inﬂ uence 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

understanding digital gaps 33
also vertical time-distance comparisons of classiﬁ cations 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 deter-
mine 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 cluster-
ing 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:
′ = ′+xx=xx x+
ij
x
ij
x
jj
x
ij ij
x,
j
xif h01 ′x=x,then .
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

b. zhang and r. d. taylor34
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 ﬁ ve
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 twenty-
four categories, which were then clustered. The provinces were then ranked
in accordance with the distance between the samples.
Case2 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 infor-
matization 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.

understanding digital gaps 35
• 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 catego-
ries, representing three ICT application levels in each region: high, medium,
and low.
According to the classiﬁ cation ta ble for the digital divide for 2002–2007,
the total classiﬁ cation ta ble of those six years was produced. Beijing and
Shanghai, as the leading regions in ICT application, have been in the ﬁ rst
class, and their digital divide index rankings have stayed ﬁ rst 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 ﬁ ne-tuned based on the
actual situation. Then the ch aracteristics of each type were summarized.
The clustering results can be used as the reference for regional classiﬁ cation.
When the ch aracteristics 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 speciﬁ c 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 ﬂ uctuations in Hainan’s digital divide index rankings,
but not large. As to the classiﬁ cation 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

b. zhang and r. d. taylor36
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.
During2003 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.
From2004 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 signiﬁ cant 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
in2005–2006, the strong ranking of Government and Society indicators
and the increase of other indicators’ rankings, especially the signiﬁ cant
increase in its Economy indicator ranking, made Hainan’s digital divide
index ranking rise two positions. The important inﬂ uence of economic
factors on Hainan’s digital divide index ranking started to appear.
From2006 to 2007 the Government indicator ranking increased slightly,
and the Economy indicator ranking continued to show a signiﬁ cant increase.
The economic factors had become the most important factors to enhance
Hainan’s digital divide index ranking. At the same time, the Education indi-
cator ranking advanced one position, but was still relatively backward. The
Society indicator ranking dropped one position, while the digital divide
index ranking remain unchanged.
During2002 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, scientiﬁ c and educational input and innovation, and eco-
nomic 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.

understanding digital gaps 37
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 infor-
mation 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 reﬂ ecting 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, efﬁ ciency 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
figure 2-6. The relations between growth, efﬁ ciency and inequality when
based on a dynamic concept of overall degree of disparity.
100
1000
Per capita income (log scale)
10000
1960 1970 1980 1990
S
122(t)
S
122(t)
S121(t)
S121(t)
R
12(t)
R
12(t)
2000
Time
2010 2020 2030
Unit 1
Scenario A
Scenario B
Unit 1
Unit 2
Unit 2
2040

b. zhang and r. d. taylor38
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 com-
pared 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 signiﬁ cant digital divide between the thirty-one prov-
inces 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 distribu-
tion 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 prov-
inces and the national average. According to the indicator system and corre-
sponding 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: ﬁ rst, 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 two-
dimensional 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
from2002 to 2007 was calculated. It showed that the digital divide is gradu-
ally 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
in2002 and by four to eight years in 2007. The majority of backward prov-
inces fall behind the national average by two years in 2002, but it becomes

understanding digital gaps 39
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 Ta ble 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 eq ual 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 Fi gure 2-7.
The developing trend is obvious. The two polygonal lines are distributed in
the ﬁ rst 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
ﬁ rst quadrant is better than the national average.
It can be seen from the polygonal line that Zhejiang is in the ﬁ rst 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 2007Trend of digital
divide
Main provinces
District1More than 0 Positively widen Beijing, Shanghai, Tianjin,
Guangdong, etc.
District2More than –1, less
than1
In the range of plus
or minus
Hainan, Shanxi, Jilin,
Xinjiang, etc.
District3Less than 0 Negatively widen Neimeng, Jiangxi, Hebei,
Xizang, etc.

b. zhang and r. d. taylor40
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 ﬂ uctuation 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 speciﬁ c 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 ﬁ ve
sub-indexes: Technology, Economy, Government, Education, and Society.
And these indexes consist separately of several speciﬁ c 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 ﬁ ve subindexes of Zhejiang and the national average.
The digital divide index of Zhejiang led the national average by 1.36 years
in2002, 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 aver-
age 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
figure 2-7. 2D polygonal lines show the time distance comparison of ob-
served provinces (Yunnan, Zhejiang), 2002–2007; positive ﬁ gures
mean ahead of the national average.
−0.1
−2−4 0
Absolute difference 2
20022004
2004
2005
2005
2006
2006
2007
2007
2003
2003
Time distance (year)
4
0.1
0
0.2
Yunnan Zhejiang
2002

understanding digital gaps 41
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
by3.65 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 efﬁ ciency evaluation method based
on the concept of relative efﬁ ciency. As an efﬁ ciency evaluation method,
DEA is a uniﬁ ed 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
efﬁ ciency value for each unit.
In case 4, DEA is used to study the digital divide based on ﬁ ve core com-
ponents: Technology, Economy, Government, Education, and Society. In
order to better analyze the impact of the various factors on the digital divide,
the efﬁ ciency of input-output was divided into two levels for comparison:
the analysis of the conversion efﬁ ciency of technology and of the conversion
efﬁ ciency of social applications. Then, based on a production-proﬁ tability
matrix, all the efﬁ ciency values are plotted into two dimensions in four
quadrants, divided by the two reference points of the national average efﬁ -
ciency in each year. In each quadrant, representative provinces are selected
for study, and for each, recommendations are proposed for improvement.

b. zhang and r. d. taylor42
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 digi-
tal divide of the various factors, we divide the efﬁ ciency of input-output into
two levels to provide a comparison, leading to a further analysis of the conver-
sion efﬁ ciency 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 inﬂ uence factors; Technology is used as a ﬁ rst-class output indica-
tor to inspect the efﬁ ciency of input factors into the related indicators. Social
factors are treated as a second output indicator, estimating whether eco-
nomic 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

understanding digital gaps 43
The geometric mean is used instead of the algebraic average because the
geometric mean represents more balanced value judgments, not the develop-
ment of indexes of inequality.
The index is:
ˆ
II
tt
II
i
i
n
1
e
I
t
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
theith indicators. For technology, n=10; for the economy n=4; for the
governmentn=5; for education n=3; for society n=7.
evaluation results and analysis
TheC
2
R 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 efﬁ ciency.
Based on the adoption of a DEA efﬁ cient deﬁ ned model, we can see:
• When technical efﬁ ciency or social efﬁ ciency is 100 percent, and the
remaining variables or the slack variables of the inﬂ uencing factors
are0, the result is referred to as technical DEA efﬁ cient or social DEA
efﬁ cient, 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 efﬁ ciency or social efﬁ ciency is 100 percent, but the
remaining variables or slack variables of the inﬂ uencing factors are
not0, the result is referred to as technical DEA weakly efﬁ cient or
social DEA weakly efﬁ cient. However, in this study, as in the model,
there is a relatively large number of variables as to the decision-
making units for provinces, so there are no examples of the DEA
weakly efﬁ cient 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 efﬁ ciency or social efﬁ ciency are less than 100
percent, it is referred to as technical non-DEA efﬁ cient or social
non-DEA efﬁ cient, 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

b. zhang and r. d. taylor44
analyzed to ﬁ nd the root causes of the result being non-DEA
efﬁ cient.
Analysis shows that in recent years, technical efﬁ ciency is slightly higher
than social efﬁ ciency, that is, the input of the inﬂ uencing factors is better
able to transfer into technology, while the application in the transformation
of society is less certain. At the same time, the efﬁ ciency levels are increas-
ing each year, and social efﬁ ciency levels are increasing faster than technical
efﬁ ciency levels. In fact, for all provinces, the annual values of efﬁ ciency
levels have increased in varying degrees, as a result of actively taking meas-
ures to enhance the level of ICT technology and the corresponding level of
social application.
For technical efﬁ ciency of output, the top provinces are Beijing, Jiangsu,
Shanghai, Shandong, and Anhui. However, for efﬁ ciency 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 efﬁ ciency; Guangdong
province, although a little behind on technical efﬁ ciency, has a high efﬁ -
ciency of social applications. Jiangsu, Shandong, Anhui, and other prov-
inces, 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 beneﬁ t
of the infrastructure construction.
analysis of the relative position of technical
efficiency and social efficiency
The relationship between the efﬁ ciency scores obtained from both the
technical efﬁ ciency and social efﬁ ciency assessments can be explored by
means of a production-proﬁ tability matrix as proposed by Boussoﬁ ane,
Dyson, and Thanassoulis. We use social efﬁ ciency as abscissa and technical
efﬁ ciency as the longitudinal coordinates, and the average point of technical
efﬁ ciency and social efﬁ ciency as the criteria for the classiﬁ cation, to divide
the ﬁ gure into four quadrants. These useful scores can also be plotted in two
dimensions.
The overall trends show that the criteria for classiﬁ cation gradually move
to the right, and the technical and social efﬁ ciency of the average is increas-
ing on a year-by-year basis. For all provinces, technical and social efﬁ ciency
correspondingly increased more quickly. Judging from the details, each
quadrant has its own c haracteristics. Two c haracteristic provinces from two
of the four quadrants are selected and compared. The remaining variable

Random documents with unrelated
content Scribd suggests to you:

resident of the plateau above. In fact many of the wealthy mine
owners of Potosí in former days, if not at the present time, made
their homes here, where life is much more enjoyable. Made the
capital of Bolivia in 1826 it still has the name, though now it is
the seat only of the Supreme Court and of the Archiepiscopal
See; the Legislative and Executive Departments of Government
being at La Paz. The Legislative Palace of Sucre with handsomely
decorated halls still remains, there is a stately new Government
Palace, a Palace of Justice, the University of San Francisco
Xavier, and other important buildings. Among the churches, the
Metropolitan Cathedral is the richest in Bolivia. The Virgin of
Guadalupe, an image of solid gold, with its rich adornment of
jewels, is said to be worth a million. Among the nine plazas, that
of the 25 de Mayo has a special mark of distinction in the fact
that it has two streams, one on each side, one of which flowing
northward joins the Mamoré so reaching the Amazon, while the
other turning southeast goes on to the Pilcomayo and at last to
the estuary of La Plata.
One who sees only the plateau region of Bolivia knows but a
small part of the country; the section east of the Andes, now
becoming accessible, is far more attractive and within a half
century may have the larger part of the population.
From La Paz to Antofagasta. The remaining route from La
Paz to the sea will be followed by those who have visited any of
the three cities last mentioned, the old road by way of Oruro to
the southern port, Antofagasta, though not until 1908 was the
railway opened between Oruro and Viacha. Many in the past
have groaned over the journey which formerly involved two days
by diligence to Oruro and three by rail to Antofagasta, but since
the introduction of sleeping cars on the old section and the
completion of the new the trip may be made in comfort and
even with pleasure in 48 hours. Within the year the road has
been prolonged from Viacha down to La Paz, another great
improvement.

Except for the fine view of Illimani on the left in the early part
of the journey, the ride to Oruro is of no great interest. Some tall
mud built piers may excite curiosity: a few remaining from those
erected three centuries ago which formerly, it is said, marked the
entire route from Lima to Potosí. Before reaching Oruro, a ride of
about seven hours, a snow-crowned volcanic peak may be seen
at the southeast, Sajama, with an alleged altitude of 22,700 feet.
A possibility, is therefore presented of its overtopping
Aconcagua, or like Coropuna turning out 1000 feet lower.
INDIANS TRANSPORTING FREIGHT

PLAZA AND GOVERNMENT PALACE, ORURO
At a station called Patacamaya a halt is made for almuerzo.
Strange to say, the restaurant, where a fair meal is served, is
kept by an American and his wife who have been living there
about twenty years. The gentleman remarked that he was
contented, doing well, and had no desire to return to the States.
Fortunate it is that all have not the same tastes, some enjoying
the warm tropics, some the desert, some the cool plateau, some
happy only in large cities, and others whom the solitary places
please. Many who go down to engage in railroad construction, to
work in mines or smelter, or even to fill office positions in cities,
soon become tired and return; others are fascinated with the
life, being successful, and persons of more importance than they

would be at home, and they are glad to settle permanently in
those countries.
Oruro is an important mining town of about 20,000 people,
with a very good hotel, the Unión, facing the pretty Plaza.
Arriving on Wednesday or Saturday at Oruro, one may the same
evening at 7.30 take the express train for Antofagasta, a ride of
36 hours. The plateau seems rather dreary and only those who
have an interest in mining matters will care to stay over. The
various mines on the outskirts of the city produce both silver and
tin. There are many foreign residents with several clubs and life
is not so dismal as may at first glance appear, although the
climate at this altitude of 12,500 feet in the exposed position on
the plain is a trifle raw. The Government Palace and the
University building face the Plaza, and the city boasts of a
theater, a public library and a mineralogical museum, as well as
the usual churches, hospitals, and schools. Oruro was noted
during the colonial period as next to Potosí in the richness and
production of its mines and in 1678 is said to have had 76,000
inhabitants. In the immediate vicinity are half a dozen mines,
formerly great silver producers, but now worked chiefly though
not entirely for tin. The San José mine, two miles from the town,
several years ago was yielding $55,000 a month in tin and silver.
It is an interesting place to visit, employing 1000 or more people
and equipped with the best of modern machinery. There are
workings 1000 feet deep. The Socavón de la Virgen, nearer the
city, is one of the oldest of Bolivia. In all four provinces of this
Department are rich tin mines. The ore is treated by grinding
and concentration, the product exported averaging about 64 per
cent tin. Copper also is found, and farther south borax, and
metals of almost every kind.
For the through journey to Antofagasta, staterooms should be
engaged in advance at La Paz and in the best possible car; as I
was informed that there was considerable difference. Some
persons complain about everything and I had heard much of the
discomfort of the journey. But the accommodations which I

enjoyed were decidedly superior to those of an ordinary Pullman
and I never experienced a more comfortable railway ride. The
road is of very narrow gauge, 2½ feet, so that an aisle passes
along one side of the car with staterooms in a row on the other.
In these the berths are not crosswise of the car as in Argentina
but lengthwise. My room had two very comfortable leather-
covered armchairs, facing each other, on which the berth was
later arranged with none above it. A wash-basin with running
water was at the side, a small mirror, and several nails on which
to hang clothing. In a dining car good meals, dinner and
almuerzo were served at a fair price, morning coffee in one’s
own stateroom. Traveling from Oruro at night one misses the
sight of Lake Poopo. Poopo is a curious shallow, salt, and turbid
lake with no visible outlet, fed by the Desaguadero River from
Lake Titicaca. Although 24 by 53 miles in extent it is at most but
9 feet deep, often less than 5, and seems to be shrinking. In this
dry air and strong sunshine the water may in time disappear,
leaving only a bed of salt. Uyuni, from which the railway is now
being continued to Tupiza, 125 miles beyond on the Pan
American route to Argentina, is also passed in the night. From
Tupiza it is hardly 60 miles to La Quiaca which was reached by
the Argentine Railway several years ago. A few miles from Uyuni
are the Pulacayo and Huanchaca mines which have produced
within the last quarter century about 5000 tons of silver, thus
taking rank as the second silver district in the world (the first is
Broken Bow, Australia). Electricity is here the motor power;
Corliss engines render service; several thousand men and
women are employed, the latter sorting ore with wonderful
accuracy. The day following is spent among the desert
mountains. The hills are red, yellow, white, and gray, dotted with
black cinders. Volcanoes are numerous, mostly extinct but
showing perfect cones against the blue of the sky. Large level
sheets of saline material are frequent. Some jagged hills have
streaks, blood-red or chrome-yellow. The volcano San Pedro,
17,170 feet may be smoking. From a smaller cone, Poruña, at its
side, stretches a great stream of lava, like a glacier, half a mile

wide and several long, through which in a cutting the railroad
passes. Just before dark, close to the Conchi station, the train
crosses a viaduct 336 feet above the Loa River, more than twice
as high as the celebrated Forth Bridge. It is a graceful steel
structure with six lattice girder spans of 80 feet each, on steel
towers. Early the second morning one arrives at Antofagasta.

CHAPTER XVI

THE CHILIAN COAST—ARICA TO VALPARAISO
Arica. Arriving at Arica by sea, or departing as well, one may
observe in great white letters on the rocky Morro, Vive Battalion
No. 4, commemorating the Chilian victory with its massacre of
Peruvians, June 7, 1880. The 1700 Peruvians here stationed,
whose cannon were directed towards the sea, suffered an
assault in the rear from 4000 Chilians who had landed at night
several miles below. Short of small arms and ammunition, after
an heroic defense for one hour, the commander, Col. Bolognesi,
perished having used his last cartridge, and many soldiers leaped
to the rocks by the sea, who preferred this death to having their
throats cut by the Chilians. Others were crowded off by Chilian
bayonets, and for months the bodies were seen below. No
prisoners were taken, the entire garrison of 1700 being
slaughtered.
The harbor, one of the best south of Callao, is called by one
writer the emerald gem of the West Coast, on account of its
green trees and other verdure. The line of railway may be seen
among the cliffs, and a great cross on the highest hilltop. The
town is called by one writer very squalid, by another a neat,
attractive place in comparison with most of the port cities, the
houses of various colors, blue, green, orange, etc., many with
arched entrances affording pleasing views of an inner patio. On
account of earthquakes the buildings are chiefly of one story,
many of corrugated iron. The most noted of the ’quakes was

that of 1868 when two United States frigates were in the harbor.
One of these, the Freedonia, was lost with all on board; the
other, the Wateree, by a wave 60 feet high, was carried over
houses a mile inland, suffering a loss of half the crew. The ship
there became the home of several Indian families, until the next
earthquake and wave carried it back to the beach without doing
injury to the occupants. Rarely from the harbor may be had a
beautiful sunset view of snow-crowned Mt. Tacora, 19,000 feet,
though other mountains are frequently seen. In this port
Hernando Pizarro built ships for the invasion of Chile. On the
broad beach is a prehistoric cemetery with embalmed mummies,
said to be equal to those of Egypt. Some of the eyes are
translucent with a rich amber tint, which scientists say are of
squid or cuttle-fish here numerous, substituted for the eyes of
the dead. It is said that when some of these were sent to
Tiffany’s in New York to be polished, the workmen suffered a
violent irritation of the eyes, lips, nostrils, and throat. Though all
recovered, the work was not resumed. An analysis showed
animal matter with saltpeter and unknown minerals.
It is believed that along here is a subterranean outlet of Lake
Poopo, as the fresh water fish of Lake Titicaca, peccajay, are
caught in the ocean, and driftwood of the mountain vegetation
appears. Formerly Arica was a great market for vicuña skins,
which were brought down from the interior, but their number has
now greatly diminished. A highway constructed by the Incas
1000 years ago, called the camino real, has been in use ever
since, the Bolivians, even after the construction of the railroad to
Mollendo, still using it to bring down ore by means of llamas and
burros and to carry up supplies. The new railroad may not cause
a complete disuse of the old route, as the carriage of freight by a
road of so heavy grade is likely to be expensive.
Tacna, 38 miles distant, capital of the province, connected by
rail with Arica, is a pretty and a larger city, worthy a visit. The
prosperity of this section has been delayed by the friction and
hostile feeling between the Peruvian and Chilian Governments

and peoples, resulting from the unfortunate war 1879-1883, and
the unsettled conditions following. The Tacna-Arica question has
been one of greater bitterness than that of Alsace-Lorraine; the
present arrangement, to postpone the plebiscite twenty-one
years, will be greatly to the advantage of both countries. On the
desert between the two cities is often an unusual effect of
mirage, and from Tacna there is a mountain view of much
grandeur.
Iquique. The next important port south of Arica is Iquique,
but between the two is Pisagua where many boats call, affording
opportunity for any who desire, to disembark and go 124 miles
by rail to Iquique, thus to see without loss of time something of
the rich nitrate lands of Tarapacá. This part of the coast may not
look very different from some of the Peruvian, yet it is still more
of a desert; for the Peruvian will blossom like a rose, with a
sufficient water supply, while this is less easily transformed. In
Iquique, gardens and plazas have been made by bringing from a
distance artificial soil for the trees, shrubs, and plants, which
must be carefully nurtured. The nitrate ports are said to look like
western mining towns, with wide streets, and one-story houses
made from Oregon lumber, with iron roofs. There are many
shops selling much liquor and canned stuffs. The streets were
formerly dusty, the air full of sand. Unnatural tastes were
developed by the conditions. Two miners in earlier days, wishing
to enjoy a feast, sat down with two cans of pâté de foie gras, a
loaf of bread, a bottle of brandy, and two cans of condensed
milk, the last being eaten with spoons as dessert.
Hotels, Phoenix 8 to 15 pesos, Europa 7 to 15, Grand, 6 to 10,
all A. P. Iquique, the principal Chilian port except Valparaiso, is
the most important center of the nitrate industry. With a
population of 50,000, called a fine city, it has an enormous
commerce for its size, not merely from the export of nitrates but
because it is unique in having all its supplies brought in by sea,
food, fuel, and formerly water. The port receives more than 1000
vessels a year. The population is rather rough and hard to

govern, though with a circle of aristocratic society, with the usual
accessories. The Arturo Prat Plaza with a statue of the hero in
the center is an attractive place. One may here first observe
women conductors on the street cars, many of whom will be
seen in other Chilian cities.
Water, formerly, when brought by sea, 10 cents a gallon and
at times $2.00 when the supply boat was overdue, now comes
from the mountains, a distance of 148 miles, by a 10 or 12-inch
pipe, partly on the surface of the desert, or buried two or three
feet. To Antofagasta water is brought 173 miles from a point
10,700 feet above the sea; to Taltal, 102 miles. Though
expensive, costing millions, it has proved profitable. The streets
of Iquique are now piped, hydrants protect against fire, the dust
is laid by sprinklers, some people have bathrooms, a few,
fountains in patios, a costly luxury. It was once said that people
drank champagne because water was too expensive. It is an
enterprising community with a good portion of Anglo-Saxons;
there are broad streets, fine churches, schools, hospitals, a large
theater, pleasant homes, and good Clubs. Some of the people
entertain sumptuously, with dinner parties as in London. A broad
driveway along the beach leads to Cavancha, an attractive resort
with a dancing pavilion, and a choice flower garden tended with
utmost care. Halfway is the Jockey Club-house, with race track,
tennis, and bowling.
A railway climbs the variously colored mountain back of
Iquique to the Pampa of Tamarugal, where it branches to various
officinas, interesting to visit if time allows. People who are born
and have lived in this section can hardly believe stories about
grass that has to be cut, and of trees and flowers. A girl of
sixteen who had visited Santiago on her return said, “Trees,
trees, everywhere, grass growing in a thick mat, and hundreds
of flowers! A perfect paradise!”
The valuable nitrate lands which, previous to the war,
belonged to Peru and Bolivia are now the chief source of Chile’s

wealth. Yet it is a curious fact that though Chile receives from
her export tax on nitrates the large sum of $13,700,000 annually,
the finances of the country, if they may be judged by the
currency, are in a poorer condition than those of Peru, where
with a firm gold basis gold and silver coins are used, while in
Chile there is paper money of low and fluctuating value.
The nitrate deposits are found in the three provinces of
Tarapacá, Antofagasta, and Atacama, along from Pisagua to
Coquimbo, about 300 miles. The deposits with an average width
of 2½ miles are between the coast hills and the Andes, 10 to 80
miles from the sea, and from 2000 to 5000 or more feet above
its level, covering a tract of about 250,000 acres. The deposits,
sometimes on the surface, are oftener overlaid with strata of
earth varying in thickness and character, occasionally with
guano. They are not continuous, but separated by other
deposits, in some places salt. The raw material called caliche
carries usually from 20 to 65 per cent of nitrate of soda. It is
pickled in tanks from eight to twelve hours, the sand and refuse
dropping to the bottom. The liquid called calso runs off into vats.
The salt by-product is used or discarded. When treated and
ready for export the article carries 15 to 16 per cent of nitrogen
and 36 per cent of sodium. The amount of production is
regulated by a syndicate, according to the needs of the world.
About 35,000 men are employed, the laborers earning from
$1.00 to $2.00 a day. These establishments, called officinas, are
interesting to visit, but it is a gloomy, depressing region for most
persons. The superintendents, doctors, and other officials
receive good salaries and are supplied with comfortable quarters.
$100,000,000 or more of British capital and some German, is
invested here and large fortunes have been made. New nitrate
fields recently discovered are held at $2000 an acre.
The nitrate of commerce is a white cheese-like substance from
which the highest grade gunpowder is made; it is also used in
chemical works to produce nitric and sulphuric acid, etc., but the
bulk of it is employed as a fertilizer, doubling or tripling the

harvest. A mineral substance, it is distinguished from guano, the
excrement of birds. As to its origin there are various theories,
but none is generally accepted. A byproduct, a yellow liquor,
which in its preparation is drawn off from the nitrate into a
crucible, is then chemically treated, poured into smaller pans,
and on cooling leaves on the dish a blue crystal, the iodine of
commerce, which costs as much per ounce as saltpetre per 100
lbs. The casks in which it is placed are covered with green hides
which shrink and keep out the moisture. Worth $700 to $800 a
cask, the iodine is shipped in the treasure vaults with bullion.
About 40 per cent of the nitrate goes to Germany, 30 to the
United States, 20 to France, the rest to Great Britain and
Belgium.
Antofagasta. The next port, 200 miles below Iquique, at
which express boats call, is Antofagasta, the terminus of the
other railway from Bolivia, via Oruro. Here are sea-lions, diving
birds, and a considerable town, but no sheltered harbor, in spite
of which much commerce is carried on. This, with Iquique, as a
poor port, almost rivals Mollendo. One writer says it is an ugly
dun-colored place, another that it is the prettiest town since
leaving Panama. It has an air of prosperity with good shops and
business houses, a comfortable hotel, the Grand, A. P., 7 to 20
pesos, well furnished rooms, and real milk; another says the
hotel is very bad. Much depends upon one’s disposition, point of
view, what he expects, and where he has come from; and you
may read exactly opposite opinions of many places and people,
as happens even of cities in the United States.
Some steamers call at Caldera, 207 miles south of
Antofagasta, with a sheltered harbor, and the oldest railway in
South America connecting it with the town of Copiapó; the
express boats call only at Coquimbo nearly 200 miles farther and
200 north of Valparaiso.
Coquimbo, at the end of the desert country, a busy port,
shipping more copper than any other in South America, is

situated at the foot and up the side of cliffs. The country around
is very rich in fossils. At Herradura on Horseshoe Bay was found
a petrified icthyosaurus 20 feet long, which visitors are taken to
see; they are informed that it is 12,000 years old. Above in the
mountains, at an altitude of 4000 feet, is a very sacred shrine, a
Virgin of the Rosary, at a small village called Andacollo. During
Christmas week pilgrims come by thousands from all parts of the
country, even from Peru and Argentina, some walking hundreds
of miles. Precious gifts and jewels valued at hundreds of
thousands of dollars have been presented at various times.
One day’s sail from Coquimbo is Valparaiso.
Chile. The country of Chile is very peculiar; let me hasten to
add, in nothing more serious than its shape. It is indeed
excessively long and narrow, its great extent from north to
south, 18° to 56° S. Lat., a distance of nearly 3000 miles, giving
it a remarkable variety of productions and making it larger than
any European country except Russia, although it is only from 100
to 300 miles wide. It is peculiar also that in spite of its scanty
width, it is divided into three narrower strips, a low Coast Range,
a longitudinal valley or plateau, and the high range of the Andes.
With practically no rain in the north, it has a gradually increasing
rainfall towards the south, till near the extremity there is rather
too much. The northern part is the nitrate and mining section:
the central and larger part is an agricultural zone of great
possibilities, with good pasturage area; while farther south is an
excellent forest region. There must obviously be a great variety
of scenery as well as of climate, so that in one section or
another all tastes may be gratified.
Chile was first invaded by Europeans soon after the founding of Lima in
1535. To Pizarro, Charles V, on hearing of the conquest, had given the
country seventy leagues south of that previously bestowed; to Diego de
Almagro the two hundred leagues beyond. In which section lay Cuzco
was a matter of dispute. Pending its settlement Almagro decided to
conquer the remainder of his province. That this region was richer in gold
and silver than Peru was doubtless a tale of the Incas to distract the

conquerors for their own advantage. However, with an army of Spaniards
and some Indian captives, Almagro set out over the Bolivian plateau to
investigate and take possession of the unknown country. On the barren
heights they suffered hunger, cold, and mountain sickness, the difficulties
of this terrible journey in many ways surpassing those of Hannibal and
Napoleon in crossing the Alps. Failure and disappointment were the only
results of the expedition, which was followed by the execution of the
gallant leader after his return to Cuzco.
In spite of Almagro’s disastrous experience, a second expedition was
inaugurated by Pedro de Valdivia, who proceeded along the desert shore,
instead of over the plateau, and after arriving at Arica, there constructed
vessels to pursue the journey. With no great loss, in December, 1540, he
reached the valley of the Mapocho, and selecting a favorable site, on
February 12, 1541, he proclaimed a new city: Santiago, for Spain’s patron
saint, de la Nueva Estremadura, from his native province. On the Plaza
de Armas was built a small chapel and a Cabildo or Municipal Council
Chamber, as well.
Still unsatisfied Valdivia pursued his explorations southward, beyond
the Bio-Bio River. In his absence the small garrison he had left behind
barely escaped destruction, being saved only by the valor and boldness
of the solitary woman in the party, Doña Ines de Suarez. The
Araucanians, the most powerful tribe in this section, were of different
caliber from the Quichuas, and long and fiercely they continued the
struggle against the invaders, who treated them with barbarous severity.
After the founding of Concepción, Imperial, Villa Rica, and Valdivia, and
the settling of the conqueror himself at the town of Concepción, the
Indians under the command of Lautaro, who as a servant of Valdivia had
learned something of Spanish methods, attacked and defeated the
Spaniards, capturing Valdivia and putting him to death with tortures.
After long-continued warfare a truce was established, with the Bio-Bio
River as the boundary line, but for two hundred and fifty years the
contest went on for the subjugation of the natives. At last, when the
Chilians rose against Spain, the Araucanians lent assistance, and
friendliness was established. As in other lands, however, civilization of a
sort proved too much for the Indians and few of pure blood remain.
On the 16th of June, 1810, the movement for independence began
with the abdication of the Governor, Carrasco, on account of difficulty
between himself and the Real Audiencia. September 18, 1810, the
Cabildo or City Council in open session elected a Junta to govern until a
National Congress should be convoked, ostensibly for the purpose of
holding the dominion for King Ferdinand, deposed by Napoleon. The

people regarding this as the birth of their independence were filled with
joy. An army subsequently sent from Spain landed at Concepción,
marched northward recruiting royalists, and after several engagements
finally put to rout the patriots, who were commanded by Bernardo
O’Higgins assisted by Colonel Juan Mackenna. October 16, 1814, General
Osorio with the Spanish army entered Santiago and there maintained
Spanish rule for three years longer. General O’Higgins meanwhile fled to
Mendoza in Argentina to join the army which was being organized in that
city by General San Martin for the expulsion of the Spanish power from
the entire continent. Three years were required for this work. In January,
1817, the invasion of Chile from Argentina was begun by a well-drilled
army of 5000 men, 1600 horses, and many pack mules. One division
came by the Uspallata Pass, along the coach route across the Cordilleras,
and the one followed by Almagro almost three centuries earlier. A second
division under San Martin came by the lower Los Patos Pass. The two
divisions, having united on February 12, gained a complete victory over
the royalists in the famous battle of Chacabuco, and February 14 entered
Santiago. The enthusiastic and grateful Chilians now offered to San
Martin the governorship of the country. This unselfish patriot declining
the honor, an assembly, February 17, appointed General O’Higgins
Dictator, thus concluding the so-called Reconquista or Reconquest of
Chile. However, troubles were not over. The Viceroy of Peru sent General
Osorio again to Chile. Landing at Talcahuano in the south he was able to
advance with his army, after defeating O’Higgins, until he approached
Santiago; but on the plain of Maipo, April 5, 1818, San Martin again
gained a decisive victory. Meanwhile on the anniversary of the battle of
Chacabuco the Act of Independence was read in the Plaza of Santiago,
and the oath was taken by the leaders. The United States was the first
nation to recognize the Republic. A navy was soon formed and with the
aid of Admiral Lord Cochrane, a squadron of eight warships and sixteen
transports in 1820 carried north the army of San Martin for the conquest
of Peru.
In 1823 General O’Higgins was obliged to resign his Dictatorship and a
period of confusion followed. In 1833 a constitution was adopted. In the
administration of Manuel Montt in the fifties railway construction was
inaugurated. In that of President Pinto occurred the War of the Pacific
with Peru and Bolivia, 1879-81, though the treaty of peace was not
signed till 1883, when the province of Tarapacá was ceded to Chile, and
occupation for ten years was arranged for Tacna and Arica. About the
same time a boundary treaty was concluded with Argentina, with which
nation Chile had been on the verge of war. Balmaceda, elected President
in 1886, instituted many reforms but by his arbitrary methods brought on

civil war. A victory by the constitutional party was followed by
Balmaceda’s suicide.
The unfortunate death of the able President Montt in 1910 was
succeeded by the election of the present incumbent, Ramon Bárros Luco.
Among the prominent Chilian families (it has been said that one hundred
of these govern the country), are many British names, the forbears of
these having married into the best Spanish American families and
become patriotic citizens of their adopted country.

CHAPTER XVII

VALPARAISO
Hotels. Royal, A. P., 12 to 25 pesos; Grand, A. P., 10-15 pesos; Palace,
about the same or a trifle less; Colon, 6 to 12 pesos. All higher in
summer.
Money. The Chilian peso, paper, varies in value from 20 to 30 cents or
more; recently it was 22 cents.
Chief Points of Interest. The Monument and the Government Palace
near the landing; the business streets; Plaza Victoria and the church
facing it; the Av. Brazil with the British Monument; the Naval School; the
English and the Spanish American Cemeteries on the heights, these with
the Naval School reached by ascensors; and the suburb Viña del Mar.
Valparaiso, Vale of Paradise, the largest and busiest port on
the Pacific south of San Francisco, like many others along this
coast, has no real harbor. The spacious semi-circular roadstead
lies open to the wintry northers which occasionally bring terrific
storms. On such occasions, ships at anchor in the bay to escape
the fury of the waves often steam for the open sea, lest they be
driven ashore or be overwhelmed in the deep, as has several
times happened to ocean steamers. In the summer there is no
danger, and after the completion of the breakwater designed to
protect the bay from the savage force of the tempestuous sea, it
will be safe at any time. The great depth of the water a short
distance from shore renders the construction difficult, but
satisfactory plans at length were devised and in October, 1912,
work was begun on the port improvements which besides the
breakwater 945 feet long will include additional docks. Those in

existence are sadly inadequate for the vast commercial
movement at this port.
While from a business point of view the harbor at present is
poor, from the deck of a steamer or from the hills above the
town, there is a busy and beautiful scene. Scattered over the
waters are hundreds of vessels of various shapes and of every
size, some from the farthest corners of the globe, showing flags
of many nations (probably none of the United States), others
mere lighters or rowboats to transport freight or passengers
from ship to shore. Around the bay, a few rods back from the
water, rise in a semi-circle steep hills or cliffs to a height of 1000
feet or more. Farther back, more lofty ridges are seen, and it is
said that on a clear day in the far distance may be descried, in
the sharp toothed ridge which forms the backbone of the
continent, the snow-flecked peak of Aconcagua.
On the narrow strip of shore between the sea and the hills,
varying in width from two blocks to half a mile, is the substantial
business section of the city; while climbing up the slopes and
crowning the hilltops is most of the residence portion, both the
fine dwellings of the prosperous and the humble homes of the
poor.
The arriving steamers are as usual beset by a throng of
boatmen, and wary must be the tourist who is not exorbitantly
fleeced, unless he has a friend on board to guide, or one from
the city to greet him. As the Chilian peso, of somewhat variable
value, is generally worth less than a quarter of a dollar, the tariff
price is not so high as it sounds; one peso for each person or
considerable piece of baggage is a suitable fee, though much
more is likely to be demanded.

VALPARAISO HARBOR

MONUMENT TO ARTURO PRAT, PLAZA INDEPENDENCIA
At the landing, arrangements may be made for the transport
of the heavy baggage by cart, while you go with hand baggage
to the hotel in a carriage; or a tram car may serve you. The
hotels, the Royal, Palace, and Grand, are all within half a mile of
the landing. The Royal Hotel, 65 Esmeralda street, which is
sometimes full to overflowing, will be found amply satisfactory.
The American proprietors, Mr. and Miss Kehle, have made it
more like a hotel in the United States than are any others that I
have seen on the West Coast. Located on one of the principal
business streets, it affords excellent meals in several large
dining-rooms; and handsomely furnished chambers, with modern
equipment including red satin puffs for the beds in addition to
fine blankets. The price is from 12 to 15 pesos and up according
to the room. The Grand Hotel is said to afford similar

accommodations at about the same prices. The Palace, a little
cheaper, is well situated on the Plaza de los Bomberos, and
others less pretentious, as the Colon, 87 Esmeralda, are called
clean and good.
In Valparaiso, a city of nearly 200,000, it seems to be the
fashion for the residents to reply, when asked what there is to
see, “O, nothing at all.” This is by no means true, though at least
twice as much time should be devoted to Santiago. First there is
the large square near the landing on which is the handsome
Casa del Gobierno. In the center of the plaza is a fine
monument, The Country to the Heroes of the 21st of May, and
at one corner near the docks is the railway station to Santiago.
The air seems crisp and the city more European than any
previously seen.
The business streets have many handsome buildings two or
three stories high, a few even more, looking fresh and clean,
since the greater part of this district was laid low by the terrible
earthquake of 1906. A twelve-month of unusual shrinkage, of
adjustment of the earth’s surface, and of consequent calamity
was practically coincident with this year. In April 1906 occurred
the catastrophe at San Francisco, August 16, the practical
destruction of Valparaiso, and in January, 1907, the disaster at
Kingston. Some buildings in Valparaiso withstood the shocks, but
with the ’quakes and the resulting fires little of the lower part of
the city remained undamaged. The upper town was to a great
extent uninjured and the shipping in the bay received no harm.
Few traces of the calamity are now left, as like San Francisco the
town was soon rebuilt in a superior manner. While slight
earthquakes are frequent they are not fearsome, as heavy
shocks are usually half a century apart. Besides earthquakes,
Valparaiso has experienced other calamities. Founded in 1536, in
its earlier days it was three times captured and sacked by
pirates; in 1858, it was destroyed by fire; in 1866, bombarded by
a Spanish fleet; and in 1890 it suffered considerable injury from
the Balmaceda revolution. It is to be hoped that after all these

vicissitudes it may enjoy a peaceful existence. A stroll along the
principal streets to the office of the American consul, Mr. Alfred
Winslow, to the banking house of W. R. Grace, and to gaze at
the handsome shop windows is the pastime of an hour or two.
Between the hills and the water it is impossible to lose one’s
way. The double-decked tram cars are an imposing sight, and
rather curious objects are the women conductors. Having heard
of these before arriving, I was expecting to see some trim young
women, with possibly a coquettish eye turning at times upon
some of the gentlemen patrons, as occasionally happens in some
of our cheap restaurants; but no! Staid indeed are the women
conductors in Valparaiso and Santiago, and far from handsome.
Plainly dressed in a sort of blue uniform with white aprons, they
are obviously of the so-called laboring class, of rather stolid
appearance, perhaps the mothers of families, and closely intent
upon their duties. It appears that during the war of ’79–’81, so
many young men joined the army that women were drafted into
this service. Performing it in a satisfactory manner they
continued to be so employed though not to the total exclusion of
men. They mount to the upper story to collect fares and in
Santiago swing along the sides of the open cars quite in man
fashion, though necessarily hampered by their voluminous skirts.
Manifestly competent for the labor, less difficult than other duties
like scrubbing floors, supposed to lie more within their sphere, it
would seem that bifurcated garments, even knickerbockers,
would enable them to perform either service more easily. If men
and women were to exchange garments for a hundred years it is
conceivable that the idea as to which is the weaker sex might be
changed also.
A few car rides may be taken to advantage, the greater if
sitting above; but among the natives of the upper class this is
taboo, as the price is only half of that below; the fares being five
and ten centavos respectively. A gentleman in Santiago
remarked to me that although he preferred riding outside it
would never do except in the evening, when he could not be

recognized from the street or from the upper windows of houses
in passing.
Not far from the Royal Hotel is the Plaza Victoria on one side
of which is the Espiritu Santo Church, the most fashionable in
the city, though with an ordinary exterior. A flower market is
passed on the way, where beautiful roses and other flowers may
be purchased in quantities for a single peso. The general market
as a matter of course is worth seeing, especially in the season of
fruits, as Chile rivals California in the excellence and variety of
these, and surpasses it in cheapness. The fruits of the
Temperate Zone, cherries, peaches, apples, pears, and grapes,
luscious in quality and, they say, unrivaled in any part of the
world, in their summer and fall, tempt the tourist on every hand.
It is important to ascend the hills in two or three different
places, both for the view going up and for what is to be seen at
the top. The ascensors are similar to those of Cincinnati, one
being carried up by cable as another is coming down; but the
inclines seemed steeper and one appeared rather rickety. There
have been fatal accidents. However,—I went as do others. Near
the top of one of the inclines which is but a short distance from
the Hotel Royal is a cemetery where chapel-like tombs and pretty
head stones and monuments are closely packed together among
shaded walks on the very edge of the precipitous bluff. One has
here a magnificent view of the city below fringing the semi-
circular shore, of the blue waters, alive with ships, and of the
surrounding hills. Through cañons here and there separating the
various hills and bluffs, a few carriage roads wind steeply upward
and more footpaths, by which some pedestrians climb; but most
persons will prefer to save time and strength by taking their
chances in an ascensor. Perched on these steep inclines are
houses of the poor, while at the top are many fine villas occupied
by native and foreign residents. Close to the Chilian cemetery on
the bluff is the English burial ground surrounded by a high wall.
In a far corner of this enclosure is a small marble tomb on a
concrete foundation with a marble cross above, the whole about

five feet high, in which Americans will have a special interest.
The inscription reads:
“In memory of the officers and seamen slain on board the
United States frigate Essex in this harbor in an engagement with
H. R. Majesty’s frigate Phoebe and brig Cherub, February 28,
1814.” A list of 52 names follows and the statement that it was
erected by officers of four ships of the United States Navy.
This ship, the Essex, commanded by Capt. David Porter, after
inflicting much damage on British property, capturing 360
seamen and 100 cannon, was surprised in this harbor by two
British ships. Though disabled by a squall she made a splendid
fight until more than three-fifths of the crew were killed or
wounded, and the ship was on fire in several places, when she
struck her colors. A more conspicuous monument for the gallant
dead might seem appropriate.
By another ascensor, a trip should be made to the Naval
School, which crowns a splendid height nearer the outer edge of
the harbor. A fine large building, well equipped in the best
modern English fashion, stands back of a pretty garden. There
are good class rooms, laboratories, machine shops with guns
mounted as on board ship, and all essentials for a thorough and
practical course of study. In the rear patios are athletic fields
with bathing facilities. The cadets are generally from the best
families, and the program of study is based on that of English
schools; the fleet is organized on the British model, and the
ships are constructed in British shipyards. There is, further, a
training ship for sailors, where if unable to read and write they
receive instruction, as do soldiers in a corresponding institution
in Santiago.
On the fine broad Avenue Brazil is a handsome arch with the
British Lion above, presented to the city by the British colony
here, at the Centennial in 1910.

AVENIDA BRAZIL, WITH BRITISH MONUMENT

RESIDENCE VIÑA DEL MAR
Viña del Mar. An excursion should by all means be made to
this suburb; to Miramar if time allows. The former may be
reached by tram or train in half an hour or so. It is pleasant to
go by one and return by the other. The tracks, nearly parallel,
pass several pretty suburbs and give several glimpses of the sea
beyond the harbor before reaching the destination. Viña del Mar
is not only a suburb of Valparaiso whither many Englishmen and
others go in the afternoon for sports, and where many business
men of Valparaiso have homes, but it is also a fashionable
summer resort for the wealthy residents of Santiago and other
parts of Chile. It is a charming place with a pretty railway station
near a large and attractive plaza. Many carriages stand near, in
one of which for a few pesos a pleasant drive may be taken

around the town and out to the hippodrome or race track, a mile
or more outside the city. Within the track enclosure, a pretty spot
surrounded by green hills, the foreigners have laid out a golf
course, grounds for cricket, and for football. The place is thus
visited, especially on Sundays, by many, not only for the races,
to which the Chilians are as devoted as the Argentines, but for
athletics of various kinds. The Chilian horses seem very large
after those of Peru, and trotting is their specialty. Some of them
do this so well that their gentle trot is as easy as the lope or
canter of most other animals.
A pretty and commodious clubhouse faces the Plaza, and near
by are many charming villas of attractive architecture surrounded
by luxuriant vegetation of tropical and temperate climes,
beautiful flower beds, trees, and shrubbery. Half a mile from the
center of the town is a fine beach bordered by jutting rock
promontories. Large bathing establishments, cafés for ices and
tea, and splendid villas with well laid out grounds recall our own
shore resorts. A good pedestrian may be tempted to climb over
the steep enclosing hill and descend on the other side to the
electric car track for his return to the city. The Grand Hotel with
beautiful grounds is the leading hostelry of the place.
Miramar is a small but popular bathing resort in the opposite
direction from Valparaiso, reached by electric cars; but the
bathing is here more dangerous, as not far from shore the
bottom drops suddenly to a great depth.
From Valparaiso to Santiago by rail is a ride of 3½ or 4 hours
by express trains and about two more by accommodation. The
price of tickets for the express is 12.80 pesos, 4 extra for seat in
Pullman; 8.50 pesos by slower train. It is a pleasant ride, for a
few miles near the shore, passing Viña del Mar, then east
through the Coast Range to the Central Plain, at Llai-Llai leaving
the Andine Railway to turn southward to Santiago.

CHAPTER XVIII

SANTIAGO
Hoíels. Oddó, A. P., 12 to 40 pesos; Grand, A. P., 12 to 18 pesos;
Francia, and Royal, about the same; others at lower prices.
Chief Points of Interest. Plaza de Armas; Cathedral and other buildings
around; the Capitol; the Moneda; the Alameda; Parque Cousiño; most
important, Santa Lucia Park and the Cemetery; the Art Gallery.
Santiago, the capital and largest city of Chile, the third or
fourth in size in South America, considered by some travelers to
have the most beautiful location of any capital in the world
except Rio de Janeiro, is situated on the river Mapocho in the
long central valley of Chile, at an elevation of 2000 feet. Founded
by the doughty warrior and Spanish invader, friend and almost
counterpart of Francisco Pizarro, Pedro de Valdivia, it was by him
planned and laid out in 1541 after he had first built a fort on
Santa Lucia hill, an excellent site for the purpose, recalling the
ancient Greek Acropolis or some of the mediæval strongholds.
On account of the too great dispersion of the invaders, the
settlement for some years had a hard struggle for existence, but
during its century of independence it has grown rapidly. Its
population, now approaching 400,000, is ten times as great as
when independence was declared in 1810.
The site is indisputably one of remarkable beauty and
picturesque charm, without any interference with the
convenience of a large city. The hills in and on the edge of the
city, rising like small islands abruptly from the plain, do not

preclude long level streets, yet form a peculiar and admirable
embellishment, while east and west, the mountains of the Great
Cordillera and of the Coast Range, which a few miles away rise
as lofty ramparts to the ethereal blue, are an ever sublime and
noble contrast to the verdant smiling plain.
The climate of Santiago, which at 33° S. has about the same
latitude as Charleston and San Diego N., is considered excellent;
though the three winter months, in dwellings destitute of heating
apparatus, seem rather cool indoors to residents of the United
States. In the summer, though not extremely hot, it is very
dusty, so that wealthy residents at this season escape to Viña del
Mar or other seashore resorts, to the beautiful lake region, to the
springs and baths among the mountains, or even to the fjords in
the distant south. An amusing mot of a German is related by one
who did not seem to appreciate it. “The climate of Santiago is
good but it is very unhealthy.” And both statements have been
quite true, the latter inexcusably so, resulting from the fact that
ordinary sanitary measures have been neglected. The medical
congress in 1911 was held in the midst of an epidemic of
smallpox. There has been a woeful lack of sewerage. But happily
the officials have at last come to realize the importance of
sanitation, an adequate system of sewerage is now installed, and
doubtless other deficiencies will soon be remedied.
From the fine large railway station on the outskirts of the city,
a carriage or tram car may be taken to one of the hotels near
the center, a mile or more distant. To secure rooms at the Oddó,
for many years regarded as the leading hotel of Santiago, it is
often necessary to engage rooms in advance, as both main
building and annexes are generally crowded. The Oddó, near the
Plaza de Armas, is on one of the principal streets, the Ahumada,
327, the annexes on another at right angles with this, the
Huérfanos, 976 and 1012, all three in the heart of the city. The
Grand Hotel, preferred by some, is close by, Huérfanos 1164.
Other hotels approximating these are the Hotel Francia, finely
located on the south side of the Plaza, and the Royal. Prices at

Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com

Beyond Broadband Access Developing Databased Information Policy Strategies Richard D Taylor Amit M Schejter

About This Presentation

Slide Content

Tags

Categories

Download

Quick Actions

Statistics

Related Slideshows

Beyond Broadband Access Developing Databased Information Policy Strategies Richard D Taylor Amit M Schejter

About This Presentation

Slide Content

Slide 1

Slide 2

Slide 3

Slide 5

Slide 6

Slide 7

Slide 8

Slide 9

Slide 10

Slide 11

Slide 12

Slide 13

Slide 14

Slide 15

Slide 16

Slide 17

Slide 18

Slide 19

Slide 20

Slide 21

Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27

Slide 28

Slide 29

Slide 30

Slide 31

Slide 32

Slide 33

Slide 34

Slide 35

Slide 36

Slide 37

Slide 38

Slide 39

Slide 40

Slide 41

Slide 42

Slide 43

Slide 44

Slide 45

Slide 46

Slide 47

Slide 48

Slide 49

Slide 50

Slide 51

Slide 52

Slide 53

Slide 54

Slide 55

Slide 56

Slide 57

Slide 58

Slide 59

Slide 60

Slide 61

Slide 63

Slide 64

Slide 65

Slide 66

Slide 67

Slide 68

Slide 69

Slide 70

Slide 71

Slide 72

Slide 74

Slide 75

Slide 76

Slide 77

Slide 78

Slide 79

Slide 80