Information Technology Policy and the Digital Divide: Lessons for Developing Countries

Kagami 00 prelims 2/15/04 2:54 PM Page i Information Technology Policy and the Digital Divide Kagami 00 prelims ...

Author: Mitsuhiro Kagami | Masatsugu Tsuji | Emanuele Giovannetti

191 downloads 1215 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Kagami 00 prelims

2/15/04

2:54 PM

Page i

Information Technology Policy and the Digital Divide

Kagami 00 prelims

2/15/04

2:54 PM

Page ii

Kagami 00 prelims

2/15/04

2:54 PM

Page iii

Information Technology Policy and the Digital Divide Lessons for Developing Countries

Edited by

Mitsuhiro Kagami Ambassador for Nicaragua, Embassy of Japan, Nicaragua and formerly Executive Vice President, Institute of Developing Economies (IDE), JETRO, Japan

Masatsugu Tsuji Professor of Economics, Osaka School of International Public Policy, Osaka University, Japan

Emanuele Giovannetti Research Associate, Department of Applied Economics, University of Cambridge, UK and Tenure Research Fellow, University of Rome ‘La Sapienza’, Italy

Edward Elgar Cheltenham, UK • Northampton, MA, USA

Kagami 00 prelims

2/15/04

2:54 PM

Page iv

© Mitsuhiro Kagami, Masatsugu Tsuji, Emanuele Giovannetti 2004 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited Glensanda House Montpellier Parade Cheltenham Glos GL50 1UA UK Edward Elgar Publishing, Inc. 136 West Street Suite 202 Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Cataloguing in Publication Data Information technology policy and the digital divide : lessons for developing countries / edited by Mitsuhiro Kagami, Masatsugu Tsuji, Emanuele Giovannetti. p. cm. Includes bibliographical references and index. 1. Information technology—Government policy—Developing countries. 2. Information technology—Government policy. I. Kagami, Mitsuhiro. II. Tsuji, Masatsugu. III. Giovannetti, Emanuele. HC59.72.I55I545 2004 338.9'26'091724—dc21 2003049217

ISBN 1 84376 413 X Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

Kagami 00 prelims

2/15/04

2:54 PM

Page v

Contents List of contributors

ix

1. Introduction Mitsuhiro Kagami, Masatsugu Tsuji and Emanuele Giovannetti PART I

1

COUNTRY/AREA STUDIES

2. Beyond the IT revolution: the Japanese broadband strategy Masatsugu Tsuji 1. Introduction 2. Lessons learned from the net bubble 3. Broadband network infrastructure 4. Possible applications of broadband: case studies 5. Conclusion 3. Internet upstream connectivity and competition policy: Western Europe and Southern Africa Emanuele Giovannetti 1. Introduction 2. The supply side 3. Internet pricing 4. The European Internet backbones 5. The price of Internet connectivity 6. Antitrust analysis for the backbone market 7. Exploring the backbone through cyber-geography 8. South Africa Appendix 4. IT policies and issues: US and the Americas Andrew B. Whinston 1. Introduction 2. Internet infrastructure and Internet access: a snapshot 3. Telecommunications reforms and Internet usage 4. Competition in local loop and broadband service 5. Costs of access 6. The digital divide in the US

v

15 15 16 18 26 32 35 35 37 38 39 41 48 51 53 60 62 62 64 70 74 79 81

Kagami 00 prelims

vi

5.

6.

7.

8.

9.

2/15/04

2:54 PM

Page vi

Contents

7. US and the Americas: telecommunications connection 8. Conclusions Software in India: development implications of globalization and the international division of labour Paul Kattuman and Arnab Bhattacharjee 1. Introduction 2. Globalization: international competition and firm level responses 3. The Indian software industry 4. Conclusions Jumping up to the Internet-based society: lessons from South Korea Yasushi Ueki 1. Introduction 2. Rapid diffusion of new technologies 3. Accelerators of rapid penetration of new technologies 4. Contrastive effects of accelerators on e-commerce 5. Social issues caused by diffusion of the Internet 6. Concluding remarks Information technology: some implications for Thailand Chanin Mephokee 1. Introduction 2. Current IT status 3. Impact on firms and employment in Thailand 4. Some policy issues 5. Conclusions Information policy and information technology in Central and Eastern Europe with emphasis on Estonia Tanga McDaniel 1. Introduction 2. CEE in context 3. Online activity: transactions, services and rural initiatives 4. Case study: IT adoption, expansion and policy in Estonia 5. Conclusions Annex Internet and telecommunications outlook in Latin America Andrew B. Whinston and Soon-Yong Choi 1. Introduction 2. Current status of Internet development 3. E-commerce in Latin America 4. Telecommunications deregulation and policy issues 5. Expanding Internet access in Latin America 6. Outlook for future Internet growth

87 89 92 92 94 96 110 114 114 115 118 123 129 132 135 135 137 150 153 158 160 160 161 169 172 180 184 185 185 185 192 194 198 203

Kagami 00 prelims

2/15/04

2:54 PM

Page vii

Contents

10. Policies for Internet access: cases of Mexico and Argentina Soon-Yong Choi 1. Introduction 2. Telecommunications in Mexico 3. Mexican telecommunications reforms 4. Internet and e-commerce in Mexico 5. Telecommunications in Argentina 6. Initiatives for Internet growth: Argentina 7. Lessons learned PART II

vii

206 206 207 212 215 218 221 224

CHALLENGING ISSUES

11. Tipping, standardization and convergence: catch-up and failure in Japan’s standards strategy Mitsuhiro Kagami 1. Introduction 2. Tipping and de facto standards 3. De jure standards 4. Japanese defeats in the standards war 5. Technology and satisfaction 6. Convergence of technology 7. Concluding remarks 12. Is the Japanese press a dinosaur in the 21st century?: the IT revolution and newspapers in Japan Kojiro Shiraishi 1. Introduction 2. Newspapers in Japan 3. Newspapers and television 4. Outlook of new businesses 5. Conclusions 13. PKI solutions for trusted e-commerce: survey of the de facto standard competition in PKI industries Atsuho Maeda 1. Introduction 2. Analysis of major PKI vendors 3. Strategy for Asian market 4. Future prospects 5. Closing remarks 14. The interconnection and pricing of the Internet Takanori Ida and Masashi Ueda 1. Introduction 2. What is the Internet?

229 229 230 232 234 236 239 243 247 247 248 251 252 257 260 260 263 270 275 280 282 282 283

Kagami 00 prelims

2/15/04

viii

2:54 PM

Page viii

Contents

3. Pricing the Internet and its problems 4. Industrial structure of the Internet and its interconnection problem 5. The model analysis of the Internet 6. Development of the one-way and two-way model 7. Further discussion 15. Conclusion Mitsuhiro Kagami, Masatsugu Tsuji and Emanuele Giovannetti Index

285 287 290 295 299 303

309

Kagami 00 prelims

2/15/04

2:54 PM

Page ix

Contributors Arnab Bhattacharjee, Department of Applied Economics and Robinson College, University of Cambridge, UK Soon-Yong Choi, Center for Research in Electronic Commerce (CREC), The University of Texas at Austin, USA Emanuele Giovannetti, Department of Applied Economics, University of Cambridge, UK and Department of Public Economics, University of Rome ‘La Sapienza’, Italy Takanori Ida, Faculty of Economics, Kyoto University, Japan Mitsuhiro Kagami, Embassy of Japan, Nicaragua and formerly Institute of Developing Economies (IDE), Japan External Trade Organization (JETRO), Japan Paul Kattuman, Judge Institute of Management and Corpus Christi College, University of Cambridge, UK Atsuho Maeda, Japan External Trade Organization (JETRO), Japan Tanga McDaniel, Department of Applied Economics, University of Cambridge, UK Chanin Mephokee, Faculty of Economics, Thammasat University, Thailand Kojiro Shiraishi, The Yomiuri Shimbun, Japan Masatsugu Tsuji, Osaka School of International Public Policy, Osaka University, Japan Masashi Ueda, Research Centre of Socionetwork Strategies, Kansai University, Japan

ix

Kagami 00 prelims

x

2/15/04

2:54 PM

Page x

Contributors

Yasushi Ueki, Institute of Developing Economies (IDE), Japan External Trade Organization (JETRO), Japan and Economic Commission for Latin America and the Caribbean (ECLAC), Chile Andrew B. Whinston, Department of Management Science and Information Systems (MSIS), Economics and Computer Science, and Center for Research in Electronic Commerce (CREC), The University of Texas at Austin, USA

Kagami 01 chap 01

2/15/04

2:55 PM

Page 1

1. Introduction Mitsuhiro Kagami, Masatsugu Tsuji and Emanuele Giovannetti 1.

QUICK CHANGES IN ‘IT’

Information technology (IT) has greatly changed our society and this change is indeed worth calling the ‘third industrial revolution’. Basic trends can be observed in the IT revolution in terms of structural changes. That is (1) from telephony to the Internet; (2) from fixed to mobile telephone; and (3) from narrowband to broadband. In the near future, the next trend will be multimedia convergence, in other words (4) from TVs to personal computers (PCs) if TVs have the digital broadcasting function. Also, it has been pointed out that diffusion of the Internet depends on three factors: availability; accessibility; and affordability. Availability means the network’s penetration into households. Accessibility focuses on providing access points such as public telephones and community Internet access sites. Affordability is measured in terms of price of connection. Technological changes in IT have been quickly advancing beyond people’s imagination. Demand for broadband access (digital subscriber line or DSL, asymmetric digital subscriber line or ADSL, and CATV) is intense and growing while lower technologies such as Integrated Services Digital Network (ISDN), have quickly become obsolete. The third generation mobile phone is now on sale and people can enjoy photos and moving pictures through these cellular phones. Wireless networks such as Bluetooth, home servers and wireless LAN make our lives more convenient as electronic home appliances can be connected to mobile phones. The convergence of broadcasting and the home computer is taking place and TVs will function as PCs when terrestrial digital broadcasting begins. Such possibilities, which only a decade ago seemed fantasy, are actually within our reach. Under such circumstances, several fundamental factors should be grasped. First, IT-related infrastructure in both physical and software terms should be established. In the future optical fibre networks – especially fibre to the home or FTTH, which is the most convenient and speediest, but rather expensive – are expected to be deployed. Authentication and certification on the Internet is 1

Kagami 01 chap 01

2/15/04

2:55 PM

Page 2

2

Introduction

another area of importance, not only for e-government but also for all kinds of e-businesses. Here, suitable systems such as digital signature and certificate authorities are necessary and safe encrypted data transfer by the public key method should be established. This is called public key infrastructure (PKI). Second, connectivity issues should be addressed, especially affordability. The Internet route has become more complex reflecting worldwide competition and technical changes in routers and software. There are basically three layers: end-users; Internet service providers (ISPs); and major Internet backbone providers (IBPs). IBPs are usually global class telecommunications conglomerates. Message senders and final receivers on the Internet are identified but the intermediate routes upon which messages travel are not certain because messages are divided into several packets and sent to final receivers through different vacant routers and network interconnection points (or ‘hops’) so exact paths cannot be traced. Connectivity pricing (or wholesale prices) is thus difficult to determine. Therefore we have to clarify what rational prices are for Internet pricing. In addition, the possibility of collusive activities by oligopolistic backbone companies should be monitored and controlled. Third, standards in IT have also become complex and fierce competition is going on in such fields as cellular phones, digital TVs, coding and decoding, and business models. Standards are closely related to intellectual property rights. However, open source software such as Linux and TRON1 has gained popularity even for commercial use. Because technological changes are so quick, new standards swiftly replace existing standards. Therefore it is difficult to draw a clear line between market-induced standards and mandatory standards. Fourth, the digital divide has been growing not only within advanced countries but also between developed and developing countries. Basically information networks use existing telephone lines such as DSL and ADSL and this basic infrastructure is lacking in developing countries. Thus, the gap between rich and poor countries is expanding, especially in the broadband era. For example, the diffusion of the Internet and broadband in Latin America is quite slow (Argentina is an exception in terms of possible broadband penetration due to its high CATV diffusion) as compared with East Asian countries. Chile, which has the highest Internet penetration (1155 per ten thousand inhabitants in 2000) in Latin America, is less diffused than Malaysia (1590), the sixth ranked in East Asia.

2.

‘IT’ RECESSION

The IT revolution has been overshadowed by the US economic downturn, especially after the dot.com bubble burst at the beginning of 2001. Rapid decreases in IT-related stock prices and venture capital brought about recession

Kagami 01 chap 01

2/15/04

2:55 PM

Page 3

Introduction

3

in IT industries. This in turn resulted in further stock price declines. Indeed, people who speak ill of IT thought that the IT revolution was over. There are several reasons why the IT recession came to the fore. First, excess capital and investment in this sector resulted in oversupply of IT-related equipment and inventories, while demand reached saturation point. PC and mobile phone use quickly spread to almost all necessary consumers and the market went flat. In a sense this sector, too, cannot be free from business cycles. Second, stock prices relating to IT were excessively overvalued. After the IT enthusiasm was over, inflated prices had to shrink. The US economy itself had been on a buying spree reflecting the arrival of the ‘new economy’. This monetary bubble led the IT sector to instant boom but equally its collapse brought the sector crashing down. Third, the IT sector lacking a traditional foundation got itself into trouble since new business models sometimes lacked marketing and distribution infrastructure. For example, Amazon.com at first did not have its own distribution systems including distribution centres and warehouses. It had to establish its own infrastructure (‘click or brick’) and these investments were accompanied by chronic red tape. Fourth, contents were not so attractive. Although kids can enjoy interactive games and the older generation take pleasure in digital high definition TVs, most ordinary working people cannot spare much time for IT-related entertainment and the poor of course cannot afford expensive IT equipment while at the same time most people remain skeptical of an insecure e-business environment. Fifth, the terrorist atrocities in the US on 11 September and succeeding air strikes on Afghanistan threw a dark shadow over the world economy and froze consumer demand including for IT-related products. However, the importance of the Internet as a communication and business tool has not diminished so the IT revolution continues. In particular, future utilization of broadband is a guide to the second IT boom.

3.

BEYOND THE ‘IT’ RECESSION

Recent phenomena show signs that a new wave in IT business is coming. Current broadband accesses are: (a) CATV (cable modems), (b) DSL/ADSL, (c) FTTH, and (d) FWA (fixed wireless access). Speed for downloading is up to 20 Mbps for CATV, 8 Mbps for DSL, and around 100 Mbps for optical fibre access (FTTH). Due to the compression technologies of digital information, people can enjoy movies/videos and interactive games through broadband services. Since DSL/ADSL uses traditional telephone lines and prices for its

Kagami 01 chap 01

4

2/15/04

2:55 PM

Page 4

Introduction

facilitation are not so expensive as other broadband accesses, DSL/ADSL users are increasing. In Korea, the number of high speed Internet subscribers reached 7.04 million inhabitants in September 2001. The penetration ratio is almost 14 per cent, which is one of the world’s highest ratios. Out of these, four million are ADSL subscribers and 2.4 million are CATV subscribers. In Japan, DSL/ADSL subscribers dramatically increased from 211 subscribers in March 2000 to 112 thousand in March 2001 and reached 1.20 million in November 2001, while CATV grew from 216 thousand in March 2000 to 1.15 million in September 2001. These figures show that broadband is widely accepted and that its deployment globally continues. FTTH access is very expensive in terms of infrastructure investment. However, the realization of the information highway through FTTH is the most widely expected and the construction of its networks depends on government policy and private sector initiatives supported by users’ demand. The diffusion of broadband can act as a kind of ignition switch against the present IT downturn. Furthermore, quality contents should be devised and increased to enhance broadband services.

4.

FRAMEWORK OF THE BOOK

This volume consists of two parts: Part I, ‘Country/area studies’; and Part II, ‘Challenging issues’. Part I surveys country/area situations which cover Asia (India, Japan, South Korea and Thailand), Europe (EU and Eastern Europe, especially Estonia), the Americas (USA and Latin America, especially Argentina and Mexico), and South Africa. Part II deals with contemporary issues such as standardization, convergence between TVs and PCs, IT uses in newspapers, PKI, and upstream Internet pricing. Finally, policy implications, particularly for developing countries, will be drawn from these analyses. Chapter 2 titled ‘Beyond the IT revolution: the Japanese broadband strategy’ by M. Tsuji focuses on the characteristics and issues of the Japanese broadband strategy, and analyses how the second IT revolution in Japan can be promoted through broadband services as an engine of growth. He explains the current situation of broadband infrastructures in Japan such as FTTH, DSL, ADSL, CATV, and FWA. Also, government policies such as the National Broadband Network Initiative of 2001 are introduced. He clarifies the fact that the recent increase in CATV and DSL subscribers has been triggered by a ‘technologypush’ and ‘cost-push’ but consumers are not necessarily satisfied with the existing contents. Without introducing ‘killer contents’, competition between broadband and existing e-commerce will end up with a zero-sum game, meaning that broadband will only replace e-commerce. Further development of

Kagami 01 chap 01

2/15/04

2:55 PM

Page 5

Introduction

5

broadband requires a ‘demand-pull’. He concludes that a wide variety of broadband uses and new applications such as ‘telecare’ and e-education can overcome the so-called IT recession and enrich our immediate future life. E. Giovannetti in ‘Internet upstream connectivity and competition policy: Western Europe and South Africa’ in Chapter 3 analyses wholesale prices of Internet connectivity. Internet connection has three tiers: final users, ISPs (Internet Service Providers), and IBPs (Internet Backbone Providers). Price structure is divided into two: retail prices which users pay for ISPs and telephone companies (prices for the ‘last mile’), and wholesale prices which ISPs pay for IBPs (sometimes called ‘upstream’ connectivity prices). He points out that while a rich and varied economic debate has developed on the issues of liberalization of the telecom market in terms of cost structure and pricing, little has been done on the less known side of the Internet: the upstream connectivity offered by IBPs. The main problem is the elusive nature of the commodity traded, wholesale transmission of information packets along routes which are often recalculated at each step of the transmission process. Even a single e-mail message is decomposed into many sub-messages (packets) which may, or may not, reach the final destination traveling across different routes and networks and while some of the networks crossed are for a fee others are free. Thus, he emphasizes that we need measures to monitor and clarify the paths that data packets take through the Internet, recording all the ‘hops’ (routers) along the way. He discusses antitrust issues for multinational telecom companies or IBPs through mergers and acquisitions. He also illustrates the relevance of cyber-geography or Internet connectivity maps of Europe as well as South Africa. In Chapter 4, ‘IT policies and issues: US and the Americas’ A. Whinston shows that US Internet users are expanding and connection speed is upgrading. For example, the US Internet penetration rate is 58.1 per cent, second to only Sweden (61.8 per cent) according to Nielsen Net Ratings and broadband users reached 12 million at the end of 2000. Such expansion is backed by the Telecommunications Act of 1996 which promoted and liberalized the following major fields: (1) relaxation of FCC (Federal Communications Commission) regulatory intervention; (2) a more favourable environment for mergers/acquisitions across the telecommunications sector; (3) expanding competition to achieve lowered price and better service; (4) promotion of new technologies and businesses; (5) promotion of the convergence and integration among telephony, cable and computer industries; and (6) minimizing state regulatory powers. He argues that despite such strengths, however, the ‘digital divide’ still plagues many groups of the US population, mainly those with low income and low education levels. Especially, inner city neighborhoods have the lowest Internet penetration rates. He points out that low Internet penetration rates for Latin American countries are mainly due to the following three factors: first, inadequate infra-

Kagami 01 chap 01

6

2/15/04

2:55 PM

Page 6

Introduction

structures such as electricity, telephone connections, and postal services for ebusiness; second, the low level of income along with the associated problem of uneven income distribution; and third, a significant number of Internet users connect to the Internet from various locations (away-from-home Internet access such as work places, libraries, schools and Internet cafés). In ‘Software in India: development implications of globalization and the international division of labour’ in Chapter 5, P. Kattuman and A. Bhattacharjee demonstrate the possibility of ‘leapfrogging’ developments through knowledge intensive industries such as software in India. They carry out a firm-level survey of the software industry and explore differences and similarities between foreign firms and domestic firms (with and without foreign subsidiaries) in terms of their inputs and strategies (as regards markets, specialization and application areas), and their resultant output variables. The international location tendency, the learning process and evidence of a move up the value chain by Indian software firms show that domestic firms have a route available to them to overcome the difficulties of growth and join the high value club by leveraging their relationships with foreign high value end firms. They argue that one advantage of software as a sector is its absence of backward linkages, making it possible to maintain the software industry as ‘an island of competitiveness’ notwithstanding the rest of the economy. It means that once the sector has developed, it could contribute substantially to domestic development in the more general sense. Therefore, they conclude that the software sector offers a pathway to development for middle group countries such as India in which the preconditions exist in the form of substantial investment in education. In Chapter 6 Y. Ueki explores broadband diffusion and the e-commerce situation in Korea ‘Jumping up to the Internet-based society: lessons from Korea’. He describes the fact that the number of Internet users reached 24.4 million in December 2001. The number of high speed Internet subscribers was 7.8 million in December 2001, and among them 4.5 million were subscribing to ADSL and 2.7 million to cable modem. Moreover, it is said that there were more than 25 000 ‘PC-bangs’ (Internet cafés) in 2001 in Korea where the youth enjoy interactive games and Web surfing. He analyses the reasons why this rapid expansion takes place, identifying the following factors: (1) strong government support toward IT-based society (deregulation of the telecommunications sector, infrastructure building, guiding with national plans and projects, and human resource development in terms of Internet literacy); (2) collective housing, especially, in urban areas, coupled with an eagerness for education; (3) people’s homogeneous mentality and a sense of alienation in the case of differences. He also indicates negative side effects of the IT society such as high tech crimes and distortions in daily life and culture.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 7

Introduction

7

C. Mephokee explains the Thai case in ‘Information technology: some implications for Thailand’ in Chapter 7. He states that Thailand plays the role of the IT consumer as well as producer. The Internet penetration rate was around 6 per cent in Thailand in 2001. This low penetration rate comes from three factors: first, high prices due to monopoly of the state enterprise (Communication Authority of Thailand); second, the lack of qualified IT personnel; and third, the language barrier (English). On the other hand, Thailand plays a role in the production and export of IT-related products (HDDs, PCBs, ICs, keyboards, printers, and monitors). Computers and computer parts and ICs have become the two most important exporting items in Thailand, accounting for about 18 per cent of total exports. In order to reduce the ‘digital divide’ and create a literate IT society, he recommends the government liberalize the telecommunications market to allow more competition from new entrants, domestic and foreign. He points out that market liberalization has been included in the Telecommunications Master Plan 1997–2006; however, it has not shown any progress yet. He also recommends promoting human resource development throughout the country. This includes formal education from elementary schools to colleges and informal education for workers and people of all ages, under the slogan ‘IT for all’. Chapter 8, ‘Information policy and information technology in Central and Eastern Europe with emphasis on Estonia’, written by T. McDaniel, discusses various IT policies and programmes relating to Central and Eastern Europe (CEE). In most cases, data are provided on the following ten countries: Albania, Bulgaria, Czech Republic, Estonia, Latvia, Lithuania, Poland, Romania, and Slovenia. In these regions, IT-related changes are also quite rapid. Monopoly concessions in the telecommunications sector that have helped to keep the price of local phone service from falling are ending. The monopoly agreements granted to the dominant telecommunications operators in the Czech Republic, Estonia and Slovenia ended on 31 December 2000. In the remaining seven countries the agreements were scheduled to end in 2002 and 2003. Combined with the existing policies of liberalization in wireless technologies and in Internet service provision, liberalization of local voice telephony could help to provide more depth to the IT sector (for example by making existing connections faster). Achieving greater network range requires forward-looking policies involving both the public and private sectors. Many CEE countries have national policies for expanding and enhancing IT use; these policies often begin with the liberalization of national telecommunication companies and plans for bringing government itself into the information age. Policies for educational and rural communities have also been adopted, and the use of ‘telecentres’ has helped to make these national policies feasible. The author emphasizes Estonia because it has adopted an aggressive national programme for IT expansion, has the highest per capita Internet use in the region, and has an innovative financial

Kagami 01 chap 01

8

2/15/04

2:55 PM

Page 8

Introduction

sector with most banking transactions already occurring online. The use of ‘telecentres’ has helped to provide greater access to school networks and rural communities (programmes called ‘Village Road’ and ‘Tiger Leap’ in Estonia). In this context, Hungary has its own programme called Telecottages. She concludes that the needs of countries that have already embarked on extensive liberalization and IT development programmes (most of the countries in this chapter) are very different from countries that are not only developing their infrastructure but are also rebuilding their infrastructure after periods of sustained civil and international conflict. Albania, Bulgaria and the countries of South East Europe may find it more beneficial to altogether leapfrog conventional PSTN (Public Switched Telephone Networks) and find outside investment for increasing wireless and satellite technologies. A. Whinston and S.Y. Choi in ‘Internet and telecommunications outlook in Latin America’ in Chapter 9 focus on the current status of telecommunications markets, Internet access and e-business in Latin America, and investigate pricing, telecommunications policies and technological factors that affect current and future growth rates in Internet penetration. Most countries in Latin America during the 1990s opened up their telecommunications markets, privatized government-owned monopolies, and liberalized regulatory policies relying more on market forces than decrees. Such reforms coincided with the introduction of the Internet and the World Wide Web, offering them an opportunity to participate in the new information age as a full-fledged member. Still, Latin America as a region lags far behind North America, Europe and AsiaPacific in terms of Internet penetration. Latin America in general needs better telecommunications infrastructure such as traditional telephone networks, advanced telecommunications networks, Internet backbone, interconnection points and access points. They argue that in most of these measures, Latin America is ill equipped to sustain future growth. Another limitation in the region is the lack of interconnection between major network access points within the region. Most international traffic is routed to and from US network access points adding considerable costs to the users. Policies to expand access to telephone service and the Internet can be evaluated by three broad measures: availability, accessibility and affordability. In these three factors the region is left behind. It is, however, worth mentioning that Argentina is best poised for broadband expansion due to its high cable penetration rate (more than 50 per cent). Finally, the low level of income is a critical factor that overshadows any future development in Latin America. Simple privatization, which is often made ineffective by monopolized private companies, may not bring about substantial gains that enable the majority of their population to participate in the digital revolution. Latin American countries, thus, must focus on providing Internet access through schools, work places and other public access sites such as telecentres.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 9

Introduction

9

S. Choi analyses the cases of Mexico and Argentina in ‘Policies for Internet access: cases of Mexico and Argentina’ in Chapter 10. Both Argentina and Mexico have a 10 per cent or lower rate of Internet penetration while some Asian countries such as Hong Kong, Korea, Taiwan, and Singapore surpass the 50 per cent rate. According to Choi, this disparity is due to the inadequacy in basic telecommunications infrastructure, high costs for telephone/Internet access, limited network service options, and uneven income distribution. Nevertheless, several policy initiatives were put in place during the last ten years, mainly the privatization of state-owned telephone monopolies and the introduction of a competitive, open telecommunications policy framework. As a result, substantial progress has been reported in terms of telephone and network availability. Mexico represents in many ways a typical case of Latin American political, economic and regulatory environment where a large privatized company, Telmex, dominates the telecommunications industry. Argentina, offering a more pro-competitive market, has instituted several innovative strategies, such as a special area code for Internet dial-up access (dial 0610) with discount pricing and community technological centres. However, fears remain that past rapid growth in telecommunications services and the Internet may have affected mostly the unfulfilled demand by upper and middle income households, given the economic problems and income distribution in Latin America. There is a significant skepticism about a sustained penetration beyond these segments in the future. M. Kagami in ‘Tipping, standardization and convergence: catch-up and failure in Japan’s standards strategy’ in Chapter 11, Part II highlights the importance of standardization. Technical standards have principally three categories: de facto standards; forum standards; and de jure standards. De facto standards are the result of competition. Markets decide the popularity of products. A slight change in the function or character of a product’s features may produce a change in consumer response to the said product and result in falling sales and a loss of market dominance. This is called ‘tipping’. A famous example is the VHS vs. Betamax rivalry for the home videotape market. Tipping is, in consequence, related to de facto standards. Group efforts of manufacturers bring about forum standards which can avoid losses and risks which come from severe competition through a de facto standards war. De jure standards come from public or international organizations to regulate and harmonize technical specifications from the point of view of public and/or international concerns. According to the author, Japan has not taken any initiative in the discussion of international de jure standards and, as a result, failed to set up some key standards such as the ISO 9000 series and ISO 14 000 series although Japan was recognized as a world leader in the above-mentioned topics relating to quality control and environmental management. He further explains the recent technological advancement on convergence of digital TVs and PCs where

Kagami 01 chap 01

10

2/15/04

2:55 PM

Page 10

Introduction

standards of digitization are integrated in both technologies. Kagami claims that since Japan’s technical levels and management skills are satisfactory as compared with those of other advanced countries, Japan should contribute more to international standards formations. In Chapter 12, K. Shiraishi analyses the Japanese press in the age of the IT revolution in ‘Is the Japanese press a dinosaur in the 21st century?: the IT revolution and newspapers in Japan’, focusing on one of the leading national dailies, The Yomiuri, which the Guinness Book of World Records lists as the world’s biggest commercial newspaper with 10.2 million printed copies every morning. In November 2000, The Yomiuri launched a new department called the News Distribution Centre to efficiently disseminate breaking news to the Internet, satellite broadcasting and cell phones. Japanese reporters and editors have begun to change their newspaper-deadlines-first policy and become more cooperative toward news distribution services for electronic and electric wave media. An opinion widely shared by Japanese newspaper publishers and editors is that the Internet would deliver a direct blow to advertising and editing, the lifeline of newspapers, and that if proper measures were not taken, newspapers would be driven out of the market within a few years. According to Shiraishi, however, there is another opinion: ‘There are many people who believe that newspapers are dinosaurs. We believe exactly the opposite. We believe that newspapers can in fact evolve into a new form of media that blends the old familiar aspects of newspapers with the new technologies that are emerging. So that you have the ability to read, browse and scan, and at the same time being able to interact with the newspapers, to interact with advertisers through your newspapers in ways that are not possible through media today.’ He concludes that like it or not, the Japanese press is moving in this direction. The transaction basis has been shifting from face-to-face commerce to commerce via the Internet. How to ensure mutual trust between those involved in such transactions has become an important issue. One measure to cope with this is system solution enterprises aiming to offer a trusted communication technology platform. Some of the systems established so far are a Digital Signature (DS) as a means of verifying the true identity of another party, and a Certificate Authority (CA) to verify registration of the communication partner. These are both based on technologies of encrypted data transfer by the public key method. These are called Public Key Infrastructure (PKI). A. Maeda in ‘PKI solutions for trusted e-commerce: survey of the de facto standard competition in PKI industries’ in Chapter 13 analyses competition and strategies of major PKI vendors. There are four major companies offering authentication and certificate technologies: RSA Security Inc. (US), VeriSign Inc. (US), Entrust Inc. (Canada), and Baltimore Technologies plc. (Ireland). These four companies are fiercely competing with each other to establish de facto standards. Maeda explains their strategies for making inroads into Asian markets, except

Kagami 01 chap 01

2/15/04

2:55 PM

Page 11

Introduction

11

China which tries to avoid Western standards as much as possible. He worries that these four will take over global electronic certification markets and argues that much wider perspectives are necessary for the public sector with respect to a trust system. The trust framework of each society is determined in accordance with the status of each society respectively. It is not a simple matter of business tools, but of complicated political as well as security issues that face society as a whole. Chapter 14, titled ‘The interconnection and pricing of the Internet’ by T. Ida and M. Ueda, provides a model framework to analyse the industrial structure of the Internet. They emphasize pricing of the Internet analysing three main elements of network costs: (1) the cost of connecting to the network; (2) the cost of providing additional network capacity; and (3) the social cost of congestion. They especially focus on the vertical features and interconnection among end-users, ISPs and IBPs and build a ‘components model’ in order to consider the one-way model and the two-way model respectively. The oneway model of connection is the network structure where one company needs access to another but the reverse does not hold. One example is that some ISPs provide access services to end-users in a retail market and at the same time serve as IBPs that provide transit services to ISPs in a wholesale market, while other ISPs that do not serve as IBPs have to buy transit services from IBPs to provide final services to customers. The two-way model of connection demonstrates that customers calling each other belong to two different local networks and each carrier must buy termination access from the other network. One example is that IBPs have to interconnect at a point of interconnection so that end-users on each network can exchange information. They set up mathematical models and compare the social welfare between the one-way and two-way systems. This is only a first step trial and actual applications for policy-making in the network economy remain. Lastly, the ‘Conclusion’ (Chapter 15) by the three editors follows, emphasizing three negative aspects and two positive aspects in the rapidly changing IT revolution, especially from the developing countries’ point of view. The former include: (1) digital divide and universal service; (2) monopoly and hegemony; and (3) demand consideration, while the latter contain: (4) leapfrogging industrialization; and (5) broadband expansion. This joint research project was organized and financially supported by the Institute of Developing Economies (IDE), Japan External Trade Organization (JETRO), Tokyo in 2001. Three teams from IDE (Japan), the University of Cambridge (UK) and the University of Texas at Austin (US) were set up to undertake this project. An international workshop was held for the project at IDE, Chiba on 4–5 December 2001. We are grateful to all participants for their stimulating comments and contributions on the occasion. Special thanks go to each team leader: Professors Masatsugu Tsuji, Emanuele Giovannetti and

Kagami 01 chap 01

2/15/04

12

2:55 PM

Page 12

Introduction

Andrew Whinston. Our thanks also go to Mr John Gallagher for his English proofreading. Any opinions expressed in this volume are those of the authors and not of the organizations they are affiliated to.

NOTE 1. TRON stands for The Real-time Operating system Nucleus, which was invented by Professor Ken Sakamura, University of Tokyo as open source OS languages (see, for example, Mitsuhiro Kagami (2003) ‘IT Revolution and its Meaning to Society’, in E. Giovannetti, M. Kagami and M. Tsuji (eds), The Internet Revolution: A Global Perspective, Cambridge University Press).

Kagami 01 chap 01

2/15/04

2:55 PM

Page 13

PART I

Country/area studies

Kagami 01 chap 01

2/15/04

2:55 PM

Page 14

Kagami 01 chap 01

2/15/04

2:55 PM

Page 15

2. Beyond the IT revolution: the Japanese broadband strategy Masatsugu Tsuji 1.

INTRODUCTION

The remarkable recovery achieved in the early 1990s and the resulting prosperity of the US economy are founded on the development of information technology (IT). IT has brought the US economy to an entirely new stage, which is referred to as the ‘New Economy’, where not only have inventory cycles such as short-run economic fluctuations disappeared, but growth without inflation has also become possible by increased productivity. In early 2000, the stock prices of Net businesses or ‘dot.coms’ started to fall, and this was the beginning of the end of the Net bubble. Currently, approximately one-tenth of Net businesses in the US have met with bankruptcy. In Japan, on the other hand, the year 2000 marked the beginning of the Japanese IT revolution initiated by then Prime Minister Mori. The Japanese IT revolution was vital to the Japanese economy not only for it to catch up with other IT advanced economies including other East Asian economies such as South Korea, Singapore, and Hong Kong, but also to promote growth during the long period of economic stagnation since the bursting of the bubble economy in 1991. Public funds had been invested so far in various IT projects and the construction of IT infrastructure. Japan today, however, due to the effect of the bursting of the Net bubble in the US, has been suffering a recession in IT-related industries. Household appliances companies and PC manufacturers have already laid off several tens of thousands of employees and a large number of Net businesses such as e-marketplaces, on-line security companies, and Net-banks are in danger of going out of business. Then has the IT revolution come to an end? Definitely not. A second phase sooner or later will arrive, and when it does, broadband will be the key factor for driving the revolution, since broadband will be able to transmit a huge volume of data such as motion pictures at ultra high speed.1 One reason for the dot.coms shakeout is the technological shortcomings of the Internet, which will be discussed later. The current stage appears to be a preparatory one for this 15

Kagami 01 chap 01

2/15/04

2:55 PM

16

Page 16

Country/area studies

second revolution, and telecommunications infrastructures such as optical fibres, wireless access, and DSL (digital subscribers line) compete for subscribers, and great efforts have been made to develop new technology, contents, and business models which will make broadband a reality. Competition among these infrastructures has led to a decrease in user charges, and this, in turn, has increased the number of users. This chapter focuses on the characteristics and issues of the Japanese broadband strategy, and analyses how the second IT revolution in Japan can be promoted. In the next section, we analyse why the bubble burst, and the kind of lessons to be learned from this. Section Three provides the current situation of broadband infrastructures such as FTTH (fibre to the home), DSL, CATV, and FWA (fixed wireless access), and discusses network (infrastructure) competition among them. This section also presents the National Broadband Network Initiative designed to make Japan the most advanced nation in terms of broadband. Future applications of broadband such as medical care and education will be presented in Section Four. Issues to promote the second IT revolution in Japan will be discussed in the conclusion.

2.

LESSONS LEARNED FROM THE NET BUBBLE

Here we present the current situation of the IT recession, and discuss the causes of the Net bubble. In order to prevent this in the coming broadband age, the kind of lessons to be learned from the bubble will be presented. 2.1

The IT Recession

In the 1990s, the US economy was driven by the rapid diffusion of IT technology in such a way that there was an increase in the capability of IT equipment and this caused its prices to be lowered. In addition, IT created new demand for telecommunications and information hardware such as PCs, equipment, and new business models. The growth of the information and communications industry can explain the 30 per cent economic growth in the US since 1995.2 This process can be called a ‘revolution’ of the same significance as that of the 18th century. Firms promoting this revolution were mainly venture businesses related to e-commerce, ISPs (Internet service providers), and contents. In 2000, however, the demand for Net businesses was expected to decrease, and this resulted in a pessimistic outlook in their future stock prices. This led to a retreat in funds investing in IT-related venture businesses, and this again caused lower stock prices, resulting in a vicious cycle. The exact causes will be discussed later, but the IT recession was triggered by these financial phenomena.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 17

Beyond the IT revolution: the Japanese broadband strategy

17

This can be clearly described by the trend of stock prices, that is, the Bloomberg US Internet Index, which focuses on stocks of Net businesses, started to head lower in March 2000, and in October 2001 it dropped by 87 per cent, which was lower than that prior to the bubble. NASDAQ stock prices also showed a similar trend, whereas that of the New York Stock Exchange, on which traditional big companies are listed, remained at the same level during this period. This implies that the IT recession was initiated by financial phenomena. The aftermath of the bursting of the Net bubble in the US can be seen in the fact that in 2000, 225 and in 2001, 537, altogether 762 Net businesses and dot.coms were shut down or declared bankrupt within the period of two years. It is said that there are 7000 to 100 000 Net businesses in the US, thus nearly ten per cent went bankrupt. Among them, 226 were related to access, 284 content, 207 e-commerce, 130 infrastructure, and 46 professional services.3 Since 2001, IT-related manufacturing has also been affected, as is evident from the shipment of PCs, for instance, which showed negative growth, and the layoffs which began in the IT industry. Telecommunications carriers, ISPs, and DSL companies have also been suffering from lower demand and resulting lower profits. 2.2

The Causes for the Bursting of the Net Bubble

The bursting of the Net bubble can be summarized as follows: (a) Overestimation of scale and pace. Since IT was an entirely new technology, it created an overestimation of the scale and speed of its diffusion. This overestimation of market participants, including investors and venture capitalists, created the bubble. IT must be supported by its consumers, but sooner or later the final demand reaches a saturation point. Let us take ecommerce as an example, which is one of the typical applications of IT and the Internet. Amazon.com, e-toy, and e-Bay are the most popular dot.coms in this area, and they attracted a large number of customers. In the US, more than half of Internet users purchase goods from such companies. Approximately 8.9 per cent of the users shop every week, and heavy users totaling 4.5 per cent account for nearly one-third of the total amount purchased (UCLA Internet Report). This implies that not all consumers are enthusiastic about Net shopping. (b) Robustness of existing systems. All systems have their reasons and strong basis for their existence, and it is impossible to replace them overnight. E-commerce, for instance, is supposed to offer goods at a much cheaper price than established suppliers by getting rid of existing distribution channels. E-businesses, however, could not bypass them all. Moreover,

Kagami 01 chap 01

18

2/15/04

2:55 PM

Page 18

Country/area studies

they had to invest much more in the process of bypassing such channels, that is, they had to build their own distribution facilities. Amazon.com spent much in this regard, and thus its profit has always been negative in spite of its high stock price. Another example is the so-called ‘combini banks’, which have ATMs (automatic transactions machines) installed at convenience stores and offer a 24-hour cashing service. They were mistaken in their projection of demand and they have not been attracting as many customers as expected. The reasons are simple; there are already enough ATMs everywhere that do not have the extra service charges seen when cashing at convenience stores. In addition, combini banks have had to establish their own networks and supporting facilities such as back offices. Again, these are costly. (c) Immature technology. IT is new but its technological level is also far lower than what would meet customers’ satisfaction. Internet telephony as well as video on demand via the Internet is not at a satisfactory level. Another example is e-commerce. One Japanese survey presents the problem of shopping on the Internet. Among 3641 samples, 78.5 per cent of the responses were ‘Can’t examine product by hand,’ and 49.8 per cent were ‘Uncertainty about problems with payment’. 4 It is obvious that IT cannot display the real goods, but it is certain that current IT fails to provide images with the same effect as face-to-face communication. In sum, IT creates the demand for final goods, but it does it in a zero-sum way. That is, some of the final users change to e-commerce use, but this decreases the same demand in traditional markets. In order to gain a net increase in final demand, IT needs further development. In what follows, we will show that this is broadband technology.

3.

BROADBAND NETWORK INFRASTRUCTURE

As stated above, it is broadband that will initiate the second IT revolution, or make a breakthrough in the current IT recession. In this section, we will present the broadband initiative which is designated to promote the Japanese IT revolution. 3.1.

Current State of Broadband in Japan

In order to discuss the broadband strategy, let us present the current situation of Japanese broadband. Japan lagged behind in Internet penetration, but due to the great effort made recently in promoting broadband and competition among carriers as well as ISPs, not only has the number of subscribers shown a

Kagami 01 chap 01

2/15/04

2:55 PM

Page 19

Beyond the IT revolution: the Japanese broadband strategy

19

remarkable increase, but user charges have also decreased. Current broadband infrastructures are summarized in the following four accesses as being most important: (a) FTTH; (b) DSL; (c) CATV; and (d) FWA. All of these have the characteristics of flat rate charges and connecting to the Internet for 24 hours a day as well as high speed. In what follows, we examine these four types of access in more detail. 3.1.1 FTTH FTTH aims to deploy optical fibres to each home, and its speed is the fastest among broadband alternatives such as 10Mbps to 100Mbps. Among the many alternatives of broadband infrastructures in Japan, telecommunications carriers as well as the public sector tend to focus on deploying optical fibres, whereas other economies give more attention to DSL and CATV (cable modems), which are much cheaper than optical fibres. The current total number of business firms and individual subscribers of optical fibres in Japan is approximately 230 000 (as of September 2001), which is the second highest to the US’s 376 000 (as of December 2000). As a result, broadband user charges have been decreasing due to competition among carriers. In early 2001, a venture business called Yusen Broadnetworks began service and in August 2003, NTT started FTTH service. As examined in detail by Tsuji (2001), the density of the Japanese fibre optics network is at the highest level, namely, 95 per cent in business, and 30 per cent in residential areas, and the national average is 35 per cent. Current subscribers are, however, limited mainly to businesses and SOHO (Small Office Home Office) in the metropolitan areas, but monthly charges have become low enough to be affordable to most households; namely, NTT locals offer FTTH called ‘B FLET’S’ from US$30 to US$70 depending on the speed and type of house such as a condominium or single house.5 Since the deployment of the optical fibre network is one of the key elements of the Japanese broadband initiative, it has been promoted and subsidized by the government.6 3.1.2 DSL DSL is technology which uses existing telecommunications copper lines for Internet access. NTT originally had adhered to ISDN due to its business strategy as well as the technological difficulty of DSL.7 After abolishing this policy and converting its facilities so that they are suitable for DSL, NTT locals now cover 85 per cent of city areas for DSL service. Thus, giant NTT has been promoting its sale, and as a result competition among DSL operators has become intense, especially after Yahoo BB, an affiliated company of the largest search engine Yahoo Japan, entered this market with low charges such as US$20 for 8Mbps. NTT locals have also lowered their charges (see Figure 2.1). The number of subscribers has seen a remarkable increase (see Table 2.1).

Kagami 01 chap 01

2/15/04

2:55 PM

20

Page 20

Country/area studies

Rate (Yen)

Wireless service

7000

Wire line service CATV 10Mbps ¥6,000

6000 5000 FLET’S ISDN 64Kbps: ¥2,800+ Charges for ISP (¥1,000–2,000)

4000

PHSInternet 64Kbps ¥3,000

3000

FLET’S ADSL 1.5Mbps: ¥2,900+ Charges for ISP (¥1,000–2,000) NTT–ME(condominium) 1Mbps: ¥3,500 So-net(ADSL) 1.5M ¥3,167 8M ¥3,467

64K Source:

B-FLET’S family (10M) ¥5,000 Condo (100M) ¥3,500 + Charges for ISP ¥3,800–4,800

Yusen BB 100M ¥6,100 FTTH (future)

2.4GHz Wireless 3Mbps ¥3,800 Internet Condominium 10Mbps Yahoo(ADSL) ¥3,000–¥4,000 8M ¥3,017*

1M

10M

Speed

Author’s review of carriers and providers

Figure 2.1 Table 2.1

Subscribers

Broadband charges in Japan Number of DSL subscribers in Japan Jan. 2000

June

December

June 2001

November

19

1235

9727

291 333

1 204 564

Source: Ministry of Public Management, Home Affairs, Post and Telecommunications (MPHPT) (2001).

The areas covered by DSL and the number of DSL subscribers are increasing, but it still faces the following issues: (a) distance from telephone station. Users must be situated within 3km of the station. In rural areas, since the telephone station network is relatively small in number, it cannot cover the entire area;8 (b) copper lines. DSL requires copper wire lines, and newly built condominiums and large apartment complexes in metropolitan areas are connected to telephone stations via optical fibres, thus DSL is not available; and (c) speed. DSL is of the best effort type, and its speed depends on many factors such as type of copper wire line, noise from other electrical equipment, and PCs. Thus, DSL is convenient but requires further technological development. 3.1.3 CATV CATV was originally used for broadcasting, and its network can be utilized for Internet access. It is independent from NTT local loops. There are currently

Kagami 01 chap 01

2/15/04

2:55 PM

Page 21

Beyond the IT revolution: the Japanese broadband strategy

21

243 CATV companies that offer Internet connection services. CATV utilizes coax in general, not optical fibres, so its speed is at maximum 8Mbps, and its user charges are about US$40. Thus, cable modems are rather more expensive than DSL, so that the growth rate is smaller than DSL. The number of cable modems is shown in Table 2.2. The Japanese figures are internationally low, compared with those of the US and South Korea; the former has about four million, and the latter 2.3 million. This is due to the fact that the size of Japanese CATV companies is small, and they cannot afford to invest in new access service. Recent deregulation has been promoting the merger of CATV companies to strengthen their financial basis. Table 2.2

Number of cable modems in Japan (Unit: 1000) Mar. 2000 June

Modems Source:

216

329

Sept. Dec. Mar. 2001 June 463

625

784

967

Sept. 1151

MPHPT (2001).

3.1.4 FWA FWA does not require subscribers’ wire lines such as fibre optics or metallic cables but replaces them with wireless local loops.9 The speed of access is about 3Mbps, and users’ charges are about US$30, including charges for the provider. Access via FWA does not depend on the distance from the operational base, unlike DSL. So far, three major companies have started this service, and the number of subscribers is not large. One of the most important characteristics is that FWA does not require subscribers’ lines so it is rather inexpensive to deploy, and its use can be expected to be suitable for rural areas. 3.1.5 User charges A marked increase in broadband subscribers is mainly due to the decrease in user charges. Figure 2.1 indicates the charges of the different broadband alternatives. At present, there is no difference between ISDN and DSL. The low rates of DSL are symbolized by Yahoo BB, whose monthly charges are about US$20.10 This is too low to cover costs at the current number of subscribers, but companies like Yahoo BB lower charges to attract enough subscribers in order to make a positive profit. In addition, Figure 2.2 indicates that charges for FTTH and DSL have been lowered during the last two years due to increased competition and technological development. This figure also shows that Japanese charges are no longer very expensive compared to those of other countries, especially for DSL which

Kagami 01 chap 01

2/15/04

2:55 PM

22

Page 22

Country/area studies

US$

FTTH

140

Internet access charges Telecommunications charges

142

120 100 DSL

80 60 40

40

20

56

55

50

46 35

48 31

20

0 Dec. Aug. Aug. Feb. Oct. Oct. Sept. Oct. Oct. Oct. 2000 2001 2001 2000 2001 2001 2001 2001 2001 2001 10Mbps 10Mbps 100Mbps 1.5Mbps 1.5Mbps 8Mbps 788Kbps 1.5Mbps 500Kbps 500Kbps NTT NTT YUSEN NTT NTT YAHOO US US UK France EAST EAST EAST EAST BB

Source:

MPHPT (2001).

Figure 2.2

International comparison of broadband charges

are lower than those of the US, where charges have tended to rise because of increased bankruptcy and the resulting decrease in competition in this industry. 3.2

National Broadband Network Initiative

3.2.1 Estimation of future broadband access Thus, the four basic broadband infrastructures discussed have both merits as well as demerits, namely, FTTH transmits a huge volume of data at high speed, but it is costly, whereas DSL, CATV, and FWA are of a slower speed, but rather economical. Regarding the costs of deployment of infrastructure, the first requires high costs, on the other hand, the latter three such as FWA are expected to be made available at low costs with the use of existing facilities. In order to establish infrastructures, we have to have a mixture of these four. As one such example, we present the e-Japan project which aims to provide broadband to all sectors of the economy. The e-Japan project, which was announced in March 2001, sets the following objectives so as to establish the most advanced IT economy in the world: (a) by the end of 2005, at least 30 million households will have 24-hour connection to high speed Internet access networks, and ten million to super high speed Internet access networks; (b) efforts will be made to prevent a digital divide

Kagami 01 chap 01

2/15/04

2:55 PM

Page 23

Beyond the IT revolution: the Japanese broadband strategy

23

due to geographical factors; and (c) by the end of 2005, regional public networks will be established that will be connected to all schools, libraries, public facilities, and hospitals. According to this plan, the government has set forth the ‘National Broadband Networks Initiative’, which identifies the abovementioned four accesses – FTTH, DSL, CATV, and FWA – as being the most important for achieving those goals.11 It also provides a schedule to realize broadband, and states the respective roles of government and the private sector, and the social benefits expected to result from broadband such as receiving high quality public services, including those related to medical care, welfare, education, and culture. The National Broadband Initiative has the following estimates as indicated in Figure 2.3, regarding the number of households which will be connected to the Internet with high speed access as mentioned above.

Number of households subscribed (10 thousand)

2500

2000

FWA CATV DSL FTTH

1500

1000

500

0 Jun. 2001 Dec. 2001 Dec. 2002 Dec. 2003 Dec. 2004 Dec. 2005 Source:

MPHPT (2001).

Figure 2.3

Estimation of broadband by the National Broadband Initiative

In more concrete terms, by March 2003, almost all metropolitan areas including the wards of Tokyo and prefectural capitals, and by March 2005, almost all other 623 cities were due to be covered by high speed access networks to the Internet, and at that time, the share of the four broadband infrastructures is expected to be FTTH 7 730 000, DSL 6 950 000, cable modem 4 290 000, and FWA 800 0000. The estimation of these figures is obtained by regression analysis on the basis of users’ past trends.

Kagami 01 chap 01

24

2/15/04

2:55 PM

Page 24

Country/area studies

The basic framework of this deployment deploys the same idea as the US’s ‘National Information Initiative’ (NII), that is, it is achieved through competition in the private sector. This however creates the issue of ‘market failure’. Private firms cannot engage in business in unprofitable regions, thus a digital divide may be created. Particularly, depopulated and isolated areas characterized by geographical constraints will find it difficult to attract private firms to participate in the deployment of network infrastructure. In such cases, the public sector is required to take that role. The deployment of public networks such as public LANs will interconnect all public facilities and institutions in these areas for services related to education, administration, welfare, medical care, and disaster prevention.12 Larger scale LANs for a region are also referred to as a WAN (wide area network), which interconnects all agents in the region, whether public or private. Without public assistance, the above National Broadband Initiative cannot be achieved. 3.2.2 Issues of the National Broadband Initiative One of the key factors for realizing the National Broadband Initiative is whether FTTH, CATV, DSL, and FWA will be given sufficient priority for realizing the project. DSL has issues regarding its availability. It is well known that DSL cannot be used for households which are located about two miles away from local NTT offices.13 DSL is also available only for copper cables, not for ISDN or FTTH. It is sometimes affected by noise emitted from electric ovens or other electrical appliances. Another issue related to FTTH, CATV, and DSL is the installation of networks in apartment complexes and condominiums. Although those recently built are already equipped to connect to the Internet, those without this capacity are burdened with additional installation costs. For this reason, there are cases where residents are unable to reach a consensus regarding installation. Digital divide is not a concept applicable only to depopulated areas, it occurs in metropolitan areas as well. New legislation or subsidy schemes are required to cope with this issue. Another issue is that the National Broadband Initiative provides only for the deployment of infrastructures, and is less related to applications. Without being utilized by the final users, broadband infrastructures have no meaning. FTTH has a maximum speed of 100Mbps, but there are only a few software products and contents so far which can fully utilize such speed. Possible broadband applications such as contents, software, and business models will be discussed later. 3.3

Recent Broadband Development: From Last One Mile to Last Quarter Mile

Here we will examine the recent development of four broadband infrastructures in terms of competition. FTTH, CATV, and FWA have their own

Kagami 01 chap 01

2/15/04

2:55 PM

Page 25

Beyond the IT revolution: the Japanese broadband strategy

25

networks, and they are independent from NTT’s local subscribers’ line. Thus, the situation is referred to as network competition. Through competition of the four networks, broadband infrastructures have become closer to their users, so the issue of the ‘last one mile’ now becomes that of the ‘last quarter mile’. 3.3.1 NTT (FTTH) In 2000, NTT locals in Tokyo started high speed Internet access service through the optical IP communication network, which was built specially for Internet connection service. The Internet is of the best effort type, and its quality is not guaranteed 100 per cent, but it is designated for high quality service at 100 per cent connection with fixed user charges. This service is called ‘B FLET’S’. Since then, this service has been expanded to other large cities such as Osaka, Nagoya, and Kobe. The speed is from 10Mbps to a maximum of 100Mbps, depending on the region. This service makes it possible to realize the FTTH dream, that is, each house is connected directly to an optical fibre network. It costs about US$300 per month for business users, US$75 for the basic type, US$40 for the family type, and US$30 for the condominium type.14 These prices are almost the same as those of DSL. Although this network still has technological and content problems, Japanese Internet usage leads many other economies. NTT locals are also making the effort to promote B FLET’S in such a way that if there are at least thirty subscribers in an area, NTT will expand optical fibres from their terminals (nodes). This is a big shift from their past marketing strategy which deemed such service to be appropriate only if a sufficient amount of users, much larger than 30, could be identified around the NTT local office. 3.3.2 Yusen Broadnetworks (FTTH) Yusen Broadnetworks is another firm which provides FTTH with a maximum 100Mbps (best effort), and it started service in Tokyo in March 2001.15 At the end of 2001, it was due to begin providing services in other metropolitan areas. Yusen Broadnetworks has a unique marketing strategy for providing service, that is, it concentrates only on populated areas such as those containing more than 3000–4000 households within a radius of 2km. Yusen is able to provide services at rather low cost, since it makes use of existing telephone poles and power lines. It also is not required to provide universal service, since it is a private firm; that is, it provides service only to profitable areas. 3.3.3 K-Opticom (FWA) Electric companies own power plants, distribution networks, and other facilities, and the aim of the networks is to control and operate their systems related to the generation, transmission, and distribution of electric power, and securing a stable supply. Since these networks have a sufficient level of quality and reli-

Kagami 01 chap 01

26

2/15/04

2:55 PM

Page 26

Country/area studies

ability, they can be utilized as the public switched network for telecommunications. The characteristic of this network is that it can play the role of subscriber lines, and this is why electric companies have started providing telecommunications service.16 K-Opticom is an affiliated company of Kansai Electric Company, which is a regional monopoly in the Kansai area. In June 2001, K-Opticom started a new service of high speed Internet access through their electric wire lines. Its service is based on the optical fibre networks of Kansai Electric Company, which currently owns 10 000km of backbone networks and 6000km of access networks. It was due to expand the network to 36 000km by March 2002. KOpticom has a very simple IP network consisting of one Network Operation Centre, eight regional centres, and 37 local centres. All are connected by double links for security. Business firms can use this for IP-VPN (Virtual Private Network), and utilize high speed communication at low cost. The Internet access service for individual subscribers is called ‘EO’ and its speed is from 64Kbps (EO 64 Air) to 10Mbps (EO Mega-fibre). From November 2001, it was due to start a new wireless service called ‘EO Mega-air’, which is 2.4 GH.17 Service charges are US$25 for EO 64 Air, US$24–US$32 for EO Megafibre, and $36 for EO Mega-air. There is no charge for IPS, which is a characteristic of K-Opticom. 3.3.4 iTS Communications (CATV) CATV companies are also strengthening competitiveness by increasing access speed to the Internet, but their speed is limited to about maximum 8Mbps. iTS Communications located in Yokohama, formerly Tokyu Cable Television, has been providing various broadband services since 1998. It will soon start high speed Internet access services such as 30Mbps, and this is about four times faster than DSL. This high speed is realized by adopting the optical-coax hybrid network such that optical fibres are used for the trunk lines and coax for access lines. This enables it to distinguish its services from DSL, which is its competitor. DSL’s shortcoming is that the quality level of images is not very high, particularly for HDTV (high definition TV), which requires at least 20Mbps.

4.

POSSIBLE APPLICATIONS OF BROADBAND: CASE STUDIES

One of the most important issues of broadband is application. Entertainment such as games, Karaoke, movies, animation, and e-commerce like shopping are the main targets. Sony Pictures, for instance, plans to distribute movies through the Internet, and users can enjoy movies at home by downloading them.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 27

Beyond the IT revolution: the Japanese broadband strategy

27

Here, we present possible applications of broadband in the areas of medicine and education as case studies. The cases discussed here pertain to the application of narrowband, not broadband, but the projects have been successful. Some of them are considered rare examples in the world. If they are replaced by broadband, much more fruitful results can be expected. 4.1

Telecare18

4.1.1 Definition of telecare Telecare is one category of telemedicine and implies the use of electronic signals to transmit medical information on patients living in remote areas. This is a real-time and two-way interactive transmission of information of large capacity such as images and data. Telecare differs from telemedicine in the sense that the persons who transmit and receive medical information are not medical doctors, but the patients themselves and their families, nurses, carers, home helpers, medical technical experts, and so on. Consequently, under the current level of technology, telecare cannot provide advanced medical treatment and services, and focus is on primary care and mental health care such as diagnosis of the patient at home by examining the images on PC or TV screens and by observing health data transmitted by the system. The telecare system in effect today in Japan can be broadly categorized into three groups in terms of aim, nature of medical information, equipment, and type of network as follows: (a) tele-home-care; (b) tele-health; and (c) community health and welfare management type. In what follows, let us examine the first two as the future application of broadband. 4.1.2 Tele-home-care system This system aims at providing telecare, for example, for bedridden patients and patients stricken with terminal diseases who require medical care. The characteristics of this system are the real-time and two-way interactive transmission of motion pictures via videoconference systems or videophones. This system is classified into three subcategories according to network type: (a) CATVBroadcast; (b) CATV-LAN; and (c) ISDN. The CATV network is utilized by (a) and (b) which can transmit a high definition motion picture of 30 cells per second using a color digital (CCD) camera with 360 000 elements. As far as the system is concerned, (a) uses the broadcasting network, and (b) the LAN network. As for working examples of (a), the systems of Goshiki Town in Hyogo Prefecture and Kamaishi City in Iwate Prefecture are two in operation. As for (b), there is the care-at-home support system called ‘Anshin-netto’ in Minami-Shinano Village in Nagano Prefecture which is the only working example of the LAN type.

Kagami 01 chap 01

28

2/15/04

2:55 PM

Page 28

Country/area studies

The ISDN type (c) of tele-home-care utilizes ISDN 64kbps as its network and image information is transmitted by the videophone system. As for quality of image on videophone screens, the motion picture is 10 to 25 cells per second, and it is inferior to that of the CATV type. On the other hand, in the case of ISDN, correspondence such as the exchange of messages among patients can be easily accomplished, and families of the patients exchange information through ISDN. This type of system is in operation in 20 regions such as Bekkai Town in Hokkaido, Mogami Town in Yamagata Prefecture, and Mitoyo Region in Kagawa Prefecture. 4.1.3 Tele-health type system The tele-health type and community health and welfare management type systems differ from the previous tele-home-care type in that both do not use image information. The aim of the health system is not to treat patients’ illnesses but to regularly observe the health condition of elderly residents or patients, for instance, after they have been released from hospital. The system consists of the following devices. First, at the patient’s home, a camera, PC, and remote monitoring, which is also called a remote sensor, are installed, which measure temperature, blood pressure, pulse, heartbeat, electrocardiogram, and amount of blood oxygen as part of patient observation. The medical information obtained through these instruments is then transmitted to medical institutions such as the local health centre via the telecommunications network which includes public telephone lines, leased circuits, ISDN, and the CATV network. This system is a simple device, but when it is used continuously, the condition of the illness such as a chronic disease is shown in graphs, which are then used for diagnosis and consultation. The system is also effective for encouraging patients to take more interest in their health condition. Some of the terminals are equipped with a simple voice function, and the doctor can examine the condition of the patient’s health by talking with the patient. There are 76 local governments across Japan which operate the tele-health system including Tadami Town, and Nishi-Aizu Town in Fukushima Prefecture, and Manmoku Village in Gunma Prefecture. The total number of devices amount to more than 8100 as of August 2000. 4.1.4 Effectiveness of telecare We conducted a field survey on the users of a tele-health system in Nanmoku Village, Kamaishi City, and Katsura Village, and obtained the following results regarding the effects of tele-health systems: (a) stabilizes the condition of diseases; (b) raises health consciousness; (c) decreases anxiety about health; and (d) decreases medical expenditures. Health data sent to the medical institution are simple, but basic. By examining the medical data each day, medical

Kagami 01 chap 01

2/15/04

2:55 PM

Page 29

Beyond the IT revolution: the Japanese broadband strategy

29

staff are able to recognize changes in health condition, and give advice to the users. By reading their data records, users begin to want to improve the data. Thus, they pay more attention to their own health. Users can communicate with medical staff via the system, and as they realize they are connected to the medical staff 24 hours a day, this decreases their feelings of anxiety. According to replies to our questionnaires, about 20 per cent of the users claim their medical expenditures declined after their use of the tele-health system. This is a rather surprising result, and we have to prove this hypothesis by other methods of research. Regarding assessment by users in our survey, in three regions more than 90 per cent of users recognize the system to be useful, and they want to continue its use. More than two-thirds of the users were satisfied with the function of the device and replied that there was no need for improvement. Most of the users are elderly people and they replied that the devices are easy to learn how to operate. In sum, we can conclude that the three systems surveyed are supported by their users. 4.2

Distance Learning: Setouchi Town, Kagoshima Prefecture19

4.2.1 Objective of distance learning In 1995, the Ministry of Education started to apply telecommunications and multimedia to further develop education in rural areas. The aim of this project is to interconnect schools, from elementary to high school, in isolated islands or mountainous regions with urban areas, via ISDN and satellites. In addition, another project for hospitalized school-age children has been started to interconnect the hospitals and the schools they are registered at. These projects interconnect two to three schools via optical fibres with the speed of transmission of either 64Kbps or 1.5Mbps. The schools share the same lectures and participating schools are the principal and branch schools, or schools in the same prefecture. There are currently 22 schools in ten prefectures involved in these projects. As an interesting example among the author’s field research, let us introduce the project at Setouchi Town, Kagoshima Prefecture. Setouchi Town is located at the southern tip of Amami-Oshima Island, Kagoshima Prefecture, and consists of three other isolated small islands, namely, Kakeroma, Koishima, and Yoro Island. The population of the town is 12 017, and it has been showing a decreasing trend since there is no big industry in the town, except for fishery. 4.2.2 ISDN network The tele-education system of Setouchi Town interconnects three town elementary schools and the Education Research Centre of Kagoshima Prefecture by ISDN (64Kbps and 1.5Mbps). They are Koniya School on the main island,

Kagami 01 chap 01

30

2/15/04

2:55 PM

Page 30

Country/area studies

Hyo School on Kakeroma Island, Ikeji School in Uke Island, and Yoro School on Yoro Island. Each elementary school is equipped with a videoconference system, with 29- and 50-inch monitor TVs, an electric board, and a camera to show the textbook and educational materials on the teacher’s desk, which is operated from the teacher’s table. There are six other cameras at Koniya School to show the classroom to other schools. One camera shows the teacher, the other the entire classroom, and the other four, which are fixed in the classroom, show the students in the classroom. The screen of the 50-inch monitor TV can be divided into either four or nine frames according to the situation; for instance, to see the classrooms of the other schools at the same time. All schools are thus interconnected via fibre optics with the three other schools, and they share the same lecture together. In addition, the speed of transmission can be selected from 128Kbps, 384Kbps, and 1.5Mbps: for instance, for transmission of the classroom, a high speed with a high quality of images and voice can be selected, and for a videoconference of a teachers’ meeting, a low speed. 4.2.3 Effectiveness and issues of tele-education in Setouchi Town The systems are designed to be operated by either the teachers or pupils. The main aim of tele-education in this town is to unify the feeling of the teachers and pupils on the isolated islands. On the other hand, there are the following problems related to the system. When the 50-inch monitor TV is divided into nine frames, each frame is too small to see clearly; in other words, the figures of the pupils and teachers are too small. Also, when many schools participate, the speed of transmission becomes slow, and the voices lag behind the images. 4.3

International Distance Learning20

Here, an experimental international distance learning project via ISDN is presented as a case study, which is called the Human Resource Development Project for Vietnam. 4.3.1 The Project Three foundations – the Hoso Bunka Foundation, or Broadcasting Culture Foundation (HBF), the International Communication Foundation (ICF), and the Telecommunications Advancement Foundation (TAF) (all three referred to hereafter as HIT) have been engaged in carrying out joint projects for human resource development (HRD) in both the telecommunications and broadcasting fields. The International Distance Learning Project (hereinafter, the Project) was planned as one of the HRD projects, using a TV conference system via international ISDN between the Global Information and Telecommunication Institute of Waseda University and Training Centre No. 1 of the Posts and

Kagami 01 chap 01

2/15/04

2:55 PM

Page 31

Beyond the IT revolution: the Japanese broadband strategy

31

Telecommunications Institute of Technology (PTIT) of Vietnam Posts and Telecommunications. The Project was successfully implemented in 1998 with three series of lectures and terminated in February 2001. Courses related to multimedia and telecommunications were offered from Waseda University, and each session lasted 90 minutes, including interactive questions and answers. Lecture documents were distributed one week in advance and participants were expected to prepare for the lecture. 4.3.2 Network configuration of the Project The international ISDN link between Vietnam and Japan was established at the end of 1997 for the TV conference service. ISDN was considered very attractive from the viewpoint of cost and timesaving. Figure 2.4 shows the actual implemented network configuration of the Project. The ISDN circuit between Waseda University and KDD is connected through NTT’s network in Japan. It extends to Vietnam via KDD’s International ISDN. The circuit between Japan and Vietnam is routed through Asia Pacific Cable (APC), Asia Pacific Cable Network (APCN), and Thailand-Vietnam-Hong Kong Cable (TVH). These fibre optic cables are interconnected via Hong Kong. The visual signals transmitted from Waseda University were received by the PC-based TV conference system in PTIT Hanoi, and projected on an enlarged screen by a liquid crystal display (LCD) projector connected to the PC. Screen PC-based TV conference system

1.75m

PC-based TV conference system

1.75m

LCD* Projector PC/Personal computer PTIT (Posts and Telecommunications Institute of Technology) TC 1 (Training Centre 1/Hanoi) Lecture Room

International ISDN 128Kbps 2B + D

*LCD: Liquid Crystal Display

Figure 2.4

System of international distance learning

Waseda University/ Tokyo GITI (Global Information and Telecommunication Institute)

Kagami 01 chap 01

2/15/04

2:55 PM

32

Page 32

Country/area studies

4.3.3 Assessment of the Project The results were highly evaluated by recipient institutions in Vietnam, and considered more satisfactory than initially expected. The factors which led to such a conclusion may be analysed as follows: 1. Compared with a satellite circuit set up by VSAT, ISDN, if available, can much more simply and economically implement an international distance learning program between two points if lectures are transmitted at regular intervals such as one week or so. 2. The recent PC-based TV conference system enables high quality transmission and voice presentation together with clear video images of class scenes and lecture documents in a sufficient resolution to the receiving side. If lecture documents are prepared so as to fulfill their functions, quite satisfactory results will be achieved in international distance learning over ISDN circuits even at a low speed of 128Kbps. 3. With the combination of a projector with a high degree of brightness and PC, on-line distance learning can be easily realized in a class of 20 students or so. 4. The contents and subjects of lectures were negotiated and selected so as to interest recipient institutions and to contribute to resolving the regional information gap.

5.

CONCLUSION

Thus far, we discussed how broadband has been introduced to the Japanese economy, and how due to technological development and competition among Internet access companies, broadband charges have been decreasing. Therefore, it can be said that the recent increase in CATV and DSL subscribers has been triggered by a ‘technology-push’ and ‘cost-push’. Consumers, however, are not necessarily satisfied with the existing content. According to the Nikkei Survey conducted in April 2001 asking what will be the driving force of broadband contents, 45.5 per cent replied visuals (excluding movies), 38.8 per cent games, 31.8 per cent movies, 29.5 per cent education, 27.3 per cent music, and so on. At the current stage, none of them are being widely provided. Items purchased through the Internet are mostly conventional commodities such as books, clothing, computers, foods, music CDs, DVDs, video cassettes, accessories, and so forth.21 Without introducing the ‘killer contents’ mentioned earlier, competition between broadband and existing e-commerce will end up with a zero-sum game, that is, broadband will only replace e-commerce. Further development of broadband requires a ‘demand-pull’.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 33

Beyond the IT revolution: the Japanese broadband strategy

33

Measures to cope with the digital divide are also necessary for, as discussed in detail, it will occur not only in rural but also urban areas. According to MPHPT, more than one-third of the total 44 210 000 households in Japan live in apartment complexes or condominiums; particularly in the Tokyo and Osaka metropolitan areas, more than half live in such housing. Those newly built are already equipped with broadband; on the other hand, existing ones are not. They must bear the cost of conversion. Since there is no firm legal basis to introduce broadband to those buildings, suitable legal as well as financial schemes are necessary. Japan has the densest network of optical fibres, as pointed out by Tsuji (2003), and in order to fully utilize these networks, it plans to make it possible for everyone to be able to have equal access. Theoretically, this is ideal, however, if it is to be realized, who will construct the network? Even if Japan has the densest fibre optic networks, they are not sufficient. An incentive scheme is still required to deploy optical fibres at the current stage of broadband development.

NOTES 1. Broadband is generally defined as having a high speed of more than 200Kbps in the reports of various governments as well as international organizations, but in business it is much faster reaching speeds of more than 1Mbps. 2. For economic analyses as to how IT contributed to the growth and economic activities of the US, see Choi and Whinston (2001), and US Department of Commerce (2000), for example. 3. See http://www.webmergers.com 4. InfoCom (2001), p. 114. 5. In addition to monthly charges, US$250 to cover initial service fees are required. 6. In the fiscal 2002 budget, for example, funds are planned to subsidize half of the costs of local governments for deploying optical fibre networks around public facilities such as schools, town halls, and hospitals. 7. Regarding these issues, see Tsuji (2003), pp. 47–9. 8. In the case of NTT East, there are about 3000 telephone stations, and less than one-third have been converted to DSL. 9. The term ‘fixed’ implies that receiving terminals such as wireless antennas are fixed to houses, for instance. If terminals are mobile, then they are called mobile access. 10. It is estimated that in order for Yahoo BB to break even at this amount, it has to attract a couple of million subscribers. This is based on the theory of dynamic pricing. 11. For more details, see http://www.soumu.go.jp/joho_tsushin/eng/Release/Telecommunications/news/011016_1.htm 12. Japan has a tradition of deploying infrastructures by public funds in such a way that the central government subsidizes the local government. This situation is referred to as ‘public-public partnership’ by Tsuji (2002a). 13. The areas where DSL is available is indicated on the homepage of NTT locals, but it is said that nearly 50 per cent of households within an available region are unable to have access to the Internet due to one reason or another. 14. In addition to these costs, users have to bear the initial costs for installation which amount to about US$230. 15. Yusen Broadnetworks listed its stock on Nasdaq JAPAN in April 2001.

Kagami 01 chap 01

2/15/04

2:55 PM

34

Page 34

Country/area studies

16. Tsuji (2002a) discusses their telephony service in the local call market using TTNet as an example. 17. EO Mega-air is categorized as FWA, since it is utilized inside a house or office and access to the Internet is through wireless LAN. 18. This part is based on Tsuji (2002b). 19. This part is based on Tsuji et al. (1999). 20. This part is based on Tsuji et al. (2002). 21. Survey conducted by InfoCom Research; see InfoCom (2001), p. 111.

REFERENCES Choi, S.-Y. and A. Whinston (2001), ‘IT Revolution in the US: Current Situation and Problems’, in M. Kagami and M. Tsuji (eds), The IT Revolution and Developing Countries: Late-Comer Advantage?, Institute of Developing Economies/Japan External Trade Organization, pp. 270–92. InfoCom (2001), Information and Communications in Japan 2001, InfoCom Research, Inc., Tokyo. MPHPT (2001), National Broadband Initiative Towards the Most Advanced IT Nation in the World, October. Tsuji, M. (2002a), ‘Infrastructure-Building in the Japanese Telecommunications Sector: From Public-Public Partnership to Public-Private Partnership’, in S. Berg, M. Pollitt and M. Tsuji (eds), Private Initiatives in Infrastructure: Priorities, Incentives, and Performance, Edward Elgar. Tsuji, M. (2002b), ‘The Tele-Home-Care/Tele-Health Systems in Japan’, Global HealthCare, January, pp. 13–29. Tsuji, M. (2003), ‘The IT Revolution and Telecommunications Infrastructure’, in E. Giovannetti, M. Kagami and M. Tsuji (eds), The Internet Revolution, Cambridge University Press, pp. 39–55. Tsuji, M., D. Kubo, F. Taoka and M. Teshima (2002), ‘The Role and Issues of International Tele-Education: Lessons Learned from the Japanese Experience’, Proceedings of the 15th Annual Conference of the Asian Association of Open Universities, Indira Gandhi National Open University, New Delhi, India. Tsuji, M., F. Taoka and M. Teshima (1999), ‘Multimedia Technology and TeleEducation: An International Comparison’, Proceedings of the International Conference on Distance Education and Distance Learning, Tsinghua University, Beijing, China, 15–21 April. US Department of Commerce (2001), Digital Economy 2001, Washington, DC.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 35

3. Internet upstream connectivity and competition policy: Western Europe and Southern Africa Emanuele Giovannetti* It was just a length of cable dangling from a ship off the coast of South Africa that ran along the bottom of the sea and then up onto shore here on June 19, like a mooring line ... It established a key beachhead for an advanced undersea communications cable that by year’s end will give sub-Saharan Africa its first world-class connections to the global telephone network and the Internet. Hiawatha Bray, From the Boston Globe, 22 July 2001.

1.

INTRODUCTION

While a rich and varied economic debate has developed on the issues of liberalization of the telecom market and its relation to the costs, and therefore diffusion, of the Internet, little has been done on its less known side: the upstream connectivity required by Internet service providers to reach the rest of the Internet. While in its infancy Internet connectivity between different networks was mainly a technical problem and was taking place on a cooperative base at public exchanges, commercialization of the Internet, and a progressive wave of mergers and acquisitions, have deeply changed the way information packets travel across the different networks composing the Internet. In particular, while major backbones still exchange traffic free of charge amongst themselves, they started charging for transporting information from smaller providers, who can only achieve global Internet connectivity by using these backbones. Also, the evolution toward quality and delay sensitive Internet applications, in other words, video on demand, voice over IP and secure e-commerce transactions, is prioritizing the quality aspects of Internet connectivity. While quality in the Internet is expressed by an array of properties of the information transmission process, it is strongly related to the type and number of connections each network has with the dominant backbones. 35

Kagami 01 chap 01

36

2/15/04

2:55 PM

Page 36

Country/area studies

Many are the issues competition authorities need to scrutinize before understanding if there is any scope for antitrust action or regulation. The preliminary tasks of defining the relevant market, both in terms of product space and of geographical extension, ascertaining the degree of market power of single competitors, monitoring the evolution of prices and their geographic differentials, evaluating the existence of entry barriers and detect anti-competitive behavior or quality based non-price discrimination are particularly challenging for the market of Internet connectivity. The main problem is the elusive nature of the commodity traded, wholesale transmission of information packets, along routes which are often recalculated at each step, hop, of the transmission process. While, indeed, it is easy to calculate the traffic exchanges for traditional telephony, which travel along a dedicated circuit, and to verify the associated economic transactions, new and different tools are required to trace Internet traffic flows when, for example, even a single e-mail from Cambridge to the IDE in Japan is decomposed into many sub-messages which may, or may not, reach the final destination traveling across different routes and networks while some of the network crossings are for a fee and others take place for free. A study of the connectivity of the Internet is therefore an important preliminary step since it describes the actual chain of interconnections for Internet traffic. An empirical analysis of this traffic can be based on software that allows the visualization of the paths that data packets take through the Internet, recording all the ‘hops’ (routers) along the way. We argue that an essential source of relevant information to explore the backbone market is to be found in studies of cyber-geography, which provide the traffic-tracing devices and the tools for their interpretation, and in online Internet connectivity trading places which, in turn, provide a basis for the evaluation of Internet connectivity prices. Recently, antitrust hearings are starting to include evidence from studies in cyber-geography to obtain a clearer picture of the evolution of the Internet backbone industry. The following sections attempt to introduce these complex issues and to suggest a direction of study requiring a common effort both by economist and network specialist, to be able to map the physical/virtual world of cyberspace and, on the basis of these maps, provide an understanding of the network hierarchical layers and of the likely course of action to maintain a competitive market. In Section 2 we briefly describe the supply side of the Internet, in Section 3 we introduce the theoretical debate on Internet pricing. Section 4 considers the European Internet backbones describing their recent evolution and Oftel’s assessment of the UK’s Internet connectivity market. Section 5 deals with the empirical evidence on the different elements composing Internet connectivity. Section 6 describes the EU Commission investigation of the proposed, and rejected, merger between MCI-WorldCom and Sprint. Section 7 introduces

Kagami 01 chap 01

2/15/04

2:55 PM

Page 37

Internet connectivity and competition policy

37

some examples of techniques used in cyber-geography to understand the degree of connectivity of the Internet and therefore its usefulness to explore market power. Finally Section 8 reviews the current state of the Internet backbone market in South Africa, its connectivity map and the competition policy issues at stake.

2.

THE SUPPLY SIDE

The supply side of the Internet has players that can be divided into functional categories:1 • Internet service providers (ISPs). These provide access to the Net through personal, business, or institutional accounts. • Portals. These are sites intended to be the first place people access when using the Web. Typically a ‘portal site’ has a catalog of websites, a search engine, or both. The variety of portals is closely related to the dispersion of tastes and preferences among consumers. • Content providers. These are websites providing the information users are looking for. But they can also be seen as users themselves, and therefore part of the demand side of the industry using the Internet to cast their contents. • Network access points (NAPs), and Internet exchange points (IXPs). These form the physical interfaces between networks. • Internet backbone providers (IBPs). They carry data traffic, across long distances on fibre optic cables. At each node they provide routing of the information packages to direct each incoming message to the next step of its path. Each single network is connected along two dimensions with the rest of the Net: by sharing the same transmission control protocol/internet protocol (TCP/IP), the communication protocol providing a common language for computers to exchange information, and through the physical network interconnection points. Traffic growth and commercialization have led much of this interconnection from being carried at NAP’s to move toward exchanges at bilateral interfaces. 2.1

Linkages

The Internet is used to exchange information, through a common protocol, the TCP/IP,2 which manages both the routing, decomposition and recombination, of the information packets composing the building blocks of the information

Kagami 01 chap 01

38

2/15/04

2:55 PM

Page 38

Country/area studies

exchanged. Consider an end user, A, who connects to a point of presence (POP) of its Internet service provider. This connection can take place through dialup, ADSL, cable TV, or dedicated access. If the information exchange, communication, is between A and another end user and/or website, B, connected to a different ISP, then A’s ISP needs to lease lines to transfer A’s message from its POP to either a NAP or an Internet exchange point. If such a point is shared by the sender’s and the receiver’s ISPs, then the path is decided, otherwise the IP protocol will evaluate the path, for each single information packet, to reach B’s ISP’s point of presence. This structure of direct or indirect connections between ISPs, NAPs and IXPs can be repeated many times depending on the actual distance in the network topology between the two users, A and B. Long distance connection is finally provided by backbone operators. The entire route is recalculated at each router, and the actual transmission process is carried by many different ISPs located at different hierarchical levels. The multi-ownership of different segments of the routes used to exchange information between different end points of the Internet is one of the features of the Internet architecture generating difficulties in understanding pricing in the Internet. On the other hand, the rapidly expanding number of IXPs, particularly in Europe, by reducing the number of links and the average distance traveled, simplifies the information package routing; however the multiplexing and connectionless nature of the Internet still makes the price formation process extremely more complex than in the dedicated connection structure used in traditional telephony.

3.

INTERNET PRICING

The brief history of the Internet, see for example Hobbes (2001), had a major turning point in 1995 with the shutdown of network financing by the National Science Foundation. By 1993, more than 50 per cent of Internet traffic was non-academic. The transition between a fully academic and a predominantly commercial network opened a sparkling debate on how to price the use of the Internet. The first issue analysed was the problem of congestion pricing, that is, how to develop a price system that would efficiently allocate network resources among the users and prevent the possibility of network collapse because of congestion. The analysis of the free riding problem, associated with the characteristics of an academic transmission protocol, now used by commercial players, has also been described as the tragedy of the telecommons.3 Following Reichl et al. (1999), the literature on congestion pricing can be classified according to broad categories suggesting different price mechanisms, these being: usage sensitive, packet or flow auctions, flat rate,

Kagami 01 chap 01

2/15/04

2:55 PM

Page 39

Internet connectivity and competition policy

39

service pricing and edge pricing. Usage based pricing is desirable for MacKieMason and Varian (1995) to prioritize the use of a congested resource, so that higher priority is given to those who most value access. In particular they describe the properties of a pricing design, the smart market approach, which has since been seminal in the theoretical debate on Internet congestion pricing. Schenker et al. (1996) criticized the possibility of using optimal pricing based on congestion costs since communication exchanges often involve the packet transfers across different networks, with different interconnection, access, peering,4 and pricing agreements. They suggested instead the use of edge pricing, where the entire computation of the charges for the user takes place at the access point. With edge pricing the ISP charges the users of its network and then makes a sequence of bilateral agreements with adjacent providers. Pricing is therefore entirely local and can be based on any combination of access and usage based charges: the two extreme being per packet charge, at one end of the spectrum, and capacity purchase at the other. A different stream of literature analysed Internet pricing as the equilibria emerging from the oligopolistic interaction among ISPs. Mason (2000), for example, studies duopolistic competition among Internet access providers and shows that using two part tariffs, a price made up by a subscription fee plus a usage based component, generates tougher competition among providers, than if using flat rates. Similar problems have been addressed in the economic literature on two-way access for telephone networks; see Armstrong (1998) and Laffont et al. (1998), or Chapter 5 of Laffont and Tirole (2000). However a different analysis is required to model interconnection in the Internet since packet switched networks, such as the Internet, do not route packets along a dedicated line as circuit switched networks do; an attempt along these lines can be found in Giovannetti (2002).

4.

THE EUROPEAN INTERNET BACKBONES

In the following we concentrate on Western European Internet connectivity, by studying its backbone industry and the related emerging policy issues faced by competition authorities. In December 1999 the European Commission launched the eEurope initiative with the objective of speeding up the process of bringing Europe online.5 The eEurope plan defined its three main objectives, of a cheaper, faster and secure Internet and the European Parliament endorsed the eEurope action plan identifying unbundled access to the local loop as a short term priority to bring about a substantial reduction in the costs of using the Internet. However high speed access to the local loop is not sufficient to achieve, per se, the eEurope objectives. Indeed one of the main problems in securing fast and cheap Internet access arises not in the final connections

Kagami 01 chap 01

2/15/04

2:55 PM

Page 40

40

Country/area studies

between users and ISPs but in the costs and quality of the connection between ISPs and the rest of the Internet. The association of European ISPs, EuroISPA, pointed out in 1998 that it was in fact common for many European ISPs to lease bandwidth to the United States to route intra-European traffic, as this was often commercially convenient6 though technically inefficient. This unfortunate state of affairs,7 which implied a net transfer of resources from European ISPs to US carriers, was the result of strong regional (route) asymmetries in the price for bandwidth. Indeed, bandwidth is not relevant on its own: since the objective is to transfer data between different geographical locations, what matters is bandwidth along a given route, for example, the transatlantic route London–NewYork. The last complicating element is that information transfers do need to pass through several different routes, crossing them at different interconnection points. These junctions themselves may have different rules and charging structures. In recent years there has been a rapid transformation in long distance telecommunications: data traffic represents now more than 50 per cent of the overall traffic and it is increasingly transmitted over the IP network. Entrants in the backbone market have deployed in the last four years more than 10 000 route miles of fibre network and the amount of bandwidth which can be provided on a given strand of fibre is increasing enormously because of the new ways of exploiting fibre such as dense wave division multiplexing (DWDM). Table 3.1 provides the number of completed and planned data traffic route-miles outside the US for some of the major backbone competitors. Table 3.1

Completed and planned non-US route miles

Network

Completed non-US route miles

Planned non-US route miles

31 000 NA 3591 37 825 NA 22 000

42 700 80 000 4750 37 825 31 000 22 000

360networks Global Crossing Level 3 Qwest Williams Communications WorldCom Source:

4.1

Dain Rauscher Wessels (2001).

The Economic Debate

Interconnection charges among backbone operators have been predominantly of the settlements-free type, for peering,8 while money is usually paid for transit arrangements9 (see Kende (2000)). The transformations taking place in the

Kagami 01 chap 01

2/15/04

2:55 PM

Page 41

Internet connectivity and competition policy

41

backbone market in terms of capacity, switching technology and mergers and alliances are changing the incentive compatibility of existing pricing systems and regulatory authorities are becoming increasingly interested both in the possible emergence of market failures and in the interaction between existing regulation and interconnection pricing. In particular the United Kingdom telecommunications regulatory authority, Oftel, has recently published its Effective Competition Review of Internet Connectivity10 investigating the state of competitiveness in the market for Internet connectivity. The first step of the analysis requires the definition of the relevant market, both in terms of the economic space of substitute products and of its geographic extension. The market for Internet connectivity is defined as the wholesale market for an essential input required for the provision of Internet retail access. The exact boundaries of the market have been drafted by using the test of the hypothetical monopolist11 leading Oftel to the conclusion that Internet connectivity constitutes a market on its own. The definition of the geographic extension of the market for Internet connectivity is also of particular interest. This starts by considering the geographical scope with which ISP retailers may purchase Internet connectivity. Three possibilities are identified for an ISP based in the UK to buy connectivity: in the UK, elsewhere in Europe or in the US. The total price paid by the ISP is given by the sum of the link to the interconnection point and the transit charges paid to the backbone operator for interconnection and the additional costs faced by a British ISP to buy connectivity outside the UK was considered high enough12 to make the UK Internet connectivity market competitive. This finding, conflicting with the earlier EuroISPA statements that the European market was extremely expensive is, most likely, due to the recent evolution of the European backbone industry and the new wave of investment which has led to what has been described as fibre glut. In particular in the UK there are over 20 suppliers of Internet connectivity whose market shares were assessed and failed to identify an operator as having a dominant position in terms of market volume. A second important aspect, in analysing the market structure, is the possibility of new entry. New European networks have been built by entrants like Level3, Carrier1, 360networks, Global Crossing, Interoute and KPNQWest. Table 3.2, derived from the Oftel review and integrated with companies’ Website information, provides brief profiles of the large firms supplying Internet connectivity within the UK.

5.

THE PRICE OF INTERNET CONNECTIVITY

The upstream connectivity costs for an ISP can be divided into two main classes, the bandwidth leasing costs required to connect to peering or exchange points

Kagami 01 chap 01

2/15/04

42

Table 3.2

2:55 PM

Page 42

Country/area studies

Main Internet connectivity operators in Europe

Provider

Network coverage

360networks

360networks was constructing a network to link 11 European countries, to provide long-haul connectivity among 35 European cities and to extend its network 18 000km by mid-2001. The company was also buying 330km of dark fibre strands on rings in six cities: Geneva, Lyon, Marseilles, Milan, Paris and Strasbourg. AboveNet had 197 GBS total capacity, 497 peering agreements. European presence in Vienna, London, Paris, Frankfurt and Amsterdam. A GBP 4 billion investment was due in 2000–2003 to extend the existing 50 000km pan-European IP backbone by 20 000km, using Cisco 12 000 series routers to provide multiple 10Gbit/s trunks. The network covered Denmark, Ireland, Norway, Spain and Sweden by January 2001. CWG was currently investing around US$1.5 billion in its network. The backbone was being connected with 11 000 miles of dark fibre, integrating national IP networks. The company aimed to have a total of 200 European POPs in operation by 2002, on a network reaching more than 40 cities in 13 countries. In the first half of 2000, the company acquired 22 ISPs, providing itself with local presence and customer bases in 12 European countries. Carrier1 operated a pan-European fibre network connecting 12 countries, extending over 10 000km and connecting POPs in 20 European cities. It was in the process of constructing six city ring fibre networks and planned to build at least another 14 in 2000–2001. The COLT Internet Backbone provided high-speed connectivity between 12 European cities and on to the US and the rest of the world. The European peering points where Colt was present were: LINX (London), DE-CIX (Frankfurt), MAE-FFM (Frankfurt), AMS-IX (Amsterdam), BNIX (Brussels), PARIX (Paris), SFINX (Paris), ESPANIX (Madrid), CIXP (Geneva), TIX (Zurich), MIX (Milan), VIX (Vienna), DGIX (Stockholm). The existing 2.4Gbit/s backbone network covered 21 cities in 17 countries. AT&T and BT had said they would invest US$3 billion in the Concert joint venture to 2005, in order to build an IP backbone linking about 60 cities by the end of 2000, and about 100 cities outside the US and UK by the end of 2001. Deutsche Telekom operated a nationwide IP backbone in Germany with 74 nodes. Following the end of its partnership with Global One, the company announced plans to invest 4 billion marks over five years on its own pan-European fibre networks, aiming to install 90 POPs in 40 countries and 150 000km of fibre.

AboveNet BT Ignite

Cable & Wireless

Carrier1

COLT Telecom

Concert

Deutsche Telekom

Kagami 01 chap 01

2/15/04

2:55 PM

Page 43

Internet connectivity and competition policy EasyNet

Energis France Telecom

Global Crossing

GTS

Interoute

KPNQwest

Level 3

PSINet

43

EasyNet is a pan-European Internet Service Provider operating in the UK, France, Germany, Belgium, Spain, Switzerland, the Netherlands, and Italy. The recent merger with ipsaris, meant that it has 3500 route kilometers of light fibre cable, with approximately 350 000km of optical fibre throughout the UK. Energis operated 6500km of fibre-optic network in the UK and a continental European backbone network of 12 000km, linking 18 POPs. France Telecom was in the process of constructing a new European backbone network, which was due to stretch 20 000km and connect 40 POPs in 16 European countries by the end of 2001. It would interconnect with other national and regional networks, giving total coverage of 250 European cities and access to the networks of France Telecom’s Equant and Global One subsidiaries. Global Crossing operates a backbone network serving cities in Belgium, France, Germany, Italy, the Netherlands, Switzerland and the UK. It connects with GTS’s Atlantic Crossing 1 (AC-1). The company intended to expand its VoIP backbone to 15 additional US cities as well as to Amsterdam, Brussels, Copenhagen, Frankfurt, London and Paris by the end of 2000. GTS operates a network consisting of three major components: a pan-European IP-optimized backbone; a number of fibre-optic MANs in Western and Eastern Europe; and Gemini AC-1, a transatlantic backbone cable. The backbone network extended 17 500km by May 2000. Interoute was constructing a 20 900km pan-European IP network that would link 70 European cities in 17 countries and have 200 POPs. The first phase of the network, due for completion in mid2001, would consist of eight rings connecting 46 cities in nine European countries with 18 000km of cable. There were plans to migrate all circuit-switched voice traffic onto the new network by the end of 2001. KPNQwest was in the process of constructing a pan-European fibre-optic backbone, based on seven EuroRings and connected via a transatlantic link to Qwest’s network in North America and Asia. The EuroRing network, when complete, would reach approximately 20 000km and link 59 business centres in Western and Eastern Europe. It would have around 450 POPs throughout the region. Level 3 was constructing a three-ring European inter-city IP network. Ring 1 covered 1 800 miles and links gateways in Amsterdam, Brussels, Frankfurt, London and Paris; Ring 2 is the inter-city loop through Germany; Ring 3 would add an additional 1 300 miles. On completion, the total network would link 13 local city networks and extend 5300km. A long-term network upgrade was currently underway in Europe, replacing PSINet’s E3 infrastructure with STM-1 and above. The final IP backbone network would reach over 21 000km and link at least 30 major European cities.

Kagami 01 chap 01

2/15/04

2:55 PM

44

Page 44

Country/area studies

Table 3.2

continued Network coverage

Provider Teleglobe

The company’s pan-European network reached around 7700km by the end of 2000. Its network construction programme involved building a number of DWDM, fibre-optic rings. By the end of 2000 Teleglobe was aiming to connect 30 European metropolitan areas with 26 000km of fibre. Telenor A pan-European IP network (Nextbone) directly connects Oslo via STM-1s to Frankfurt, New York and Stockholm. Connections at Frankfurt and Stockholm link the STM-1 network to three European rings with 45Mbps capacity. The largest connects Amsterdam, Frankfurt, London and Stockholm. The two smaller rings connect Copenhagen, Helsinki and Stockholm, and Frankfurt, Milan and Zurich. A further 45Mbps extension to Paris was planned. Telia operated a 30 000km IP network covering 15 European Telia countries. Its core ‘Viking Ring’ backbone network came into operation in early 1999, and links Frankfurt, Hamburg, London, Paris and Stockholm. In June 2000, Telia announced that it would construct a new 1400km OC-192 element to the ‘Viking Ring’ in Western France. WorldCom/UUnet WorldCom operates a wholly owned pan-European network spanning over 12 800km. WorldCom’s subsidiary UUNET operates an IP network within Belgium, France, Germany and the Netherlands. The network reached over 200 POPs. Source:

Analysys/Oftel and companies’ websites.

and the costs of interconnection when the data traffic leaves the original network. On many key routes bandwidth prices, relevant for reaching the Internet connectivity point, have been dramatically reduced as a result of technical innovation and stronger competition. On the other hand data on pricing and quality of IP connectivity are often specified in bilateral contracts and kept confidential. However, some of these prices are becoming available from online connectivity traders. Band-X is a trading exchange to buy and sell wholesale Internet capacity online. Internet connectivity is sold from New York, London and Hong Kong, so that an ISP in need of connectivity can buy it from one of these locations and its cost is determined by the sum of the online selling price and the cost of connecting to the nearest trading place. Table 3.313 provides information about the prices and quality for different bandwidth levels both on the London and New York Band-X exchanges. As one can see from Figure 3.1 the London connectivity supplied by operator K is the cheapest for all bandwidth levels, confirming the competitiveness of the UK Internet wholesale connectivity market.

45 Band-X.

212.00 180.96 168.23 168.23 NA NA 28.50

10Mbps 25Mbps 30Mbps 45Mbps 100Mbps 155Mbps Quality

153.40 153.40 143.47 143.47 133.58 128.63 84.11

K-UK Price per Mbps (US$) per month 735.06 NA NA 737.89 NA 392.17 109.48

R-UK Price per Mbps (US$) per month 225.00 180.00 180.00 180.00 180.00 150.00 46.93

X-US Price per Mbps (US$) per month 230.00 225.00 225.00 183.00 177.50 177.50 30.41

N-US Price per Mbps (US$) per month

300.00 270.00 230.00 195.00 180.00 180.00 47.59

L-US Price per Mbps (US$) per month

2:55 PM

Source:

A-UK Price per Mbps (US$) per month

Sample IP connectivity prices and qualities

2/15/04

Seller IP capacity

Table 3.3

Kagami 01 chap 01 Page 45

Kagami 01 chap 01

2/15/04

2:55 PM

Page 46

46

Country/area studies

800

A-UK K-UK R-UK X-US N-US L-US

700 600

US$

500 400 300 200 100 0 10

25

30

45

100

155

Mbps Source:

Band-X website.

Figure 3.1

IP transit prices

Commoditization, facilitated by the emergence of a transparent trading place, usually provides higher incentives toward product differentiation. This shows, in particular, with the efforts to improve quality and reliability of the connectivity supplied. Quality depends on many aspects of a network like its capacity, architecture and the number of peering, private and public agreements. There are, however, some simple ways of testing the quality of a connection and build quality indexes. In particular Band-X provides a quality index for IP connectivity based on the network statistics described below. The monitoring metrics used are: ‘Traceroute’ which measures the number of hops (or routers) which traffic passes through to get to a destination and back. This figure should ideally be as small as possible. ‘Ping’ is used to provide packet loss information, which indicates how much traffic is being lost, usually an indication of congestion or problems occurring on the network. This should be zero in a network performing properly. This metric also delivers the round trip time in milliseconds for the traffic to travel to and from a remote site. Again, the shorter the time, the better. Throughput – the rate at which information travels across the IP network – is measured by examining the transfer rate for replies to HTTP requests for information on specific Websites. (Source: http://www. Band-X.com)

Kagami 01 chap 01

2/15/04

2:55 PM

Page 47

Internet connectivity and competition policy

47

The quality index for the providers whose IP connectivity prices have been seen before is represented in Figure 3.2. It is interesting to see that the London based operator K, the cheapest connectivity provider, also has a high quality index. 120 100 80 60 40 20 0 A-UK Source:

K-UK

R-UK

X-US

N-US

L-US

Band-X.

Figure 3.2

IP quality index

The quality of IP connectivity provided by a backbone is used more and more frequently to asses the potential degree of market power by different competition authorities, since degraded quality of interconnection is seen as a powerful form of non-price discrimination in a concentrated industry. Cremer et al. (2000) indeed model these aspects of quality competition for the backbone market and show that a larger backbone prefers a lower quality interconnection with a smaller one, identifying the incentives for a ‘targeted degradation’ strategy where the larger backbone lowers the quality of interconnection to its smaller rivals. Similar considerations have been central in antitrust investigations as we shall see in the analysis of the proposed merger by MCI-WorldCom and Sprint discussed below. In its review,14 however, Oftel reached the conclusions that the wholesale IP transit market in the UK is effectively competitive, since wholesale prices are falling and there is publicly available information on connectivity quality, while switching costs for unbundled IP connectivity services

Kagami 01 chap 01

48

2/15/04

2:55 PM

Page 48

Country/area studies

are not too high and there is no clear dominant market player in terms of traffic volume, also because of the considerable entry in the industry.

6. ANTITRUST ANALYSIS FOR THE BACKBONE MARKET The backbone market is changing rapidly via mergers and acquisitions. One growing concern relates to the possibility of abuse of market power, one expression of which has been identified in the different interconnection charges levied on small and larger providers, a process started in 1997 by UUNET’s decision of setting minimum traffic requirements for free peering with smaller ISPs. Cave (1999) analysed the possible problems and/or desirability of having some degree of market power and a hierarchical structure in the backbone. The denial of free peering to small ISPs is in fact at the same time solving a free riding attitude leading to possible inefficiencies and congestion, while posing a threat of anti-competitive behavior. Possible regulatory measures discussed by Cave include: obligation to peering, regulation of transit prices, monitoring/prohibition of vertical integration between IBPs and ISPs and price transparency. Of course these types of heavy interventions may induce distortions and are particularly worrying in the case of the Internet that has shown spectacular growth and diffusion in the absence of regulation. It is however of paramount relevance to assess the effects of the changes in the backbone market and their social welfare and regulatory implications. 6.1

Rejection of the Proposed MCI-WorldCom and Sprint Merger

The two most relevant antitrust cases discussed in the industry have been the merger between MCI and WorldCom in 1998 and the rejected proposed merger between MCI-WorldCom and Sprint in 2000. On 11 January 2000, the European Commission received a notification by which MCI-WorldCom would merge with Sprint by an exchange of shares. After an extensive investigation into the merger proposal, on 28 June 2000 the Commission adopted the decision that ‘The notified concentration consisting of the merger between MCIWorldCom and Sprint is declared incompatible with the common market and the functioning of the EEA Agreement’ (Official Journal of the European Commission (2000)). The three relevant markets affected by the proposed merger were identified as being: (1) the provision of host to point of presence connectivity, (2) the provision of Internet access services and (3) the provision of top-level or universal connectivity. The investigation concentrated on this last product market.15 One of the main issues at stake, and a major source of disagreement between the Commission and the two defendant companies, concerned the hier-

Kagami 01 chap 01

2/15/04

2:55 PM

Page 49

Internet connectivity and competition policy

49

archical nature of the Internet. The Commission stressed that a hierarchical structure was clearly exposed by the evidence that top-level providers achieve their connectivity entirely by settlement-free peering mainly at private peering points, whereas smaller providers need to purchase transit from a top-tier network to achieve global connectivity. The present dominant position of WorldCom has been attained through a very active acquisition policy. In the civil action brought by Department of Justice of the United States (US Department of Justice (2000)) against the acquisition of Intermedia Communications by WorldCom are described some of the more than 60 acquisitions operated by this company. In 1995 WorldCom acquired the network service operations of Williams Telecommunications, with its 11 000-mile fibre optic network; in 1996, through the acquisition of MFS Communications Company, WorldCom obtained control of UUNET, the world’s largest Internet backbone provider. In 1998 WorldCom acquired Compuserve, a leading Internet provider and ANS, AOL’s primary Internet backbones network. Other acquired backbones were GridNet, Unicom-Pipex, InNet, NL Net and Metrix Interlink. As a result of the leadership position reached in these years the WorldCom acquisition of MCI in September 1998 was accompanied by the imposition, by the US Department of Justice and the EU Commission, that MCI divest its Internet assets to Cable & Wireless. Having defined the relevant market as the one composed by the providers equipped with a set of peering agreements with 100 per cent settlements-free connectivity across the Internet, the Commission’s investigation found that only five top-level networks, MCI-WorldCom, Sprint, AT&T, Cable & Wireless and GTE, satisfied these criteria and defined the market participants as those who peer both with MCI and Sprint. 16 By adding networks accessible directly rather than through a third party, four more companies were added as market participants: Exodus, Digex, Abovenet and Epoch. The two groups made a total of seventeen players in the market for top-level Internet connectivity. Any other Internet provider would be required to purchase transit from at least one of the top five providers to achieve global connectivity. If connectivity is crucial in defining market leadership the physical expression of market shares is given by traffic flows. These are both composed, for each network, by the internal traffic flows and by the flows exchanged with other networks. The ratio between traffic flows of the different networks have therefore been used to evaluate market shares in the top level Internet connectivity market. These are reported in Table 3.4, together with a revenue based market share distribution obtained by Probe Research (Pappalardo (2001)). From these estimates the new merged entity would have had a market share between 37 and 51 per cent against the next competitor’s one not being larger than 15 per cent.

Kagami 01 chap 01

2/15/04

2:55 PM

50

Table 3.4

Country/area studies

Backbone market shares

Top tier backbones

GTE Sprint C&W MCI WorldCom AT&T *

Page 50

Market shares (revenues year 2000) Source: Probe Research.*

Market shares (traffic ratios) Source: EU.

6.30% 6.50% 3.50% 27.90% 10.00%

[0–10]% [5–15]% [0–10]% [32–36]% [5–15]%

Pappalardo (2001).

The Commission concluded that the proposed merger would have led to the emergence of a top level network provider able to act almost independently of its competitors and customers and to determine its own, and its competitors’, prices and technical developments in the industry. Specifically, the size of the proposed merged company would have allowed a threat of selective service degradation of the Internet connectivity supplied by its competitors. In particular the Commission estimated that the traffic remaining on Net for the newly merged company would be between 40 and 80 per cent compared to a percentage of no more than 32 for the other connectivity providers who would then be forced to exchange around 20 per cent of their traffic with the new dominant player. These size asymmetries imply that a degradation of the peering interface would have a worse effect on the smaller size networks than on the larger one. Furthermore, a degraded quality of competing networks would endogenously affect the already unbalanced market shares, inducing more customers to switch to the larger network, possibly even pushing the competitors outside the peering status which, as we have seen, usually depends on company size. Losing peering status would imply that competitors’ traffic would need to pass through a larger number of hops with the ensuing service degradation due to higher latency and packet loss. Another relevant issue decisive in appraising the competitive effects of the merger was its effects on potential entry in the industry. Since the peering rules require an entrant to be of considerable size, the Commission found that the merger would have generated a formidable barrier for potential entrants in the top tier backbone market. Finally the price effects of the merger would be definitively significant since better connectivity implies better ability in providing service level guarantees. Already without the merger MCI-WorldCom customers pay prices which are on average 20 per cent higher than those charged by the nearest competitors. On the basis of these considerations, finally, in July 2000, the proposed merger

Kagami 01 chap 01

2/15/04

2:55 PM

Page 51

Internet connectivity and competition policy

51

between MCI-WorldCom and Sprint was abandoned following the block imposed by both the Department of Justice and the EU Commission.

7.

EXPLORING THE BACKBONE THROUGH CYBER-GEOGRAPHY

Analysis of the Internet backbone industry, understanding its boundaries, traffic flows, prices, market shares and revenues are particularly challenging both because of the lack of satisfactory statistical data and their elusiveness, due to the non-dedicated connection modes of Internet protocols and traffic exchanged among operators. The public nature of the Internet and the routing protocols on which it is based allow, however, the analysis of the paths followed by information packets from origin to destination through the Internet. An entire branch of research, cyber-geography, is devoted to the mapping of this physical-virtual world. An empirical analysis of traffic flows can be based, for example, on software like traceroute or tracemaps which allow the visualization of the paths that data packets take through the Internet, recording all the ‘hops’ (routers) along the way. This would help the understanding of the hierarchies of Internet interconnection, since ‘Traceroute reveals the hidden complexity of data flows, traversing ten, twenty or more nodes, usually owned and operated by competing companies, to reach a given destination’ (Dodge (2000)). In the Appendix below we show an example of the Tracemap output displaying the route followed by a 64 bytes packet between Matrix.net in Austin, Texas and the Department of Applied Economics in Cambridge. The Cooperative Association for Internet Data Analysis (CAIDA) plays a relevant role in developing tools to analyse and visualize data about connectivity and performance on the Internet. The aim is to construct a global Internet topology and to measure the performance of specific paths through the Internet. We have seen that the first relevant question assessed in the antitrust hearings was about connectivity and, by sending out packets of data from a source to different destinations, it is possible to verify many aspects of the actual connectivity of the Internet. Among the most interesting characteristics of a route taken by a given group of packets are the round trip time (RTT) and the data path, describing how a packet reaches the destination from its origin. On the basis of this information it is possible to analyse ‘frequency and pattern of routing changes’. This sort of analysis is of particular relevance for the Internet since it measures when and how often alternative paths are used for the same source to destination route. An extensive analysis of a relevant number of these path data enables the understanding of the specific role played by a given

Kagami 01 chap 01

2/15/04

2:55 PM

52

Page 52

Country/area studies

backbone or a traffic exchange point. Table 3.5, derived from Claffy et al. (1999), describes the information retrievable using this method. The authors, using samples covering 20 588 end destinations, determined the frequency with which an individual backbone provider (identified by an autonomous system number, AS) appeared in a path and the relative depth of those appearances, both in terms of number of backbones and the number of hops crossed from the source. Table 3.5

Backbone frequencies

Operator

AS1 AS Depth2 IP depth3 Frequency4

CERFnet 1740 Cable & Wireless USA 3561 Sprint 1239 UUNET Technologies, Inc. 701 Internet Systems, Inc. 6196 Compass Communications 7336 Defense Research and Engineering Network 668 UUNET Technologies, Inc. 702 Verio 2914 Los Nettos 226 BBN Planet 1 Telia Network Services 1833 IBM 2685 European Unix Network (EU.net) 286 AT&T 7018 TeliaNet Sweden 3301 Sprint International 4000 Primenet Services 3549 PSINet Inc. 174 European Unix Network (EU.net) 194

0.0 1.9 1.2 1.3 1.2 2.8

2.0 5.6 5.1 5.8 4.2 9.2

18 941 8028 6363 6071 3608 2561

0.8 2.3 1.3 1.0 1.7 2.6 0.6

4.3 12.2 7.5 4.1 8.0 10.6 4.3

1326 1285 1057 910 804 758 697

2.0 1.4 3.4 3.1 2.0 1.6 0.1

6.8 6.5 12.2 7.5 5.4 6.5 0.4

639 553 500 477 463 444 441

Notes: 1. Dominant Autonomous Systems. 2. Average number of ASes a message passes through before reaching this AS. 3. Average number of IP addresses a message passes through before reaching (the destination machine in) this AS. 4. Total number of times this AS occurred in the data (from the set of all destinations and intermediate routers). Source:

Claffy et al. (1999).

Kagami 01 chap 01

2/15/04

2:55 PM

Page 53

Internet connectivity and competition policy

53

In this example, CERFnet/AT&T, Cable & Wireless (which purchased InternetMCI’s backbone in 1998), Sprint, and UUNET play a major role in transporting packets across the Internet. It is clear how this type of network statistical exploration is becoming an essential tool in understanding the hierarchical structure of the backbone industry required for antitrust investigations.

8.

SOUTH AFRICA

8.1.

From the First Internet Link to Present Backbone Industry

The early history of the Internet in South Africa has been reconstructed in Lawrie (1997). The first sustainable e-mail link to the Internet was established in 1988 between Rhodes University and a home in Oregon, US. It was a significant day in our lives when the Fidonet gateway was still operating in production three weeks after it was opened for general campus use at Rhodes, in February 1989. ... We had broken through the sanctions barrier on a zero budget operation without having to resort to any cloak and dagger exercises. We had also beaten the South African government of the day, which was desperately trying to control every last form of communication in and out of the country.

In 1990 the first Internet link was established between Rhodes University and the University of Cape Town. The first Cisco Routers arrived in 1992 following the lifting of the apartheid sanctions. The first leased line running TCP/IP Internet protocol was installed in November 1991 by the PSTN monopoly, Telkom. This leased line connected the existing academic network, Uninet, to the rest of the Internet. Also in November 1991, the first leased line connection at 14.4Kbps with the US was made. By the end of 1999 there was about 90Mbps in total of international bandwidth. At present, South Africa has its Internet Service Providers Association (ISPA) which has established two peering points for its members, one in Johannesburg and the other in Cape Town (JINX and CINX) which are used by most of the ISPs, resulting in extensive savings in international bandwidth costs. ISPA is supporting the maintenance of a map, adapted in Figure 3.3, of South African ISPs developed by Gregory Masel. As shown on the map the South African connectivity between the local ISPs and the global Internet takes place through four main backbone providers: Internet Solution, UUNET South Africa, SAIX and Satellite Data Network. Below we describe their characteristics based on information retrieved from these companies’ websites. Satellite Data Networks (SDN) was launched in October 1998 as the only local ISP in South Africa that independently designed, implemented and managed a satellite facility for international Internet access. Licensed by the

Kagami 01 chap 01

2/15/04

2:55 PM

54

Page 54

Country/area studies Dorea ⎫ 3 Virtual E Solutions ⎬ ⎭

CiTEC (MAE-East) CiTEC

Artslink

CiTEC (London)

InTouch

UUNET SA (Boston)

46Mb/s

UUNET SA (New York)

42Mb/s

IS (New York)

166Mb/s

UUNET (SA)

Vouzi

Internet Solution

Netactive Interpacket

Storm Internet DataPro Cyberhost

BTGnet

@lantic Alternet

SAIX (London) SAIX (New York)

Digex

34Mb/s

IBI

SAIX

ICSweb

8.5Mb/s

IID

45Mb/s

Satellite Data Networks ZSD Posix

AT&T Global Network

Genuity

6Mb/s

AT&T Global Network

Xnet

⎫ ⎪ ⎪ ⎬6 ⎪ ⎪ ⎭ ⎫ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎬ 37 ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎭ ⎫ ⎬3 ⎭

Absolute Solutions ⎫ ⎪ eNetworks

ECnet

Uninet

Source: Updated 21 January 2001. Copyright © G. Massel, 1995–2001 Reproduced from http://www.ispmap.org.za/

Figure 3.3

⎫ ⎪ ⎪ ⎪ ⎪ ⎬ 15 ⎪ ⎪ ⎪ ⎪ ⎭

South African connectivity map

⎪ ⎪ ⎬8 ⎪ ⎪ ⎪ ⎭

CIS CompuDoc Digital Synergy

G-Net Limpopo

CHiPS

Kagami 01 chap 01

2/15/04

2:55 PM

Page 55

Internet connectivity and competition policy

55

Independent Communications Authority of South Africa (ICASA), the network connects directly to the Internet by using a high speed, fully redundant international link via Intelsat to a first-tier Internet carrier in the United States. In addition to connecting to all major public network access points (NAPs), there are private interconnection agreements with major Internet carriers worldwide. At national level, the SDN network incorporates a fault-tolerant national backbone composed of multiple E1 circuits. SDN’s national network also connects to both the ISPA peering points, CINX and JINX. UUNET South Africa is a leading service and wholesale connectivity provider in Southern Africa providing network services both nationally and internationally. UUNET South Africa is expanding to become a major competitor in the entire African continent through UUNET Africa, a joint venture with Africa Online. This newly formed company, UUNET Africa, is extending its presence in Kenya, Egypt, Uganda, Tanzania, Swaziland, Zimbabwe, Zambia, Ghana and the Ivory Coast. This joint venture was expected to include 18 countries across Africa by the end of 2002. Quoting Mike Jensen,17 UUNET Africa could become the dominant pan-African ISP now rivalled only by M-Web, operating in South Africa, Namibia, Zimbabwe and Uganda. The main problem in setting up a continental provider is, however, the lack of interconnect points between the operators in the different African

Oshakati Tsumeb Otjiwarongo

Boston New York Washington

JINX

Gaborone

Okahandja Windkoek

Walvis Bay

Pretoria Victory Park

Koelmanshoop

Lüderitz

Harrowdene Rosebank New Doomfontein Bloemfontein

Durban

CINX N1 City Port Elizabeth Newlands

2Mbit/s 4Mbit/s Source:

8Mbit/s 12Mbit/s

UUNET website.

Figure 3.4

UUNET SA network

14Mbit/s 34Mbit/s

45Mbit/s

Kagami 01 chap 01

56

2/15/04

2:55 PM

Page 56

Country/area studies

countries. For this reason UUNET Africa is investing in an IP-based satellite network across the continent. UUNET South Africa is a subsidiary of MCIWorldCom, whose influence in the top tier international backbone market has been discussed before. From Figure 3.4, showing the network map of UUNET South Africa, one can see the role of international connectivity of this network with its 34Mbps links to Boston, New York and Washington, WorldCom’s international network. SAIX is the IP Network of the PSTN monopolist, Telkom S.A., and has the largest geographical IP network in South Africa with more than 100 POPs and 40Mbps of international bandwidth from South Africa into the US and UK. A new undersea cable, the SAT-3/WASC, with a total length of 14 000km, once completed will link Portugal, Spain, Canary Islands, Senegal, Ivory Coast, Ghana, Benin, Nigeria, Cameroon, Gabon, Angola and the Republic of South Africa where it will interconnect with a second undersea cable linking South Africa to South East Asia via India (SAFE). Once operational there will be a total capacity of 120Gbits for the SAT-3/WASC segment and 80Gbits for the SAFE segment. This new infrastructure will both facilitate inter-African traffic and ease African access to global markets. Furthermore since the infrastructure is mainly owned by African telecom groups, the revenues generated will remain in Africa, reversing the current situation in which around 80 per cent of Africa’s intracontinental telecommunication revenues end up with foreign operators because of the lack of intra-continental connectivity. Internet Solutions (IS) was founded in 1993 and now connects over 1500 organizations to the Internet. Connectivity to the IS backbone is achieved through a Telkom circuit at a fixed cost made by two components: the Telkom’s fixed monthly rental fee, which increases with distance, and the access to the IS network, the proper Internet connectivity. This connectivity cost is proportional to the amount of bandwidth required. In recent years IS built an international infrastructure reaching global connectivity by creating points of presence in Europe and America along with peering agreements with the biggest service providers in the world, running over private lines to Singapore, London and Frankfurt. IS’s top connectivity offer is based on a 4Mbps fibre optic line that terminates directly into their node in London. IS also offers a satellite connection currently available to New York, but soon to be extended to the UK and Europe. IS has peering agreements with UUNET and SAIX in Johannesburg, Cape Town and Durban so that internal South African Internet traffic is exchanged locally improving Internet access speeds and service. Furthermore peering with other second tier local ISPs takes place at the two public peering points, Johannesburg Internet Exchange (JINX) and Cape Town Internet Exchange (CINX).

Kagami 01 chap 01

2/15/04

2:55 PM

Page 57

Internet connectivity and competition policy

57

Together with the construction of the SAT-3 link, mainly financed by Telkom, and with the expansion of the UUNET Africa network, another relevant backbone infrastructure project, Africa one, is under study, for improving Internet connectivity in the African continent. This project entails the laying out of a 32 000 kilometer submarine fibre optic cable surrounding the African continent. The design includes 20–30 landing points in African coastal cities, Europe and the Middle East. This African ring is meant to connect with Global Crossing’s worldwide network and directly link the continent to more than 250 major cities. Countries in the African interior and those coastal states without landing points will eventually be connected to the Africa one network via terrestrial fibre, microwave or satellite facilities. 8.2

Telecommunication Policy Considerations

The wave of investment in Internet connectivity in Southern Africa is paralleled by the active regulatory debate in the telecommunication field. On 16 November 2001, the South African parliament adopted a Telecommunications Amendment Bill to prepare for the licensing of a second fixed-line operator to compete against the incumbent monopoly and still partly state-owned Telkom. This amendment also prepares for the listing of a substantial percentage of shares of Telkom, otherwise owned by the state with a 30 per cent participation of Telkom Malaysia and SBC of the US. This legislative framework opened the perspective of reaching an end to the existing monopoly by mid-2002. The relevance of introducing competition into telephone connectivity is evident from the role that Telkom charges have in forming the overall costs of Internet connectivity. These are the downstream costs suffered by users to connect to their ISPs and by ISPs to connect either to a peering point or to a backbone provider. There is strong pressure from South African ISPs and IBPs to proceed with the opening to competition of the telecom industry particularly after Telkom’s recent service increases which together with the progressive weakening of the national currency have led to a considerable increase in the costs of international bandwidth faced by ISPs, making Internet access in South Africa more expensive than in the US and EU. Finally while liberalization of the downstream telecom market is important for reducing the costs of Internet connectivity in Southern Africa a close scrutiny in the connectivity sector in terms of antitrust and competition policy is also required. Indeed, we described a South African backbone market dominated by four major operators: SAIX, owned by Telkom, UUNET S.A., Internet Solutions and the Satellite Data Network, SDN. However this already thin backbone industry is facing a consolidation phase. In February 2000 UUNET S.A. offered to buy Conlog Holdings’ 75 per cent interest in SDN and on 16 May the South African Competition Board ruled in favour of the acquisition leading to a combined UUNET S.A./SDN capacity of 128Mbps

Kagami 01 chap 01

2/15/04

58

2:55 PM

Page 58

Country/area studies

of international bandwidth with multiple 34Mbps access lines to peering points in South Africa. As one can see the problems faced by the competition authorities in Western Europe and in Southern Africa have common features and a great deal of analysis is required to assess the final decisions and the trade-offs between having financially strong players able to invest in upgraded infrastructures and maintaining a competitive backbone market.

NOTES *

1. 2.

3. 4. 5. 6.

7. 8.

9. 10. 11.

12.

13. 14. 15. 16. 17.

The author gratefully acknowledges the generous support of the Institute of Developing Economies (IDE-JETRO) and the support from the British Economic and Social Research Council under the grant reference number: R000238563. It is however important to keep in mind that overlaps of these functions are often observed, also because of the recent trend in vertical integration, and mergers in the industry. The focus of the debate on protocols is that TCP/IP, reflecting its academic origins, may be inadequate to manage the enormous growth both in traffic and in commercial applications of the Internet, and it can be inefficient at managing different delay and packets drop tolerance being based on the principle of first come first served. Paraphrasing the famous term ‘tragedy of the commons’ concerning the problem of free ride in the utilization of common property resources (Hardin (1968)). We will discuss these concepts at length in the following paragraphs. See Giovannetti (2003). In 1998, for example, the monthly cost for a 2Mbps connection between London and Paris was US$38 while the same capacity connection between London and Virginia (the closest extra-European exchange point) was US$30 even though Virginia is almost 25 times further away from London than Paris (EuroISPA (1998)). Which is more dramatic for the African traffic. ‘Peering has a number of distinctive characteristics. First, peering partners only exchange traffic that originates with the customer of one backbone and terminates with the customer of the other peered backbone. … As part of a peering arrangement, a backbone would not, however, act as an intermediary and accept the traffic of one peering partner and transit this traffic to another peering partner.’ Kende (2000) Transit arrangements occur when one backbone pays another backbone to deliver traffic between its customers and the customers of other backbones. See Oftel (2001). From the Oftel document: ‘A product is considered to constitute a separate market if a hypothetical monopoly supplier could impose a small but significant, non-transitory price increase without losing sales to such a degree as to make this unprofitable. If such a price rise would be unprofitable, because consumers would switch to other products, or because suppliers of other products would begin to compete with the monopolist, then the market definition should be expanded to include the substitute products.’ Oftel (2001) Some 25 per cent of the connectivity price, based on Band-X data comparing the price of a 45Mbits circuit to Paris, with the average price of 45Mbits of IP transit, used to represent Internet connectivity (Oftel (2001)). The different operators are anonymously defined by letters: A, K, R for the London trading place and X, N, L for the New York one. Oftel’s review focused on the intermediate level of Internet connectivity. This is a layer above the case studied by Oftel. The list is however kept confidential. As reported in ‘UUNET’s Pan-African Network Takes Shape’ by Lance Harris, Telecommunication Week, 13 August 2001.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 59

Internet connectivity and competition policy

59

18. A subsequent connection to the South African PSTN website www.telkom.co.za required 22 hops, had 50 per cent packet loss, started with the backbone provider Performance Systems International (NET-PSINETA), then, through a peering exchange, moved to Cable and Wireless USA, and crossed three different networks in the UK before reaching the SAIX network in South Africa. Trying to reach La Repubblica newspaper in Italy the packets went from Virginia to Chicago on (NET-PSINETA), where they hopped to BBN Planet (NET-SATNET), then visited New York, London, Paris and Milan where at the 22nd hop the Italian provider Inet dropped them because it was unable to reach the website of La Repubblica in Rome.

REFERENCES Armstrong, M. (1998), ‘Network Interconnection in Telecommunications’, The Economic Journal, Vol. 108, pp. 545–64. Band-X, http://www.band-x.com Bray, H. (2001), ‘The Wiring of a Continent, Africa Goes Online’, The Boston Globe, 22 July 2001. Cave, M. (1999), ‘Interconnection and the Internet: Competition and Regulation Issues at Local Access and Backbone Levels’, in OECD, Working Party on Telecommunication and Information Services. Claffy, K., T. Monk and D. McRobb (1999), ‘Internet Tomography’, Nature: Webmatters, 7 January 1999, http://helix.nature.com/webmatters/tomog/tomog.html Cremer, J., P. Rey and J. Tirole (2000), ‘Connectivity in the Commercial Internet’, Journal of Industrial Economics, Vol. 48(4), December 2000, pp. 433–72. Dain Rauscher Wessels (2001), ‘An Overview: The Emerging Long Haul Carrier Market’, http://www.tamirfishman.com/download/BANK_US_LongHaul0601.pdf Dodge, M (2000), ‘Mapping How the Data Flow’, Mappa Mundi Magazine http://mappa.mundi.net/maps/maps_004/maps_004p2.html EuroISPA (1998), ‘Good Intentions, the Effects of Telecoms Pricing Policies on the European Internet’, http://www.euroispa.org/telecoms1.html Giovannetti, E. (2003), ‘IT Revolution, Internet and Telecommunications: the Transition towards a Competitive Industry in the European Union’, in E. Giovannetti, M. Kagami and M. Tsuji (eds), The Internet Revolution: A Global Perspective, Cambridge University Press, Cambridge, UK. Giovannetti, E. (2002), ‘Interconnection, Differentiation and Bottlenecks in the Internet’, Information Economics and Policy, Vol. 14, pp. 385–404. Hardin, G. (1968), ‘The Tragedy of the Commons’, Science, Vol. 162, pp. 1243–8. Harris, L. (2001), ‘UUNET’s Pan-African Network Takes Shape’, Telecommunication Week, 13 August 2001. Hobbes, R. Zakon (2001), ‘Hobbes’ Internet Timeline v4’, http://www.isoc.org/guest/ zakon/Internet/History/HIT.html Internet Solutions website, http://www.is.co.za Kende, M. (2000), ‘The Digital Handshake: Connecting Internet Backbones’, Federal Communications Commission OPP Working Paper, No. 32. Laffont, J.-J., P. Rey and J. Tirole (1998), ‘Network Competition: I. Overview and Nondiscriminatory Pricing’, Rand Journal of Economics, Vol. 29, 1, pp. 1–37. Laffont, J.-J. and J. Tirole (2000), Competition in Telecommunications, MIT Press, Cambridge, MA.

Kagami 01 chap 01

2/15/04

60

2:55 PM

Page 60

Country/area studies

Lawrie, Mike (1997), ‘The History of the Internet in South Africa: How it Began’, http://apies.frd.ac.za/uninet/history/ MacKie-Mason, J.K. and H.R. Varian (1995), ‘Pricing the Internet’, in B. Kahin and J. Keller (eds), Public Access to the Internet, MIT Press, Cambridge, MA, pp. 269–314. Mason, R. (2000), ‘Simple Competitive Internet Pricing’, European Economic Review, 44, pp. 1045–56. Official Journal of the European Commission (2000), Regulation (EEC) N 4046/89, Merger Procedure, Brussels, European Commission, DGXIII. Oftel (2001), Effective Competition Review of Internet Connectivity, 23 August 2001. Pappalardo, D. (2001), ‘The ISP Top Dogs’, Network World Fusion, 30 May 2001. Reichl, P., S. Leinen and B. Stiller (1999), ‘A Practical Review of Pricing and Cost Recovery for Internet Services’, http://www.berlecon.de/iew2/program.html Satellite Data Network website, http://www.sdn.co.za Saix website, http://www.saix.co.za Schenker, S. (1995) ‘Service Models and Pricing Policies for an Integrated Services Internet’, in B. Kahin and J. Keller (eds), Public Access to the Internet, Prentice-Hall, Englewood-Cliffs, NJ. Schenker, S., D. Clark, D. Estrin and S. Herzog (1996), ‘Pricing in Computer Networks: Reshaping the Research Agenda’, Telecommunications Policy, Vol. 20, pp. 183–201. US Department of Justice (2000), Civil Action No. 1: 00CV02789 Against WorldCom and Intermedia.

APPENDIX An example of a traceroute map is given in Table 3.A1 for a connection from http://visualroute.datametrics.com/ in Virginia, US to the Department of Applied Economics, University of Cambridge.18 The site was found in 17 hops with zero packet loss.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 14 15 16 17

0

Hops Loss

Table 3.A1

Herndon, VA, US

Location

Ms

61

University of Cambridge University of Cambridge

The JANET IP Service, EU-GW connections The JANET IP Service University of London Computer Centre

2:55 PM

Teleglobe Inc.

Teleglobe Inc

Performance Systems International

Network

2/15/04

0 0 0 0 Se.peering.tier1.us.psi.net 4 Troy, NY,US 0 If-21.core.Washington.Teleglobe.net Washington, DC, US 4 If-5-0.core1. Newark.Teleglobe.net 10 if-9-1.core2.NewYork.Teleglobe.net New York, NY, US 10 if-6-0.bb8.NewYork.Teleglobe.net New York, NY, US 10 ix-9-2.bb8.newyork.teleglobe.net 10 us-gw.ja.net 90 london-bar1.ja.net London, UK 109 pos9-0.lond-scr.ja.net 110 cambridge-bar.ja.net 111 route-sj4.cam.ac.uk 112 route-sidg-3.cam.ac.uk 110 econ-2.econ.cam.ac.uk 110

Visualroute.datametrics.com

Node name

A traceroute path

Kagami 01 chap 01 Page 61

Kagami 01 chap 01

2/15/04

2:55 PM

Page 62

4. IT policies and issues: US and the Americas Andrew B. Whinston 1.

INTRODUCTION

Indicators for Internet infrastructure and online users worldwide have shown phenomenal growth in every aspect of the new telecommunications technology since the introduction of the World Wide Web in the early 1990s. During the last decade, the number of people who have access to the Internet has increased from a virtually negligible number to a range of 300–500 million, depending on various surveys. NUA’s estimate put the figure at 513 million in August 2001 (NUA (2001)). Some 426 million were online in July 2001 according to Nielsen/NetRatings; of these, 236 million people were active online users during the month of July 2001 (Nielsen/NetRatings (2001)). At the same time, the number of Internet hosts that provide contents and services increased from just over a million in 1993 to 110 million in January 2001 (ISC (2001)). Despite such extraordinary successes, adequate and effective access to the Internet continues to be limited to a few countries and, within each country, to those who are in middle to upper income brackets. A critical measure of Internet diffusion is the share of the US among the Internet users. In 2000, about half of Internet users were located in the US and Canada, slightly down from 57 per cent in 1999, according to CommerceNet. Worldwide, top-ten countries accounted for 85 per cent of the total Internet population in 2001 (see Table 4.1). The growing disparity in Internet access among countries or socio-economic groups is called ‘the digital divide’. A deepening digital divide in the Internet age is a critical policy issue because the Internet as a general purpose technology has become essential not only for communications needs but also in economic, social and political arenas. To promote better Internet access, the US and other countries in the Western hemisphere have implemented significant changes in telecommunications policy, infrastructure management, and e-business promotional strategies. In this and the following two chapters, we evaluate the current status of such policies and their effects on increasing access to the Internet. In this chapter, we take up the case of the US telecommunications policy and the question of universal access. In Chapter 9, we review current phases of 62

Kagami 01 chap 01

2/15/04

2:55 PM

Page 63

IT policies and issues: US and the Americas

63

Internet usage in Latin America and investigate the connection between Latin American telecommunications policies and the digital divide. Finally in Chapter 10, we focus on Mexico and Argentina as a case study to evaluate the effectiveness of telecommunications reforms and discuss success factors. Table 4.1 Country

Internet population and penetration rate, July 2001 Internet Penetration Internet users Average # Hours population rate (%) during July of sessions online

US 165 180 807 Japan 46 659 923 Germany 27 914 911 Korea 26 590 004 UK 23 870 341 Italy 18 697 197 Canada 14 445 047 Brazil 11 937 559 Taiwan 11 602 523 France 11 107 974 Australia 9 674 157 Netherlands 8 671 316 Spain 7 384 966 Sweden 5 543 193 Hong Kong 3 935 769 Argentina 3 882 526 Belgium 3 663 437 Mexico 3 419 075 Switzerland 3 415 278 Austria 2 995 651 Denmark 2 930 032 Norway 2 452 772 Singapore 2 103 331 Finland 1 977 637 Israel 1 939 084 New Zealand 1 747 203 South Africa 1 499 186 Ireland 1 250 404 Total 426 491 303 Note:

58.1 36.8 33.9 54.3 39.8 32.3 46.5 6.9 51.3 18.8 49.8 54.4 18.6 61.8 56.5 10.0 35.6 3.4 47.2 36.9 53.7 54.4 51.2 38.1 29.7 44.6 3.4 32.9

102 077 288 20 061 849 15 144 455 16 835 227 13 098 235 8 321 314 8 754 653 6 038 867 5 021 109 5 468 447 5 640 427 4 526 370 3 934 630 3 048 001 1 804 016 1 872 249 1 579 445 1 670 201 1 811 677 1 326 513 1 617 277 1 389 599 955 824 1 096 792 976 261 1 015 577 611 467 560 842 236 258 612

20 19 17 26 13 12 19 13 13 15 13 15 11 12 19 13 12 10 13 13 12 11 15 10 14 15 10 9 18

10:19:06 9:27:04 7:49:23 19:20:17 6:22:21 5:48:17 9:08:10 8:10:48 8:04:21 6:58:58 7:41:43 6:44:11 7:00:14 5:29:16 12:12:19 7:21:47 6:22:27 7:37:26 5:31:32 6:16:39 5:13:55 4:37:47 8:48:06 4:10:11 7:15:36 6:59:51 4:20:43 3:56:15 9:36:12

Penetration rate is calculated as (Internet population)/(total population).

Source: Nielsen/NetRatings, Global Internet Index, 2001. Population figures: Mid-2001 estimates by Population Reference Bureau.

Kagami 01 chap 01

64

2/15/04

2:55 PM

Page 64

Country/area studies

2.

INTERNET INFRASTRUCTURE AND INTERNET ACCESS: A SNAPSHOT

2.1

Internet Access: Penetration Rate

Despite the fact that estimating Internet usage is still an imprecise science, most available data indicate that the access to the Internet is distributed unequally across nations, as discussed above. But even among those who are leaders in Internet usage, Internet penetration rates vary widely. The Internet penetration rate is calculated by dividing the total Internet population by a country’s total population. Among the 28 countries for which Nielsen Net Ratings provide detailed data, the penetration rate varies from 3.4 per cent (South Africa and Mexico) to 61.8 per cent (Sweden) (see Table 4.1). Especially notable is the fact that three Latin American countries included in the survey – Brazil, Argentina and Mexico – have the lowest penetration rates except for South Africa. This is the first indication of their serious problem in terms of expanding Internet opportunities to the majority of their populations. On the other hand, penetration rates for the leading countries in North America, Europe and Asia-Pacific are surpassing or approaching the 50 per cent mark, which is used as an indicator of the maturity of a given technology. When an innovative technology appears, a sequence of innovators (2.5 per cent), early adopters (13.5 per cent) and early majority (34 per cent) adopt the new technology (Rogers (1983)). Early majority adopters are usually risk-averse persons with above average wealth and education, and their adoption signals a general acceptance of the technology. Early technologies took a number of years before they reached 50 per cent of households: newspapers, 100 years; telephone, 75 years; phonograph, 55 years; automobiles, 90 years; electricity, 50 years. Estimates vary according to sources, but recent technologies have been adopted rather quickly: colour TV, 18 years; radio, 10 years; VCR, 10 years; PCs, 20 years. The Internet has grown faster than any other technology, being adopted by more than 50 per cent of households in less than ten years. Figure 4.1 shows the Internet penetration trend in the US. Low penetration rates for Latin American countries are due to several factors. Foremost is limited access to basic telecommunications infrastructure such as telephone and mobile phones. Recent reforms and efforts to privatize the telecommunications sector are indications that governments in these countries are seriously tackling the issue. Whether they have been as effective as expected will be analysed in later chapters. Secondly, the low level of income, along with the associated problem of uneven income distribution, is another factor that hinders rapid adoption of the Internet. Thirdly, survey data on Internet usage typically focus on household users. However, a significant number of

Kagami 01 chap 01

2/15/04

2:55 PM

Page 65

IT policies and issues: US and the Americas

65

80 70 60 50 %

40 30 20 10 2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

0

Sources: US online users: Jupiter Media Metrix; population estimates and projections: US Census Bureau.

Figure 4.1

Internet penetration rate in the US

Internet users in Latin America connect to the Internet from various locations. Nielsen Net Ratings’ Global Internet Trends Report for second quarter of 2001 showed that 51 per cent of the Internet population in Brazil, Mexico and Argentina accessed the Internet away from home. In comparison, between 23 per cent and 33 per cent of users in Europe and Asia-Pacific countries log on the Internet away from home. Such away-from-home Internet access – for example at Internet café, library or work – highlights the critical nature of such problems as low PC penetration at home and the policies to provide more public access points. 2.2

Basic Infrastructure Availability and Internet Access

During the next five to ten years, most Internet services will be moving toward formats tailored for broadband – and wireless – platforms. In the US, the majority of Internet users depend on 56K modems for connection. However, broadband users – including cable modems, DSL (Digital Subscriber Line), satellite and ISDN connections – have increased rapidly from 1.3 million in September 1999 to 12 million at the end of 2000 (Parks Associates and Nielsen data). Among athome Internet users, 12 per cent were reported to have broadband connection

Kagami 01 chap 01

2/15/04

2:55 PM

66

Page 66

Country/area studies

in 2000 (Arbitron/Edison Media Research (2001)). Currently, US Internet users are upgrading Internet connection through 56K modem but the most explosive growth is seen in the broadband adoption (Table 4.2). Table 4.2

Internet connection speed upgrades, US home users (Users in thousand)

Connection 14.4K modem 28.8/33.6K modem 56K modem Broadband Source:

Dec. 1999 6085 31 602 30 912 4725

Dec. 2000 4842 22 960 57 898 11705

(Unit: %) Change –20.4 –27.3 87.3 147.7

Nielsen/NetRatings.

Broadband Internet access is exceptionally strong in the Asia-Pacific region where there are over six million users with high speed access, usually through DSL, according to a Gartner Dataquest report. During the first six months in 2001, there were 4.4 million new DSL subscribers worldwide, half of which were in the Asia-Pacific region, according to Point Topic (www.pointtopic.com), a DSL subscriber analysis firm. Out of 10 million DSL subscribers in June 2001, the Asia-Pacific region accounted for 4.6 million users. Korea, the leader in DSL lines, had 3.2 million DSL subscribers. A few European countries are also increasingly connecting to the Internet via cable or DSL high speed lines (Table 4.3). In contrast, broadband Internet access in the US is dominated by cable operators. In June 2001, almost 70 per cent of broadband users in the US relied on cable modems (6.5 million cable broadband users vs. 2.9 million DSL subscribers). Cable operators began offering broadband services early, and 66 per cent of US households have access to cable, compared to only 45 per cent who are serviced by DSL. But telephone companies who offer DSL services are also hampered by the technical limitations of DSL. For example, DSL services work reliably only within 12 000 feet of a terminal of a telephone network. In addition, unlike cable television services that can be bundled with high speed Internet service, DSL lines do not carry video signals, limiting revenue options. Telephone companies are currently pushing for a complete deregulation regarding long distance and high speed Internet services. H.R. 1542 (TauzinDingell) introduced in 2001 addresses several regulatory provisions such as long distance services, network unbundling and resale requirements (CRS (2001)). Telephone companies argue that such restrictions are limiting their abilities to deploy broadband services more rapidly while the dominant cable

Kagami 01 chap 01

2/15/04

2:55 PM

Page 67

IT policies and issues: US and the Americas

67

players are unregulated. Some versions of the Tauzin-Dingell bill were due to pass in 2002. Nevertheless, the question remains as to whether the sub-par performance of DSL services is primarily due to regulation or non-competitive pricing practices of telephone companies. Table 4.3

Broadband access in selected European countries, August 2001 (Unit: %)

Country

Sweden Denmark Germany France Spain Norway UK Italy Source:

Broadband penetration 13.8 13.2 7.8 6.4 6.2 5.1 2.3 0.9

Cable

Type of broadband Satellite ADSL

3.3 6.6 3.6 3.2 2.7 3.5 1.7 0.5

0.0 0.0 0.0 0.4 0.0 0.2 0.1 0.1

5.5 6.6 4.2 2.8 3.5 1.4 0.5 0.3

T1/ Leased line 5.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

NetValue, http://www.netvalue.com/

Argentina, Brazil and Mexico are far behind in broadband access. Taking into account such factors as PC penetration, availability of cable, competition in the local loop, government policies and others, eMarketer categorized 29 countries into five tiers according to the broadband readiness and potential (Table 4.4). The prospect ranges from very high (tier one) to very low (tier five). Tier four and five countries generally suffer from low PC penetration rates, lack of cable or DSL infrastructure, and an economic condition which is least conducive to broadband growth. Table 4.4

Broadband readiness and potential, 2001

Tier One

Tier Two

Tier Three

Tier Four

Tier Five

US Canada Korea Sweden Denmark Germany

Netherlands Finland Japan Hong Kong Singapore Norway

Taiwan Austria Belgium UK Switzerland Australia

France Spain Portugal Italy Argentina

India Poland Brazil Mexico China Russia

Source:

eMarketer (Macklin (2001)).

Kagami 01 chap 01

2/15/04

68

2:55 PM

Page 68

Country/area studies

Access to the Internet is determined by several factors, including the adequacy of basic telecommunications infrastructure, which is a critical reason why many Latin American countries face serious challenges in catching up with other Internet leaders. A comparison of two key variables that determine Internet access is presented in Table 4.5. Japan and Korea are shown here because the average figures for Asia are not representative of those leaders. Low telephone penetration rates for Latin American Internet leaders are notable if governments are to formulate policies to promote IT and Internet within these countries. Without an adequate telephone infrastructure and available PCs, a wider diffusion of the Internet will be difficult. Reforms in the telecommunications industry have focused on upgrading and expanding the basic infrastructure. But whether privatization and competition policies have brought about noticeable impacts or not is still uncertain. In some cases, a rapid increase in Internet usage among a selected few has resulted in a wider digital divide while the majority of the population are disconnected from the digital revolution. Improving basic telecommunications infrastructure and access remains the number one concern in Latin American IT policy. Table 4.5 Area

World US Brazil Argentina Mexico Europe Asia Japan Korea Africa Source:

Basic infrastructure and internet usage, 2000 Telephones per 100

PCs per 100

Internet users per 10 000

16.3 70.0 18.2 21.3 12.5 39.3 9.8 65.3 46.4 2.5

7.7 58.5 4.4 5.1 5.1 16.8 2.9 31.5 19.0 0.9

588 3466 294 675 274 1264 325 3709 4025 53

International Telecommunication Union (2001).

In certain senses, low telephone penetration rates may not matter if policies are chosen to bypass or leapfrog technological developments. For example, although the diffusion of DSL broadband networks suffers from inadequate telephone networks, access using cable modem, satellite or wireless can be an alternative. However, except for Argentina, the prospect for cable broadband is also low due to low cable penetration rates (Table 4.6).

Kagami 01 chap 01

2/15/04

2:55 PM

Page 69

IT policies and issues: US and the Americas

Table 4.6

Cable penetration rates, 2000

Country

%

Argentina Brazil Chile Colombia Mexico Peru Puerto Rico Uruguay Venezuela

57.1 7.2 17.9 39.8 16.2 10.1 26.0 42.6 10.5

Average

18.6

Source:

2.3

69

Deutsche Bank (2000).

Public Access to the Internet

American online users accessed the Internet from home (37 per cent), work (35 per cent) or school (27 per cent) in 1999, according to Nielsen Home Technology Report. The remaining 28 per cent accessed it from somewhere else. These access sites include someone else’s home or public places such as libraries or community technology centres. In Asia and Europe, public access sites are primarily commercial operations such as Internet cafés or cybercafés. Particularly in Asia, commercial access sites have great success in expanding Internet access to the masses. In Korea, there were about 16 000 ‘PC-bang’ (meaning PC-room) by 2000 according to Donga.com analysis. There were 2000 similar Internet cafés in Taiwan. These public access sites offer a very inexpensive alternative to owning a PC and subscribing to an ISP. For example, PC-bangs in Korea are open 24 hours and offer DSL-level connectivity, at a price of about US$1.00–$2.00 an hour. The great number of PC-bangs may be an odd phenomenon in Korea, where residential access and broadband availability are already outstripping the level seen in North America or Europe. Most users are young males and primary activities are interactive games and web surfing. Users are also not limited to those with low incomes or other disadvantaged groups. Rather, PCbangs are concentrated in the areas near colleges and business offices. In the US and most other countries, public access sites are established by government grants or community initiatives for the expressed purpose of expanding Internet access to those who will be otherwise excluded. But publicly

Kagami 01 chap 01

70

2/15/04

2:55 PM

Page 70

Country/area studies

funded Internet centres may not be as inviting as PC-bangs where users can do whatever or go whenever they wish.

3.

TELECOMMUNICATIONS REFORMS AND INTERNET USAGE

The telecommunications industry worldwide has undergone major changes during the last two decades toward a more open and competitive market. Digitization, computers and the Internet, and the convergence in analog, digital, and wireless technologies are the market forces that made reforms imperative in many countries. In the US, the long struggle among AT&T, the Justice Department and new technology companies ultimately led to the breakup of AT&T and the Telecommunications Act of 1996 which aimed at reducing federal regulation, creating competition and promoting economic convergence in telephony, cable and computer industries. Internationally, World Trade Organization (WTO) members implemented their agreement on basic telecommunications, opening their markets and allowing competitive services trade (the Fourth Protocol to the General Agreement on Trade in Services (GATS), came into force in 1998). Such reform-minded actions stem from the recognition of the fact that telecommunications technologies, computers and software are converging into an essential infrastructure which acts not only as a basic communications medium but also as a platform for international trade – especially in valueadded services and digitized products – and economic development. Besides voice and data services through wired and wireless networks, the convergence in various communications markets has affected almost all aspects of economic activities and international trade. Exports in value-added services such as messaging, data transfer, distance learning, remote call centres and other services all depend on open access to competitive telecommunications markets around the world. However, regulatory reforms do not create market forces that bring about competition unless they are accompanied by efforts by private industries. And markets often fail to safeguard marginal consumers. The increasing digital divide in the US is evidence that the long-held principle of universal access is in jeopardy when it comes to access to the Internet. At the same time, despite reforms and privatization, open and cost-effective telecommunications markets are still not in evidence in many Latin American countries. The success or failure of any reform must be evaluated by increased or decreased levels of access and the cost of access. In this section, we examine the current status of telecommunications reform and policy-making in the US focusing on whether

Kagami 01 chap 01

2/15/04

2:55 PM

Page 71

IT policies and issues: US and the Americas

71

such efforts have increased public access to the Internet. This section will offer a general background for later discussion on the digital divide in the US, policy issues for broadband Internet access, and the relationship between the US and Latin American countries in the IT, Internet and telecommunications industries. 3.1

Brief Overview of US Deregulation Efforts

Invented in 1876, the telephone as the forerunner of telecommunications media exemplifies the difficulty in promoting and managing an essential infrastructure that tends to be a natural monopoly. During the formative years of Bell Telephone and later AT&T, its growth was aided by the simple economics of network externality, the economies of scale and scope. These factors favoured one giant network firm instead of many smaller providers, which AT&T exploited to acquire competitors. The heavy investment required to build a network also meant that only a monopoly would be able to provide such an essential service without failing in the market. By linking universal service with economics of efficient networking, AT&T was able to amass a vast network of telephone providers. The history of telecommunications regulation and reform in the US can be summarized as a struggle between the economics of AT&T’s natural monopoly and the market forces that favoured new technologies and competition. Major reform initiatives were introduced with the development of new technologies which reshaped the telecommunications industry. After divesting Western Union’s telegraph service in the 1910s, AT&T and the Justice Department entered into an agreement that maintained a government-sanctioned monopoly. As later economists pointed out, a natural monopolist could ‘capture’ regulators in such a way that protected the monopolist’s interest (Stigler (1971)). But with the advent of broadcasting through the use of the electromagnetic radio spectrum, AT&T asserted itself in the areas of newly developing technologies and markets. The Communications Act of 1934, which created the Federal Communications Commission (FCC) and put in place major regulatory guidelines for broadcasting, telephony, and telecommunications equipment markets, was a direct result of new communications technologies and the increasing concerns among new technology firms and government regulators. Nevertheless, AT&T retained its monopoly in three key areas of the telecommunications industry: equipment, long distance service, and local exchange service. From the post-World War II years to the divestiture agreement in 1982, regulatory and antitrust actions against AT&T continued as new technologies and markets developed. In 1956, a federal appeals court ruled against AT&T and the FCC in the case of a ‘terminal equipment’ manufacturer, Husha-Phone (Wilson (2000)). Hush-a-Phone was a small plastic snap-on device that users could place over the mouthpiece of the telephone to reduce

Kagami 01 chap 01

72

2/15/04

2:55 PM

Page 72

Country/area studies

background noise. AT&T petitioned the FCC that the device could potentially lower AT&T’s service quality and therefore should not be allowed. This was a strategic move by AT&T who strived to maintain their end-to-end monopoly in the telephony market. But the court invalidated AT&T and FCC’s views, in part establishing the consumer’s right to use terminal devices. It is noteworthy that the FCC, although being a regulatory agency, essentially shared the same opinion with AT&T. While Hush-a-Phone was a rudimentary, non-powered device, the case of Carterfone in 1968 brought the new radio technology and the question of interconnection to the forefront. Since 1959, Carter Electronics Corporation had marketed a two-way radio device that allowed remote, mobile users to be linked to the telephone network. Carterfone was essentially an analog-to-radio converter that relayed the signal from the telephone headset to a mobile Carterfone unit. In its response, AT&T introduced a rule that prohibited any device that established a ‘direct electrical connection’ to its network. After two years of public hearings, the FCC ruled against AT&T. However, terminal equipment and interconnected device markets did not emerge from this case. Even though AT&T implemented stricter rules regarding terminal devices and interconnection to stave off further competition, real challenges appeared in long distance services through new technologies. Various privately owned point-to-point networks had been authorized by the FCC during the post-war period. But the allocation of new frequencies above 890MHz for microwave communications and subsequent developments in satellite-based communications and associated services brought final challenges to AT&T’s monopoly in local exchange networks and its monopoly over the common carrier telephony service. Microwave Communications Incorporated (MCI) began its ‘shared private network’ service for long distance telephony in 1972, ten years after it first petitioned the FCC for such a service. MCI’s service consisted of microwave communications between selected cities across the US, while local interconnections were left to the customers. Nevertheless, the question of interconnection between MCI and other long distance service providers and AT&T’s local exchange networks was largely settled in favor of competition by the end of the 1970s. Following the cases of Carterfone, MCI and new technology providers, AT&T agreed (or proposed) to divest itself from equipment and local operating companies in 1982, giving birth to regional Bell operating companies (RBOCs). RBOCs would continue to be local monopolies providing common carrier services for both telecommunications equipment vendors and long distance service providers. Expcept in the local exchange market, the US telecommunications markets were competitive by the end of the 1980s.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 73

IT policies and issues: US and the Americas

3.2

73

Telecommunications Act of 1996

The Telecommunications Act of 1996 was a major revision to the Communications Act of 1934. While encompassing past developments in the AT&T telephony case, cable television legislations, and new technologies of microwave, the Internet and wireless, the 1996 Act was an affirmation of the new reality evident in the rapidly converging telecommunications sector and computer-based networks. As such, it included provisions dealing with the broadcasting industry (spectrum usage and cross-industry ownership), local access issues in telephone networks, cable franchise regulation, universal access, and obscenity and violence issues. Major economic components of the 1996 Act include: • Relaxation of FCC regulatory intervention; • More favourable environment for mergers/acquisitions across telecommunications sectors; • Expanding competition to achieve lowered price and better service; • Promotion of new technologies and businesses; • Promotion of the convergence and integration among telephony, cable and computer industries; • Minimizing state regulatory powers. These changes have the greatest effect on market structure in the broad categories of industries that fall under the telecommunications industry. These include basic telecommunication services such as voice, packet- or circuitswitched data services, telex and telegraph, FAX and private leased line services. In addition, various types of value-added services such as e-mail, voice mail, EDI and online data processing services are affected. The immediate effect of the Act on the telecommunications industry was that of relaxing market ownership restrictions and lowering entry barriers. Broadcasting companies are no longer restricted in cross-ownership; any telecommunications company may enter into any other market; and private sector initiatives on new communications technologies and resulting business ventures are generally promoted. In terms of Internet access, the 1996 Act is the final step toward opening up the local access market. Since the breakup of AT&T, markets for long distance and other enhanced services became competitive, where AT&T’s share dropped from 90 per cent in 1984 to 38 per cent by 2000 (see Figure 4.2). At the time of the legislation, AT&T’s share was already below 50 per cent of the market. The intent of the Act was to encourage the same kind of competition in the local access market that connects end users where local Bell operating companies (LBOCs) maintained a monopoly in the ‘final mile’.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 74

74

Country/area studies

100 90 80 70 60 % 50 40 30 20 10

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

0

Note: Figures include long distance carriers only, excluding local exchange carriers’ long distance revenues. Source:

US FCC, Trends in Telephone Service (2001).

Figure 4.2

AT&T’s share of total long distance service revenues

At the same time, increased market competition brought down long distance charges significantly, continuing the long term trend that began with AT&T’s divestiture in 1983 (Figure 4.3). By 1996, per minute charges had fallen by half from those of 1984. Most significantly for market competition AT&T’s access charges for local loop dropped from an average of US$60 to less than US$20.

4.

COMPETITION IN LOCAL LOOP AND BROADBAND SERVICE

Despite the intention of the 1996 Act to make local networks more competitive, consumers have not been offered a significant number of choices for local telephone providers. To the contrary, the 1996 Act has resulted in a significant consolidation in the number of local telephone companies, known as regional Bell operating companies (RBOCs), who were spun off from AT&T as a result of its break-up in 1984. Out of the original seven RBOCs (organized from 18 existing regional companies), there are now four entities: SBC Communications (Southwestern Bell, Pacific Bell and Ameritech), Bell South, Qwest

Kagami 01 chap 01

2/15/04

2:55 PM

Page 75

IT policies and issues: US and the Americas

75

1.200 1.000 0.800 $ 0.600 0.400

ARPM Source:

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

0.000

1984

0.200

ARPM, net of access

Economides (1998).

Figure 4.3 Average revenue per minute for AT&T (ARPM Index 1990 = 100) Communications International (US West), and Verizon Communications (Bell Atlantic, NYNEX and GTE). While choices were reduced due to consolidations, the loosening of crossmarket entry restriction meant that competitive local exchange carriers could build new networks or lease lines from incumbent carriers. Cable operators and satellite communications enterprises are now free to offer voice service in addition to video, data and other enhanced services. As a result, the latter players face a lower barrier to entry into local network markets. More competition and choices for Internet access are expected at the local level. 4.1

Local Competition

The US Federal Communications Commission now collects data on the level of competition in local exchange markets by zip codes. According to the latest report, 56 per cent of the areas reported at least one competitive local exchange provider or more. Some 46 per cent of the zip codes still rely on one incumbent provider (Figure 4.4). However, the FCC also reports that 88 per cent of US subscribers live in competitive market areas since many of the non-competitive zip codes are rural areas. Among competitive areas, about 40 per cent reported

Kagami 01 chap 01

2/15/04

2:55 PM

76

Page 76

Country/area studies

from one to four competitive local exchange carriers (CLECs), shown as ‘low’ competition areas in Figure 4.4. Areas with a medium level of competition include those with five to nine CLECs, while high competition areas reported 10 or more competitors. Areas with medium and high levels of competition are growing in number, but still account for less than 16 per cent. Although the number of competitive local exchange carriers is growing rapidly, effects on end users and consumers are not clear. CLECs, instead of building their own network infrastructure that includes the last mile, largely rely on leasing existing lines from incumbent carriers. For example, about 35 per cent of their customers were served by CLECs’ own last mile or local loop facilities in 2000. The rest were served by leased lines. From an economic efficiency point of view, unnecessary duplication of network infrastructure should be discouraged. However, competitive carriers leasing incumbents’ lines are constrained by the incumbent’s pricing policy, which will reduce price effects from competition. 50 45 40 35 30 % 25 20 15 10 5 0 No provider

Low

Medium

Jun. 2000 Source:

High

Dec. 2000

US FCC, Local Telephone Competition (2001).

Figure 4.4

Competition in local exchange markets

Furthermore, new entrants tend to focus on medium and large businesses, institutions and large organizations with voluminous demand. About 60 per cent of CLEC lines are targeted on large users while 80 per cent of the customers

Kagami 01 chap 01

2/15/04

2:55 PM

Page 77

IT policies and issues: US and the Americas

77

served by incumbents are small businesses and consumers. Thus, the majority of residential consumers are largely ignored by the increasingly competitive market players. 4.2

High Speed Internet Access

High speed Internet access in the local exchange market is provided via digital subscriber lines (DSLs), coaxial cable lines, satellite and microwave connections. During the year 2000, broadband services penetrated to 75 per cent of the zip code areas reported by the FCC, growing from 56 per cent the year before (Figure 4.5). The largest broadband connection is through coaxial cable (3.6 million lines at the end of 2000). Although DSL subscriber lines are smaller than cable at 2 million in 2000, they are growing three times faster than coaxial cable lines. Still, a quarter of US zip code areas have no high speed Internet access provider and another quarter are served by only one provider. Low competition areas with two to four providers cover the most geographical areas with 35.5 per cent. Medium (five to nine providers) and high (ten or more) competition accounts for 15 per cent. Although half of the US has no or only one provider, 98 per cent of zip code areas with high population density (that is, 3147 persons or more per square mile) report at least one broadband Internet access provider. This implies that the majority of the population are reached by high speed Internet access. Nevertheless, the lack of competition in high speed Internet access providers and their focus on high density, urban areas is creating unequal access to the Internet in selected areas and population segments. Compared to the top 10 per cent of the most populous zip code areas, where high speed Internet access is available at 98 per cent, low density, rural areas lag significantly behind. For example, only 37 per cent of the bottom 10 per cent of zip code areas in terms of population density (fewer than six per square mile) reported at least one broadband subscriber to the Internet (Table 4.7). Broadband penetration at the bottom 30 per cent of the US is low. The majority of sparsely populated regions have no high speed Internet users. The gap between the rich and the poor is also evident (Table 4.8). In the lowest income areas, only 56 per cent of the zip codes reported at least one subscriber, compared to 96 per cent for the highest income areas. In the lowest five deciles, high speed connection is available for less than 70 per cent of the areas. Nevertheless, even in this lowest income decile, 92 per cent of the population lived in the zip code areas where high speed service was available. In other words, the disparity in high speed Internet access is more prominent in terms of population density than household income.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 78

78

Country/area studies

50 45 40 35 30 % 25 20 15 10 5 0 No provider

One

Low

Dec. 1999 Source:

Table 4.7

90–100 80–90 70–80 60–70 50–60 40–50 30–40 20–30 10–20 0–10

Jun. 2000

Dec. 2000

Competition in high-speed Internet access Broadband access by population density, by zip codes, 2000 Density More than 3147 947–3147 268–947 118–268 67–118 41–67 25–41 15–25 6–15 Fewer than 6

Subscribers* 98.2 96.8 95.4 91.7 85.8 76.8 66.4 54.3 44.5 37.2

Note: * Percent of zip codes in decile with at least one high-speed subscriber. High speed is defined as over 200 kbps in at least one direction. Source:

High

US FCC, High-Speed Services for Internet Access (2001).

Figure 4.5

Deciles

Medium

US FCC, High-Speed Services for Internet Access, 2001.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 79

IT policies and issues: US and the Americas

5.

79

COSTS OF ACCESS

Innovative and pro-competitive telecommunications policies are designed to attract new entrants into local access markets, which promotes more choices and lower prices for consumers. For consumers, Internet access costs consist of two elements: local access charges (dial-up charges) and fees levied by the Internet service provider (ISP). Broadband access often does not require ISP charges, although some telephone operators may require existing telephone service and cable operators may demand a subscription to basic cable television service. Table 4.8

Broadband access by household income, 2000

Deciles

Median HH income

90–100 80–90 70–80 60–70 50–60 40–50 30–40 20–30 10–20 0–10

Over $54K $44K–$54K $39K–$44K $35K–$39K $32K–$35K $30K–$32K $28K–$30K $24K–$28K $22K–$24K Below $22K

Subscribers* 96.1 90.4 82.4 78.7 74.6 69.8 69.4 67.1 62.6 56.0

Note: * Percent of zip codes in decile with at least one high speed subscriber. High speed is defined as over 200Kbps in at least one direction. Source:

US FCC (2001).

Despite increasing competition in the local telephone service, average monthly charges as well as initial connection charges for new service have not shown any significant change over the last ten years (Figure 4.6). In 2000, an average monthly charge for telephone connection was US$20.78. In addition, an average Internet user also pays for ISP. According to the OECD, ISP charges constituted a larger portion of total Internet access charges (that is, telephone access charges + ISP charges) in 1995. By 1998, increased competition in the ISP market had lowered ISP charges significantly (OECD (1999)). However, local telephone charges did not decrease substantially due to the market’s noncompetitive nature. In 1998, a survey of OECD countries showed that on average 65 per cent (57 per cent off-peak) of total Internet access fees were paid to local telephone access providers. The difference between traditional telephone calls and Internet usage highlights the need to restructure telephone tariff schedules. Based on the

Kagami 01 chap 01

2/15/04

80

2:55 PM

Page 80

Country/area studies

average duration of a telephone call during peak and off-peak hours, telephone providers instituted various tariff structures, using measured rates, flat rates and time-of-usage differentiation schemes. Different providers in each country also utilize different tariff schedules, making comparisons difficult. But telephone tariff structures have an immense effect on accessibility to the Internet. Measured tariff models result in excessively high Internet access costs while unmeasured, unlimited access charges often reduce users’ incentive to disconnect even when idling. Nevertheless, a general cost comparison is possible by focusing on off-peak Internet access at an average usage of 20 hours per month. A typical dial-up charge for an ISP in 1998 was US$19.85 in the US where most dial-up services are charged at a low price for initial hours and a flat rate schedule that varies by usage. Together with charges for local telephone access, a typical dial-up Internet access cost just below US$40, a lower than average figure compared to an OECD average of US$46.62. Figure 4.7 shows a comparison of monthly telephone charges and ISP charges in selected countries. In this figure, ISP charges were calculated for 20 hours per month usage except for those using a fixed rate schedule – Finland, Turkey, Korea, Mexico, Greece and Switzerland – where the figure represents a charge 50 45 40 35 US$

30 25 20 15 10 5 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Monthly charge Note: Source:

Connection charge

Data are for urban areas only. USS FCC, Statistics of Communications Common Carriers (2001).

Figure 4.6

Average residential rates for local telephone service

Kagami 01 chap 01

2/15/04

2:55 PM

Page 81

IT policies and issues: US and the Americas

81

for unlimited access. Internet access costs vary widely from Finland at US$19.77 to a high of US$68.44 for Germany. The share of ISP in total cost also varies widely, from the low of 18.6 per cent (Mexico) to 81.5 per cent (Switzerland). Finland (as well as Denmark and Norway) offers the least expensive Internet access. These countries also use a flat-fee schedule. But Internet access in Switzerland and Greece, despite offering a flat rate plan, is relatively expensive. 80 70 60 US$

50 40 30 20

ISP

Germany

Switzerland

Greece

Japan

Mexico

Korea

UK

US

Turkey

Canada

0

Finland

10

Telephone

Notes: Based on the largest ISP in each country, for 20 hours online access a month. In US$ purchasing power parity. Source:

OECD (1999).

Figure 4.7 Comparison of total internet access charges, off-peak rates, 1998

6.

THE DIGITAL DIVIDE IN THE US

By the numbers of Internet users and IP hosts, the US is one of the leaders in Internet access and usage. Basic Internet indicators on PC ownership (59 per 100 inhabitants) and Internet usage (3466 per 10 000 inhabitants) are well above world averages (7.7 and 588, respectively) according to the International Telecommunication Union. The rapid diffusion of PCs and Internet continues. In 2000, half of all US households had a computer at home, and 80 per cent of them accessed the Internet (Figure 4.8). Despite such strengths, however, the

Kagami 01 chap 01

2/15/04

2:55 PM

82

Page 82

Country/area studies

digital divide is apparent among geographical regions, income and education levels and racial groups. 60 50 40 % 30 20 10 0 1998

1999 Computer

Source:

Internet

US Department of Commerce (2000).

Figure 4.8 6.1

2000

US households with a computer and internet access

Persistent Disparities in Income and Education Groups

The digital divide exists between rural and urban areas, although the extent is minor compared to other variables (Figure 4.9). Central cities are particularly affected by the disparity. But a more prominent digital divide runs along education and income levels. Households in the highest income group (over US$75 000) are six times more likely to have Internet access than those in the lowest income group (less than US$15 000). The disparity in terms of education levels is as severe. The significant gap between racial groups, and to a minor degree the plight of central city areas, are strongly affected by the fact that minority races (Blacks and Hispanics) tend to dominated low income, low education groups. 6.2

Digital Divide or Digital Inclusion

As PCs and Internet usage penetrate into wider groups of US households, a US Department of Commerce report characterizes the current status of Internet

Kagami 01 chap 01

2/15/04

2:55 PM

Page 83

IT policies and issues: US and the Americas Geography

Income

Education

83 Race

90 80 % Internet access

70 60 50 40 30 20

0

Source:

Urban Central City Rural Under 15K 15k–25K 25K–35K 35K–50K 50K–75K 75K+ LTHS HS Some Col College Grad White Black Asian/Pl Hispanic

10

US Department of Commerce (2000).

Figure 4.9

Digital divide in the US, 2000

access as a ‘digital inclusion’ rather than a ‘digital divide’. Such a rosy description is based on the fact that lower income groups registered higher rates of growth than higher income groups. For example, households with less than US$15 000 annual income experienced 78 per cent growth in Internet access between 1998 and 2000 (from 7.1 per cent to 12.7 per cent). Those in the US$15 000–$25 000 bracket registered 94 per cent growth while the highest income bracket (over US$75 000) showed only a 29 per cent growth rate. Similarly rapid gains in lower education households are noted. Given such accomplishments, the government report devotes a considerable amount of attention to the issue of Internet access by people with physical disabilities. However, such a positive interpretation of Internet diffusion in the US ignores the consistent fact that only one in eight in the low income group – who are also likely to be in central cities, belong to a minority racial group and have low education – has Internet access while more than three out of four have the same privilege at the other end of the socio-economic spectrum. High growth rates, when starting with a low base percentage, often obscure the level of disparity. At the same time, a survey of a few years may not correctly reflect a

Kagami 01 chap 01

84

2/15/04

2:55 PM

Page 84

Country/area studies

long-term trend. High growth rates, if not sustained over a long period of time, may fail to expand the opportunity to the majority of households in the disadvantaged groups. 6.3

US Framework for Universal Service

The initial idea of universal service in the US appeared as a marketing scheme of AT&T during the early 20th century. As AT&T’s patent expired in 1908, it faced more competition and a possible nationalization of the telephone network such as many other nations had implemented. To fend off competition and government intervention, AT&T launched a sustained series of advertisements, touting its effort to provide universal service that could only be achieved through a single telephone monopoly (Marchand (1999)). Beatty (1999) noted: In a series of monthly magazine advertisements in a homey populist style, AT&T defended its goal of monopolizing the phone system as a natural one, the necessary guarantor of ‘universal service’ through a ‘single system’. Other ads followed, depicting heroic telephone linemen fighting blizzards, and comely telephone operators weaving strands of speech through ‘the magic loom of the Bell System’, to quote from the ad’s lush copy. One ad plunged so deeply into bathos as to show a widow and her children opening the envelopes containing their AT&T dividend checks. So successful was Vail’s ad campaign at changing AT&T’s image that after the federal government took control of the phones in the war emergency of 1918 there was no talk of making the arrangement permanent. AT&T got its sacrosanct monopoly back as soon as the war was over. Why should the public take over a company already dedicated to public service?

But the idea of universal service is now entrenched in the policies of most governments as well as social and political groups’ agendas. While universal service is a political or social goal, economists would argue that interventionist pricing policies often fail to increase social welfare. Given existing costs to extend telephone services to each and everyone, market-determined prices are bound to be higher than the level needed to guarantee universal access. To increase social welfare, economic models favour raising prices to the cost of service, which invariably leaves some consumers without access to the service. But political mechanisms seldom work the same way. Subsidies and monopoly rights are granted to special groups and providers in order to extend the service. AT&T’s long monopoly over local exchange markets was first and foremost based on the desire to guarantee universal service. Universal access to the Internet now has the same level of urgency among policy-makers in the age of information. The Telecommunications Act of 1996 specifically stipulated that ‘access to advanced telecommunications and information services should be provided in all regions of the Nation’ (Section

Kagami 01 chap 01

2/15/04

2:55 PM

Page 85

IT policies and issues: US and the Americas

85

254(b)(2)), thereby making Internet access one of its defined goals of universal access. The Act also put a new emphasis on explicitly funded projects to promote universal access in contrast to the previous approach that relied mostly on implicit support through rate balancing and tariffs. Instead of subsidizing rural areas from high cost urban service prices, equitable and non-discriminatory contributions are distributed to target programmes such as lifeline and link-up programmes and to specific groups such as healthcare providers in rural areas, educational institutions and libraries, and low income customers. The collected contributions are deposited in the Universal Service Fund, administered by Universal Service Administrative Company (see http://www.universalservice.org). 6.4

Extending Universal Service to Internet Access

Universal service goals and policies regarding Internet access in the US revolve around the main objective of extending universal service into the information age, put forward in the National Information Infrastructure (NII) Agenda for Action (NTIA (1993)). The Universal Service Administrative Company focuses on four areas: high cost, low income, rural health care, and schools and libraries projects as mandated by the 1996 Act. A major funding effort is carried out through the so-called E-rate grants, which provide discounts to public and private schools, libraries and consortia. The majority of E-rate funding involved 70 per cent to 90 per cent discount levels, and was distributed mainly for internal networking needs (57 per cent) with 33 per cent on telecommunications services, and 10 per cent on Internet access (Figure 4.10). Relatively speaking, grants are primarily used to install and enhance basic telecommunications needs rather than for increasing Internet access. Grantees are located in both rural and urban areas but emphasis has been on low income urban areas. In its third year (2000–2001), 25 per cent of E-rate grants were given to rural applicants, compared to 62 per cent urban and 12 per cent unknown. A similar information infrastructure support program is run by the Department of Commerce’s National Telecommunications and Information Administration (NTIA). Since 1994, NTIA’s Technology Opportunities Program (TOP) has awarded US$150 million to 456 programmes, which were also matched by US$221 million in local funds. The type of funded projects and target institutions are similar to the E-rate programme as they aim at extending access to telecommunications and the Internet to underserved and disadvantaged communities (poor, rural and inner city). Major types of organizations receiving funds include educational, public safety, governmental, community and healthcare organizations.

Kagami 01 chap 01

2/15/04

2:55 PM

86

Page 86

Country/area studies

1200

1200 1000 700

US$ million

800 600 400

220 200 0 Telecom Source:

Internet

Internal

Universal Service Administrative Company, Annual Report 2001.

Figure 4.10

Type of E-rate grants, 2000–2001

TOP’s main concern is to lower technological barriers to accessing advanced telecommunications technology (that is, Internet). Out of 42 TOP projects surveyed by NTIA in 2001, 83 per cent proposed and implemented an access site where community members could access the Internet (Table 4.9). Other activities included launching resource sites or web-based information sites, supporting alliances and community networks that enhanced Internet access, and improving network services to extend healthcare and governmental services. Access sites and resource centres established by TOP are mostly located in non-profit entities, K-12 schools and school districts, and in colleges. As such, TOP projects strengthened community access points in underserved and disadvantaged areas. But, along with E-rate discount programmes that also support similar educational and community centres, US initiatives for universal service primarily focus on establishing non-residential access points. This is similar to the telephone universal service framework which emphasized public telephones in low income, underserved areas. To the extent that these programmes do not improve low PC ownership (due to low income) and high residential access fees that prevent residential Internet access among disadvantaged groups, improvements in the actual number of users may not coincide with increased numbers of community access centres.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 87

IT policies and issues: US and the Americas

Table 4.9

Types of activities in Technology Opportunities Program projects

Activity

Percentage of projects with the activity

Establish access sites Establish resource centres Provide information or services via Web Develop an alliance for better access to technology Establish a network to provide educational services Provide Internet services through an established ISP Establish a network to provide health services Establish a network to provide government services Create a network to refurbish or distributed donated computers Create a new entity to provide telecommunications services Source:

7.

87

83% 67% 67% 62% 57% 48% 31% 26% 21% 21%

US Department of Commerce (2001).

US AND THE AMERICAS: TELECOMMUNICATIONS CONNECTION

Although we have highlighted the digital divide in the US among various income groups, there is an immense, insurmountable disparity in terms of Internet access and the use of advanced telecommunications network between the US and Latin American countries. More detailed discussion on Latin American countries will be presented in Chapter 9. However, computer and information industries in the US are closely connected to their counterparts in Latin American countries. With telecommunications reforms, privatization and ‘opening up’ domestic markets to foreign ownership and competition, US and European telecommunications players are increasingly active in these countries. While IT and Internet adoption rates are relatively low in Latin America, telecommunications exports to these countries represent billions of dollars each year, and are growing rapidly. Mexico alone imported US$2.8 billion worth of US telecommunications equipment in 2000, a 20 per cent increase from 1999 (Figure 4.11). The top ten markets including Mexico, Brazil and Argentina accounted for US exports of more than US$5 billion in 2000. Mexico alone accounted for 13 per cent of total US telecommunications exports in 2000 of US$22 billion (Figure 4.12). With more competitive and open telecommunications markets, US players are expected to invest and compete directly in Latin American

Kagami 01 chap 01

2/15/04

2:55 PM

88

Page 88

Country/area studies

3000

US$ million

2500 2000 1500 1000 500 0 1999

1998 Mexico

Brazil

2000 Argentina

Source: US Office of Telecommunications Technologies, (2001) ‘Latin America 1998–2000: Top ten markets for US telecommunications equipment’.

Figure 4.11

US telecommunications equipment exports, 1998–2000

countries. In Mexico, the giant Telmex has entered into a US$3.5 billion alliance with Bell Canada International and SBC communications. Telmex also operates the Latin Internet portal in partnership with Microsoft. On the Internet, AOL Time Warner (AOL Latin America) and Yahoo (Yahoo Latin America) are significant players while BellSouth and SBC Communications are investing in cellular networks and partnering with local and international companies. The telecommunications industry, particularly the Internet and e-commerce, in Latin America is seen as an important market for US companies. Although Internet penetration is limited to the top 10 or 20 per cent of income groups, the potential is in the liberalized market conditions that present more favourable investment and operational opportunities in those countries. Whether past telecommunications reforms and market privatization efforts will lead to increased Internet access for the majority of the Latin American population is still a question. But increasingly, US telecommunications corporations and their business models will be impacting the way the Internet is accessed in Latin American countries.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 89

IT policies and issues: US and the Americas

89

7. Colombia 6. Dominican Republic 5. Chile

9. Guatemala 8. Peru

10. El Salvador

4. Venezuela 3. Argentina

1. Mexico

2. Brazil Source: US Office of Telecommunications Technologies, (2001) ‘Latin America 1998–2000: Top ten markets for US telecommunications equipment’.

Figure 4.12 Top ten Latin American markets for telecommunications equipment exports from the US, 2000

8.

CONCLUSIONS

As the Internet expands in other countries, the US’s share in key variables of Internet usage is shrinking. Nevertheless, PC and Internet penetration rates and the availability of the Internet in the US are growing steadily. Telecommunications reforms during the 1990s accelerated this trend by recognizing the innovative characteristics of new telecommunications media and evolving market conditions and by formulating policies to further enhance such developments. Despite such strengths, however, the digital divide still plagues many groups of the US population, mainly those with low income and low education levels. These groups are also predominantly ethnic minorities and more likely reside in urban, inner city areas or remote rural regions. In particular, inner city neighborhoods have the lowest Internet penetration rates. Economic or market forces are often inadequate to address their interests. Universal service policies formulated under the Telecommunications Act of 1996 adopted market-based approaches rather than government intervention in rate-pricing or access guarantee. The primary tools are equitably collected funds that are distributed as grants or cost-sharing measures such as discounts on telephone charges. As we will discuss in Chapter 9, such community-oriented

Kagami 01 chap 01

2/15/04

2:55 PM

90

Page 90

Country/area studies

projects are also becoming more prevalent in Latin American countries. While these are in line with the dominant trend in the telecommunications industry and political leadership, in other words market liberalization and reduction in active government intervention in telephone tariff settlement, their effect on increasing Internet access to the underserved and disadvantaged population will be minimal.

REFERENCES Arbitron/Edison Media Research (2001), ‘Internet VI: Streaming at a Crossroads’, http://www.arbitron.com. Beatty, J. (1999), ‘How Big Business Got a Soul’, in Atlantic Unbound, Available at Atlantic Online, http://www.theatlantic.com/unbound/polipro/pp9903.htm Congressional Research Service (CRS) (2001), ‘Broadband Internet Access: Background and Issues’, CRS Issue Brief (IB10045), http://www.cnie.org/nle/st-49.html Deutsche Bank (2000), ‘Latin America Media: Reshaping the Cable-Broadband Landscape’, Equity Research, 28 June 2000. Economides, N. (1998), ‘The Telecommunications Act of 1996 and Its Impact’, http://raven.stern.nyu.edu/networks/telco96.html International Telecommunication Union (2001), World Telecommunication Indicators. Internet Software Consortium (ISC) (2001), ‘Internet Domain Survey’, http://www.isc.org/ds/WWW-200101/index.html Macklin, B. (2001), Global Prospects for Broadband Internet, eMarketer.com. Marchand, R. (1999), Creating the Corporate Soul: The Rise of Public Relations and Corporate Imagery in American Big Business, University of California Press, Berkeley, CA. McConnoughey, J. (1997), ‘New Universal Service: NTIA’s Guide for Users’, http://www.ntia.doc.gov/opadhome/uniserve/univweb.htm National Telecommunications and Information Administration (NTIA) (1993), ‘NII Agenda for Action’, http://www.ibiblio.org/nii/NII-Agenda-for-Action.html Nielsen/NetRatings (2001), Global Internet Index. NUA (2001), ‘How Many Online’, http://www.nua.ie/surveys/how_many_online/ index.html OECD (1999), Communications Outlook 1999. Population Reference Bureau (2001), ‘Population Estimates, 2001’, http://www.prb.org/ Content/NavigationMenu/Other_reports/2000–2002/sheet1.html Pastore, M. (2001), ‘Residential High-Speed Access Takes Big Step in 2000’, Cyber Atlas, Internet.com. Rogers, E.M. (1983), Diffusion of Innovations, Free Press, New York. Shaw, J.K. (2001), Telecommunications Deregulation and the Information Economy, second edition, Artech House, Boston. Stigler, J. (1971), ‘The Theory of Economic Regulation’, The Bell Journal of Economics, Vol. 2: pp. 3–21. US Census Bureau (2001), ‘National Population Estimates’, http://www.census.gov/ population/www/estimates/popest.html US Census Bureau (2001), ‘National Population Projections’, http://www.census.gov/ population/www/projections/natproj.html

Kagami 01 chap 01

2/15/04

2:55 PM

Page 91

IT policies and issues: US and the Americas

91

US Department of Commerce (2000), Falling Through the Net: A Report on Americans’ Access to Technology Tools. US Department of Commerce, NTIA (2001), Evaluation Report: Technology Opportunities Program. US Federal Communications Commission (2001), High-Speed Services for Internet Access: Subscribership as of December 31, 2000. US Federal Communications Commission (2001), Local Telephone Competition: Status as of December 31, 2000. US Federal Communications Commission (2001), Statistics of Communications Common Carriers, 2000/2001 edition. US Federal Communications Commission (2001), Trends in Telephone Service. Vogelsang, I. and B.M. Compaine (eds) (2000), The Internet Upheaval: Raising Questions, Seeking Answers in Communications Policy, The MIT Press, Cambridge, MA. Wilson, K.G. (2000), Deregulating Telecommunications: U.S. and Canadian Telecommunications, 1840–1997, Rowman & Littlefield Publishers, Inc., Lanham, MA. World Trade Organization (2001), ‘Telecommunication Services’, http://www.wto.org/ english/tratop_e/serv_e/telecom_e/telecom_e.html

Kagami 01 chap 01

2/15/04

2:55 PM

Page 92

5. Software in India: development implications of globalization and the international division of labour Paul Kattuman and Arnab Bhattacharjee 1.

INTRODUCTION

The traditional view of economic growth is that it is driven by savings, investment and capital intensity. The growth of East Asian countries has been attributed to these factors. A counter-view argues that the micro foundations of growth, and thus the proximate cause of differences between national economic performances, lie in capabilities of firms for technological advances. Accumulation of technical knowledge is crucial in economic development; cumulative increases in firm level capabilities attract investment and drive growth. If capital markets work well, capital will flow toward capable firms, and toward the economies that have more capable firms. In each region, firms would compete to enter into gainful activities. If there were wage differentials, and if labour were not free to move, firms would shift their activities to low wage regions. High capability firms would be distributed uniformly across regions. In practice, however, there are stable differences across countries; real wage differentials persist, more capable firms do not move readily to low wage regions, and firms in poorer regions do not readily acquire higher capability. Will the race for capability enhancement lead to convergence between national economies? What conditions will lead to economic polarization of the world? The convergence/non-convergence issue has gained increasing attention in both developed and developing countries, particularly in the context of internationalization of economic activity and international division of labour. What are the implications of globalization for developing countries? At the same time, what are the implications for the North of the increasing competition from the South for high quality employment and value addition, one reflection of which is de-industrialization. The explosive growth of the economic role of IT technologies worldwide has highlighted these questions. IT has facilitated globalization, in the forms of international outsourcing, FDI and co-production. Do the diffusion of the 92

Kagami 01 chap 01

2/15/04

2:55 PM

Page 93

Software in India: globalization and the division of labour

93

general purpose technology implicit in IT, and the growth of software industry mean fierce competition to developed countries from developing countries such as India and China? In the case of the US, it has been argued that firms have been able to derive substantial benefits from outsourcing software development and services, particularly essential maintenance and development services and system integration services to subcontractors, leaving their own IT staff to focus on higher value-added work. Indian firms competed fiercely among themselves for contracts, leaving most of the gains from trade to US firms (Arora et al. (2000)). At a more fundamental level, from the point of view of technological knowledge advance and capability building, there are now more transnational joint research and strategic technology alliances (OECD (1997)). The multidisciplinary nature of new technologies and high costs/risks has encouraged international partnerships and alliances (Hicks et al. (1996)), facilitated by new communication technologies. Recent trends in R&D activity indicate increasing interaction between producers, suppliers, centres of learning (universities, and so forth) and R&D organizations, often through communication networks (Howells (1995)). But developing countries have poor prospects of participating in these networks. While the traditional literature on FDI and on export-led growth provides some explanation for limited foreign economic presence in low wage countries, the geography and trade literature (Fujita et al. (1999)) highlights the powerful reasons for the prevalence of geographic industry clusters. There may however be specific ways in which the IT industry is different from other traditional (commodity-type) industries. OECD (op. cit) reports that the trend toward globalization in the IT industry has been matched by, and to a large extent driven by, a shift in MNC activity in the industry (particularly, US-based MNCs) to Asia, and this shift has been prompted by labour cost differentials (ibid.).1 Any positive impact of such globalization is not likely to be sustainable, since internationalization should increase competition and lead to an equalization of wages across countries (Forge (1995)). But are there more permanent effects of such globalization? Does co-production (even when the initial relationship is severely unequal) of skill intensive services involving firms in low wage economies and in advanced economies offer a route for developing country firms to step permanently up the value chain? India is a good ‘case’ to explore these issues in detail as it is there, if anywhere, that software appears to have shown promise in beneficial international division of labour (Fernandes et al. (2000); Lakha (1994); Lateef (1997)). There is substantial foreign presence in the software industry in India, and many foreign firms have set up software development centres. Over a period of time, several of the domestic companies have also set up subsidiaries abroad engaged in production of software and services, well beyond marketing. It would appear that

Kagami 01 chap 01

94

2/15/04

2:55 PM

Page 94

Country/area studies

longer lasting advantages are beginning to emerge from the ongoing process of globalization in the Indian software and services industry than what one would expect of a typically transitory relationship sustained only by low wage levels.2 This chapter builds on Kattuman and Iyer (2001) and Kattuman and Bhattacharjee (2001), to explore differences and similarities between foreign firms and domestic firms (with and without foreign subsidiaries) in terms of their inputs and strategies (in terms of markets, specialization and application areas), and their resultant output. We attempt to identify aspects of performance in which (India-domiciled) foreign firms lead, and where Indian firms lead; this helps to draw out the impact of internationalization on firm performance.

2.

GLOBALIZATION: INTERNATIONAL COMPETITION AND FIRM LEVEL RESPONSES3

A key implication of trade liberalization is increased competition. Efficient firms should be expected to respond to increases in competition by investing in building up capability, broadly defined, encompassing both productivity and quality levels. In low technology markets in developing countries’ firms oriented purely domestically may not find it worthwhile to invest in raising their capability.4 It may not be worth sinking fixed costs in product development and process improvement, if the demand for older generation products is sustained. But in high technology markets5 firms in both developing and developed countries have little choice. To survive, a firm will need to position itself in a narrow window in the range of capabilities; all firms cannot succeed. The key argument is that in these markets, the growing but more competitive global market does not draw increasing numbers of active firms, but spurs the efforts and fixed outlays of a smaller number of survivors. The optimal strategy for weaker firms may be to quit the race, and the long-term consequences of globalization may be consolidation by some and exit by the others. In the race for increasing capability, firms can be expected to explore all avenues, and advanced firms in high wage countries may find it feasible and economical to locate some of their activities in lower wage developing countries. If it works well, developed country firms can reduce costs; from the perspective of developing countries, firms have at least the potential to begin enhancing capabilities from such sub-contracted or outsourced work. If that potential materializes and if the learning is cumulative, at least some of these firms may in turn, find it worthwhile to migrate part of their operations to locations in advanced countries, not only for marketing, but to take advantage of ‘socialization’ and ‘local collective learning’ (Maskell and Malmberg (1999); and Cohendet et al. (1999)). The ‘tacit knowledge’ accumulated by these firms should increase

Kagami 01 chap 01

2/15/04

2:55 PM

Page 95

Software in India: globalization and the division of labour

95

their productivity in their home locations, and this process, if pronounced enough, could mitigate the strong arguments in favour of nonconvergence. These conjectured benefits would depend very much on the nature of the industry in question. One set of reasons why this does not happen, and why high capability firms cluster spatially and interact among themselves is given by the geography and trade literature (Fujita et al., op. cit). Supply-side inputoutput linkages among firms in any region can sustain spatial clusters of high capability firms. Capable firms enjoy the positive externalities from there being other capable firms in their neighbourhood, reducing costs of transport, coordination, monitoring and contracting (see also Maskell and Malmberg op. cit; and Cohendet et al., op. cit). If the backward and forward linkages are strong, in stable polarized equilibrium, some countries will be always high wage, and some, low wage: a capable firm will not find it profitable to migrate to a lower wage region because of the cost of the loss of high quality local supply. The coordination failure in migration in search of high quality workers with lower wages will forever consign poor regions to have low wages and less capable firms. Such an argument has been used to explain the existence of high tech clusters, like Silicon Valley, Route 128, and Silicon Fen in Cambridge, UK. Sutton (2001) highlights another mechanism that contributes to such polarization. A firm’s capability is embodied in large part in the ‘tacit knowledge’ possessed jointly by the firm’s workforce. Poor mobility of even some individual workers may imply that relocation of the firm’s activities would involve costs that outweigh gains from lower wages. Moving the firm with the loss of a significant fraction of ‘immobile’ individuals would imply costly loss of collective tacit knowledge. Once capabilities are embodied in domestic employees, a firm is no longer perfectly mobile in the face of real wage differentials. Empirical evidence on multinational firms demonstrates their propensity to shift only some kinds of activities to low wage countries, while retaining core competencies in their ‘home’ location. However, there may be a distinction to be made between ‘commodity-type’ industries and the IT industry, which can operate effectively with teams who may be distributed across different geographical areas, and which is less reliant on inputs other than well trained and learning capable software professionals. A distinctive aspect of digital products is the nature of separability of the many activities that are necessary for their production and delivery. In the popular model of software development, the waterfall model (Royce (1970)) development of software is set out in terms of a hierarchy of sequential steps. The first stages (conceptualization, requirement analysis and high level design) are the high value-added stages, while the later stages (low level analysis, coding and so on) are low value-added segments. These different activities are separable in production across space, so long as there are mechanisms for continuous coordination through information exchange. One consequence of the separa-

Kagami 01 chap 01

96

2/15/04

2:55 PM

Page 96

Country/area studies

bility of activities in production is the potential for product differentiation that is common in software to a greater degree than in all other physical products. The delivery and transmission mechanisms of the industry complement the separability. Digital networks that make up the transmission infrastructure allow reliable and real-time transfer of digitial files that comprise both work in progress as well as final products. This makes it possible to work in geographically separated locations in co-production. The above features reinforce a third aspect (which it shares with other high technology industries): continuous technological change. Consequently, production activities are not stable, they change rapidly, and firms themselves must undergo continuous restructuring. If production is geographically dispersed, suppliers and partners in developing countries also participate to varying extents in the process of change. There is inherent potential in the industry for learning by doing, and this could serve as the basis for enhancing the capability of vendors. Producers and users of software may outsource projects to developing countries for a variety of reasons: primarily to access software engineering skills in sufficient quality and scale at lower operating cost, as well as to insure against the risk of cost escalation, through fixed price contracts and staged payments. While relationships typically start through well defined projects that require a low level of coding, there are demonstrated growth routes up the value chain. As projects get completed, vendors gain trust of clients, and aided by their certification, could move on to projects that call for somewhat more responsibility and accountability. By paying attention to quality, instituting industry standard quality practices, at least some firms can move from backoffice work to focus more on strategic business projects, higher level (and higher paying) strategic business products and services. The key issue in the move up the value chain is that while developing country vendors may have the aptitude to master advanced business models and practices they are not going to learn it operating purely in the domestic domain. To develop international business experience, they need to locate internationally. In the case of the Indian vendors, a significant number of firms have moved on to set up production subsidiaries abroad. If these firms acquire business capabilities as quickly and as aggressively as they did their high level quality certification status, then it is not unreasonable to conjecture that they might grow in productivity and market share, particularly in the business services segment.

3.

THE INDIAN SOFTWARE INDUSTRY

The origins of the Indian software and services industry go back three decades to the founding of Tata Consultancy Services (TCS), a spin-off within the

Kagami 01 chap 01

2/15/04

2:55 PM

Page 97

Software in India: globalization and the division of labour

97

premier private sector TATA business group. The initial market was domestic, mainly software development for a few public sector firms and fewer large private sector firms. The early entrants into the industry were cross-entrants. A few large firms in other sectors, including computer hardware firms, diversified into software: notably, Tata, Hindustan Computers Limited (HCL) and Wipro. Many of these cross-entrants spun off in-house computing service divisions into independent business units, to serve the limited domestic market in a regime of import substitution. From the beginning of the 1980s, the industry demonstrated export potential, and encouraged by this the government put together an export policy that allowed liberal imports of hardware and software, allowing entry of wholly owned foreign firms, and setting up of software technology parks as export processing zones (Heeks (1996)). The liberalized policy toward foreign direct investment (FDI) was seized upon by NRIs as well as multinational corporations (MNCs). Citicorp Overseas Ltd, a wholly owned subsidiary of Citibank, set up shop at Mumbai in 1985. A group of non-resident Indian executives in Texas Instruments (TI) promoted the setting up of a subsidiary in 1986 in Bangalore. Other MNCs followed, notably HP, Novell and Oracle. By the 90s many foreign firms had set up offices and subsidiaries in India, often with domestic partners. While their initial objective was to sell their own software and hardware products in the Indian market, as advantages of locating software development in India became evident, these firms moved to establish significant software development centres in India; some well known firms in this class are Oracle, Texas Instruments, Motorola, Siemens, and Microsoft. Some of these software development centres do fairly sophisticated work.6 A number of firms followed this lead and have established significant operations to take advantage of the pool of relatively cheap skilled workforce to compete in the international market. With the advent of high quality communications, one advantage of India that has come to the fore is the time zone, making the location a prime site for 24-hour continuity in the development process. The same advantage is enjoyed, as far as the US market is concerned, by Indian firms who develop in India and test software among clients in the US. As is typical of young industries with technology embodied in human rather than physical capital, and comprised of many market niches, established firms served as incubators for entrepreneurs. In the late 80s de-novo entrants began to spring up in large numbers in the industry; one of the best known of which is Infosys. Many of these firms were started by breakaway groups of managers from established software firms. For example, Infosys was started by a group of managers who left another early entrant, Patni Computer Systems, sustained initially by a maintenance contract from a client of the parent firm. In this selfreinforcing process, the growing population of firms served as the growing set

Kagami 01 chap 01

2/15/04

2:55 PM

98

Page 98

Country/area studies

of incubators of future firms. A small but potentially significant variant within this class are the firms set up by NRIs, some (not all) relocating in India having gained some experience after either long periods or shorter secondments abroad. On the other hand, some of the domestic firms, particularly among the market leaders, have established subsidiaries abroad that are engaged in software production and services. Starting with 38 members in 1988, the industry association NASSCOM grew to have nearly 1000 members by the year 2001. As of 2000, the industry was estimated to employ close to 300 000 employees. The growth rate of the industry, over the five years 1995–96 to 2000–01 has been in excess of 50 per cent. 3.1

Typology of Firms

Traditionally, higher coordination and communication costs, and the importance of geographical and cultural proximity (Cohendet et al., op. cit.) in the development of core competencies have deterred MNCs from subcontracting their critical software development jobs to geographically distant locations. Indian companies, for the most part, have had to stay content with low level routine and repetitive tasks with limited learning potential, and low opportunity for leapfrogging.7 However some US companies have outsourced significant shares of high end software development to their own subsidiaries in India (Arora et al. (2000)). At the same time, some Indian software companies have opened subsidiaries/development centres abroad. Notwithstanding the higher labour costs at these overseas centres, they provide the opportunity for accumulating the essential tacit knowledge that comes out of engaging in the design and production of high value complex services and products. Of the thousand and more Indian software producers registered with the NASSCOM, we collected data on ownership patterns of 293 Indian software firms, as of end-March 2001,8 and the empirical analysis in this chapter is based on these data. These 293 companies accounted for 73.7 per cent of the market share in 1999–2000, as compared with 74.1 per cent market share registered by the 482 companies listed on the NASSCOM directory (NASSCOM (2000)). Of these 293 companies, 155 were of domestic ownership, while the remaining 138 were either subsidiaries/offices/software development centres of foreign firms, or had at least 30 per cent foreign equity ownership. Of the 155 domestic companies, 59 have established foreign subsidiaries that are engaged in actual software services activities overseas, as opposed to purely marketing activities. These 59 domestic companies with foreign subsidiaries (about 20 per cent of the 293 companies in number) accounted for as much as 48 per cent of the total software revenues earned in 1999–2000 (Figure 5.1). By contrast, the foreign firms, nearly half the 293, cornered only 27 per cent

Kagami 01 chap 01

2/15/04

2:55 PM

Page 99

Software in India: globalization and the division of labour

99

of the revenues. At first sight it would appear that domestic companies with foreign subsidiaries are by far the size leaders in the Indian software market. No. of firms, 2001 Domestic companies Foreign with foreign subsidiaries companies 20% 48%

Domestic companies without foreign subsidiaries 32%

Figure 5.1

Revenues, 1999–2000 Domestic companies with foreign subsidiaries 48%

Foreign companies 27%

Domestic companies without foreign subsidiaries 25%

Distribution of Indian software firms, number and revenues

In this section, we analyse the cross-sectional distributions of several important characteristics of these three segments of firms, encompassing strategy, inputs and observable performance criteria, in an attempt to understand what the essential differences between the three categories are. As alluded to earlier, considerable attention has been focused in the literature on the nature of the relationship between MNCs and Indian software companies, in terms of the contractual agreements (Arora et al. (2000); Banerjee and Duflo (2001)) and export orientation of Indian domestic software producers (Heeks, op. cit.; Arora et al., op. cit.), as well as the activities of MNCs in the ICT industry and their subsidiaries in India (D’Costa (2000)). As an aid to our understanding of the performance and operations of foreign and domestic software companies in India, we first examine some of the broad characteristics of these foreign companies. The geographical area-wise ownership pattern of the 138 foreign companies is quite concentrated (Table 5.1). A majority of these companies (97) are owned by corporate entities in North America; 27 are European, and the remaining 14 Asian. Country-wise, the US owned the maximum number of these companies (95), followed by the UK (10), Germany (6) and Japan (5). Either expatriate Indians, or foreign nationals of Indian origin promoted 17 of these 138 foreign companies (MNCs); 15 of these companies were from the US. Some 16 of the 138 foreign companies began operations under domestic ownership, and were later on taken over. Activity-wise, the 138 foreign companies cover a wide range of types: some (like the subsidiaries of financial services companies such as Citibank, Deutsche Bank, Churchill Insurance, Phoenix Life Mutual, and so on) are almost entirely engaged in catering for software services to their parent companies, some (at least partially) work as software development centres of software MNCs, and

Kagami 01 chap 01

2/15/04

100

2:55 PM

Page 100

Country/area studies

Table 5.1

Ownership of foreign software firms in India

Continent / Country North America US Others Europe UK Germany Netherlands Denmark Others Asia Japan Singapore Others Total

No. of firms

Revenues, 99–00* (Rs.mn.)

Paid-up capital* (Rs.mn.)

97 (70) 95 2 27 (20) 10 6 4 3 4 14 (10) 5 4 5

48 811 (75) 48 599 212 14 531 (22) 9944 792 2404 172 1219 1750 (3) 742 386 622

8196 (78) 8175 21 1436 (14) 892 70 140 138 196 855 (8) 495 160 200

138 (100)

65 092 (100)

10 487 (100)

Figures in parentheses are percentage of total. Note:

*

For some firms for which requisite data were not available, estimated figures are used.

some (such as IBM with Tata Group, British Aerospace with Hindustan Aeronautics, Bell South with Telecommunication Corporation of India, British Telecom with Mahindra Group) are joint ventures between MNCs and local enterprises. A preliminary classification of firms among these (not necessarily mutually exclusive) typologies reveals that about one-third of the companies were joint ventures between MNCs and domestic enterprises. Also, approximately 70 per cent of these foreign companies actually engage in (high level) software development work in India, as opposed to low level coding. The predominance of the above types would suggest a strong symbiotic relationship (also noted by Arora et al. (2000)) between MNCs (particularly based in the US) and their Indian counterparts.This would also partially vindicate the notion that geographical proximity (Cohendet et al., op. cit.), which is noted to be very important in other hi-tech industries, and higher coordination costs would discourage delegation of high end software development work to Indian firms. On the other hand, most of the prominent domestic Indian companies have established overseas subsidiaries/offices engaged in production activities. In fact, these Indian companies, which constitute 39 per cent of the 153 companies

Kagami 01 chap 01

2/15/04

2:55 PM

Page 101

Software in India: globalization and the division of labour

101

in numbers, accounted for as much as 66 per cent of their aggregate software revenues in 1999–2000. This indicates the eagerness of these companies to circumvent the limited scope for ‘local collective learning’ in the domestic market and set up shop abroad, notwithstanding higher labour and coordination costs. Thus, even though they largely operate in the same market geographically, it appears possible that the three classes of firms, namely, foreign firms located in India (138 firms), purely domestic firms (155), and of which, those having ‘production’ subsidiaries abroad (59), have markedly different business strategies. How different the dimensions of these strategies are, and how far such strategies have succeeded in enabling these firms to achieve consistently higher output performance, is an issue on which the analysis in this chapter would shed some light. In terms of methodology, we shall profile the differences between these three classes of firms, in terms of their inputs, activities/strategies and outputs, and integrate these quantifiable characteristics by building appropriate linkages. Factor inputs in software production are human capital, physical and financial capital, probably in that order. Ex ante, size, and age of the firms may also matter to productivity. The most significant input and the main source of the competitiveness of the Indian software industry has been the inexpensive skilled manpower generated by Indian higher education and technology institutions (Fernandes et al., op. cit.; Kattuman and Iyer, op. cit.). As a measure of factor endowments of the three classes of software firms in India with respect to human capital, we consider the number of software employees in March 2000. As regards physical and financial capital, it is well recognized that the software service industry does not require substantial upfront investment. However, software development requires substantial investment in both physical and financial capital, and most Indian software developers face constraints in this regard. Most of these firms either rely heavily on equity financing from their parent companies or other companies in their business group, or have adopted a strategy of using services to finance product development (Arora et al. (2001)). In this chapter, we have used equity capital as an indicator of input endowments. Beside these two major inputs, and access to infrastructure,9 size and age profile are the other major inputs in the software production process. We have considered annual software revenues and exports (1997–98 to 2000–01) as measures of size. In the face of stiff competition in the Indian software industry, reputation is important in the software market. Both age (Banerjee and Duflo, op. cit.) and exports (Shy op. cit.) are acknowledged as important indicators for reputation, and so is the adoption of quality standards (Arora and Asundi (1999)).

Kagami 01 chap 01

102

2/15/04

2:55 PM

Page 102

Country/area studies

Closely related to the factor inputs, and important contributors to production are the activities/strategies adopted by software companies. Indian software firms are widely diversified in their choice of activities/strategies, be that in the form of adoption of quality standards, or choice of activities (in terms of areas of specialization and application industries), export markets, and businesses (products, projects and services10) (Kattuman and Iyer, op. cit.). Arora and Asundi, (op. cit.) find that adoption of quality standards is an important determinant of growth and productivity in the Indian software industry. Initially, ISO was the preferred quality certification in India, but the SEI CMM certification is popular now. Similarly, these firms have shown substantial variation in the distribution of their areas of specialization and industries of application (D’Costa, op. cit.; Kattuman and Iyer, op. cit.). Kattuman and Iyer (ibid.) find that, while larger firms serve more industries as well as engage in larger technical portfolios (areas of specialization), a significant proportion of the smaller firms, though serving a lesser number of industries, have chosen to diversify their technical portfolios. Substantial variation is also evident in the orientation to markets, in terms of the export markets that the firms choose to serve, and in the choice of business (products, projects and services). In an earlier paper (Kattuman and Bhattacharjee, op. cit.) we have explored the effect of choice of activities/strategies in determining growth and market share in the Indian software industry. Here, we shall explore how far the distribution of these various strategies and activities has varied across the three classes of firms, and how such choices have impacted upon their output. Probably, the output variable of the Indian software industry that has received most attention in the literature is productivity. The productivity of the industry (in US$ terms) is markedly lower compared to the ICT industries in other comparable export competitive economies (Arora et al. (2000)).11 On the other hand, being an export competitive industry in an emerging market economy with low inter-industry linkages has enabled the Indian software industry to attain higher productivity, as compared with other domestic industries. The industry is also characterized by the incentives provided to employees to encourage higher productivity and possible innovations (Patibandla and Chandra (1998)). Unlike productivity, the Indian software industry has been comparable to its competitors (particularly, Israel and Ireland) in terms of growth and profitability. However, as Patibandla et al. (2000) stress, the industry needs to increase its productivity through product innovations, in order to maintain high growth and profitability against a rising wage rate. Besides, while growth has so far largely been fuelled by exports, maintenance of high growth in the industry is also contingent on being able to make domestic markets/demand grow. Again, while the revenues and exports of the industry are comparable by

Kagami 01 chap 01

2/15/04

2:55 PM

Page 103

Software in India: globalization and the division of labour

103

international standards, these have been driven by a higher share of low-skill export-oriented work (Heeks, op. cit.; D’Costa, op. cit.). Is the industry on the path toward higher productivity through product innovations? Who are the potential drivers in this direction? Are the domestic companies with foreign subsidiaries more likely to carry the industry to a higher growth path, or are the foreign companies with software development centres in India more likely to take this role? How have growth and productivity translated into market shares for the different classes of firms? 3.2

Profile of Differences: Foreign vs. Domestic Firms

Tables 5.2A, 5.2B and 5.2C profile the cross-sectional distributions of usage of factor inputs, activities/strategies, and output measures respectively, for foreign firms, domestic firms and domestic firms with foreign subsidiaries, in the Indian software industry. The results indicate the inputs of the three categories of firms, their choices on the activity/strategy matrix, and the resultant distributions of their output parameters. The three categories display substantial differences in several features of their input usage, activities/strategies and output. Table 5.2A Parameter

Cross-sectional variation in firm characteristics* – factor inputs Foreign firms

No. of companies 138 No. of employees, 2000 298 (423) [Q1, Median, Q3] [42, 110, 396] Paid-up capital, 2000 77.7 (129) [Q1, Median, Q3] [7, 35, 82] Age (in years), 2001 8.5 (5.4) [Q1, Median, Q3] [5, 7, 10] Revenues, 97–98 (Rs.mn.) 199 (285) Exports, 97–98 (Rs.mn.) 200 (261)

Domestic firms

Dom. firms having foreign subsidiaries

153 538 (1,425) [55, 141, 340] 131.2 (230) [22, 57, 132] 12.3 (8.8) [6, 10, 15] 446 (1,251) 424 (1,180)

59 812 (1,436) [90, 262, 842] 181.8 (206) [50, 100, 181] 12.4 (6.4) [7, 12, 16] 650 (1,144) 518 (1,009)

Figures reported are sample averages (standard deviations in first brackets). Quartiles and median are reported in square brackets. Note: * The data for Tables 2A, 2B and 2C are largely drawn from the NASSCOM Directory (NASSCOM, 2000), and supplemented with additional information from the India Infoline and Myiris investor information services, articles and special issues in several news dailies and magazines, as well as firm web-sites. The distribution of activities/strategies (Table 5.2B) pertain to March, 1998.

Kagami 01 chap 01

104

2/15/04

2:55 PM

Page 104

Country/area studies

In terms of human and financial capital used, on the average, the domestic firms are substantially larger than their foreign counterparts; domestic firms with foreign subsidiaries are the largest of the three groups. The medians, and first and third quartiles of the use of human and financial capital were also substantially higher for domestic companies. A similar observation holds also for the size (revenue and exports) and age (in years till 2001) of foreign and domestic firms. As far as the reputation implication of age goes (Banerjee and Duflo, op. cit.), the older domestic firms may be better off; on the other hand, reputation may not really be an issue for the subsidiaries of established foreign companies. It is not surprising then, that the Indian companies with foreign subsidiaries corner the highest share of the revenues and exports of Indian software firms. But, whether their production technologies compare favourably with the foreign companies, and whether they are higher on the learning curve and poised to make profitable investments in the high end of the business will depend, to a large extent, on their position on the activity/strategy matrix. Table 5.2B reflects several similarities and differences between the three categories of firms in their choice of position on the activity/strategy matrix. The preferences of foreign firms, domestic firms and domestic firms with foreign subsidiaries are almost the same, as regards application areas and areas of specialization, (except that, domestic companies with foreign subsidiaries have a high involvement in ‘maintenance’ as an area of specialization). However, the portfolios of the domestic companies are considerably more diversified as compared to the foreign firms, in terms of the number of areas each firm specializes in, and the number of industries served by them. In this respect, the portfolios of the domestic companies with foreign subsidiaries are even more diversified. It would, thus, appear that the essential difference between the choices of foreign and domestic firms is in the degree of specialization versus diversification; domestic firms prefer a considerably more diverse portfolio. As compared with the foreign firms, a substantially higher percentage of domestic firms with foreign subsidiaries get quality certified; this may reflect, to some extent the necessity for these companies to acquire reputation. The choices of the domestic companies with foreign subsidiaries with respect to export markets and business areas are also considerably more diversified, as compared with the foreign firms. Even though, like the foreign firms, a majority of them operate in the US market, unlike the foreign firms, almost an equal proportion of these domestic companies prefer to operate also in European markets.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 105

Software in India: globalization and the division of labour

105

Table 5.2B Cross-sectional variation in firm characteristics – activities/strategies Parameter

Foreign firms

No. of companies 138 Quality certified (no.), 1998 ISO (percent of total) 50 [36] CMM (ISO or CMM, percent of total) 21 [41] Export markets, 1998 – No. of companies (percent) US 109 [79] Europe 91 [66] Asia 87 [63] Application areas, 1998 (percent of cos.) High Web appl. [61] Manufacturing [55] Finance/Bkg. [51] Library mgmt. [11] Printing/Pub. [11] Low Textiles [8] Specialization, 1998 (percent of cos.) High Web technology [62] S/W Prod. Dev. [55] E-Commerce [47] CD-ROM/Mult. [4] Data Processing [4] Low Games/Graphics [1] Business areas, 1998 (percent of companies) Products 102 [74] Projects 108 [78] Services 94 [68]

Domestic firms

Dom. firms having foreign subsidiaries

153

59

64 [42]

30 [51]

23 [46]

14 [56]

133 [87] 118 [77] 112 [73]

55 [93] 53 [90] 46 [78]

Web appl. [73] Manufacturing [71] Finance/Bkg. [64] Printing/Pub. [17] Defence [16] Textiles [9]

Manufacturing [85] Finance/Bkg. [85] Web appl. [78] Library mgmt. [19] Printing/Pub. [14] Textiles [12]

Web tech. [80] S/W Prod. dev. [68] E-Commerce [68] CD-ROM/Mult. [13] Chip/Microproc.[10] Games/Graphics [3]

Maintenance [85] Web technology [83] S/W Prod. dev. [76] CD-ROM/Mult. [10] Games/Graphics [3] Chip/Microproc.[3]

130 [85] 142 [93] 135 [88]

51 [86] 58 [98] 53 [90]

Figures reported are number of companies. Percentages of total number of firms are reported in square brackets.

Projects dominate the choice of business areas for all the three categories of firms. However, for foreign firms, the second choice is products, followed by services; for domestic companies, it is the other way round. Even so, a larger proportion of domestic firms prefer to work in product development than foreign companies. As suggested by Arora et al. (2001), it is possible that these companies prefer to work in projects and services, in order to generate capital to plough back into their product development activities. This hypothesis also fits in well with the observation that ‘maintenance’ is the most preferred area

Kagami 01 chap 01

106

2/15/04

2:55 PM

Page 106

Country/area studies

of specialization for the domestic firms with foreign subsidiaries. In so far as working in high end product development is associated with the highest degree of knowledge enhancement and innovations, it is encouraging that a comparatively high proportion of both domestic and foreign firms are engaged in this business activity. The most significant difference in the cross-sectional profiles of output performance of Indian software firms is in the productivity of these three categories of firms (Table 5.2C). The average productivity (revenue per software employee) of domestic firms is substantially higher (32 per cent) than that of foreign firms, though the standard deviation is also high. The average productivity of the domestic companies with foreign subsidiaries is slightly higher than that of all the domestic companies. While the average long-run growth rate (annual compounded growth rate, 2000–01 over 1997–98) is significantly higher for domestic companies, the average growth rate of domestic companies with foreign subsidiaries is lower than that of the foreign firms. However, the cross-sectional distribution of growth rates for domestic companies has a thicker upper tail than that of the foreign companies, as evident from the substantially higher third quartile. The export intensity of foreign companies is higher, on an average, and the third quartile at 100 per cent indicates that more than a quarter of these companies export 100 per cent of their software products and services. Consequent upon the higher revenues of the domestic companies, their market share is also higher; the market share of the domestic companies who have foreign subsidiaries is even higher. On the basis of this profile, it would seem appropriate to conclude that the superior factor endowments of the domestic companies (particularly those with foreign subsidiaries), combined with their diversified positioning with respect to application industries and areas of specialization, and probably a higher focus on product development, have enabled them to perform better than their foreignowned counterparts, on average. In particular, there is a 33 per cent productivity gap in favour of the domestic companies with foreign subsidiaries vis-à-vis the foreign firms, on average. The question is, does this performance gap hold prospects for a move up the value chain?12 As noted earlier, the variation amongst the domestic firms with foreign subsidiaries over the cross-section is considerably higher than that of the foreign companies. It is possible therefore, that within-industry dynamics may restrict substantial possibilities of learning and value creation. To analyse this issue further, we explore the entire cross-sectional distributions of the output variables. Figure 5.2 displays kernel density estimates of the cross-sectional distributions for foreign firms, domestic firms without foreign subsidiaries and those with foreign subsidiaries, with respect to some of these variables where significant differences were found.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 107

Software in India: globalization and the division of labour

Table 5.2C Parameter

107

Cross-sectional variation in firm characteristics – output Foreign firms

Domestic firms

No. of companies 138 153 Revenues (Rs.mn.) 2000–01 809 (1507) 1498 (3957) [Q1, Median, Q3] [90, 200, 726] [100, 235, 825] 1999–00 483 (822) 937 (2486) 1998–99 316 (508) 650 (1866) 1997–98 199 (285) 446 (1251) Exports (Rs.mn.) 2000–01 741 (1351) 1258 (3733) [Q1, Median, Q3] [78, 150, 647] [57, 135, 611] 1999–00 423 (756) 753 (2238) 1998–99 304 (469) 602 (1784) 1997–98 200 (261) 424 (1180) Long run gr. rate 85.8 (224) 103.5 (151) [Q1, Median, Q3] [26, 47, 85] [39, 62, 105] Productivity 2.16 (1.60) 2.86 (3.5) [Q1, Median, Q3] [1.2, 1.8, 2.5] [1.2, 1.9, 3.0] Exp. intensity, 99–00 88.2 (19) 65.2 (35) [Q1, Median, Q3] [86, 98, 100] [33, 81, 96] Market share 2000–01 0.21 (0.4) 0.40 (1.0) [Q1, Median, Q3] [0.02, 0.05, 0.21] [0.02, 0.06, 0.18] 1999–00 0.20 (0.3) 0.38 (1.0) 1998–99 0.20 (0.3) 0.41 (1.2) 1997–98 0.20 (0.3) 0.44 (1.2)

Dom. firms having foreign subsidiaries 59 2348 (4266) [150, 603, 2457] 1537 (2581) 1029 (1769) 650 (1144) 1921 (4026) [80, 390, 1256] 1168 (2336) 817 (1581) 518 (1009) 79.8 (70) [39, 62, 107] 2.88 (2.5) [1.5, 2.2, 3.1] 71.4 (32) [51, 85, 97] 0.62 (1.1) [0.03, 0.14, 0.68] 0.63 (1.1) 0.65 (1.1) 0.65 (1.1)

Notes: Figures reported are sample averages (standard deviations in first brackets). Quartiles and median are reported in square brackets.

The kernal density estimates of the output variables (particularly those relating to log-revenues, long-term growth rate and productivity) for domestic companies with foreign subsidiaries reflect a comparatively thicker upper tail, and probably a hint of bimodality.13 This feature would appear to be the resultant of some of these firms tearing away from the mainstream and moving toward the high end of the market. The move up the value chain for these companies is made possible through participating actively in accumulating tacit knowledge essential for increasing returns, through local collective learning at more established hi-tech clusters around the world. If this were indeed true, this would be welcome news for the Indian software industry. How far this is indeed true will be clear in the near future.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 108

108

Country/area studies

Revenues (Rs.mn.), 2000–01 0.0020

Density

0.0016 0.0012 0.0008 0.0004 0.0000

0

Foreign cos.

500

1000

1500

2000

Domestic cos. without forgn. subs.

2500

3000

Domestic cos. with forgn. subs.

Density

Log revenues, 2000–01 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00

1

Foreign cos.

2

3

4

5

6

Domestic cos. without forgn. subs.

7

8

9

10

Domestic cos. with forgn. subs.

Market share, 2000–01 8.00

Density

6.00 4.00 2.00 0.00

0

Foreign cos.

1

2

3

Domestic cos. without forgn. subs.

4

5

6

Domestic cos. with forgn. subs.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 109

Software in India: globalization and the division of labour

109

Age (in years), 2001

Density

0.12 0.08 0.04 0.00

0

Foreign cos.

5

10

15

Domestic cos. without forgn. subs.

20

25

Domestic cos. with forgn. subs.

Growth rate, 1997–98 to 2000–01 0.010

Density

0.008 0.006 0.004 0.002 0.000

1

Foreign cos.

50

100

150

Domestic cos. without forgn. subs.

200

Domestic cos. with forgn. subs.

Productivity (Rs.mn./Employee) 0.40

Density

0.30 0.20 0.10 0.00

0

Foreign cos.

1

2

3

Domestic cos. without forgn. subs.

4

5

6

7

Domestic cos. with forgn. subs.

Figure 5.2 Kernal density estimates: foreign firms and domestic firms with and without foreign subsidiaries

Kagami 01 chap 01

2/15/04

2:55 PM

110

4.

Page 110

Country/area studies

CONCLUSIONS

Has the rise of IT as a general purpose technology afforded more equal opportunities in development to developing countries such as India, at least those with large pools of human capital? We conclude with two points. First, as Sutton (op. cit.) points out, one of the most important handicaps that developing country firms face lies in the costs they face due to unhealthy regulation and corruption. Inefficiencies in infrastructure (power supply, communications) add to these. In most developing country industries, where firms draw inputs from other domestic firms, these costs are too high to be internationally competitive. One advantage of software as a sector is the virtual absence of backward linkages, making it possible to maintain the software industry as ‘an island of competitiveness’ notwithstanding the rest of the economy (Ghemawat and Patibandla (1997)). Secondly, the central question in a view of development that places technology and knowledge centre-stage is whether firms in poor countries have a route available to them to increase their learning capabilities and become dynamically efficient. In all production technologies there are natural hierarchies of capabilities, ranging from the simpler to the more difficult. Even in industries where international co-production is logistically feasible, the division of international labour specializes difficult, skill-intensive activities to high-wage workers in advanced country firms, and easier routine tasks to firms and workers in developing countries. The question is whether developing country firms can promote themselves in the value chain by leveraging their relationships with international high value firms. In the context of the IT industry, where the nature of the product admits, more than in any other, conditions for distributed production relationships between firms from developed and developing countries, there is a greater potential for firms to learn through continuous international interaction. But it is clear that the learning process bears more fruit when developing country firms also pursue a strategy of international location: we have presented evidence of the returns to Indian firms from promoting production subsidiaries abroad. The causality in this loop can run either way and both: firms that locate production abroad may gain in productivity; the most productive firms may be the ones to locate production abroad. Either way, these results underscore the importance and the rewards, in a knowledge intensive industry, of engaging in learning in the most proactive way.

NOTES 1. OECD (1997) reports that average employee compensation in Asia in 1993 was only half that in Europe (and only a third or a quarter that of Europe in certain countries), and labour

Kagami 01 chap 01

2/15/04

2:55 PM

Page 111

Software in India: globalization and the division of labour

2.

3. 4.

5.

6. 7. 8. 9. 10.

11. 12.

13.

111

productivity in Asia was 73 per cent that in Europe. On the other hand, the average annual growth rate was almost thrice as high as in Europe, and while in the early 1990s, profit ratios in Europe turned negative, the IT industry in Asia was registering moderate profitability around the same time. Ghemawat and Patibandla ((1997), (1999)) offer the hypothesis that prospering hi-tech industries in LDCs that are international trade-oriented contribute more to the domestic economy compared to inward-looking industry clusters. On the other hand there is also the problem of misallocation of human capital from other sectors of the economy (Balasubramanyam and Balasubramanyam (1997), (1998)). The argument in this section draws upon Sutton (2001). Sutton (op. cit.) points to a ‘low capability’ trap and draws on the example of the Indian machine tools market, where most users find CNC machine tools uneconomical given the wage levels. Only if the general level of industrial development advances (with corresponding increase in wages, and costs of other factors of production, as well as incomes), will demand shift toward ‘second generation’ varieties. Producers may get trapped with obsolete technology unless they invest in new technology ahead of demand. Sutton (op. cit.) distinguishes between two types of markets. In ‘high tech’ industries, competition in ‘capability-building’ will shake out all but a limited number of competitors. The key driver of competition is the degree to which products associated with different technical trajectories are good substitutes in demand. If the effectiveness of firms’ investments is high, the number of viable firms in a competitive market will be small. Consequently, if entry in such an industry is high, not all firms will survive. This is the type of market where across all technical trajectories, or sub-markets, demand is substantially based on either a single criterion, or only a limited set of criteria (for example, CPUs in computers). A variant of such a market structure arises when the trajectories are linked strongly, not in terms of product substitutability in demand as in the above case, but from the supply side. Here, there are significant scope economies in capability-building (advances in capability in one trajectory automatically enhance capability along another trajectory). This is particularly true of the information technology industry, the choice of technical trajectories (activities/strategies) in which, and the resultant impact on performance, will be analysed in this chapter. For example, the operating system for the ‘network computer’ introduced by ORACLE is said to have been designed entirely in India (OECD (2000)). Bajpai and Shastri (1998) provide a classification of software specialization according to levels of skill/tacit knowledge required. Our sources included the NASSCOM directory (NASSCOM (2000)), the India Infoline and Myiris investor information services, articles and special issues in several news dailies and magazines, as well as firm websites. Since the comparison is only within Indian companies, input endowments with respect to communication and other infrastructure may be more homogeneous. Software services have traditionally formed the backbone of the business (particularly export business) of Indian software firms. This trend could partly be due to the reluctance of overseas clients to have their software developed at far-away locations (D’Costa (2000)), and has been partly constrained by the availability of substantial capital required for software development (Arora et al. (2001)). In terms of revenues per employee, Arora et al. (2000) report that, while the Israeli and Irish software industries earn as much as US$100 000 per year per employee or more, firms in the Indian software industry earn only about US$15 000–$20 000. According to Patibandla et al. (2000), a part of this competitive advantage derived from higher productivity may be wiped off, in time, by increases in labour cost. Patibandla and Chandra (1998), and Patibandla et al., (op. cit.) feel that the road to the move up the value chain is to create appropriate employee management policies to further enhance productivity and reduce information asymmetries between employees and employer. Bimodality is often an indication that the population may be a mixture of two sub-populations, with well separated sub-population modes. If the sub-population modes are not sufficiently separated, the overall distribution can be unimodal, but may have a characteristically drawn out modal region.

Kagami 01 chap 01

2/15/04

2:55 PM

112

Page 112

Country/area studies

REFERENCES Arora, A., V.S. Arunachalam, J.M. Asundi and R.J. Fernandes (2000), The Globalization of Software: The Case of The Indian Software Industry, Report submitted to the Sloan Foundation, Software Industry Center, Heinz School of Engineering and Public Policy, Carnegie Mellon University, February. ——, ——, —— and —— (2001), ‘The Indian software services industry’, Research Policy, Vol. 30, 8, pp. 1267–87. —— and J.M. Asundi (1999), ‘Quality Certification and the Economics of Contract Software Development: A Study of the Indian Software Service Companies’, presented at the NBER Conference on ‘Organizational Change and Performance’ at Santa Rosa, CA, April 1999 and NBER working paper No. 7260, NBER, Cambridge, MA. Bajpai, N. and V. Shastri (1998), ‘Software Industry in India: A Case Study’, Development Discussion Paper No. 667, Harvard Institute for International Development, Harvard University, December. Balasubramanyam, A. and V.N. Balasubramanyam (1997), ‘Singer, Services and Software’, World Development, Vol. 25, 11, pp. 1857–61. Balasubramanyam, V.N. and A. Balasubramanyam (1998), ‘The Distribution of Gains between Investing and Borrowing Countries Revisited: The Case of India’s Computer Software Sector’, in D. Sapsford and J. Chen (eds), Development Economics and Policy: The Conference Volume to Celebrate the 85th Birthday of Professor Sir Hans Singer, Macmillan Press, London, pp. 287–99. Banerjee, A.V. and E. Duflo (2000), ‘Reputation Effects and the Limits of Contracting: A Study of the Indian Software Industry’, Quarterly Journal of Economics, Vol. 115, 3, pp. 989–1017. Cohendet, P., F. Kern, B. Mehmanpazir and F. Munier (1999), ‘Knowledge Coordination, Competence Creation and Integrated Networks in Globalized Firms’, Cambridge Journal of Economics, Vol. 23, 2, pp. 225–41. D’Costa, A.P. (2000), ‘Export Growth and Path-Dependence: The Locking-in of Innovations in the Software Industry’, Paper presented at the 4th International Conference on Technology Policy and Innovation at Paraná, Brazil, August. Fernandes, R.J., A. Arora and J.M. Asundi (2000), ‘Supply and Demand for Software Developers in India’, Working Paper, Software Industry Center, Heinz School of Engineering and Public Policy, Carnegie Mellon University, September. Forge, S. (1995), ‘The Consequences of Current Telecommunications Trends for the Competitiveness of Developing Countries’, Report to the World Bank, January. Fujita, M., P. Krugman and A.J. Venables (1999), The Spatial Economy: Cities, Regions, and International Trade, MIT Press, Cambridge, MA and London. Ghemawat, P. and M. Patibandla (1997), ‘India’s Exports Since the Reforms: Three Analytic Industry Studies’, mimeo, Harvard Business School, November. —— and —— (1999), ‘India’s Exports Since the Reforms’, in J. Sachs, A. Varshney and N. Bajpai (eds), India in the Era of Economic Reforms: A Political Economy, Oxford University Press, New Delhi. Heeks, R. (1996), ‘India’s Software Industry: State Policy, Liberalisation and Industrial Development’, Sage Publications, New Delhi. Hicks, D.M., P.A. Isard and B.R. Martin (1996), ‘A Morphology of Japanese and European Corporate Research Networks’, Research Policy, Vol. 25, pp. 359–78. Howells, J.R. (1995), ‘Going Global: The Use of ICT Networks in Research and Development’, Research Policy, Vol. 24, pp. 169–84.

Kagami 01 chap 01

2/15/04

2:55 PM

Page 113

Software in India: globalization and the division of labour

113

Kattuman, P. and A. Bhattacharjee (2001), ‘Firm Growth and Market Structure in Indian Software and Services Industry’, Paper presented at the Workshop on ‘The Indian Software Industry in a Global Context’ at Ahmedabad, India, August. —— and Iyer, K. (2001), ‘Human Capital in the Move Up the Value Chain: The Case of Software and Services Industry’, in M. Kagami and M. Tsuji (eds), The ‘IT’ Revolution and Developing Countries: Late-comer Advantage?, Institute of Developing Economics/JETRO, pp. 208–27. Lakha, S. (1994), ‘The New International Division of Labor and the Indian Computer Software Industry’, Modern Asian Studies, Vol. 28, 2, pp. 381–408. Lateef, A. (1997), ‘Linking Up with the Global Economy: A Case Study of the Bangalore Software Industry’, New Industrial Organization Program DP/96/1997, International Institute for Labor Studies. Maskell, P. and A. Malmberg (1999) ‘Localised Learning and Industrial Competitiveness’, Cambridge Journal of Economics, Vol. 23, 2, pp. 167–85. NASSCOM (2000), A Directory of Indian Software and Services Companies, National Association of Software and Service Companies (NASSCOM), New Delhi. OECD (1997), Information Technology Outlook, 1997, Organisation for Economic Cooperation and Development (OECD), Paris. —— (2000), Information Technology Outlook, 2000, Organisation for Economic Cooperation and Development (OECD), Paris. Patibandla, M. and P. Chandra (1998), ‘Organisational Practices and Employee Performance’, Journal of Economic Behaviour and Organsation, Vol. 37, pp. 431–42. ——, D. Kapur and B. Petersen (2000), ‘Import Substitution with Free Trade: Case of India’s Software Industry’, Economic and Political Weekly, 8 April 2000, pp. 1263–70. Royce, W.W. (1970), ‘Managing the Development of Large Software Systems: Concepts and Techniques’, WESCON Technical Papers, Vol. 14, p. 723. Shy, O. (2000), ‘Exporting as a Signal for Product Quality’, Economica, Vol. 67, pp. 79–90. Sutton, J. (2001), ‘Rich Trades, Scarce Capabilities: Industrial Development Revisited’, Discussion Paper No. EI/28, The Toyota Centre, Suntory and Toyota International Centres for Economics and Related Disciplines, London School of Economics and Political Science, September.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 114

6. Jumping up to the Internet-based society: lessons from South Korea Yasushi Ueki 1.

INTRODUCTION

Asian countries have already started making efforts to diffuse the Internet nationwide. But they are facing a lot of hurdles such as building infrastructure, deregulation of telecommunications, high price of equipment and systems, security and digital divide. Even if they could succeed in encouraging IT, rapid diffusion causes frictions economically and socially. This is mainly because all of the people and social systems cannot react and change themselves flexibly in accordance with rapid technological changes. In addition, IT will also change the sense of values especially among younger generations, which brings cultural problems and ethic conflicts between generations. Some of the troubles cannot be coped with using existing laws and dispute settlement processes. As a pioneer of broadband Internet and mobile phones, South Korea (hereafter Korea) has made efforts to encourage IT and overcome problems. The Korean government has shown initiatives to promote the introduction of IT. As a result of the efforts, Korea emerged as one of the most advanced countries in the world in the introduction of broadband Internet, especially Asymmetric Digital Subscriber Line (ADSL), and mobile phones. But some problems still remain and new social and economic issues are emerging. Although there are differences in preconditions between Korea and other developing countries for diffusion of IT, for example income levels and infrastructure, there are common factors and policy issues for developing countries. In this sense, other developing countries can garner lots of information to learn from Korea’s experiences. This chapter firstly reviews the process of rapid diffusion of IT and explains its effects on Korea. Then problems in the business area are mentioned. After that, social issues are made clear. Finally some implications for developing countries will be derived.

114

Kagami 02 chap 06

2/15/04

2:54 PM

Page 115

The Internet-based society: lessons from South Korea

115

2.

RAPID DIFFUSION OF NEW TECHNOLOGIES

2.1

Rapid Increase in Users of Internet and Broadband Internet

The number of Internet users in Korea has increased rapidly since the commercial Internet access service started in 1994. It increased from 138 thousand in 1994 to 1634 thousand in 1997, and the average annual growth rate was more than 100 per cent. Although a credit crunch happened in Korea in 1997, the number of users increased to 1634 thousand and the growth rate was 220 per cent in 1997. Although the number of users reached 3103 thousand the penetration rate for the Internet was still 6.7 per cent (Figure 6.1). The turning point was in 1999. The growth rate jumped to 250 per cent and penetration to 23.2 per cent in 1999. In 2000 the number of Internet users reached 19 million and penetration 40.5 per cent. Technologically, diffusion of broadband accelerated in 2000. The number of high speed Internet subscribers was 49.6 thousand in January 2000 and over four million in December. Categorizing by technologies, approximately 2.1 million households were subscribing to ADSL Internet and 1.4 million to cable modem Internet. The share of ADSL was above 50 per cent and cable modem was about 35 per cent (Table 6.1). Table 6.1

The number of high speed Internet subscribers (Unit: 1000)

Country/type

1999.12 2000.3 2000.6 2000.9 2000.12 2001.3 2001.6 2001.9 2001.12

Korea Total ADSL+CATV ADSL CATV Other Japan DSL+CATV DSL CATV

374 365 171 193 9 154 n.a. 154

864 847 546 301 16 216 0 216

1575 1243 679 564 332 330 1 329

2626 2142 1268 874 484 466 3 463

4017 3460 2074 1386 557 635 10 625

5095 4563 2806 1757 532 855 71 784

6251 5458 3505 1953 793 1258 291 967

7039 6437 4009 2428 602 1802 651 1151

7806 7176 4453 2723 630 2827 1524 1303

Source: MIC (Korea), Ministry of Public Management, Home Affairs, Posts and Telecommunications (Japan).

2.2

Preceding in the Introduction of New Technologies

In addition to broadband Internet, Korea is preceding in the introduction of new information technologies. A variety of contents and applications encourage the usage of broadband although the relation between the two is a chicken-andegg problem.

116

1994.12

138 1995.12

366 1996.12

731

The number of Internet users in Korea

KRNIC.

Figure 6.1

Source:

0

5000

10 000

1997.12

1634

1998.12

3103

1999.12

10 860

2000.12

19 040

2001.12

24 380

2:54 PM

15 000

20 000

25 000

30 000

2/15/04

(Unit: 1000)

Kagami 02 chap 06 Page 116

Kagami 02 chap 06

2/15/04

2:54 PM

Page 117

The Internet-based society: lessons from South Korea

117

2.2.1 Mobile Internet The number of wireless Internet subscribers was over 15 million as of December 2000. Although the number of users is not as high as Japan, Korea started providing the service early and the number of WAP/ME users is growing rapidly (Table 6.2). Table 6.2

The number of mobile phone-based Internet users (Unit: 1000)

Type ISMS WAP/ME Total Japan

2000.3

2000.9

2000.12

2001.9

2001.12

3121 1104 4225 7499

6882 5311 12 193 19 679

6965 8820 15 785 26 866

6508 15 924 22 432 44 937

5684 18 190 23 874 48 495

Source: MIC (Korea), Ministry of Public Management, Home Affairs, Posts and Telecommunications (Japan).

LG Telecom was a pioneer of the mobile Internet service. This company started mobile Internet service in May 1999. It is said that this was the first commercialization of the 2.5G (cdma 2000 1x) service. LG Telecom is also the first provider of WAP in Korea and mobile JAVA in the world (National Computerization Agency (NCA) and Ministry of Information and Communication (MIC) 2001). 2.2.2 Streaming and webcasting Broad diffusion of broadband Internet enhanced needs for streaming and webcasting that resulted in an increase of contents providers. MIC (2001) said that there are about 30 companies in Korea that provide paid Internet movie services. There was a remarkable increase in webcasting stations in 2000. The number of webcasting stations, which was about 200 as of January 2000, reached about 1000 by the end of 2000. In addition to this, lots of illegal copies, even of brand new content, can be found on websites. These contents-rich situations promoted the use of broadband Internet. 2.2.3 VoIP International Telecommunication Union (ITU) (2001) indicates that 50 per cent of the Voice-over-IP (VoIP, Internet phone) market is occupied by five VoIP carriers: Serome is top with a 22 per cent share, followed by ITCX (11 per cent), Net2Phone (7 per cent), iBasis (6 per cent) and Deltathree.com (4 per cent).

Kagami 02 chap 06

118

2/15/04

2:54 PM

Page 118

Country/area studies

Serome was the first VoIP service provider in Korea. Serome launched its VoIP service named Dialpad in October 1999 in the US and in January 2000 in Korea. Serome is the largest VoIP service provider in Korea with about five million subscribers in 2001. Web2Phone and other companies followed Serome. Web2Phone was established in November 1999, launched the Wowcall pilot service in February 2000 and formally opened Wowcall in July 2000. The number of subscribers had reached over 2.2 million by the middle of 2001. The uniqueness of the Web2Phone is its payment system. Subscribers can use the services from the company’s website that is like a portal that is providing such services as web phone, video chatting and shopping mall. Subscribers need to pay by a coupon named Wow to use the VoIP service. They can buy it by paying 10 000 won for 2200 Wow (4.5 won/Wow). The fee for domestic calling in Korea nationwide is 10 Wow/minute that is 45 won and the fee for international calls from Korea to Japan is 30 Wow/minute that is 135 won. Subscribers can also get the coupon by clicking a banner advertisement or by buying something via the shopping mall linked with the website. That means that 10 Wow/minute for domestic calling is the highest price and subscribers can call free.

3.

ACCELERATORS OF RAPID PENETRATION OF NEW TECHNOLOGIES

As observed above, the Internet spread across Korea in a short time. Korea Network Information Centre (KRNIC) (2000) analysed factors in the sharp increase in Internet users from four aspects: social, government policy, technical and business (Table 6.3). By analysing information from interviews conducted in Korea and published articles, characteristic aspects for explaining the rapid diffusion of new technologies may be the deregulation of the telecommunication sector, a significant government role and the Korean mentality. 3.1

Deregulation of the Telecommunication Sector

Multinational trade negotiations in the Uruguay Round and WTO in the 1990s strongly affected the stance of the Korean government on regulations on the basic telecommunications sector (Lee and Lie (2000)). The government transformed the regulations into more pro-competition ones under agreements of multinational negotiations and even autonomously. The government has eased regulations for entry and pricing conditions, and foreign ownership limitations.1 Two significant changes in the regulatory regime happened in 1997. One was a change in the classification of service providers in accordance with the

Kagami 02 chap 06

2/15/04

2:54 PM

Page 119

The Internet-based society: lessons from South Korea

Table 6.3

119

Factors in Internet diffusion in Korea

Aspect

Change

Social aspects

Change in social perception Geopolitical characteristics Cultural characteristics Government policy aspects Policy change in service charge Internet promotion policies Change in education policies Nationwide promotion Start-up company support Openness of Internet policies Technical aspects Enhanced Internet Infrastructure High speed national information infrastructure project Intensive technical support to IT industry Business aspects Investment on Internet infrastructure Emergence of SOHO Change in perception of the IT industry Increase in IT businesses Source:

KRNIC (2000).

amendment of the Telecommunication Business Act. According to commitments made in the WTO agreement on basic telecommunications, the government newly introduced a category of special service provider that is a leased line-based service provider and clarified the positions of new services including VoIP for regulation. Broadband Internet was classified as a valueadded service, where entry and pricing were deregulated. This enabled entry into new markets and promoted competition in services. The second was the introduction of competition in local call services that had been exclusively dominated by Korea Telecom (KT). As a result, Hanaro Telecom, the pioneer of ADSL in Korea, was permitted to enter the local call market. Hanaro launched local call and high speed Internet commercial services in April 1999. It is said that Hanaro started its ADSL business for differentiation of service. KT followed Hanaro. As a pioneer in ADSL, Hanaro acquired 156.1 thousand subscribers and 91 per cent of the ADSL market by December 1999. KT had only 7.3 thousand subscribers and 4.3 per cent. In December 2000, the market share of KT overtook Hanaro’s although both companies increased the number of subscribers (KT: 1322.5 thousand subscribers and 63.8 per cent, Hanaro: 651.8 thousand and 31.4 per cent).2

Kagami 02 chap 06

120

2/15/04

2:54 PM

Page 120

Country/area studies

One of the reasons for the reverse seems to be the absence of an unbundling rule for other facility-based service providers (FSPs).3 KT could use its own local loop network to provide ADSL nationwide. In addition the distance issue in ADSL does not apply to KT because more than 90 per cent of total households are located within a four-kilometre radius of its branch offices (UBS Warburg (2001)).4 But Hanaro invested in local loop, or leased networks from Powercomm, a network operator and subsidiary of Korea Electric Power Corporation. However, the cost and time seemed to be a burden to Hanaro. 3.2

Significance of the Government’s Role and Support

Historically, the Korean government has taken initiatives in promoting industrial and technological policies. The Korean government promoted policies to foster industries in the old economy such as steel, shipbuilding, petrochemical, home appliances, automobile and semiconductors. The government selected target technologies or sectors and implemented industrial policies to promote exports or substitute domestic products for imports. This is true also for information technologies. Policy priorities shifted from building telecommunication infrastructure to creating the environment for R&D and new businesses. 3.2.1 Building infrastructures based on national plans The government promoted information technology policies from the 1980s. The Korean backbone computer network project (1987–1992) was launched to facilitate the use of PCs. That was followed by the Korean information infrastructure (KII) project to build a high speed network. In 1995 the government allocated a budget to the KII project and the completion date was set at 2010 but later the schedule for the project was moved up. In addition to network infrastructure, after it launched Cyber Korea 21 in March 1999, a comprehensive master plan aimed at the transition of Korea to a creative knowledge-based nation, the government gave more priority to development of human resources and input resources into projects related to education and the digital divide (MIC (1999), (2001)). These network and human resource infrastructures provided the foundations for rapid diffusion of the Internet after 1999. 3.2.2 Certification and evaluation of technologies Generally speaking, the Korean government plays a role in (1) analysis of trends, development and evaluation of policies for industrial technology; (2) support of development of industrial technology which includes contributions to funding projects for development of targeted technologies, evaluation and certification of venture industry, and (3) promotion of nurturing technology bases and technology diffusion.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 121

The Internet-based society: lessons from South Korea

121

In the research of information technologies, the Electronics and Telecommunications Research Institute (ETRI) under MIC has played a role. ETRI is a non-profit government-funded research organization that was established in 1976. The Institute has succeeded in the development of information technologies such as high density semiconductors and CDMA. In 2000, to develop core technologies for information and communication, the government selected six priority fields and two base fields, for which the government would provide financial support (Table 6.4). Table 6.4

Six priority goals and two base fields for technology development

Position

Strategic core technologies

6 Priority Next generation fields Internet Optical communication

2 Base fields

Source:

Fast routers, Internet protocols, electronic commerce, information protection, etc. Next generation switching technology, optical transmission technology, subscriber networks, LAN technology, etc. Digital broadcasting Integrated digital broadcasting systems, broadcasting signal processing technology, digital contents, etc. Wireless IMT-2000, B-WLL, very short-distance communication wireless communications, satellite communication technology, etc. Software Component software, information processing technology, 3-dimensional GIS technology, virtual reality, etc. Computer High-performance multimedia servers, portable information terminals, Linux, etc. Core parts IMT-2000 core parts, optical components, display elements, etc. Source bases Optical elements, human interface, nanotechnologies, etc.

MIC (2000).

3.2.3 Supporting venture businesses The government established funds to invest in venture business. The Ministry of Finance and Economy (MOFE) prepared an organization to evaluate and certificate technologies to make decisions regarding financing. After the latest recession in the US, the role of the government in financing was increased. Other governmental organizations, for example ETRI under MIC and the Korea

Kagami 02 chap 06

2/15/04

122

2:54 PM

Page 122

Country/area studies

Institute of Industrial Technology Evaluation and Planning (ITEP) have facilities to evaluate technologies. Evaluation of technologies by the governmental sector seems to have had an influence on companies’ R&D strategies, investment strategies by investors and decisions on standards from the supply side. This behaviour by the government is interesting because selection of technical standards is remitted to the private sector or decided through market competition, that is the so called de facto standard, and the contribution to such decisions by the government has decreased recently. 3.3

Mentality

3.3.1 Homogeneity and group consciousness UBS Warburg (op. cit.) analysed Korean homogenous culture as one of the key drivers that promoted the adoption of the Internet and mobile phones. Cho (2001) says that the Koreans see themselves as a single race living in a village and a country that strengthens their group consciousness compared with other races. This, he says, generates a mindset of ‘all are one’ and ‘I can do what he can do’, and creates a sense of alienation in the case of differences. This mindset can also explain the popularity of playing online games in ‘PC-bangs’ (Internet cafés) with friends and community sites for alumni in Korea.5 Based on this Korean follow-the-leader mentality and given the fact that most of the population in urban areas live in apartments,6 coupled with an eagerness for education, there is a sense of unease at the thought of being left behind in the drive toward an information society. As a result, Internet literacy is increasingly perceived as a basic skill in the knowledge-based society. And these factors in turn have pushed people to be more responsive to IT. 3.3.2 Pari pari tendency The pari pari tendency, a Korean word that means quick-quick, has fuelled a social atmosphere that encourages learning about computers and the Internet (KRNIC (op. cit.)). This word is often heard in restaurants to urge servers to bring ordered dishes. This tendency seems to promote usage of IT, broadband and services (Cho (op. cit.)). Duration of viewing time on web pages also seems to reflect this pari pari tendency and is very short in the case of Korean Internet users. The average time spent on a page was 29 seconds compared with 37 seconds in Japan in January 2001.7 Therefore it is necessary for webmasters to put up attractive sites to hold Internet users (Cho (op. cit.)). In the case of B2C e-commerce, the first priority for a Korean is to get something as early as possible. Problems are left behind and countermeasures against them tend to be thought out and made after the event.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 123

The Internet-based society: lessons from South Korea

123

3.3.3 Job consciousness Changes in job consciousness especially among the younger generation also seem to be important. Before the 1997 crisis, the highroad to success socially was only by access to a job in government or a large company. But these sectors were forced to restructure after the crisis. On the other hand, entrepreneurs in the IT sector became success stories. In addition, the younger generation has rebelled against the traditional hierarchical human relations of the society and as a result more young people are starting new enterprises and entering into venture businesses.

4.

CONTRASTIVE EFFECTS OF ACCELERATORS ON E-COMMERCE8

4.1

Brief History of E-Commerce

Electronic commerce in Korea started in 1996. What is unique is that introduction of B2C preceded B2B. E-commerce and B2C were first introduced by Lotte department store and Interpark. Interpark started in 1995 as an in-house venture business of DACOM, a large communication company in Korea. B2B was stared in 1999 by DACOM. But most of the B2B service was introduced after 2000 (Korean Development Institute (KDI) (2001)). 4.2

Policy Framework for Promotion of E-Commerce

The policy framework to promote e-commerce is constituted of two parts: legal framework and a general plan. Korea preceded other Asian countries in enactment of e-commerce-related acts. The government enacted two fundamental acts in February 1999: the Basic Act on Electronic Commerce to promote the stable spread of e-commerce and the Act on Digital Signature to guarantee the security and reliability of e-commerce. The general plan for promoting e-commerce was established with initiatives by the Ministry of Information and Communication (MIC), the Ministry of Commerce, Industry and Energy (MOCIE) and the Ministry of Finance and Economy (MOFE). This constituted the first comprehensive policy by the government to promote e-commerce. It is composed of five main objectives and 40 action programmes. The five objectives are as follows: (1) to secure the reliability of the cyber market; (2) to expand e-commerce infrastructure; (3) to accelerate e-commerce in the public sector; (4) to encourage e-commerce in the industrial sector; and (5) to establish the basis for cyber trade (Table 6.5).

124

Source: NCA and MIC (2001).

Establish basis for cyber trade

Spread industrial e-commerce

2:54 PM

Promote e-commerce acceleration in public sector

2/15/04

Continuing expansion of e-commerce infrastructure

Detailed contents

Affirm identity of e-commerce user, prevent forgery of electronic paper, prevent abuse/misuse of personal data, protect e-commerce system from hacking Strengthen refund system, clarify the responsibility relationship between consumer and supplier, clarify the duty and right of parties concerned with e-payment, consolidate international e-commerce related acts Improve e-commerce business conditions Prepare tax support policy, abolish non-proper acts for e-commerce Early completion of Information Super Construct a high speed backbone network (1.5–2 Mbps) nationwide Highway Develop core technology by cooperation of government and private sectors (64.3 Promote technology development and billion won by 2002), constantly promote standardization (10.5 billion won by standardization 2002) reflecting ISO trends Train e-commerce experts Offer greater e-commerce related courses in universities, execute national certificate systems Construct e-commerce material basis Early completion of CALS/EC (2002), complete EDI system which is currently under test, by 2000, spread to 3000 companies from 2001 National defence sector Early start of constructing e-supply system of bid contracts, compose and operate ecommerce ‘regular conference’ in construction sector Construction sector Promote expansion from simple product purchasing to construction bid, reflect promotion results to management evaluation, consolidate acts and secure experts for spreading e-supply in government companies Government-run company Complete electronization of supply business during 2000, expand use of cyber mall in government purchase (from 500 to 5000 items), support electronization of selfsupply business in public organizations, establish government supply electronization environment Expand model project Complete e-commerce model projects in nine main industrial sectors by 2002, induce civil investment by tax deduction, increase government support (51.3 billion won by 2002) Compose e-CEO council Continuous promotion of B2B e-commerce boom Amend cyber trade support act Solve errors according to ‘Cyber Trade Map,’ electronize trade customs business (50% in 2000) Establish a unified buyer development Provide One-Click service, strengthen e-commerce cooperation with advanced system (Silk Road21) countries

Main contents

Consolidate acts and policies to Obtain e-commerce reliability increase the reliability of cyber market Strengthen consumer protection

Main object

Table 6.5 Summary of the general plan for activating e-commerce, February 2000

Kagami 02 chap 06 Page 124

Kagami 02 chap 06

2/15/04

2:54 PM

Page 125

The Internet-based society: lessons from South Korea

125

The Korean government aimed to advance the country into a fully-fledged e-commerce nation by 2003. To facilitate e-commerce in the industrial sectors, MOCIE promoted pilot projects of B2B e-commerce and contributed financially to establishing infrastructures for e-commerce (MOCIE (2000)). 4.3

B2C E-Commerce and the Accelerators

4.3.1 Market conditions In accordance with rapid diffusion of the Internet, the final two years of the 20th century were a turning point for B2C. In 1999, the size of the B2C market was 246.4 billion won (US$207 million), which was almost four times larger than 1998. In 2000, it was estimated to be 1397 billion won (US$1235 million), 5.7 times larger than 1999. The projected size in 2001 was 1801 billion won (NCA and MIC (op. cit.)). The number of Internet shops was over 1800 as of August 2000. B2C constituted 1.1 per cent of the total retail sales amount in August 2000 (MOCIE (op. cit.)). Based on well founded infrastructure, high penetration of PCs and constant access to the Internet, B2C is becoming familiar to Korean people. As of December 2000, there were 1866 shopping malls. Seeing the classification of them by legal status of companies, 46 per cent of them were run by individual enterprises and 52.6 per cent by corporate bodies. Most of the malls specialize in specific products and use online together with offline (MIC (2001), p. 142). Although the environment for B2C is improving, transactions through B2C are limited to standardized items which purchasers can get information on from catalogues or in stores and can experience part of the services before purchase. Computers and accessories (35.5 per cent) are the biggest items in terms of purchased size, followed by home appliances, electronics and communications (14.5 per cent) and books (6.4 per cent). In terms of the number of purchases, books (42.2 per cent) were largest, followed by computers and accessories (28.4 per cent) and music, videos and musical instruments (27.2 per cent) (NCA and MIC (op. cit.)). 4.3.2 Existing hurdles in offline Looking at complaints about Internet shopping, the ‘other, and – no difficulty’ as a response (38.2 per cent) occupies the top spot. The greatest concern was leakage of personal information (20.6 per cent), however this concern is not outstanding compared with other complaints. Other dissatisfactions were mainly problems related to offline procedures such as delivery, customer service and product quality (Figure 6.2). What seemed to be the biggest problem for the promotion of B2C is confidence, especially in payment procedures. Most of the Internet shoppers in Korea use payment methods such as online account transfers (50.5 per cent) and credit cards (46.7 per cent).

Kagami 02 chap 06

2/15/04

2:54 PM

Page 126

126

Country/area studies

Order procedure 3%

Price 1%

Difficulty of inquiry 4% Others/No difficulty 38%

Exchange/recall 8%

Product quality 12%

Delivery 13%

Leakage of personal information 21% Source:

NCA and MIC (2001).

Figure 6.2

Internet shopping difficulties

Although anxiety about personal information is not extremely high and credit cards are used for payment, the growth in credit fraud and theft is rapidly increasing. The total number of cases was 1161 as of the end of June 2001 but this figure is more than two times larger than the total of the previous year (479).9 A few Koreans I interviewed in 2001 explained these facts as the Korean pari pari tendency and consciousness of security. That is to say, Koreans give priority to getting something sooner over security. Each Korean also feels that he/she is exceptionally secure. They seem to think that they can depend on the police when facing troubles.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 127

The Internet-based society: lessons from South Korea

4.4

127

B2B E-Commerce and the Accelerators

4.4.1 Market conditions In 2000, the market size for domestic B2B e-commerce sales was 6.69 trillion won (US$6 billion), according to a report by the Korea Information Society Development Institute (KISDI). Fifty-four per cent of it was conducted via conventional Electronic Data Interchange (EDI) and 46 per cent via the Internet. The amount of e-procurement, procurement via B2B, was about 4.67 trillion won (US$4 billion). And 14.3 per cent of it was made via conventional EDI and 81.9 per cent via the Internet (NCA and MIC (op. cit.), p. 57).10 As of March 2000, there were 153 e-marketplaces in Korea. By classification by industrial sectors, general marketplaces and B2B business was top at 28, followed by machinery, information and communication (21), textiles (19), and petrochemicals (18). Classifying them by the nature of providers, offline dealers came top at 67, followed by joint ventures with IT and e-commerce providers and offline dealers (41) and B2B specialized service providers (40). After its start in 1999, results of B2B seem to be limited to some private marketplaces. A large company or Chaebol, a Korean conglomerate or company group like the Japanese Keiretsu, initiates this type of marketplace and participants in the market seem to be limited to group companies. Large companies such as Samsung Electronics, LG Electronics and Posco are promoting eprocurement. It is said that Samsung made 85 per cent of procurements via B2B that resulted in a reduction of indirect expenses by 250 billion won (US$221 million) in 2000. On the other hand, large parts of the public marketplace have not achieved results yet. These marketplaces are open to all companies and are provided by private service providers. Although about 200 B2B companies were born by 2001, less than 25 of them have trading records.11 4.4.2 Will accelerators turn into hurdles? Just after its startup Korean B2B is facing severe competition before attaining critical mass. Based on information obtained through interviews in Korea in 2001, the reasons why public marketplaces are facing difficulties are broadly classified by three aspects: (1) technological factors, (2) economic factors and (3) mental factors. 1. Technological factors Standardization of the classification and code of goods, electronic documents and catalogues has not been promoted recently. Some problems existed in the settlement system. Although the Korean government has taken initiatives to coordinate among companies to solve the problems, the processes do not seem to be working well. These technological

Kagami 02 chap 06

128

2/15/04

2:54 PM

Page 128

Country/area studies

problems must be solved to facilitate the increase in demand for B2B and technological solutions. 2. Economic factors This will be a more serious factor for SMEs. SMEs do not have enough resources for information systems. They are too expensive for SMEs to introduce. Expenses for communications are also a burden for SMEs, because it is necessary to introduce secure systems for business use. The government has started a programme to subsidize SMEs to introduce ERP. It plans to pay 20 million won to each firm as a subsidy, that is about 10 per cent of the introduction cost of ERP, to 30 thousand companies by 2003.12 The business structure of Korean companies is also making it difficult to bridge company groups and Chaebols in order to found consortia for ecommerce by sector, like Covisint in the US automobile industry. Companies in a business sector have similar structures. In the case of the electronic industry, Korean companies concentrate their business into DRAMs and LCDs, and depend on imported materials. Of course that could provide an environment for their cooperation in procurement of materials. But that results in more intensive competition which makes it difficult to realize complementary relations especially in the sales marketplace. This situation has its root in the government’s initiated industrial policy for fostering protection against foreign countries in the era of the traditional economy. Chaebols participated in sectors selected by the government. In the era of the new economy, the government has prioritized the fostering of venture businesses and implements policies for promoting IT businesses and openness to foreign companies. But it seems to be difficult to completely deny a possibility that the problems have resurfaced especially in sectors where large companies are main players. 3. Mental factors It is said Chaebols have closer ties with group companies than Japanese Keiretsu. Their trade partners are limited to group companies. Excessive competition between Chaebols is fuelled by the follow-the-leader mentality and the rivalry among them. This state of affairs was heavily criticized because bankruptcy of a Chaebol caused by excessive investments triggered the credit crisis. However this closed hierarchical network has not completely diminished after the financial crisis. Rivalry between Chaebols is too intensive for them to cooperate with each other to acquire mutual benefits. This makes it difficult to found consortia for e-commerce by sector. B2B is based on confidence, cooperation and sharing information between partners to cultivate benefits, therefore the current conditions in Korea are serious from this point of view. One respondent I interviewed said innovative reform is necessary for Korea to activate consortia for B2B. But this factor

Kagami 02 chap 06

2/15/04

2:54 PM

Page 129

The Internet-based society: lessons from South Korea

129

seems to be the most difficult hurdle to overcome, because there is no systematic countermeasure against it.

5.

SOCIAL ISSUES CAUSED BY DIFFUSION OF THE INTERNET

5.1

Emerging Issues

Although new technologies have spread over Korean society widely, preparation for problems caused by new technologies may not be enough. Actually several issues have been raised: 1. Distortions in personal life and society • Internet addiction: someone too absorbed by the Internet and online games to care about their daily life or mental health. • Pornographic and hate sites that cast aspersions on everything and everyone: both are thought of as reactions against the closeness of Confucia society. 2. Effects of input devices on language/culture • The young generation has innovated by shortening phrases or omitting a part of a Korean character to make it easier to input using the keyboard and mobile phone. This style of input promotes the use of Hangeul, the current characters used by Koreans and less use of traditional Chinese characters – documents written even ten years ago are set to become ‘classical’ for younger generations. 3. High tech crimes • Leakage of personal information, privacy, mobile phone call records, theft of credit card numbers and so on.13 • Piracy: Korea was notorious for the piracy of brand-named watches, bags, and so on, and software. Today illegal software, even the latest up-to-date, can easily be downloaded at warez14 sites, where illegal software, music files, pornographic contents and so on are uploaded. • Hacking and computer viruses

Kagami 02 chap 06

2/15/04

2:54 PM

Page 130

130

Country/area studies

Although some of the negative side effects of the Internet and new technologies are new, other issues raised are long-standing ones that result from traditional society. The Internet can show even tacit, latent or underground issues explicitly. 5.2

Countermeasures: the Case of Cyber Terrorism

As a result of jumping up from dial-up access to the Internet to a constant access environment, the numbers of reported hacking and virus incidents and arrests are also exploding. The government has established frameworks against these problems within the Korean National Police Agency (KNPA). 5.2.1 Rapid increase of the number of cases The number of reported hacking and virus incidents started exploding from 1999. That timing corresponded to the rapid diffusion of ADSL. The number of hacking cases for 1999 was three times as large as the previous year. The total number of those arrested on cyber crime charges also jumped from 466 in 1998 to 2089 in 1999. This seems to reflect an increase in damage done as well as reflecting the expansion of the police organization responsible for such investigation mentioned in the next section. In 2001, the increase in arrests accelerated. The results for the first eight months in 2001 already exceed the total for 2000 (Table 6.6). Table 6.6 Year

Cyber crime: statistics of arrests Total Sub-total

1997 1998 1999 2000 2001.1–8 Source:

141 466 2089 2190 2724

Major cyber crimes Other cyber Hacking Virus crimes

6 21 26 363 635

6 16 23 360 624

0 5 3 3 8

135 445 2063 1827 2089

Korean National Police Agency.

Among major cyber crimes, the number of hacking cases exploded in 2000. The number increased from 23 in 1999 to 360 in 2000, or about 16 times that of the previous year (Table 6.7). The age distribution of those arrested for hacking between January and April 2001 shows that more than half were teenagers. During those months the

Kagami 02 chap 06

2/15/04

2:54 PM

Page 131

The Internet-based society: lessons from South Korea

131

number of arrests was 397 (teenagers 225 or 56.7 per cent, twenties 116 or 29.2 per cent and thirties 17 or 4.3 per cent).15 Table 6.7 Type Hacking Virus Source:

The number of reported hacking and virus incidents 1996

1997

1998

1999

2000

147 226

64 256

158 276

572 379

1943 572

NCA and MIC (2001).

5.2.2 Organizational framework for KNPA investigation The Hacker Investigation Squad was set up by the KNPA in 1995 for countermeasures against cyber terrorism. Afterward, the organization was expanded and renamed the Computer Crime Investigation Squad in 1997 and Cyber Crime Investigation Squad in 1999. In July 2000, the Cyber Crime Investigation Squad was greatly expanded into the Cyber Terror Response Centre in order to prevent and respond to cyber terrorism and all kinds of cyber crime more effectively. The organizational structure is constituted by four teams: (1) the cooperation team for planning, international cooperation and collaboration with private sectors; (2) the monitoring and alert team for patrol over cyber space, provision of consultation and issuance of warnings; (3) the investigation team for investigating major cyber crimes; and (4) the development team for R&D into countermeasures. This centre is composed of 70 staff of the KNPA and 83 personnel in local police headquarters. In addition to them, two personnel are arrayed in each of 230 police offices. Out of a total personnel of 460 staff, 20 staff work full-time on cyber crimes and the rest are engaged in both cyber and general crime work. Officers in the KNPA include experts specially recruited for computer crimes. Employment contracts are both long- and short-term. Moreover this offers the police officers opportunities to learn about various techniques both to analyse and respond to criminal acts. The Cyber Terror Response Centre makes much of education and efforts to respond to technological progress. 5.2.3 International cyber terrorism and framework for investigation Cyber terrorism and crime happen worldwide because there are no national borders in cyber space. One example that happened recently was attacks on Japanese organizations by Korean hackers. This was triggered by the Japanese Education Ministry’s screening of a new history textbook. Sending massive amounts of e-mail and web traffic to a specific site, or e-mail bombs in a denial

Kagami 02 chap 06

132

2/15/04

2:54 PM

Page 132

Country/area studies

of service (DoS) attack by Korean hackers crashed several Japanese organizations’ websites (National Infrastructure Protection Centre (NIPC) (2001)). In order to cope with cyber terrorism and crime that occur over national borders some international framework for investigation is necessary. The KNPA can investigate all crime whether the servers are located in Korea or other countries but in the case of suspects that are living overseas the KNPA needs to depend on international cooperation. There are two routes to ask for cooperation. One is the diplomatic channel, the other via Interpol. Generally the Interpol route is used because it takes three months to start investigations through diplomatic channels. Three months has a critical meaning because small businesses often do not keep logs for more than three months. There are also cases which require the help of the Japanese NPA to utilize its network with police organizations in nine Asian countries. The Cyber Terror Response Centre also has close relationships with the US’s FBI, CIA, and so on. Problems in international cooperation for investigation into cross-border crime are not only about swiftness but also reflect large differences in abilities and techniques among countries. To cope with these problems, it is necessary to deepen cooperation and the exchange of information on techniques for crime investigation between police organizations.

6.

CONCLUDING REMARKS

Technology sometimes dramatically evolves and can seem to be out of control. But the new technology will diffuse on the basis of existing physical and social infrastructure that includes social, political and economic systems, culture and ethics. Therefore the more receptive the infrastructure is to new technology, the faster the speed of diffusion in a country. That is why this chapter focused on government policies and the Korean mentality. From the case of e-commerce, these factors become the accelerator and brake. The pace of diffusion of B2C can be influenced by the mindset of users rather than business culture. But in the case of B2B corporate culture, previous relations between companies will influence penetration. What complicates the problem in developing countries is that previous and ongoing industrial policies also influence B2B. Government initiatives for fostering specific industries can bear similar business structures among corporate groups, which will result in excess competition and make cooperation difficult between them in the case of a closed domestic market. On the other hand, there are merits in building information networks within a corporate group. The key that will decide the direction of B2B development is the openness of the offline and online market.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 133

The Internet-based society: lessons from South Korea

133

But diffusion of new technologies reveals problems and social frictions that underlie a society and the economy, such as freedom of speech, sexual representation, the generation gap and the realization of history. Especially in developing countries, there will be issues suppressed for political reasons. All of these issues can be opened up on a website with substantial effect. These facts and problems mean that new technologies are restricted by existing social and economic systems, sense of values, ethics, and so on. For example the Internet is technologically fusing all media including scripts (newspaper and magazine), sound (radio and music) and image (picture, TV and movie) into IP (Internet protocol). But each society will require ways to match the character of each medium with its sense of social values. The open nature of the Internet makes it easy to expose and get access to existing issues that may be controversial, damaging or illegal in a given society. To cope with such situations in a borderless environment international cooperation is needed. A society that lays out conditions for deployment will develop by introducing IT, however a society that does not will become confused. To this end the Internet poses a challenge to society.

NOTES 1. Lee and Lie (2000) and Lee (2000), survey the process of recent regulatory reforms in Korea. 2. KT: 3340.5 thousand subscribers and 75.0 per cent, Hanaro: 1011 thousand and 22.7 per cent (as of December 2001). 3. Local loop unbundling (LLU) was introduced in 2001. Lee (2001) insists that Korea pushed facilities-based competition at first, which promoted setting the price at a low level, deploying advanced networks and positively affected related equipment industry. 4. Quality of ADSL depends on the distance. The longer the distance is, the worse the quality is. Four-kilometre radius, which depends on the technological standard of ADSL, is enough for subscribers to access via ADSL and keep a level of quality. 5. I love school (iloveschool.co.kr) that opened in 1999 and had more than nine million members in 2001. 6. Some 40 per cent of Koreans live in apartments (KRNIC (2000)). High speed Internet service providers use optical fibres installed by FSPs to the kerb near apartments, which is called fibre to the kerb (FTTC). ADSL service providers use ADSL as access network from the kerb to each household, and CATV Internet use coaxial cable (hybrid fibre coax (HFC)). 7. Nielsen/NetRatings survey released on 1 March 2001. 8. Ueki (2001) tried to indicate current conditions of e-commerce in Asia and assess the effects on the electronic industry in Asia. 9. Joong Ang Ilbo (http://japanese.joins.com/), on 22 August 2001. 10. The figures seem to be dependent on the definition of e-commerce. According to survey results announced by Korea National Statistical Office (NSO) in June 2001, the size of the transaction through e-commerce in 2000 was 57.6 trillion won and occupied 4.5 per cent of the total amount of transactions. The size of B2B was 52.3 trillion won and B2C was 733.7 billion won. Some 72.1 per cent of the B2B sales were via EDI and 62.8 per cent of the B2B procurement was via the Internet (Joong Ang Ilbo on 4 June 2001). 11. Joong Ang Ilbo, on 22 February 2001. 12. Joong Ang Ilbo, on 3 August 2001.

Kagami 02 chap 06

2/15/04

2:54 PM

134

Page 134

Country/area studies

13. In addition to illegal leakage of personal information, the number of cases of provision of information by communications service companies for investigation organizations in the first half of 2001 was 127.3 thousand. That was 74.5 thousand in the first half of the previous year and the annual growth rate was 71 per cent (donga.com, on 21 August 2001). 14. Etymology of warez is wares in the word ‘softwares’. Warez means illegal copies of software that are circulated on the Internet. 15. Joong Ang Ilbo, on 21 August 2001.

REFERENCES Cho, Chang-Eun (2001), Kankoku Internet No Waza Wo Nusume (in Japanese), Ascii, Tokyo. International Telecommunication Union (ITU) (2001), ITU Internet Reports 2001: IP Telephony. Korea Development Institute (KDI) (2001), IT in Korea: Current Situation and Policy Direction, January. Korea Network Information Centre (KRNIC) (2000), Analysis of Factors Leading to Sharp Increase Internet User in Korea, July. Lee, Nae-Chan (2000), ‘Paradigm Shift in the Telecom Industry and Korea’s Regulatory Reform’, presented at the eleventh Korea Information Society Development Institute (KISDI) International Conference on Governing Regulatory Reform and Privatization of the Telecommunciation Industry, December, Seoul, Korea, http://www.kisdi.re.kr/advertise/pdf/1201–3.PDF. Lee, Nae-Chan and Lie, Han-Young (2000), ‘Korea’s Telecom Services Reform through Trade Negotiations’, presented at the eleventh Annual East Asian Seminar on Economics, June, Seoul, Korea, http://www.nber.org/~confer/2000/easexi00/lee.pdf. National Computerization Agency (NCA) and Ministry of Information and Communication (MIC) (2001), Korea Internet White Paper 2001. Ministry of Commerce, Industry and Energy (MOCIE) (2000), e-Commerce Environment in Korea: Market and Government Policies, October. Ministry of Commerce, Industry and Energy (MOCIE) (2001), Progress Report, April. Ministry of Information and Communication (MIC) (1999), Cyber Korea 21: An Informatization Vision for Constructing a Creative, Knowledge-Based Nation, March. Ministry of Information and Communication (MIC) (2000), White Paper 2000: The Informatization Vision for Constructing a Creative, Knowledge-Based Society (in Korean). Ministry of Information and Communication (MIC) (2001), National Informatization White Paper 2001 (in Korean). National Infrastructure Protection Center (NIPC) (2001), Cyber Protests: The Threat to the US Information Infrastructure, October. UBS Warburg (2001), Korean Broadband Sector Review, February. Ueki, Y. (2001), ‘Electronic Industry in Asia: Changing Supply Chain and the Effects’, in M. Kagami and M. Tsuji (eds), The ‘IT’ Revolution and Developing Countries: Late-comer Advantage? , Institute of Developing Economies JETRO, Tokyo, March.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 135

7. Information technology: some implications for Thailand Chanin Mephokee 1.

INTRODUCTION

The IT revolution has been progressing on a global scale with the rapid advancement of computer and communications technologies. The IT revolution, like the Industrial Revolution in the 18th century starting in England, has changed world society. While the Industrial Revolution changed agricultural society to industrial society, the IT revolution is going to change society into a ‘knowledge-emergent society’. Advancements in information technologies, primarily the Internet, will change the quality of information exchanges and revolutionize relationships between individuals, between individuals and organizations, and between individuals and society, by drastically reducing the costs and time for information distribution. Geographical differences will cease to be a burden for people to communicate as everyone in the world will be living under an identical IT network. Current IT innovations can be divided into those in information processing technology and those in information transmission technology. In terms of the former, the processing capacity of semiconductors has increased by more than 10 000 times over the last 30 years. Information processing technology innovation has opened the way for miniaturization, major price reductions and staggering advances in performance, causing an explosive global-scale expansion in information-related equipment hardware and software markets. In the case of information transmission technology, innovations allowing the high speed, low cost transmission of high volume processed information are also creating new possibilities. The spread of Internet technology and fibre optic-based WDM technology means that large amounts of information can be exchanged among a wide range of people, making possible multilateral business activities on a global scale. In historical prospective, the Internet has diffused at a faster rate than earlier generations of communications technology. From 1990 to early 2000, the estimated number of Internet users grew more than tenfold to roughly 300 135

Kagami 02 chap 06

136

2/15/04

2:54 PM

Page 136

Country/area studies

million, affecting the way in which people communicate with each other, acquire information, learn, do business, and interact culturally. In Asia alone, it was expected that by 2003 Asia’s B2B e-commerce would top US$272 billion and the number of Internet users will be 228 million by 2005. Thailand has always adopted a policy to open itself to the global economy. As a small developing country, Thailand has played two roles in the IT revolution era. First, Thailand plays the role of the IT consumer who adopts this new technology for both consumption and production, to improve the quality of life of the people and to improve productivity. Thailand also plays the role of an IT producer. The country has been successful in attracting foreign direct investment in IT industry. As an IT user, Thailand realizes that IT is an effective tool to raise the competitiveness of the country. According to the National Economic and Social Development Board (NESDB) and the World Bank, many Thai industries are losing competitiveness, compared to their competitors. Most Thai industries have much lower margins because of a lower level of skilled labour, lower product quality, and higher material costs due to higher reject rates and poor material yields. Once IT is introduced into the production process, it can reduce the cost of procurement, marketing and administration. In the supply chain, IT can lower the cost of procuring materials and reduce cash-to-cash cycle time. IT also increases efficiency in managing customer relations and raises customer satisfaction (demand chain). Thus introducing IT, including e-commerce and B2B e-commerce, into the Thai manufacturing sectors is one of the most pressing issues for maintaining Thai competitiveness. As an IT producer, Thailand is one of the IT production bases for IT-related MNCs. Since labour cost is no longer a Thai advantage, in order to lure FDI, Thailand has to upgrade its production process to a high technology process that requires higher labour skills. Therefore IT investment in Thailand would increase demand for skilled labour and decrease demand for unskilled labour. Of course a shortage of skilled labour will exist in some fields, while simultaneously unemployment will rise in some sectors. Approximately 200 000 people are working in IT-related industry in Thailand, which accounts for 5 per cent of total manufactured labour. Obviously, any change in employment patterns will affect them directly. Moreover workers as a whole will be affected indirectly due to IT. When IT is in use, manufacturing will be able to produce with less employment. This decline in employment is the unavoidable result of the IT revolution. On the other hand, rapid growth in new markets such as the services sector will create new jobs. This will change the labour patterns in job creation, job destruction, and job switching in the Thai labour market. This chapter examines the current status of IT use in Thailand. To understand the impacts of the IT revolution in Thailand, the chapter will investigate the

Kagami 02 chap 06

2/15/04

2:54 PM

Page 137

Information technology: some implications for Thailand

137

current status of IT in Thailand for both the IT user and the IT producer. Then the impacts on firms and employment will be discussed. Finally, the chapter will provide some policy issues for promoting IT-literacy in Thailand.

2.

CURRENT IT STATUS

2.1

Internet in Thailand

In Thailand the first Internet workshop started up at the Asian Institute of Technology (AIT) in 1991 on the formation of an informal Internet Special Interest Group. After that, in 1992, six academic and research institutions were online using the Internet on a full-time basis. The Thai Social/Scientific, Academic and Research Network (ThaiSARN) is set up to be the InterUniversity Network. In 1994, ThaiSARN reached 27 institutes at 34 sites. The Internet became commercialized in 1995. Internet Thailand or INETTH was the first that was granted an ISP (Internet service provider) licence by the Communications Authority of Thailand (CAT) in 1994 and started commercial operation in 1995. Internet Thailand is a joint venture between the National Science and Technology Development Agency (NSTDA), CAT, and the Telephone Organization of Thailand (TOT), which have shareholdings of 34 per cent, 33 per cent, and 33 per cent respectively. Internet Thailand commercial service is administered by the National Electronics and Computer Technology Center (NECTEC). Currently Thailand has 18 commercial ISPs, four non-commercial Internet hubs, and two domestic Internet exchanges operating nationwide with 642Mbps total international bandwidth (into Thailand) and 526.5 Mbps total international bandwidth (out from Thailand). There are 2 300 000 Internet users, 71 995 Internet hosts under .th (Thailand) top-level domain and 6 282 Internet domains under .th. Among 18 commercial ISPs, the three largest are the Internet Thailand Company, KSC Comnet and the Loxinfo Company. In order to be granted an ISP licence from CAT, the new entity must be a joint venture with CAT and CAT will get 35 per cent of the total equity for free (33 per cent to CAT and 2 per cent to CAT’s staff). The new entity must buy leased circuits to the Internet through or from CAT. CAT also reserves the right to send its personnel to work in the ISP and has the right to veto the decisions made by the board of directors. Moreover, CAT also set up guideline pricing for how much an ISP can charge its customers. CAT (National Internet Exchange: NIX) and NECTEC ( Internet Information Research: IIR) are the two domestic Internet exchanges in Thailand. Therefore the domestic ISPs do not have to link internationally. They have such

Kagami 02 chap 06

2/15/04

2:54 PM

Page 138

138

Country/area studies

characteristics as a low hop-count, fat pipe conducts to all major commercial ISPs and major academic/research/government hubs (Figure 7.1).

Source:

Figure 7.1

http://www.cat.net.th/iig.html

Internet connectivity in Thailand

CAT (International Internet Gateway: IIG) is the only international Internet gateway, connecting the domestic Internet network to the global Internet network. The total international bandwidth backbone is 202Mbps and the total international bandwidth peer to peer is 15 Mbps. The international Internet providers include AT&T (US), Dacom (Korea), Nacsis (Japan), NTT (Japan), Reach (Hong Kong), Singtel (Singapore), and Teleglobe (US), for example (Figure 7.2). Monthly charges by international gateway are shown in Table 7.1. 2.2 2.2.1

As the IT Consumer The use of IT in Thailand

(a) The number of computer users and Internet subscribers According to the Association of the Thai Computer Industry (ATCI), the number of personal

Kagami 02 chap 06

2/15/04

2:54 PM

Page 139

Information technology: some implications for Thailand

Source:

Figure 7.2 Table 7.1

http://www.cat.net.th/1024x782.html

Thailand international Internet gateway Monthly charge by international Internet gateway in US$

Speed (Kbps)

1 year

3 years

5 years

64 128 192 256 384 512 768 1024 1536 2 Mbps 4 Mbps 6 Mbps 8 Mbps 16 Mbps 34 Mbps 45 Mbps

1000 1280 1550 1800 2290 2780 3740 4720 6750 8920 14 580 19 990 24 000 43 190 81 010 95 000

900 1160 1390 1620 2060 2500 3360 4240 6080 8020 13 130 18 000 21 600 38 870 72 910 84 910

800 1030 1240 1440 1840 2220 2990 3770 5400 7130 11 660 15 990 19 210 34 560 64 810 75 470

Source: CAT (2001) http://www.cat.net.th.

139

Kagami 02 chap 06

2/15/04

140

2:54 PM

Page 140

Country/area studies

computers (PCs) sold has been increasing over time. In 1995, there were 271 300 units of PCs sold in Thailand and this number jumped to 479 300 units in 2000 which is a 76.6 per cent increase (Table 7.2). Table 7.2

Numbers of PCs sold in Thailand

Year

PC (unit)

Monitor (unit)

1995 1996 1997 1998 1999 2000 2001e

271 300 312 540 289 000 174 000 300 600 479 300 580 090

938 000 119 000 101 000 70 000 160 000 263 000 315 600

Source:

Association of the Thai Computer Industry (ATCI) Annual Report (2001).

The National Statistics Office (NSO) conducted a survey on IT used by households during January–March 2001. It was found that there are 812 565 households having and using computers, or 5.04 per cent of total households. However there were 927 875 units of PCs being used that accounted for 5.75 units per 100 households or 1.48 units per 100 people. This ratio is considered low because, according to the WTO, the PC ratio per 100 inhabitants is 18 for high income countries, 2.3 for medium income and 0.1 for low income. There were 2 277 046 households having access to the Internet, accounting for 14.11 per cent of total households in the whole kingdom. However, there were only 3.04 per cent of total households having Internet access at home, with the remaining 11.07 per cent using the Internet from other sources, such as schools, work places, and Internet shops. Some 3 536 001 persons only have ever logged onto the Internet which is 5.64 per cent of the total population. This penetration rate is relatively low compared to countries in the same region. The penetration rate in Hong Kong is 56.5 per cent while these rates are 54.3 per cent in Korea, 51.3 per cent in Taiwan, 51.2 per cent in Singapore, and 36.8 per cent in Japan The world’s highest penetration rate is the US which is 58.1 per cent. By this indicator, Thailand is far behind the world leading Internet users. Not only is there the low Internet population ratio, but Thailand also faces the problem of inappropriate use of the Internet. Most of the Internet population are in their 20s or teenagers and they mainly use the Internet for entertainment such as playing games or entering ‘chat rooms’, and use it less for education or getting information.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 141

Information technology: some implications for Thailand

141

The high price of IT equipment is the main explanation of this low use of the Internet. Therefore this low rate of IT utilization has left firms unable to prepare for IT network globalization. Moreover Thai workers have little opportunity to learn how to command information technology in order to fit into this new type of industry. (b) The use of e-commerce Due to the low penetration of PCs and the Internet, e-commerce in Thailand is still in the early stage. The National Electronic and Computer Technology Centre (NECTEC) in 1999 showed that more than 81.6 per cent of Internet users in Thailand have no experience in joining B2C e-commerce as buyers of online goods and services. For B2C ecommerce, by comparing the ratio of web servers with secured socket layer protocol to one million persons, it was found that in 1998, this ratio for Thailand was 0.42 units per one million persons, compared to 19.82 in Singapore and 1.94 in Malaysia. This shows that the development of B2C e-commerce is very low, compared to countries in the same region (Table 7.3). Table 7.3

B2C e-commerce used, 1998

Country Singapore Japan Hong Kong Taiwan South Korea Malaysia Thailand Philippines Indonesia Source:

B2C e-commerce per million persons 19.82 10.73 10.32 4.94 3.80 1.94 0.42 0.10 0.05

TDRI (1997).

2.2.2 Digital divide According to the ILO, World Employment Report 2001, technological change always favours the prepared. The different speeds of change and different stages of preparedness mean that there exist differences in IT use, the so called ‘digital divides’. IT infrastructure and activities are highly concentrated in a few countries, especially the US. Currently, approximately 85 per cent of world ecommerce websites are US-based. Within countries, the digital divide often has common characteristics. Use of the Internet is more common among younger than older people, men than women, the more rather than less educated,

Kagami 02 chap 06

2/15/04

2:54 PM

Page 142

142

Country/area studies

urban rather than rural dwellers, and those with higher incomes than those with lower incomes. The most significant factors are the level of education and the level of income. The ILO reports that only 6 per cent of the world population have access to the Internet, and 85 to 90 per cent of them are in the industrialized countries. There are also great divides among Asian countries. In terms of Internet host penetration, for example, Thailand is considered less wired than Hong Kong, Singapore, Japan, South Korea and Malaysia, but more advanced than the Philippines and Indonesia (Table 7.4). Table 7.4 Number of Internet hosts per 1000 population in some selected Asian countries, 1999 Countries

Number of Internet hosts per 1000 population

Hong Kong Singapore Japan South Korea Malaysia Thailand Philippines Indonesia Source:

66.4 22.19 16.6 6.03 2.8 0.49 0.23 0.18

Tangkitvanich (2001).

In Thailand, there also exists a digital divide among regions. According to the National Statistics Office (NSO), there are differences in numbers of households using PCs among regions. As expected, the households having PCs at home are concentrated in the municipal areas (78.9 per cent) and the rest, 21.1 per cent, are outside municipal areas. Inside the municipal areas, 12.27 per cent of total municipal households have PCs at home while only 1.57 per cent of total households living outside the municipal areas have PCs at home. Among households having PCs at home, there are 48.2 per cent living in Bangkok, where we find the lowest share of population (12.3 per cent) but the highest income level. The second highest share of households having PCs at home is in the Central region, accounting for 22.7 per cent. Only 1.93 per cent of the PC population is living in the North East region where we find the lowest household income level but the highest share of population (33.37 per cent) in the country (see Table 7.5).

Kagami 02 chap 06

2/15/04

2:54 PM

Page 143

Information technology: some implications for Thailand

Table 7.5

Percentages of households having PCs at home by regions, 2001

Region

Number of Percentage Per 100 Population households households

1. Whole Kingdom – Municipal area – Non-municipal 2. Bangkok 3. Central 4. North 5. North East 6. South Source:

143

812 565 640 852 171 713 391 188 184 593 88 015 103 064 45 705

100.0 78.9 21.1 48.2 22.7 10.8 12.7 5.6

5.04 12.27 1.57 19.77 5.07 2.77 1.93 2.16

100.0 32.5 67.5 12.3 22.6 18.0 33.7 13.4

National Statistics Office (2001b).

Internet use in Thailand as shown in Table 7.6 clearly highlights the existence of a digital divide among regions. In 2001, the NSO found that there were 2.277 million households with access to the Internet, 61.7 per cent of them living in the municipal area and only 38.3 per cent living outside the municipal area. There were 29.9 per cent of households having access to the Internet living in Bangkok, compared to 24.9 per cent residing in the Central region, 16.7 per cent living in the North, 16.9 per cent living in the North East, and 11.6 per cent living in the South. As a ratio of the Internet population to 100 persons, 16 per cent of the population in Bangkok have logged on to the Internet, while only 2.64 per cent of the population in the North East have access to the Internet. Table 7.6 Percentages of households having access to the Internet by region, 2001 Region

Number of households

1. Whole Kingdom – Municipal area – Non-municipal 2. Bangkok 3. Central 4. North 5. North East 6. South Source:

2 277 048 1 404 654 872 392 680 297 566 795 380 267 384 169 265 518

National Statistics Office (2001b).

Percentage 100.0 61.7 38.3 29.9 24.9 16.7 16.9 11.6

Per 100 population 5.64 11.50 2.82 16.00 5.85 4.57 2.64 4.72

Kagami 02 chap 06

2/15/04

2:54 PM

144

Page 144

Country/area studies

According to the NSO, the use of IT is concentrated in Bangkok, and the Central and South regions. The digital divide can be explained by the differences in monthly household incomes. By the NSO survey, in 2000, households living in Greater Bangkok (including Nonthaburi, Pathumthani, and Samut Prakran) had the highest incomes of 24 690 Baht per household or 7716 Baht per person. Households living in the Central region were the second highest in monthly income per household with 13 301 Baht or 3800 Baht per person. The Southern households were ranked third with monthly income of 11 407 Baht or 3002 Baht per person. The ranking in monthly household income is compatible with the ranking of IT use by households. Therefore it may be concluded that the inequality in income distribution among regions is one major factor creating regional digital divides in Thailand (see Table 7.7). Table 7.7

Average monthly income per household by district, 2000

Region Whole Kingdom Greater Bangkok Central North Northeast South Source:

Per household (Baht)

Per person (Baht)

12 167 24 690 13 301 8649 7853 11 407

3386 7716 3800 2544 2014 3002

National Statistics Office (2001a).

On the supply side, the inequality in telecommunication infrastructure among regions is also a factor in regional digital divides. The phones per 100 people ratio in Greater Bangkok are almost five times higher than the ratio in provinces. However after TOT authorized private companies to install telephone lines throughout the country, the telecommunication infrastructure in the provinces has been improved significantly (Table 7.8). In the same region digital divides can be found as well. The NECTEC in 1999 found the existence of digital divides among age groups and between genders. About 57.5 per cent of users were in their 20s, 22.5 per cent were in their 30s, and 11 per cent were teenagers. In terms of gender, it was found that 65 per cent of users were male. There also exist the digital divides between large firms and small firms. For Thai industry, the high cost of leased lines makes it difficult for small and medium sized enterprises (SMEs) to adopt an e-commerce model that requires an always-on connection. Concerning B2B e-commerce application, few SMEs can afford to use application software provided by application service providers

Kagami 02 chap 06

2/15/04

2:54 PM

Page 145

Information technology: some implications for Thailand

Table 7.8 Year

Telephones per 100 people ratio Whole Kingdom

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

5.45 6.41 7.50 8.67 9.99 11.43 12.95 14.55 16.46 16.45

Source:

145

Greater Bangkok

Provinces

26.70 30.77 35.00 39.05 43.10 46.66 49.66 52.45 54.52 55.05

1.66 2.29 2.62 3.45 4.26 5.31 6.53 6.02 9.77 11.76

TOT (2002), http://www.tot.or.th/th/prodnet/statusofservice/status.php.

(ASPs), which requires a costly high speed leased line to function smoothly. The high leased line price also hinders SMEs from connecting with suppliers and logistics operators, a prerequisite for managing supply chains. For the same reason, B2C e-commerce operators cannot provide a real-time response for an inquiry from a customer. Therefore there exists a difference in IT use between large firms and SMEs, and this difference leads to the difference in their performance (intra-industrial digital divides). It is believed that education level is the main factor in digital divides in Thailand as well. Unfortunately, there are no data to prove this belief. 2.3

As the IT Producer

2.3.1 Firms Thailand is one of the host countries for investing in the IT industry. Many multinational firms, especially Japanese and American firms, have been investing in Thailand since the 1960s. In the early 1960s, Japan’s Matsushita Corporation set up a company under the name ‘National Thai’ to produce radio and television sets for the domestic market. After that most Japanese home electronics appliance firms, such as Mitsubishi, Toshiba, Sanyo, and Hitachi, began direct investment in the latter half of the 1960s. In this period, the only non-Japanese foreign affiliated firm was the one set up by ITT to produce telephone receivers and electronics parts. The 1970s were the development stage for integrated circuits (ICs) in Thailand, characterized by the direct investment of major IC producers such

Kagami 02 chap 06

2/15/04

2:54 PM

146

Page 146

Country/area studies

as National Semiconductor, Signetics, and Data-General. In the early 1970s, the main purpose of FDI in Thailand changed to an export-oriented one. The 1980s were the period of expansion and diversification of the IT industry. A new Thailand investment policy encouraged whole foreign ownership for direct investment projects that were export-oriented. During 1982–1986, there were 27 investors with 13 950 million Baht invested, generating 18 776 jobs. Most of them were American and Japanese companies. Minebea, a Japanese bearing producer, established three wholly owned subsidiaries to produce minibearings, keyboards, plastic parts, and other items for export. Seagate Technology, an American firm, also started to produce some parts for its computer hard disk assembly plant. Fujikura, a Japanese firm, invested in production plants for keyboards, ICs and other electronics parts. AT&T Microelectronics (Thailand) started producing PCBs for export as well. Currently, the main IT products produced in Thailand are computer peripherals (such as monitors, printers, keyboards, HDDs and parts, and PCBs) and integrated circuits (ICs). Since the Thai market is relatively small, 80 per cent of products are for export. The main producers, all foreign companies, are ADI (Thailand) for monitors, Canon for printers, Fujitsu for printers and HDDs, Minebea for keyboards and HDD parts, Seagate for Head Gimbal Assembly (HGA) and Head Stack Assembly (HAS), and IBM for HDDs. As a result, for the past ten years, computers and parts and ICs have become the two most important export items for Thailand, accounting for approximately 18 per cent of total export value (Table 7.9). Table 7.9

Export share of Thailand’s five major export items (Unit: %)

Computers and parts ICs Vehicles and parts Garments Shrimp Total export ($billion) Source:

1997

1998

1999

2000

12.2 4.2 2.7 5.4 2.6 58.3

14.3 4.2 3.1 5.5 2.6 54.5

13.8 5.0 4.1 5.0 2.2 58.5

12.4 6.5 4.4 4.5 2.2 69.9

2001 (Jan–Sept) 11.9 5.5 4.8 4.5 3.3 49.3

Ministry of Commerce (2002), http://www.ops2.moc.go.th/trade/trade.html.

Normally, the production process in Thailand is an assembling process that requires all the important components from abroad and the products are low end products. For example, Canon used to produce mainly dot matrix printers while Fujitsu produces HDDs, mainly desktop drives. Both are low end

Kagami 02 chap 06

2/15/04

2:54 PM

Page 147

Information technology: some implications for Thailand

147

products. Therefore the technology used in the Thai IT industry is low technology and labour-intensive requiring less labour skills. Since the production process is on downstream industry and products are for export, there does not exist a forward and backward linkage in the Thai IT industry. Even some components that the foreign affiliates purchase domestically are produced by the foreign affiliates. For example, Nidec, the biggest producer of spindle motors, set up its facilities in 1990 and 50 per cent of its products are used domestically by other foreign affiliates. Local IT firms play a small sub-contracting role supplying minor components for foreign firms in the IT industry and there is no positive condition for technology transfer. However it is undeniable that for the past ten years the amount of FDI in the Thai IT industry has been increasing significantly from 11.56 billion Baht in 1990 to 63.09 billion Baht in 2000. These ten years of FDI in the IT industry have created more than 200 000 jobs for Thai workers (Table 7.10). Table 7.10

Summary of FDI in IT industry in Thailand, 1990–2000

Year

Investors

Investment (Million Baht)

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

49 30 56 44 74 138 149 117 116 111 185

11 563 6323 14 119 5776 13 159 40 982 68 712 66 372 53 100 21 723 63 091

Source:

Thai labor (persons) 8917 7770 19 220 14 001 21 009 75 417 55 932 46 872 45 971 55 431 73 646

Board of Investment (2001), http://www.boi.go.th/thai/business/statistics.html.

2.3.2 Workers There are 31.39 million employed persons working in Thailand. Among them 12.6 million work in the agricultural and fishing sector and five million work in the manufacturing sector, accounting for 40.14 per cent and 15.93 per cent of the workforce, respectively. By age, 40 per cent of employed persons or about 13.79 million persons are 40 years old or older and only 16.97 per cent of employed persons are below 25 years of age. Therefore, 40 per cent of Thai workers are considered too old to be reeducated in new technology. Furthermore, more than half of them attained an insufficient education to learn new technology. According to the Labour Force Survey

Kagami 02 chap 06

2/15/04

2:54 PM

148

Page 148

Country/area studies

2001, conducted by the NSO, 65 per cent of the workforce have elementary or lower formal education. This figure may fare better if we only consider those employed in IT-related industry, as they are considered to have one of the highest education levels among manufacturing worker groups. However Thai workers in general have low levels of education and lack basic knowledge in both English and mathematics (Table 7.11). Table 7.11

Workforce by level of educational attainment, 2001

Level of education Total None Less than Elementary Elementary Lower Secondary Upper Secondary Higher level Others Source:

Workforce (thousands)

Percentage

31 388.20 1089.70 12 432.80 7003.80 4036.30 3160.40 3576.80 88.40

100.00 3.47 39.61 22.31 12.86 10.07 11.40 0.28

National Statistics Office (2001c).

Similarly, more than 50 per cent of those unemployed only have a formal education of elementary or lower. For this group, with a poor education and no income, it is almost impossible to attain new skills without help from the government. And, needless to say, without new skills they cannot reenter the job market. They were the first to suffer from the 1997 Thai financial crisis and they will be the first to suffer from the IT revolution in the near future (Table 7.12). Table 7.12

Unemployed by level of educational attainment, 2001

Level of education

Unemployed (thousands)

Total None Less than Elementary Elementary Lower Secondary Upper Secondary Higher Level Others Source:

National Statistics Office (2001c).

1188.4 13.5 282.1 372.1 193.2 167.4 159.0 1.1

Percentage 100.000 1.135 23.738 31.311 16.257 14.086 13.379 0.093

Kagami 02 chap 06

2/15/04

2:54 PM

Page 149

Information technology: some implications for Thailand

149

Therefore the hurdles facing the Thai workforce are great, for not only do they not have enough basic education to learn new technology, but the lack of industry linkage, the high cost of IT, and the high mobility of job changes also conspire against them. Indeed, this lack of favourable conditions for Thai workers is a more serious problem than the low levels of education. 2.3.3 Telecommunications infrastructure Like most nations, Thailand’s telecommunications development is under the direction of the government. At present, telecommunications infrastructure development is carried out by three organizations: the Post and Telegraph Department (PTD), the Telephone Organization of Thailand (TOT) and the Communication Authority of Thailand (CAT). The PTD was first established in 1883. It is a government department headed by a director general and directed by the Ministry of Transport and Communication. PTD is primarily responsible for policy rules and regulations and at present is charged with managing and controlling radio frequencies and regulating and coordinating domestic communications via satellite through integrated ground stations. PTD is also responsible for studying the application of advanced telecommunication technologies and preparing proposals for government consideration. TOT, a state-owned enterprise under the Ministry of Transport and Communications, was established in 1954. Since then, TOT’s main objectives have been to operate and develop national telephone services and to carry out all business relating to telephone activities. TOT is responsible for domestic services, international services to Laos and Malaysia and leased circuits for domestic point-to-point transmission of voice, telegraph, radio and television. Recently TOT has introduced many new services, including special SPC exchange system services, toll-free call 088 services, NMT 470MHz and 900MHz mobile phone services, paging services and data transmission network services (DATANET). CAT, a state-owned enterprise under the Ministry of Transport and Communications, was established in 1976. CAT’s main objectives are to operate and improve the activities of the posts and telecommunication system under the CAT ACT, 1976. CAT is responsible for the postal, telegraph, telex, telephoto and facsimile services, domestic radio links, and international leased circuits. Recently CAT began providing international database access (IDAR) services, mobile phone 800MHz and paging services, and also offering circuits to telex subscribers to use teleprinters to send and receive domestic and international telegraph messages. It can be seen that both monopoly state enterprises, CAT and TOT, are not strongly delineated, especially in value-added services, such as mobile phone services. To prepare Thailand for the new era of IT revolution, the Telecom-

Kagami 02 chap 06

2/15/04

2:54 PM

150

Page 150

Country/area studies

munications Master Plan 1997–2006 has been proposed. One of the goals of the Master Plan is to increase the competitiveness of the industry by abolishing existing state enterprise monopolies and opening markets to competition under the regulation of an independent National Communications Committee (NCC). The NCC will regulate the pricing structure and the price levels in the markets. More importantly, both TOT and CAT have to be privatized and the companies’ shares will be registered for public offerings on the stock exchange. In the process of privatization, both TOT and CAT will be broken up into two separate entities; whereby two will compete in local and value-added services, while the other two will compete in long distance services (domestic and international). As it was international pressure, mainly from the US, that initiated the Plan, so far the Master Plan has not been implemented.

3.

IMPACT ON FIRMS AND EMPLOYMENT IN THAILAND

3.1

Impact on Firms

Though developed countries account for most IT production (about 77 per cent in 1995), developing countries have also increased their share through FDI. Developing countries produce 46 per cent of the world’s consumer electronics (Reed Electronics Research (1995), The Yearbook of World Electronics Data). Consumer electronics have been shifted to lower wage economies with a relatively skilled workforce, while the producer countries move on to concentrate on products with higher value added, such as software and sophisticated components. This shift first happened from Japan to Korea, and later from Korea to second-tier newly industrializing countries, such as Malaysia and Thailand. Currently, large emerging economies such as China and India have taken on much of the world production of electronic consumer goods. Therefore the IT revolution will shift some IT production from Thailand to other countries such as China. In order to lure FDI, Thailand has to increase her IT environment to be able to produce higher technology and medium skilled labour products, instead. With e-commerce growth in developing countries, one obvious sector to benefit is the IT industry, both hardware and software. On the other hand, this could mean a faster transfer of technology, but equally, it could also mean that existing IT multinational firms alone benefit if the local industry is weak. Some developing countries like Thailand have followed a policy of initially protecting local industry by encouraging joint ventures with multinationals while keeping the duty on component imports lower than on the import of computer equipment, thus encouraging local assembly and production. So far, however

Kagami 02 chap 06

2/15/04

2:54 PM

Page 151

Information technology: some implications for Thailand

151

the policy has not created strong industrial linkage in Thailand because of the weak domestic industry. IT can also be used as a tool in other industries. Here are at least four different channels through which e-commerce may impact on Thai firms. First, IT makes it easier for firms to access B2C world markets. Secondly, IT can facilitate activity on the global market for Thai traditional products, such as handicrafts and processed food. IT allows local firms to tap into the B2B (and B2C, may be) supply chains. Finally, IT allows service-providing firms to operate more efficiently and to provide certain services directly to customers anywhere in the world. In Thailand, with the low level of basic skills of workers and lack of industrial linkage, Thai industry underutilizes the benefit from IT globalization. However, there is some evidence proving IT can improve small business performance in Thailand. For example, one rural community producing processed food products could not sell its products, but after introducing e-commerce in its marketing strategy, can now sell products overseas. In Thailand, so many cases show that community businesses failed because of their isolation from markets. At the same time, wealthy consumers willing to buy those products at a high price could not find suppliers. This case shows that e-commerce is the most costeffective way to conduct marketing worldwide for SMEs in Thailand. Moreover, there is some evidence showing that SMEs can gain benefits from the Internet; unfortunately, the few successful cases documented show that the SMEs involved received strong support from NGOs. The Samnuk-Rakbankerd Foundation, supported by the Thai telecommunication company United Communication Industry (UCOM), promotes mutual cooperation among village enterprises. Recently the foundation set up a website called Rakbankerd.com as a cyber marketplace to benefit village enterprises. The foundation also helps village enterprises set up their own online facilities. Members not only offer their products for sale or search for intermediate products in this cyber market, but they also exchange product and technology information, including working experiences among members. For example, one village enterprise in the Northern province of Kampangpetch exchanged rice grain for seafood products from another village enterprise in the southern province of Ranong, without a middleman, through the Internet. Also through the Internet they have been able to send their members to learn how to process seafood products at a fishing cooperative in the Eastern province of Samut Songkram. In this way, they can sell products at higher prices and buy intermediate products at lower prices than before. Moreover, the technology exchanges can improve both their productivity and their managerial skills. These long-distance friendly relations between village enterprises would not be possible without IT.

Kagami 02 chap 06

152

2/15/04

2:54 PM

Page 152

Country/area studies

Not only are SMEs with a web presence more likely to be discovered than those without, but Internet savvy may also itself be a signal to potential customers of a certain level of technical and commercial sophistication. A typical global supply chain involves multiple transactions: if at any point an electronic interface is not present, then the time and cost advantages of ecommerce are dissipated. Thus, international companies are unlikely to go to the expense of dealing with a non-Internet-capable supplier unless there are significant other cost advantages in doing so. There are still some arguments saying that since SMEs in Thailand serve mainly local markets and rely mostly on locally generated information, the benefit from having the Internet may be less than the costs of having it. This may be true in the short term, but in the long term when all economies merge to the world market, local marketing alone will not be a win-win solution for SMEs. 3.2

Impact on Thai Labour

3.2.1 Change in structure of labour demand Global information networks have made it possible for different processes in the production chain to be linked worldwide, resulting in sharp declines in transaction costs and changes in employment structure. The introduction of IT to businesses leads to a rise in skill and educational requirements. Computers are taking on much of the work done by low skilled workers in many firms and jobs are disappearing. In Thai manufacturing industry, the need for blue collar workers to be more educated has grown with the rising intensity of capital formation. As many traditional production processes previously carried out by labour are displaced by machines, there has been a fall in employment of production workers. This has been accompanied by a redefining of much production work, as the introduction of new technology and new forms of work organization require production workers to be more skilled and knowledgeable. Here demand for skilled labour increases while demand for low skilled labour declines. Meanwhile the development of e-commerce promotes outsourcing, which again leads to more employment for office workers and less employment on production lines. IT is expected to continue to diffuse rapidly over the next decade. The fast and continuously changing character of IT implies that today’s skills may become obsolete tomorrow. It is not always that higher skills are required, but different skills definitely are. To keep up with these fast-changing skill requirements needs continuous retraining. 3.2.2 Less numbers of workers Use of the Internet is nevertheless associated with new patterns of job creation and job loss. IT replaces old tasks and occupations through automation, such

Kagami 02 chap 06

2/15/04

2:54 PM

Page 153

Information technology: some implications for Thailand

153

as the telephone switchboard operator and the information provider in travel agencies. But IT also creates new tasks and occupations, such as webpage designers and software programmers. It is difficult to say in advance the effect of IT on employment. However after adopting IT firms require less workers. For example, Amazon.com had only 614 employees for sales of US$148 million in 1998, or US$267 000 of revenue per employee, compared to the largest US bookstore, Barnes and Noble, which had a sales force of 27 200 for sales of US$2.8 billion, or US$103 000 of revenue per employee. Once an economy applies IT at full speed large numbers of workers may lose their jobs. Here again, the least educated and lowest skilled workers will be the first to go. Furthermore, for countries like Thailand where access to training in new technologies is difficult for poorly educated workers, another problem that arises is that multinationals may switch investments to other countries that have workforces more capable of using IT, and many Thai workers would lose their jobs. The first to lose their jobs will be older workers with low levels of education and training and working in traditional industries or SMEs. These workers are those who do not have the basic education needed to learn IT or who are in an IT-unfriendly environment. Therefore they face a two-pronged attack on their livelihoods; directly from the relocation of multinationals and indirectly from being replaced by new technology. 3.2.3 Jobs relocation One effect of the IT revolution is that working patterns are becoming independent of location and this will change management practices, the nature of the employment contract, and the quality of work. The IT revolution creates a relocation of jobs from industrialized countries to developing countries, such as back-office staff located in call centres, data entry processing, and software development. Work that is independent of location has a growing share of employment in industrialized countries. For example, almost a quarter of the workforce in the United Kingdom now carries out at least some of its work at home. By 2003 it was estimated there would be 1.3 million employed in call centres in the European Union, up from an estimated 670 000 in 2001. Moreover, up to 5 per cent of all service-sector jobs in industrialized countries could be relocated to developing countries. This would mean that about 12 million jobs could be moved to developing countries (ILO (2001)).

4.

SOME POLICY ISSUES

As mentioned above, IT generates benefits for those who are prepared and harms those who are not, both firms and workers. The main problem that

Kagami 02 chap 06

154

2/15/04

2:54 PM

Page 154

Country/area studies

Thailand faces is the lack of an IT-friendly environment. There are two main issues that Thailand needs to address immediately in order to create this environment, one is the competitive environment and the second is the IT-literate society. The second one is the policy aimed at nurturing high quality human resources for the global IT network. 4.1

Liberalizing the Telecommunication Market

In order to develop Thai workers to cope with the IT regime, liberalizing the telecommunication market is a necessary condition. Telecommunication is an indispensable infrastructure for e-commerce. In many developing countries, monopoly of the sector is the hurdle preventing countries from entering the IT globalized network. In Thailand, although the retail Internet access market is quite competitive with 15 companies operating as ISPs, the whole market for international access is still monopolized by CAT. This monopoly creates many problems concerning the adoption of IT in Thailand. First, the cost of Internet access in Thailand is significantly higher than in many other Asian countries. The higher cost leads to a lower number of Internet hosts, the computers connected to the Internet. Second, the state monopoly also imposes higher costs on users. CAT requires that every ISP hand over onethird of its user fee in return for the concession to operate. This leads to higher charges for IT users. A country’s readiness for e-commerce depends fundamentally on network infrastructure, including narrowband and broadband, and on costs for Internet access. The quality and range of services available depend on the emergence of innovative ISPs. Therefore, telecommunications reform has been a major determining factor in the emergence of the new global network economy (new economy). Faster and more reliable network infrastructure, associated with new ways of pricing, both for consumers and for the leased lines used in B2B transactions, have led to increased Internet connections to homes and businesses. Low access costs are an important factor driving uptake, while competition among infrastructure providers and among ISPs has led to innovative pricing structures. Telecommunications deregulation has been gaining momentum in developing countries. More than 90 developing economies opened their telecommunications sector to competition between 1990 and 1998, transferring to the private sector more than 500 projects (OECD (2001)). Telecommunications reform can be a positive-sum game in which customers, existing and new operators, employees, domestic and foreign investors, and governments all gain. Competition should lead to lower prices and higher quality. However, poor performance by regulatory agencies may limit the benefits of reform.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 155

Information technology: some implications for Thailand

155

Beyond the physical infrastructure, providing wide Internet access requires the emergence of local Internet service providers, low cost connection to the web, development of sites with local language content, and the offering of a range of other services demanded by local Internet users. For example, it should be possible to offer access to most users at local call rates and an ISP can lease a high capacity line from the telecom service provider at a competitive rate. Just as in telecom services, competition needs to be fostered in the ISP market. To promote the use of the Internet and e-commerce, Thailand needs to promote competition in the telecommunication market by allowing new firms to enter the market, both domestic and foreign. In particular, the telecommunication regulations have to be on the basis of consumer’ interests as in the guidelines of the WTO’s Basic Telecommunications Agreement. 4.2

Human Resource Development

As mentioned above, human development is another important factor to improve the digital divide among countries and within Thailand. As the economy gradually adopts e-commerce, skilled IT workers will be in increasing demand. Since there exists a brain drain of IT personnel from low pay industry to high pay industry and from less IT-advanced countries to more IT-advanced ones, there is no incentive for firms to invest in human resource development. Moreover, there exists a digital divide between new and traditional industries, large firms and SMEs, and cities and rural areas. Since human resource is a public good that creates positive externalities, we face the problem of market failure. The Thai government has to play an important role as a main investor in IT human resource development, not only by emphasizing the teaching of IT in schools but also by fostering an IT-friendly environment. The main target is ‘IT for All’, which implies that Thai people at any age, at any income level, and in any location, are able to learn and get access to new technology equally. 4.2.1 Regional cooperation For Thailand, building human resource cannot be achieved without technical assistance from abroad. In particular, the Japanese government announced an assistance package consisting of ODA (Office Development Assistance) public funding and non-ODA or OOF (Other Official Flows) with the view to extending a total of US$15 billion over five years to Asian countries. The goal of the package is to bridge the digital divide between and within Asian countries. The funding will be used for: (a) Raising awareness of IT opportunities and contributing intellectually to policy and institution-building;

Kagami 02 chap 06

2/15/04

2:54 PM

Page 156

156

Country/area studies

(b) Developing and training human resources. In particular, Japan will provide assistance, mainly in the form of technical cooperation, for the training of over 10 000 individuals during the next five years; (c) Building IT infrastructure and providing assistance for network establishment; (d) Promoting the use of IT in development assistance. In particular, the package will promote the use of IT in distance training, distance learning and the provision of medical care. During the first phase, Japan will establish 30 core IT centres. In the case of Thailand, the assistance could be effectively used to provide computer and Internet access to schools, which is estimated to cost between 12.1 and 12.7 billion Baht a year. It can also be used to develop an effective programme to raise awareness in the private sector and train human resources. In order to utilize this fund, Thailand should design the specified development programme by herself and should use the loan projects carefully. The World Bank also has a programme concerning the digital divide. The establishment of telecentres seems to be a favoured investment for the World Bank, and various UN agencies, and regional development banks. UNDP, for example, has begun pilot projects aimed at the creation of electronic community centres as a platform for access and connectivity in rural areas. They will also serve as centres for capacity-building, skills enhancement, training, communications and content development. SMEs are encouraged to utilize these facilities and they will be assisted in the creation of websites, digital web management and the conduct of e-commerce. In India, the USAID project provided telecentres to low income womens handicraft-producer groups, equipping and training them to use digital cameras and the Internet to market their wares while showcasing their cultural richness. In the case of Thailand, this programme will be very useful if we can combine this project with the project provided by the government on the village fund and the business-in-village project, the so called ‘One Product for One Tamboon’. The main problem facing community businesses is the lack of marketing skills and use of the Internet can solve this problem. 4.2.2

Role of government

(a) Formal education In order to eliminate the digital divide in Thailand, the government should provide facilities for Internet access to schools nationwide, from elementary to junior and senior high schools. IT-driven lessons should be promoted, and instruction on IT-related ethics and manners should be introduced. English, the basic language in the Internet era, should be enriched. Besides, more importance should be placed on such subjects as mathematics and science to foster the ability of thinking logically. The environment to encourage

Kagami 02 chap 06

2/15/04

2:54 PM

Page 157

Information technology: some implications for Thailand

157

schools to interact with other schools both inside and outside Thailand via the Internet to promote knowledge exchange should be fostered by the government. One problem the country faces is that at present people mainly use the Internet for entertainment and do not utilize this technology for human development. The proper use of the Internet in schools can set an important guideline for both teachers and students for the use of the Internet in the future. However, Thai schools also face the problem of IT personnel shortage. Therefore, school teachers should be given more opportunities for IT training, and a registration and dispatch system of IT instructors should be introduced, so that human resources in businesses and colleges can be utilized as IT instructors. The college system should be reviewed and actively reformed to the new IT society. Market mechanisms should be introduced to adjust the supply of labour to the new demand for labour. Therefore, colleges should be able to exercise more autonomous management, including authority over personnel and budgetary matters, establishment of faculties, departments and curricula, and exchange of researchers with private businesses. The cooperation between educational institutes and business sectors is very important to develop human resource in Thailand. Furthermore, an IT-related certification system should be standardized internationally. (b) Training program IT is the tool that is used to increase productivity and reduce costs in almost all business sectors. However, this technology keeps changing over time. In order to keep up with technology changes, training for new skills is unavoidable. The ability of Thailand to attract successfully, absorb and benefit from FDI in IT industry and the transfer of technology, which it may bring, depend to a large extent on its own technological capabilities, of which the skills and technical knowledge of its workforce are critical components. The FDI linkage that has not been developed in Thailand can be achieved through a successful training programme. The national training programme can be realized under the strong support of the government. Singapore provides a good example of how the government anticipated training programmes. The Economic Development Board (EDB), which ensures that inward investment is forthcoming to provide the necessary capital and knowhow for the new industries, is also in charge of the human resource requirements for those industries. The EDB has to ensure that the education and training system is capable of producing the right type of skills required by new industries. As a result, both the Ministry of Trade and Industry in Singapore and the EDB provide this information to the Council for Professional and Technical Education, the main decision-making body for determining the output of the education and training system. On vocational training, targets for on-

Kagami 02 chap 06

2/15/04

2:54 PM

158

Page 158

Country/area studies

the-job and work-based training and for the level of investment in training by employers are set by the Singapore Productivity and Standards Board. Training programmes run by the Board are targeted to meet the demands of existing and new industries. Human resource development is however far more than just one-time training. It is a continuous learning process. Therefore, the government should provide a training programme for adults who are beginners and who want to catch up with the new technology.

5.

CONCLUSIONS

The use of information technology in Thailand is still in the early stages. On the consumer side, the percentage of people using the Internet or having computers is still low, compared to several other countries in the region. Only 5.64 per 100 persons logged on to the Internet in the year 2001. Furthermore, this figure hides internal digital divides among regions. The chapter showed that IT use is concentrated in the regions with high income levels, such as Bangkok. The digital divides also exist among age groups, between genders, and between large firms and SMEs. The use of the Internet is heavily biased toward entertainment rather than business or knowledge exchange. There are more than 81.6 per cent of the Internet population having no experience in buying online goods and services. As for business use, only a few domestic firms are able to apply IT to their business needs. On the producer side, Thailand is an export base for IT-related products. Mainly foreign firms produce IT products in Thailand by importing the main components from abroad and using low skilled labour for the assembly process. Domestic firms play a minor role as sub-contractors for labour-intensive components. There is neither forward nor backward linkage in the Thai IT industry. Therefore, technology transfer does not exist. Consequently, since the Thai IT industry is weak, foreign firms earn most of the benefits from the IT revolution. The main problem for Thailand concerning the IT revolution is the digital divide. This digital divide exists in all aspects for IT consumers and IT users. Only small groups of high income households and a few large firms enjoy the benefits received from applying IT. There are at least two main reasons to explain the Thai digital divide, one is the high price of utilizing IT and the other the lack of IT human resources. The chapter showed that if Thailand as a developing country could use the Internet and e-commerce fully, it would benefit all consumers and businesses, including SMEs, to obtain easy access to the global network. In order to reduce this digital divide and create an IT-literate society, the chapter recommends the

Kagami 02 chap 06

2/15/04

2:54 PM

Page 159

Information technology: some implications for Thailand

159

government liberalize the telecommunications market, to allow more competition from new entrants domestic and foreign. Currently, the Minister of Transportation and Communications tightly controls the telecommunications market. A few oligopolistic enterprises are in a position to extract high profits and forestall competition. Market liberalization has been included in the Telecommunications Master Plan 1997–2006; however, it has not shown any progress yet. The other recommendation is on IT human resource development throughout the country. This includes formal education from elementary schools to colleges, both hardware (equipment) and software (personnel), and informal education for workers and people of all ages, under the slogan ‘IT for All’.

REFERENCES Hossain, L. (2001), National Strategic Planning and Practice: The Case of Thailand’s Telecommunications Industry, Ashgate, Aldershot. International Labour Organization (2001), World Employment Report 2001, ILO, Geneva. Mephokee, C. (2000), ‘The Impacts of the ITA on Thailand Trade’, Research paper, Faculty of Economics, Thammasat University, Bangkok (in Thai). Ministry of Economy Trade and Industry (METI) (2000), ‘Basic IT Strategy’, White Paper, 27 Nov, METI. National Electronic and Computer Technology Centre (1999), The Internet User Profile of Thailand, NECTEC, Bangkok. National Statistics Office (NSO) (2001a), Household Socio-economics Survey 2000, NSO, Bangkok. National Statistics Office (2001b), Survey on the Use of IT by Households, Unpublished, NSO, Bangkok. National Statistics Office (2001c), Labour Force Survey 2001, Unpublished, NSO, Bangkok. OECD (2001), Information Technology Outlook 2001, Organization for Economic Cooperation and Development, Paris. Reed Electronics Research (1995), The Yearbook of World Electronics Data. Singh, Alwyn D. (2000), ‘Electronic Commerce: Some Implication for Firms and Workers in Developing Countries’, ILO Discussion Paper No. 123, ILO, Geneva. Tangkitvanich, S. (2001), ‘Assessing Global E-Commerce Policies: A Perspective from Thailand’, TDRI Quarterly Review, March. Tangkitvanich S., and D. Nikomborirak (1997), Internet Competition and Pricing in Thailand, TDRI, Bangkok (in Thai). Thailand Development Research Institute Foundation (TDRIF) (1997), Telecommunications Master Plan 1997–2006: Executive Summary, TDRI, Bangkok. UNCTAD Secretariat (1999), Can Electronic Commerce be an Engine for Global Growth?, UNCTAD, Geneva.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 160

8. Information policy and information technology in Central and Eastern Europe with emphasis on Estonia Tanga McDaniel* 1.

INTRODUCTION

This chapter highlights various policies and programmes that have affected the information and communications sector and the use of information products in Central and Eastern European countries. Countries in Central and Eastern Europe (CEE) are diverse in their infrastructure and policies toward information technology. Organizational affiliations and preparations to join the European Union affect these national differences since membership in international organizations requires candidates to meet threshold criteria on infrastructure development and competition legislation. As in most regions of the world changes in and changes resulting from the information sector are rapid. Monopoly concessions in the telecommunications sector that have helped to keep the price of local phone service from falling are ending; combined with the existing policies of liberalization in wireless technologies and in Internet service provision, liberalization of local voice telephony could help to provide more depth to the IT sector (for instance, by making existing connections faster). Achieving greater network range requires forward-looking policies possibly involving both the public and private sectors. Many CEE countries have national policies for expanding and enhancing IT use; these policies often begin with the liberalization of the national telecommunication companies and plans for bringing government itself into the information age. Education policies and rural initiatives have also been adopted, and the use of multipurpose communication centres has helped to make these national policies feasible. This chapter has two parts: Sections 2 and 3 provide a general discussion of the IT situation in CEE while Section 4 focuses specifically on Estonia. Section 5 concludes.

160

Kagami 02 chap 06

2/15/04

2:54 PM

Page 161

Information policy and IT in Central and Eastern Europe

2.

161

CEE IN CONTEXT

Improving the IT capability in CEE countries has benefits for individuals and for business. The cost of transmitting a trillion bits of information from Boston to Los Angeles has fallen from $150,000 in 1970 to 12 cents today. (UN (2001))

Heavy reliance on inferior technologies reduces competitiveness for companies operating in open markets, and for individuals limits their options for information and communication. CEE is a region of high skilled adaptable individuals who have gone a long way toward transforming from centralized to market-oriented economies, and their future potential for communications development is widely recognized. The 2001 United Nations Human Development Report introduced a technological achievement index ‘which aims to capture how well a country is creating and diffusing technology and building a human skill base – reflecting capacity to participate in the technological innovations of the network age’ (p. 46). To create the index the study uses indicators in the areas of technology creation, diffusion of recent innovations (such as the Internet), diffusions of old innovations (such as electricity and telephones), and human skills. Seventy-two countries are collected into four groups based on their scores: leaders, potential leaders, dynamic adopters and marginalized. The range of scores for the countries classified is 0.744 (Finland) to 0.066 (Mozambique). Several CEE countries fall into the potential leader category. Czech Republic, Hungary, and Slovenia are ranked third, fourth, and fifth among the 19 potential leaders with scores of 0.465, 0.464, and 0.458, respectively. Also, Czech Republic ranks 29th in a list of the 30 largest exporters of high tech products.1 2.1

Range Versus Depth of Access

It is obvious from the statistics on network penetration that network range in CEE is low relative to many Western countries. It is worth asking if countries with limited financing should follow a policy of ‘depth’ improvement or one of ‘range’ improvement. Depth of service would be improved by replacing old infrastructure where it exists and by leapfrogging conventional PSTN technology where infrastructure does not exist. Where any investments are made network range is improved. Table 8.1 shows penetration levels for different types of service lines in CEE and the degree of network digitization. Albania has the lowest penetration among the countries listed in the table but only Albania and Slovenia have networks that are fully digitized. Poland is not

Kagami 02 chap 06

2/15/04

2:54 PM

162

Page 162

Country/area studies

far behind with 97 per cent digitization. Penetration levels of alternative technologies such as ISDN are currently low for all countries, but broadband technology is slowly being introduced. Poland has issued three UMTS licences to mobile operators and the Czech Republic planned to issue four in 2002 and had already issued three wireless local loop licences.2 Table 8.1 Country

Access indicators, end 2000 Network Number Penetration Penetration Penetration digitization of ISPs conventional mobile per ISDN per per 100 people 100 people 100 people

Albania 100 Bosnia 85 Bulgaria 80* Czech Republic 83 Estonia 67 Hungary 79* Latvia 43 Lithuania 33 Poland 97* Romania 50* Slovenia 100 Notes: Source:

13 6 180 16 8 39 40 25 500 200 24

4.4 22.1 35.6 37.5 38.3 40.8 30.2 32.0 28.5 19.8 38.4

1 3 8 39 41 30 16 11 16 13 57

** ** ** 1.2 1.8* ** 0.8 ** 0.5 ** 9

*

1999 data. ** <0.5. ESIS (2001).

In Romania the Soros Foundation for an Open Society has played a part in Internet development through satellite technology and is one of the major noncommercial ISPs there.3 The penetration of both computers and Internet users was just over 3 per cent at the end of 2000 in Romania; nevertheless, given the technology in place there is potential for both depth and range improvements. The eight ISPs that operate their own backbone all have either satellite or fibre optic links to foreign connections. Table 8.2 places CEE in a broader global context and further illustrates that, as a group, the range of Internet penetration is very low. However, the region experienced large percentage growth between 1998 and 2000. The impressiveness of these growth rates derives from the low starting base, but if we compare Eastern Europe and CIS to other regions that were closer in 1998 in Internet usage (such as East Asian, Latin American and Arab countries) the impressiveness is more convincing.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 163

Information policy and IT in Central and Eastern Europe

Table 8.2

163

Internet users as percentage of the population

Region

1998

2000

Per cent change

US High-income OECD (excl. US) Latin America and the Caribbean East Asia and the Pacific Eastern Europe and CIS Arab States Sub-Saharan Africa South Asia World

26.30 6.90 0.80 0.50 0.80 0.20 0.10 0.04 2.40

54.3 28.2 3.2 2.3 3.9 0.6 0.4 0.4 6.7

106 309 300 360 388 200 300 900 179

Source:

UN (2001).

Greater liberalization in mobile relative to fixed line service has played a role in a substantial technological shift over recent years. Figure 8.1 shows changes in the proportion of total access lines composed of mobile and fixed

Absolute change in percentage

60 40 20 0 –20 –40 –60

Mobile

Slovenia

Romania

Poland

Lithuania

Latvia

Hungary

Estonia

Czech R.

Bulgaria

Albania

–80

Fixed

Source: ESIS, (2001a)

Figure 8.1 Change in fixed and mobile penetration as proportion of total lines, 1998–2000

Kagami 02 chap 06

164

2/15/04

2:54 PM

Page 164

Country/area studies

lines between 1998 and 2000. For all countries the proportion of fixed lines has been falling while mobile represents an increasing percentage of total access lines; this change is especially noticeable in Hungary. The near symmetry of the figure elsewhere illustrates the growth in total lines has been mainly due to increases in mobile subscriptions. 2.2 Pricing Policies The global nature of the Internet makes the sector more liberal almost by default compared to other network industries such as electricity and gas where companies have historically been able to lobby for regulatory protection against entry. The anti-competitive aspects in the Internet industry come from: local telephony pricing, access pricing and domain restrictions. Without competition for local telephony there are no market incentives for monopoly companies to offer unmetered tariffs and use of the Internet becomes restricted by what individuals are willing and able to pay in local fees. This monopoly provision of local service would be of less concern if alternative technologies were already in a position to compete (such as WAP on mobile phones or broadband capability), but because this is not substantially the case, the problems of local bottlenecks remain. A second pricing problem has to do with network access. If Internet service providers (ISPs) are restricted from owning their own networks, rights to use the network of the monopolist are arranged through access fees. Access pricing is a secondary problem for industry competitiveness as long as there is effective regulation ensuring that third party access tariffs are cost-reflective (that is, representative of the cost to the monopolist of providing network use)4 and non-discriminatory so that the monopoly telecom company is unable to give preferential treatment to its own Internet services relative to the services provided by competitors. If ISPs are not restricted from owning their own networks, access pricing will still arise and regulation of those prices may still be desirable if market power is a problem. To become EU Member States accession countries must harmonize their pricing policies according to EU legislation that requires service pricing (including access pricing) to be cost-reflective.5 Other pricing issues affect choice once service has been provided and not service provision itself. High costs for domain name registration or domain name restrictions limit the number of national websites, and this limitation could be binding on users with limited language skills. Various rules apply for domain name registration in CEE. In some countries there are no restrictions on who may apply for a name under the top registry: Czech Republic (.cz), Latvia (.lv), Lithuania (.lt), Poland (.pl) and Romania (.ro). Some require companies to be registered in their country: Estonia (.ee) and Slovenia (.si).

Kagami 02 chap 06

2/15/04

2:54 PM

Page 165

Information policy and IT in Central and Eastern Europe

165

Slovakia registerees can use the .sk domain for business in the Slovak Republic. Any Hungarian and organizations with Hungarian-granted trademarks (including foreigners) may register under the .hu domain.6 Since laws governing online transactions often rely on the country of origin principle, a lack of restrictions for domain name registration has a competitive downside in that transparency in e-business is compromised if parties are unable to distinguish where a counter party is physically located. The most rapid IT changes are occurring in liberalized sectors where regulations have not inhibited entry. Table 8.3 lists the dominant telecom operator in ten CEE countries and the date at which their monopoly concession was due to end. Some countries have partially liberalized their telecommunications industries by opening up long distance markets while maintaining monopolies in the local market. Poland oddly has an open local market with monopoly provision of long distance. Until recently, Poland had foreign ownership restrictions for telecom companies. These restrictions were abandoned when the new Telecommunications Act came into force in 2001. The Czech Republic, Estonia and Slovenia were early to end monopoly concessions; these countries have in general been very proactive in their attitudes toward IT adoption and expansion. Nevertheless, the telecom market remains dominated by a single firm in Slovenia, and as seen in Table 8.4, the mobile market is highly concentrated. The Estonian Informatic Centre reports that although local call prices are still high, ‘competition in the telecom market has still resulted in a significant reduction (50–70 per cent) of all international long distance calls, it has also unified the local long distance call price with local call prices and has reduced about 50 per cent of leased line Internet connection prices ... Internet access has been included among universal services by our new Telecommunications Act – hence it should be available at the same reasonable price to everyone and everywhere in Estonia and telephony network operators are required to provide it’ (Siil (2001), p. 3). Where liberalization is limited Internet pricing options are limited as well, and unmetered alternatives are not standard. Providing specific data on ISP charges in a study such as this has little value since the situation is changing so rapidly. Once new competitors enter the market, offering unmetered options is an obvious pricing strategy. Where telecentres have been introduced on a large scale, such as in Hungary and Estonia, service is often provided free of charge (particularly in libraries) to users who are willing to book computers in advance. Before the ending of its monopoly concession, the Estonian telecom operator, AS Eesti Telekom sold 2, 6, and 12-hour prepaid cards to clients of its Internet services so that they avoided the local call tariff while online.

Albtelecom BTC SPT Telecom AS Eesti Telefon Matav Lattelekom Lithuania Telekomas TPSA RomTelcom Telekom Slovenije

Albania Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovenia

no yes no no yes yes yes no yes yes

Local

yes yes no no yes yes yes no yes yes

Monopoly Long distance domestic yes yes no no yes yes yes yes yes yes

Long distance international 2003 31 Dec. 2002 31 Dec. 2000 31 Dec. 2000 2002 2003 31 Dec. 2002 2003 31 Dec. 2002 31 Dec. 2000

End of monopoly

2:54 PM

166

Source: ESIS (2001a) and ITU.

Telecom operator

Monopoly status of major telecom provider

2/15/04

Country

Table 8.3

Kagami 02 chap 06 Page 166

Kagami 02 chap 06

2/15/04

2:54 PM

Page 167

Information policy and IT in Central and Eastern Europe

Table 8.4

Mobile telephony: ownership

Country Operator

Number of Per cent of Number of fixed subscribers market line licences End 1999 End 2000 End 2000 End 2000

Slovenia Mobitel 640 000 Simobil 24 000 Czech Eurotel 1 070 000 Republic Radiomobil 875 000 Cesky Mobil 0 Estonia EMT 244 000 Radiolinja Esti 87 000 Ritabell 56 000 Hungary Westel 842 000 Pannon 670 000 Vodafone Poland PTC 1 700 000 Polkomtel 1 500 000 PTK digital 620 000 PTK analog 180 000 *

Note: Source:

2.3

167

1 000 000 140 000 2 000 000 1 780 000* 220 000* 292 000 107 500 63 000 1 600 000 1 200 000 190 000 2 456 000 2 135 000 1 073 000 119 200

87.7 12.3 50.0 44.5 5.5 63.0 23.0 14.0 53.5 40.1 6.4 43.9 37.1 19.2 1.7

1 8 8 14 50

estimate. ESIS (2001a).

Competition Policies

In some cases the obligation to meet an international benchmark in order to obtain organizational membership has motivated a liberal and proactive policy toward information technology. Czech Republic, Hungary, and Poland are already members of the OECD and are potential candidates for European Union membership. Other candidate countries from CEE include: Bulgaria, Estonia, Latvia, Lithuania, Slovakia, Slovenia and Romania. Future accession to the EU requires that these countries take on board the liberalization and competition policies of the EU, and the state of their telecommunications industry will influence the competitiveness of these countries once they are Member States. McDaniel (2001) discussed that Czech Republic, Hungary and Poland have among the highest prices for local calls and Internet access within the OECD, and this remains true. However, the ending of the monopoly concessions and the removal of foreign ownership restrictions in Poland will provide new opportunities for entry in local voice telephony. Also, the increasing availability of alternative sources of access (such as cable and wireless local loop) creates opportunities to leapfrog conventional PSTN networks and to increase investment in faster, more flexible networks.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 168

168

Country/area studies

Czech Republic, Estonia, Hungary, Poland, and Slovenia began negotiations for the EU accession process in March 1998; Bulgaria, Latvia, Lithuania, Romania, and Slovakia began in February 2000. The negotiation process includes 31 topics or ‘chapters’ which must be brought into line with the laws of the EU – three of these are competition policy, taxation and telecommunications. Provisional closure of a chapter of the negotiation implies that the country has made a credible commitment to align its legislation with European laws and directives and that it has the administrative capability to apply the legislation. As of June 2001 the Czech Republic, Estonia, Hungary, Poland and Slovenia have met the requirements for closure on telecommunications, but competition policy and taxation remain open chapters for all ten CEE countries. Actually, there is no harmonized taxation policy for the EU, except for protocols on who pays VAT and at what rate. For instance, when is a book a commodity and when a service? The answer determines the taxation rate that will be applied; ‘... [t]he EU recognizes digitally delivered products such as software, digital music, digital books, etc. as services.’7 Competition policy is set out in Articles 81 through 89 of the European Community Treaty and deals with collusive agreements, market power, mergers, liberalization, and subsidies to domestic companies that affect their competitive position relative to other Member States. For telecommunications, accession to the EU requires that countries achieve the goal of universal access. EU directives require ‘Member States to ensure that all persons reasonably requesting it can obtain a connection to the fixed public telephone network at an affordable price’ (Centre for Democracy and Technology (2000)). This is a challenging requirement and again brings into consideration the question of range versus depth of network. It is left to individual countries to determine what is affordable, and Estonia for one has included Internet access among its universal service obligations. But how should the universal service requirement be interpreted for countries such as Albania that are rebuilding their networks after years of war and social conflict who may have the opportunity to choose technologies such as satellite that would bring communications access to more of the population at a lower cost than could be achieved by extending its fixed line network? 2.4

Attracting Investment

Accession status enhances the attraction of CEE countries to foreign investors even before membership of the EU is granted. The adoption of EU laws and regulations implies a certain standard with respect to transparency particularly in the financial sector. Whilst opaque regulations inhibit foreign investors, standardized regulations accompanied with a well established means of legal appeal reduce the risk of property rights violations or regulatory and political insta-

Kagami 02 chap 06

2/15/04

2:54 PM

Page 169

Information policy and IT in Central and Eastern Europe

169

bilities. The annex to this chapter lists a selection of important legislation on IT and telecommunications in 1999 and 2000 for ten CEE countries. Companies investing in CEE accession countries already benefit from the free trade provisions within the EU. The 2000 progress report of candidate countries produced by the EU shows the increasing trade integration among these countries and Europe. Hungary and Estonia have especially high trade integration with Europe with more than 60 per cent of trading activity (imports and exports) being with Member States.8 Accession will also be associated with an increase in wage levels closer to those of the EU average. The relatively low wage of the high skilled labour force in CEE has drawn foreign investors to the region in the past. Other incentives are already being instituted by governments to maintain foreign interest; tax benefits and grants for job creation or personnel retraining are examples. Corporate profits which are retained or reinvested are not taxed in Estonia; ten-year corporate tax breaks are granted in the Czech Republic to first time manufacturing investors, and Slovenia has the lowest corporate tax rate in Europe. Grants are provided to companies attracting employment to target areas of the economy (usually including communications) in the Czech Republic, Hungary, Poland and Slovenia.9

3.

ONLINE ACTIVITY: TRANSACTIONS, SERVICES AND RURAL INITIATIVES

The legal environment, the state of telecommunications liberalization, Internet access fees, attitudes of the banking sector toward consumers making online payments, lack of credible payment options, and the nature of home delivery service hinder the business to consumer (B2C) element of e-commerce. The lack of credit card use in Europe compared to America has been one reason for differences in the amount of online transacting in Europe relative to the US, and in CEE, payment upon delivery for online purchases is common (such as in Poland and Hungary). Table 8.5 shows the number of credit cards per 1000 inhabitants in 1999 for six CEE countries with comparative numbers for the US and Japan. This situation is evolving, and the ‘credit card problem’ will soon be irrelevant as new payment options (such as electronic cash systems) are introduced. As e-business continues to grow, however, efforts should be taken to increase consumer awareness of online security and how they can be sure that firms and government bodies are using secure technology. Business to business (B2B) transactions are less hindered by the above concerns, and private companies have found it in their interests to set up their own trading facilities as well as to make online transactions feasible for new

Kagami 02 chap 06

2/15/04

170

2:54 PM

Page 170

Country/area studies

companies. Internet ‘incubators’ are a way of funneling investments into Internet start-up companies or providing resources for new companies to engage in B2B. E-incubator was developed in Poland by Cisco, Oracle, Sun Microsystems, DNP Polska, and RSA Security to spread e-commerce ideas as well as the means to develop them. Table 8.5

Number of credit cards, 1999*

Country Slovenia Hungary Croatia Slovakia Czech Republic Poland US** Japan**(i)

per 1000 inhabitants 509 358 304 287 208 181 1871 1936

Sources: * Internet in a Transition Economy, International Telecommunication Union (2001). ** Bank for International Settlements (2001); (i) 1998 data.

In a capitalist global economy it is worth remembering that online services are not solely associated with e-commerce. As part of the discussion on the digital divide, the lack of e-commerce in less connected economies is often cited as one of the consequences. There are many conveniences that come with the option of transacting online (particularly in rural areas where banking and postal services are more limited than urban areas) but the general access provided to local and international information is perhaps equally important. Distance learning opens new doors to individuals in remote areas where university courses are limited or classes too distant to attend regularly; information about and forums for discussions on health and agriculture can assist remotely located doctors and farmers. With communications technologies playing increasingly vital roles in economic development, education, health care and governance, the exclusion of those who are poor, illiterate, rural or non-English speaking has broad ramifications. (Centre for Democracy and Technology (2000))

Differences in IT use are apparent both between and within countries (in other words, the so called ‘urban-rural’ divide), but this is clearly not a problem which is specific to the countries of CEE. One reason for this brand of digital divide is the relative lack of infrastructure in rural areas compared to urban

Kagami 02 chap 06

2/15/04

2:54 PM

Page 171

Information policy and IT in Central and Eastern Europe

171

areas which follows from the higher cost of connecting people in sparsely populated territories and occasionally from environmental considerations, such as refusal for way-leaves to extend visible cables over pristine landscape. In liberalized markets where the lack of infrastructure is already a problem the divide may only be exacerbated without innovative programmes or policies that attract private capital to rural areas. The concept of telecentres, which began in Sweden in the mid-1980s, has become widespread as local and federal governments have sought feasible ways to digitally connect rural communities. Telecenters may be defined as strategically located facilities providing public access to ICT based services and applications ... while facilities and usage vary across telecenters, all reflect the intention to address the issues of access by providing technology, develop human capacity and encourage social and economic development. (Oestmann and Dymond (2001))

In addition to providing users with e-mail and Internet access, a telecentre may also provide access to office equipment, library facilities, distance education programmes, and valuable local information on travel, medical services, government programmes, employment services and general advertising. Two adaptations of the telecentre concept include the telecottages (telehaz) of Hungary and the access points in Estonia initiated under the Village Road programme (KülaTee). Telecottages began in Hungary in 1994 first for a brief period in the South Eastern city of Nagymágocs and then more stably in Csákberény in the North West; this Csákberény centre was the result of the initiative of the local population. The Hungarian Telecottage Association was formed in 1995 and funding for the programme increased in 1996 with help from the United States Agency for International Development (USAID); from this small beginning there is now approximately one telecottage per 50 kilometres in Hungary. In mid-2000 there were already 161 telecottages with another 61 planned (ITU (2001)). Most centres are equipped with a number of computers in addition to fax machines and photocopiers. Murray (2001) gives an in-depth analysis of Hungary’s experience with telecentres. Their use of telecentres reflects the true spirit in which they were meant to be utilized; that is, the priority for the programme has been small villages. Oestmann and Dymond (op. cit.) discuss that financing and ownership structures help determine the success of telecentre ventures. For Hungary, the initial funding from USAID was valuable for promoting and financing early centres and Murray (op. cit.) lists the following sources of support for the ongoing programme: ‘central government, Hungarian Telecottage Association, Hungarian Post Office, USAID, United States Department of Labor (USDOL), Democracy Network (DemNet), Hungarian British Embassy, The British

Kagami 02 chap 06

2/15/04

2:54 PM

172

Page 172

Country/area studies

Council, British Know-How Fund, European Union Delegation, PHARE, Soros Foundation, Hungary, Microsoft Hungary, Matáv (Hungarian Telecom) Elender (an Internet provider), Mikro Volán Elektronika Rt., Santa Cruz Operations’. This list reflects the diversity of funding options from which the Telecottage Association has drawn and the value of attracting interest from local, national and international sources. Estonia’s Village Road programme is briefly discussed in the next section which focuses more generally on Estonia’s experience with information technology expansion.

4.

CASE STUDY: IT ADOPTION, EXPANSION AND POLICY IN ESTONIA

Estonia is one of the most dynamic countries in CEE with respect to IT use and infrastructure. Hungary, Slovenia and Estonia have the highest penetration rates of conventional phone lines and only Slovenia has higher per capita mobile use. Estonia has the highest penetration of Internet users and the lowest per minute local call charge.10 In this section we survey policies and programmes in Estonia that have helped to extend IT use in the country. Three details which stand out are Estonia’s efforts to liberalize its telecommunication industry and harmonize its legislation to EU policy in preparation for membership; the national policy to build an Internet network and expand IT use and knowledge; and the partnerships between the public sector and private enterprise that have helped to make achieving the national policy a possibility. Quite apart from these active measures, Estonia has benefited greatly from its proximity to Scandinavia. The influence of Sweden and Finland has been especially noticeable financially and ideologically. 4.1

Liberalization

Estonia regained independence in 1991 and at that time was burdened with outdated and insufficient Soviet-era infrastructure. As part of the efforts to build its communication network the national telecom operator, AS Eesti Telefon (ETC) was privatized in 1992 and granted a monopoly concession with a mandate to provide, develop and improve the quality of basic telephone service. ETC is wholly owned by AS Eesti Telekom with the ownership structure shown in Figure 8.2. Between 1993 and 2000 the rate of fixed line penetration in Estonia grew from 23 to 38 per cent and the proportion of digital lines increased from 5 to almost 70 per cent. The increased digitization increased labour productivity by requiring fewer employees to maintain the lines and helped the company

Kagami 02 chap 06

2/15/04

2:54 PM

Page 173

Information policy and IT in Central and Eastern Europe

173

improve service quality by replacing older rotary dial phones. Between 1992 and 1999 ETC invested approximately 4.5 billion EEK in network infrastructure, the highest concentration of which was in the urban areas of Tallinn and Harju counties where one-third of the population live and work.11 Fibre optic cables connect the Estonian network to Finland, Sweden, Latvia and Russia. Public investors 25% Sonera Holding BV, Telia AB and Baltic Tele AB 48%

Republic of Estonia 27% Source:

Eesti Telekom.

Figure 8.2

AS Eesti Telekom ownership

In 1997 the ETC initiated a new accounting system to enable it to introduce more cost-reflective prices for its products and services and to reduce both local and international call charges. Through Atlas Data Communications it has built up its Internet service to handle increased demand for access – ‘30 per cent of all call minutes in December 1999 were for Internet use’12 – and has direct international connections to Canada (12 Mbps), Sweden (8 Mbps), Finland (4 Mbps) and Russia (4Mbps). Mobile service in Estonia is provided by three companies, shown in Table 8.4, and is mostly dominated by EMT with 63 per cent of the market. There are seven ISPs that operate their own backbone network: EENet, Uninet Data Communications, the government network, Data Telecom, Atlas Data Communications (ETC’s Internet service), EsData Ltd, and Delfi Online. Data Telecom is connected to Finland and via satellite to Amsterdam; Atlas connects

Kagami 02 chap 06

2/15/04

2:54 PM

174

Page 174

Country/area studies

to Canada, Finland, Russia and Sweden; ESData connects to Finland, and Delfi connects the Baltics. 4.2

Harmonization

In addition to liberalization Estonia has modified legislation to conform with EU standards and introduced new legislation to improve Internet security. Property rights, including intellectual and physical property as well as patents, copyrights, trade marks and industrial design are legally protected and the Estonian court functions legally, independently from government. The Digital Signature Act came into force in December 2000, and cooperation between telecommunication companies has helped to build public key infrastructure. Mobile, data communications, paging, satellite communication, leased lines, ISP and value-added services are liberalized markets, but companies require a licence to own and operate networks.13 The industry is monitored by the independent regulator, the Estonian Communications Board, under the conditions of the Telecommunications Act and the Competition Act. Under the new Telecommunications Act, Internet access has been included as part of the universal service requirements. Moreover, In Estonia the availability of data communication for all inhabitants has been declared a national priority, even one of human rights. (Estonian Village Road Programme)

4.3

National Promotion of IT

Estonia has two large Internet backbones: Estonian Education and Research Network (EENet) and EEBone. EENet was established by the Ministry of Education in 1993 and originally spanned only Tartu and Tallinn counties. As seen in Figure 8.3, EENet now spans the Republic and has a statutory obligation to provide Internet service to schools.14 The network is a member of the Central and Eastern European Networking Association (CEENet), the Trans-European Research and Education Networking Association (TERENA), and has links to Scandinavia through NORDunet and the Finnish University and Research Network (FUNET).15 EEBone is the government network and was launched in 1998. Figure 8.4 shows the penetration of Internet access in schools, government, hospitals, libraries and museums. Throughout the survey period, the National Ministry has been fully connected, and the proportion of schools and universities connected has been growing. The programmes, Village Road (KulaTree) and Tiger Leap, were initiated to facilitate the organization with financing for planned network expansions, principally within government and educational communities. The KulaTree project began in earnest in 1999, and was the outcome of data administration devel-

Kagami 02 chap 06

2/15/04

2:54 PM

Page 175

Information policy and IT in Central and Eastern Europe

Key:

175

màgistraalvõrk = main; magistraalliin = trunk line

Source:

EENet Estonia, http://www.eenet.ee/EENet/vorgukaart.html

Figure 8.3

The EENet network

opment programmes in 1995. The two parts of the programme, KulaTree I and KulaTree II, have dealt with local governments and with public libraries, respectively. Sources of funding for the projects have included: the Ministry of Culture, Eesti Telefon, Estonian Wireless Network, counties, self-governments, and regional programmes;16 the project is managed by the Estonian Informatics Centre. Cooperation between public and private enterprise has been an essential part of the programme. Computers and printers have been purchased via public procurement, making use of the Estonian company, Microlink, the largest computer company in the Baltics. The goal was to connect 95 per cent of the libraries by 2002, but financing has been a problem. The plan is proceeding, but beginning first with those libraries where the cost of connecting is lowest (in other words, not the most rural areas) and where the local governments have been able to make contributions to the cost. Figure 8.5 shows the growth in access points from October 1998 to January 2001. As many of these points are located in public libraries, the service is often free to users. However, demand is high and often booked far in advance. In a number of cases computers are set aside for unbooked service and require a user fee.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 176

176

Country/area studies

100.0 90.0 80.0 70.0 60.0 % 50.0 40.0 30.0 20.0 10.0 0.0 Oct.00

June.00

March.00

Dec.99

Oct.99

Aug.99

Primary and secondary schools Universities National ministries Regional and local authorities Hospitals/clinic (only hospitals) Museums Libraries (school libraries not included)

Time of the survey Source:

European Survey for Information Society (2000b).

Figure 8.4 The rate of organizations having web pages out of total number of organizations In 1996, President Lennart Meri launched what has become one of the best known initiatives of Estonian information policy, The Tiger Leap National Programme. The organization and financing of the programme are the responsibility of the Tiger Leap Foundation. The programmes goals ‘were declared as modernizing the Estonian educational system, and creating conditions for the formation of an open learning environment and for a better adaptation to the demands of an information society.’17 This is not the first attempt by Estonia at introducing information technology into the educational system; two previous programmes were in place from 1987–1992 and from 1992–1996. The first programme, whilst Estonia was still under Soviet control, achieved little success apart from the non-trivial one of stimulating interest and introducing a greater awareness of IT. The second programme bore more fruit and provided more modern technology to classrooms, but met with substantial funding problems. The Tiger Leap Foundation is composed of government, computer companies and private individuals. The programme faced obstacles in the beginning resulting from language, curriculum compatibility and legal status of the software on offer, but now that a foundation has been laid for meeting the goals of the national policy, attention is turning to sustainability of network service provision and training. Expansion in the IT sector is associated with a growing

Kagami 02 chap 06

2/15/04

2:54 PM

Page 177

Information policy and IT in Central and Eastern Europe

177

180 160 140 120 100 80 60 40 20 0 Oct98 Jan99 Apr99 Jul99 Oct99 Jan00 Apr00 Jul00 Oct00 Jan01 Source:

Siil (2001).

Figure 8.5

Number of public Internet access centres in Estonia

need for technical professionals. This is another factor that both the private and public sectors in Estonia have (i) taken notice of: The main problem lies in the seriously underdeveloped and under financed IT education at universities. There are very few new teachers, the amount of new doctorates per year is extremely inadequate. The low salaries – when compared to industry – do not motivate talented people to pursue a career in teaching. Also, the education given at existing IT departments does not correspond well enough to the needs of the companies. (IT College Estonia).

and (ii) taken action to deal with. In March 2000 the University of Tartu, Tallinn Technical University, AS Eesti Telekom and the Association of Estonian Computer Companies established the Estonian Information Technology Foundation which organized and finances the Information Technology College that began operating in late 2000. The college offers a three-year programme that will train approximately 150 students each year in IT systems development and IT systems administration. In 2001, tuition costs for the college were approximately US$1600 per year with scholarships provided from IT intensive companies. The Swedish government and Eesti Telkom are the largest donors to the college.18

Kagami 02 chap 06

2/15/04

2:54 PM

178

4.4

Page 178

Country/area studies

Public-Private Partnerships and Foreign Investment

Cooperation between the public and private sectors has played a significant role in Estonia’s attempts to increase ICT penetration throughout the Republic. In general both public and private investment are essential for building network infrastructure, and for Eastern Europe much of the necessary private capital is from foreign investors. Estonia’s legislation does not bestow discriminatory privilege upon international capital; instead part of its appeal is the equal treatment between foreign and domestic capital. Although foreign companies do not receive special investment incentives, they are legally given the same rights and held to the same obligations as domestic firms; this includes laws respecting property. Private property rights are legally protected and ‘there have been no cases of expropriation or nationalization since the country regained its independence in 1991’.19 The two largest foreign companies in Estonia are ETC and Estonian Mobile Telephone both of which are owned by Swedish and Finnish companies. Approximately 40 and 30 per cent of Estonia’s foreign investment come from Finland and Sweden, respectively, and in 2000 the financial and transport/communications sectors received the largest proportion of foreign investments. Between 1998 and 2000, foreign investment in transport and communications increased from approximately 300 million EEK to over 1.1 billion EEK. 4.5

Banking in Estonia

One area where Estonia stands out among its neighbours is Internet banking. The savings to both companies and consumers from telebanking present opportunities to involve the private financial sector in the financing of telecommunications infrastructure, particularly in rural areas. The benefit to consumers is the ability to handle a myriad of transactions in one location, a concept that gains value if the alternative is commuting to a nearby town or village where more services are offered. Where neighbourhoods are sufficiently connected to the Internet, banks may have the option of having fewer physical locations – a significant cost saving for companies who have only a few customers in remote areas. The risk of increased online activity with fewer physical branches, however, has an impact on the less technical members of society and those who are less able to travel to nearby villages. Estonia leads the way in Eastern Europe in Internet banking. According to the Estonian Informatics Centre, Internet banking is the most popular electronic service in Estonia with approximately 90 per cent of banking transactions handled electronically. Registration for online banking services is typically free for clients and commissions on transactions lower than at physical banks. Services vary from basic information about accounts (such as statements and

Kagami 02 chap 06

2/15/04

2:54 PM

Page 179

Information policy and IT in Central and Eastern Europe

179

balances) to more complex services involving third party transactions (such as payment of electricity bills and filing tax returns). The larger banks, Hansabank and Ühispank, offer teleshopping, and WAP services are growing as mobile telephony rapidly expands. Table 8.6 shows the number of Internet banking clients at the end of June 2001 at the four largest banks in Estonia. A comparative figure is shown for the Latvian customers of Hansapank – the largest Internet banking service provider in the Baltics.20 As elsewhere, security of online transactions is a problem in Estonia, but progress is occurring in this area. The Digital Signature Act came into force on 15 December 2000 and the telecom and mobile companies (Estonian Telephone Company, Estonian Mobile Company, Privador AS) have been involved in improving infrastructure so as to make use of their online services more secure.21 Table 8.6

Online banking customers

Bank

Estonia

Latvia

Hansapank Uhispank Nordea Sampo Pank

280 519 91 890 3262 14 126

40 000

Source:

Statistics Finland (2001a, 2001b).

As online services have grown so too has the IT capability within the financial sector. Abbate (2000) in her research on the Internet in Estonia describes how the lack of local software companies in the early days of the country’s independence led to Hansapank and Ühispank becoming the two largest software enterprises in Estonia. E-payments can be made directly to merchants using their services which means that growth in e-commerce will be less restricted by the lack of credit card use relative to other CEE countries. 4.6

Summary

This section has discussed Estonia’s goal of shaping and expanding its information and communications sector. All of the countries in Central and Eastern Europe have faced the problem of building and, for some, rebuilding their information networks during the past decade. In some areas Estonia stands out as a leader among these countries, particularly in Internet use. Much of Estonia’s success is related to having a national policy on information and significant private sector participation. Estonia has given substantial

Kagami 02 chap 06

2/15/04

2:54 PM

180

Page 180

Country/area studies

attention to building up information networks in educational facilities and has embarked upon a new programme of training IT professionals. This type of commitment is essential to a country, which along with its neighbouring countries has a declining population and a large number of emigrating youth. The idea of a national programme is not enough if not backed by a fundamental commitment to its implementation that implies a sustainable means of support until the programme is financially self-sustaining. Estonians who write about their achievements in IT advancement are matterof-fact and realistic: ‘The image of Estonia’s success in implementing ICT is a myth: Estonia’s “high” level is relative. However, our backwardness relative to developed countries is not as great as compared to former Eastern bloc countries.’22 Likewise, there is an awareness of their fortunate geographic position relative to their Baltic neighbours and more isolated Eastern European neighbours. Finally, in preparation for EU membership the country has gone a long way toward conforming its legislation and regulations to EU guidelines. This standardization increases their attractiveness to foreign investors.

5.

CONCLUSIONS

The status of information technology in Central and Eastern Europe continues to change very rapidly. This chapter does not survey the current status in the individual countries of the region, but instead attempts to highlight some of the policies and programmes that countries have adopted to aid and encourage the spread of IT use. The depth of telecommunications penetration has increased as concessions for the local monopolies have come to an end in some countries. Ending of monopoly contracts opens the door to companies who can provide faster connections through alternative infrastructures such as cable and satellite and allows foreign investment that is often necessary to increase the network range. Having granted a large number of licences to companies, Poland potentially has a very liberal telecommunications market and with the adoption of the new Telecommunications Act in 2001 has removed foreign ownership restrictions for licensed companies. Estonia has adopted an aggressive national programme for IT expansion, has the highest per capita Internet use in the area, and has an innovative financial sector with a large number of banking transactions already occurring online. The use of telecentres has helped to provide greater access to rural communities, and Hungary has been very successful at maintaining funding for its telecottage programme. Policies that promote IT expansion should not only consider international divides created by the relative lack of technologies in some countries. Also, they should not only consider the divide created by programmes that target

Kagami 02 chap 06

2/15/04

2:54 PM

Page 181

Information policy and IT in Central and Eastern Europe

181

lower cost urban communities at the expense of higher cost areas. Programmes that bring increased technological capability to communities may benefit the community as a whole, but may do so at the expense of those in the community who are least able to take advantage of them and likewise least able to find alternatives. For example, services such as Internet banking provide convenience for both suppliers and consumers of services, but an increase in online banking could harm the less technologically sophisticated and those who are less mobile (and so less able to travel further for service) if services in physical branches decline. Telecentres equipped with well trained staff would mitigate this type of problem. Where policies on property rights and copyrights are weak so will be the incentives to develop software programs in the national language. Reliance on international software applications limits IT benefits for education and limits the spread of ICT to rural communities and older residents. The needs of countries that have already embarked on extensive liberalization and IT development programmes (most of the countries in the chapter) are very different, if not qualitatively then quantitatively, from countries that are not only developing their infrastructure but are also rebuilding their infrastructure after periods of sustained civil and international conflict. Albania, Bulgaria and the countries of South Eastern Europe not specifically addressed here may find it more beneficial to altogether leapfrog conventional PSTN and find outside investment for increasing alternative technologies.

NOTES *

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

The author wishes to thank the Institute of Development Economies (IDE-JETRO) for its generous support and the British Economic and Social Research Council for research funding under grant R000238543. United Nations (2001). European Union, http://europa.eu.int/ISPO/esis The Soros Foundation Network and Open Society Institute started by philanthropist George Soros provides grants to CEE, the former Soviet Union, and many developing countries with the goal of promoting open societies. Where determining cost-reflective tariffs is difficult (either for companies or for regulators) a form of benchmarking could be established. If this is done internationally, however, nontrivial data conformity problems arise. Cost-reflective pricing for basic voice service has a short-term disadvantage for many domestic customers who have historically been cross-subsidized. In these cases, the need to make tariffs more cost-reflective leads to rebalancing and higher domestic prices. Simplynames, European Domain Rules, www.simplynames.co.uk. Baumgartner and Schulze (2001). CEEBICnet (2001a). CEEBICnet (2001a). ESIS data. Eesti Telefon. Eesti Telefon.

Kagami 02 chap 06

2/15/04

2:54 PM

182

Page 182

Country/area studies

13. ESIS. 14. European Survey of Information Society (2000). 15. Norway has assisted network development in the Baltic states through BALTNet see: http://www.eenet.ee. 16. Informatics, p. 7 and KulaTree web. 17. European Survey of Information Society (2000a). 18. IT College Estonia, http://www.itcollege.ee/inenglish/statusfacts.php. 19. CEEBICnet (2001b). 20. Siil (2001), p. 4. 21. Eesti Telefon. 22. European Survey of Information Society (2000b).

REFERENCES Abbate, J. (2000), ‘Research on the History of Computing and the Internet’, University of Maryland, http://www.wam.umd.edu/~abbate. Bank for International Settlements (2001), Statistics on Payment Systems in the Group of Ten Countries. Baumgartner, J. and C. Schulze (2001), Don’t Panic! Do E-Commerce, European Commission, Brussels. Centre for Democracy and Technology (2000), ‘Bridging the Digital Divide: Internet Access in Central and Eastern Europe’, http://www.cdt.org/international/ceeacess/ Central and Eastern Europe Business Information Centre (CEEBICnet) (2001a), EU Accession and US Business in Central and Eastern Europe, US Department of Commerce, Market Access and Compliance. Central and Eastern Europe Business Information Centre (CEEBICnet) (2001b), Estonia: Investment Climate Statement 2002, US Department of Commerce, Market Access and Compliance. Central and Eastern European Networking Association, http://www.ceenet.org/ Eesti Telefon, http://www.et.ee Estonian Informatic Centre, http://www.eik.ee/ Estonian Village Road Program, http://www.kylatee.ee/ European Survey of Information Society (ESIS), http://europa.eu.int/ISPO/esis European Survey of Information Society (2000a), ‘The Estonian Tiger Leap into the 21st Century’, http://www.esis.ee/ist2000/background/index.htm European Survey of Information Society (2000b), ‘WWW Indicators: Update Estonia’, http://eu-esis.org/esis2www/EEwww6.htm European Survey of Information Society (2001a), Information Society Indicators in the Countries of Central and Eastern Europe, European Community, Brussels. European Survey of Information Society (2001b), Regulatory Developments: Central and Eastern Europe Countries. Synthesis of Master Reports (1999–2000), European Community, Brussels. European Union, http://europa.eu.int International Telecommunication Union (ITU), http://www.itu.int/home/index.html International Telecommunication Union (ITU) (2001), ‘Internet in a Transition Economy: Hungary Case Study’, IT College Estonia, http://www.itcollege.ee IT College Estonia, http://www.itcollege.ee/inenglish/statusfacts.php McDaniel, T. (2001), ‘Globalizing Information? The IT revolution in Central and Eastern Europe’, in M. Kagami and M. Tsuji (eds), The ‘IT’ Revolution and Developing

Kagami 02 chap 06

2/15/04

2:54 PM

Page 183

Information policy and IT in Central and Eastern Europe

183

Countries: Late-comer Advantage?, Institute of Developing Economics/JETRO, Tokyo. Murray, B. (2001), ‘The Hungarian Telecottage Movement’, in C. Latchem and D. Walker (eds), Telecentres: Case Studies and Key Issues, The Common Wealth of Learning, Vancouver. OECD (2001), Communications Outlook: Information Society. Oestmann, S. and A. Dymond (2001), ‘Telecentres-Experiences, Lessons and Trends’, in C. Latchem and D. Walker (eds), Telecentres: Case Studies and Key Issues, The Common Wealth of Learning, Vancouver. Siil, I. (2001), ‘Estonia: Preparing for the Information Age’, International Council for Information Technology in Government Administration, ICA Information No. 74: General Issues, June. Statistics Finland (2001a), ‘Banks have 390,000 Internet bank clients’, International Business Statistics, 10 August. Statistics Finland (2001b), ‘40,000 users for Hansabanka Internet Bank’, International Business Statistics, 10 August. The Soros Foundation, http://www.soros.org United Nations (2001), Human Development Report.

Kagami 02 chap 06

2/15/04

184

2:54 PM

Page 184

Country/area studies

ANNEX: RECENT LAWS AND DOCUMENTS REGARDING THE INFORMATION SOCIETY Albania • June 2000: New Law on Telecommunications • July 1999: Law on the Privatization of Albanian Mobile Company • June 1999: Law on the Right of Information on Official Documents Czech Republic • October 2000: Act on Electronic Signature • July 2000: New Act on Telecommunications • June 2000: Act on Personal Data • May 2000: ‘Action Plan for implementing the State Information Policy’ • May 1999: State Information Policy, initial strategic document Estonia • 1998: ‘Principles of the Estonian Information Policy’ • 1999: Information Policy Action Plan • 2000: Digital Signature Act Hungary • 2000: ‘Hungarian Response to the Challenges of the Information Society’ Latvia • December 2000: ‘E-Latvia concept’ • August 1999 (submitted): Law on National Information Systems • 1999–2005 National Programme Informatics • March 2000: law on Personal Data Protection Lithuania • October 2000: presentation by the Ministry of Public Administration Reforms and Local Authorities of a programme on ‘Information Society in Lithuania’ in line with European directives • May 2000: ‘Lithuania’s Information Society Development Strategy’ • July 2000: Law on Electronic Signature • July 2000: Full harmonization of the Law on Legal Protection and Personal Data Poland • 2000: ‘Building Basis for the Information Society in Poland’, Parliament’s resolution • 2000: New Telecommunications Law • 2000: ‘Aims and Directions of Information Society Development in Poland’, prepared by the KBN and Ministry of Communications Romania • Since 1999: Draft Law for IT promotion ‘Code for Information Technologies Development and Use’, approved by the government, to be voted by the Parliament • 2000: Draft Laws on Electronic-signature, E-Commerce, Software Technologic Park • 2000: National Medium term Development Strategy of the Romanian Economy • 1999: Adoption by the government of the National Research, Development and Innovation Program (RDI) Slovakia • 2000: New Telecommunications Law • 2000: State Telecommunication Policy for 2000–2002 Slovenia • June 1999: New Law on Telecommunications • June 2000: Law on Electronic Commerce and Digital Signature Source:

European Survey of Information Society (2001b).

Kagami 02 chap 06

2/15/04

2:54 PM

Page 185

9. Internet and telecommunications outlook in Latin America Andrew B. Whinston and Soon-Yong Choi 1.

INTRODUCTION

The telecommunications industry in Latin America entered a new age during the 1990s as most countries opened up their telecommunications markets, privatized government-owned monopolies, and liberalized regulatory policies relying more on market forces than decrees. Such reforms coincided with the introduction of the Internet and the World Wide Web, offering them an opportunity to participate in the new information age as a full-fledged member. Still, Latin America as a region lags far behind North America, Europe and Asia/Pacific in terms of Internet penetration. User estimates in 2001 showed that Latin American users accounted for only 5 per cent of the world total (25 million out of 513 million users) (Figure 9.1). Nevertheless, by most accounts, Latin American Internet users are projected to grow steadily at a high rate (Figure 9.2). In this chapter, we focus on the current status of the telecommunications market, Internet access and e-business in Latin America, and investigate pricing, telecommunications policies and technological factors that affect current and future growth rates in Internet penetration. Throughout our discussion, economic situations (and unequal income distribution) appear as an underlying reason that severely limits future potential in many countries. Nevertheless, policy initiatives and market reforms are beginning to take effect in a few cases. The outlook for future telecommunications market will be summarized and, in the next chapter, we focus on two case studies, of Mexico and Argentina.

2.

CURRENT STATUS OF INTERNET DEVELOPMENT

By far, the best developed country in terms of Internet access is Brazil, with five million users estimated in 2000 (Figure 9.3). Mexico, Argentina and Chile round out the top four leaders in Latin America. In terms of top-level Internet 185

Kagami 02 chap 06

2/15/04

2:54 PM

Page 186

186

Country/area studies

Latin America 5%

Africa 1% Asia/Pacific 28%

US/Canada 35%

Middle East 1% Source:

Europe 30%

NUA, How Many Online, http://www.nua.ie/surveys/how_many_online/

Figure 9.1

Online Internet users, August 2001

25

Million

20

15

10

5

0 1995

1996

1997

1998

1999

2000

2001

2002

Source: Pegasus Research (1999) Latin American Web Users, http://www.pegasusresearch.net/metrics/growthla.htm

Figure 9.2

Projected growth in Internet users in Latin America

2003

Kagami 02 chap 06

2/15/04

2:54 PM

Page 187

Internet and telecommunications outlook in Latin America

187

domain, significant growth is evident in these four countries, although some Internet sites may be operating under .com, .net and .org domain names (Figure 9.4). Brazil is also ahead in terms of Internet penetration: available data for 1999 showed that 32 per cent of upper and middle income groups had access to the Internet in Brazil, compared to 14 per cent for Mexico, 12 per cent for Argentina and 11 per cent for Chile (US DOC, ITA (2000)). Active Internet users who have online accounts also indicate that Brazil leads other nations. Some 40 per cent of total online subscribers in Latin America are in Brazil, compared to 22 per cent for Mexico, 9 per cent for Argentina, and 6 per cent for Chile. 6000 5000

4000 3000

2712

2500

2000

1757

Source:

400

370

Peru

Uruguay

Chile

Argentina

Mexico

Brazil

0

878

Colombia

950

1000

Venezuela

Thousand users

5000

International Telecommunication Union (2001).

Figure 9.3

Internet users in largest Latin American markets, 2000

However, the correspondence between the number of existing Internet users and future growth potential is a tenuous one. Brazil’s large shares are attributed to its large population (170 million, 2000) and a relatively sound participation by upper and middle income groups. Nevertheless, its basic telephone infrastructure is no better than other leading Latin American countries in terms of per capita basis (Figure 9.5). Furthermore, figures on IT spending per capita indicate that Argentina, Chile and Brazil are more successful in building up basic infrastructure (Figure 9.6), although this level is far below the US per capita spending of US$1198.

Kagami 02 chap 06

2/15/04

2:54 PM

188

Page 188

Country/area studies

1 200 000 1 000 000 800 000 600 000 400 000 200 000 0 1997

1998

Brazil Source:

Mexico

1999 Argentina

2000

2001

Chile

Colombia

Internet Software Consortium (2001), Internet Domain Survey, http://www.isc.org.

Figure 9.4

Top-level Internet domain growth, 1997–2001

80 69.97 65.33 56.72

60 50

46.37

40 30 21.32 16.32

22.12 18.18

16.92 10.00

10

Source:

International Telecommunication Union (2001).

Figure 9.5

Main telephone lines per 100 inhabitants, 2000

UK

Japan

Korea

Paraguay

Colombia

Chile

Brazil

Argentina

US

5.00

World

0

12.47

Mexico

20

Ecuador

Per 100 inhabitants

70

Kagami 02 chap 06

2/15/04

2:54 PM

Page 189

Internet and telecommunications outlook in Latin America

189

68.7

70

65.5

63.7

60

US$

50

43.1 38.6

38.6

40 30 20

Source:

Colombia

Mexico

Venezuela

Brazil

Argentina

0

Chile

10

US Department of Commerce, International Trade Administration (2000).

Figure 9.6

Per capita IT spending, 1997

Asia & Pacific

1172Mbps

Europe

162 250Mbps 41 820Mbps 445Mbps US & Canada 767Mbps 14 140Mbps

68Mbps

Latin America & Caribbean Source:

Telegeography (2001).

Figure 9.7

Major international Internet backbone, 2001

Africa

Kagami 02 chap 06

2/15/04

2:54 PM

Page 190

190

Country/area studies Asia & Pacific

US & Canada

Tokyo

Seattle

Seoul

San Francisco

Europe

New York

Stockholm

Copenhagen

Amsterdam Los Angeles Washington, DC

Taipei

London

Sydney

Dallas

Miami

Frankfurt

Paris Internet bandwidth

Latin America & Caribbean Madrid

125 25 5 1 (Gbps)

Source:

Mexico City Buenos Aires São Paulo

Telegeography (2001).

Figure 9.8

Major Internet network access points, 2001

The growth in Internet usage critically depends on providing basic telecommunications infrastructure. For example, dial-up Internet subscribers in developed markets such as Chile and Argentina grew by 150 per cent and 136 per cent, respectively, during the 2000–01 period (Business News America (2001)). They are comparably better equipped with main telephone lines as shown in Figure 9.5. Others endowed with low fixed line infrastructure experienced lower rates of growth: Paraguay (42 per cent) and Ecuador (55 per cent). But Latin America in general needs better telecommunications infrastructure such as traditional telephone networks, advanced telecommunications networks, Internet backbone, interconnection points and access points. On most of these measures, the region is ill equipped to sustain future growth. As shown in Figures 9.7 and 9.8, the network capacity for international traffic is relatively low compared to cross-Atlantic and cross-Pacific lines. The major region-toregion interconnection backbones exist from São Paulo (and from Buenos Aires) to Miami (3.4Gbps) and from Mexico City to Dallas. But the total bandwidth

Kagami 02 chap 06

2/15/04

2:54 PM

Page 191

Internet and telecommunications outlook in Latin America

191

Per cent of population

25 20 15 10 5 0 1999

2000

2001

Brazil Mexico Venezuela Peru Source:

2002

2003

Argentina Colombia

2004

2005

Chile Other

Jupiter Communications (2000), Research releases, http://www.jup.com

Figure 9.9

PC penetration in Latin America, 1999–2005

of 14Gbps pales in comparison with 42Gbps (Asia to US) and 162Gbps (Europe to US) (Telegeography (2001)). An encouraging sign is the fact that, during the worldwide slowdown in bandwidth growth in 2001, Latin America represented the most growth in bandwidth, logging 480 per cent growth in 2001 and 290 per cent compound growth rates between 1990 and 2001. If sustained, this trend will improve Latin America’s Internet connection capacity to the global network. Another limitation in Latin America is the lack of interconnection between major network access points (NAPs) within the region. Most international traffic is routed to and from US NAPs, adding considerable costs to the users. A more concerted investment effort for interconnection is a significant issue if it is to stimulate e-commerce initiatives in the evolving regional economy. Finally, Latin America’s PC ownership, although growing steadily, was well below 10 per cent of population in 2000, compared to 55 per cent in the US. Projections indicate that Chile and Argentina will approach the 20 per cent level by 2005 (Figure 9.9). But others will struggle to reach the 15 per cent level. PC ownership is greatly influenced by the national income level, and without significant growth in GNP, Latin American countries must focus on providing Internet access through schools, workplaces and other public access sites.

Kagami 02 chap 06

2/15/04

2:54 PM

192

3.

Page 192

Country/area studies

E-COMMERCE IN LATIN AMERICA

Most estimates and projections for e-commerce in Latin America focus on the Brazilian market, which accounted for over 80 per cent of all e-commerce spending in the region in 1999. Business-to-consumer (B2C) online sales in Brazil were estimated at US$121 million in 1999, 62 per cent of the total US$194 million (Figure 9.10). Although projections can vary, Jupiter Communications considers this trend will continue. Brazil will account for more than 50 per cent of total US$8.3 billion B2C sales in 2005. Most of these figures are based on Brazil’s dominant position in Internet users and online sites. However, growing Internet users in Mexico, Argentina and Chile will present new e-commerce growth areas in the next few years. 140 120 US$ million

100 80 60 40

Source:

Other

Colombia

Peru

Venezuela

Chile

Argentina

Mexico

0

Brazil

20

Jupiter Communications (2002), Research releases, http://www.jup.com

Figure 9.10

B2C revenues, 1999

Like other countries, Latin American e-commerce is dominated by businessto-business (B2B) transactions. Pyramid Research estimated that B2C accounted for only 2 per cent of US$1.14 billion e-commerce revenues in 2001. Both B2B and B2C e-commerce are expected to grow larger but the B2B sector will continue to dominate (Figure 9.11). The lack of consumer buying in Latin America stems from several reasons. A survey by TGI.net Latina revealed that Latin American consumers have similar concerns about buying online (Zona

Kagami 02 chap 06

2/15/04

2:54 PM

Page 193

Internet and telecommunications outlook in Latin America

193

Latina (2001)). The number one reason why people didn’t buy on the Internet was that they liked to examine products before buying (47 per cent). But a large number of people also indicated that they were unfamiliar with e-commerce (44 per cent) and they feared for unsafe payment methods (42 per cent). 80 70

US$ million

60 50 40 30 20 10 0 2000

2001

2002 B2C

Source:

2003

2004

B2B

eMarketeer (2001).

Figure 9.11 B2C and B2B e-commerce revenues in Latin America, 2000–2004 Other peculiarly Latin American factors also hinder rapid growth in ecommerce. First, Latin America’s mail delivery service is relatively unreliable. Catalogue shopping is not as common as in the US. Second, some studies indicate that 74 per cent of Latin America’s B2C purchases are made with companies outside of the region, primarily in the US. Customs and taxes add costs and delay delivery to users. But even with added costs, many Internet shoppers use e-commerce to buy American products without leaving their countries. Third, the low level of credit card usage in Latin America presents another roadblock to increased e-commerce sales. In addition, real and perceived security issues are still discouraging many online shoppers. Finally, the low income level of Latin American consumers is a significant deterrence to future e-commerce growth. In contrast, B2B e-commerce is driven by large international corporations that have operations in Latin America such as Volkswagen and IBM and are expanding the use of the Internet to communicate with their suppliers in the

Kagami 02 chap 06

194

2/15/04

2:54 PM

Page 194

Country/area studies

region and with branches in the US and Europe. Some analysts predict that B2B commerce will grow to US$108 billion by 2005, from US$1 billion in 2001 (Pyramid Research (2001)).

4.

TELECOMMUNICATIONS DEREGULATION AND POLICY ISSUES

Any market potential in e-commerce and the region’s participation in the global digital economy will depend on improving telecommunications infrastructure and expanding Internet access to more people in diverse geographical and income groups. Toward these goals, most Latin American countries have been at the forefront of telecommunications reforms, market liberalization and privatization efforts during the 1990s. We now need to evaluate their effects on Internet access and the growth potential in advanced telecommunications services. 4.1

Telecommunications Reform

The telecommunications industry in Latin America experienced a full swing from privatization to nationalized monopoly back to private markets during the latter half of the 20th century. Most telecommunications services were provided by the private sector prior to being nationalized (Jamison (2000)). Private companies, under heavy pressure by governments to maintain low prices, invested little to improve basic network infrastructure and telephone services in Latin America were in general extremely poor. Nationalization and other forms of government intervention, however, did not produce improved networks and services. Low prices and general economic maladies prevented them from investing in the infrastructure. Beginning in 1978, when Chile began allowing competition in the telecommunications sector, improved service was again associated with liberalizing or privatizing the sector. Between 1984 and 1997, the majority (14 out of 24) of Latin American countries privatized their state-owned telephone companies, representing an estimated value of US$27 billion (Gutierrez and Berg (2000)), in the hope of improving telephone infrastructure. In addition, the biggest telephone market player, Brazil, privatized its Telebras system in 2000. Gutierrez and Berg compared market performance in terms of main telephone line deployment and found significant benefits from privatization efforts (Table 9.1). Table 9.1 compares 11 countries that privatized state-owned telephone companies with 12 that did not (including Brazil), using simple average annual growth rates. For non-privatized countries, the decades of the 1980s and 1990s were used as a dividing line. The growth rates are in the number of main

Kagami 02 chap 06

2/15/04

2:54 PM

Page 195

Internet and telecommunications outlook in Latin America

195

telephone lines per 100 inhabitants. While non-privatized countries also generated higher growth in the liberalizing 1990s, privatization seemed to have accelerated that growth in privatized countries to much higher rates. Individually, Bolivia (privatized in 1995), Guyana (privatized in 1991) and Peru (privatized in 1994) experienced the highest growth at 29 per cent, 24.1 per cent and 23.7 per cent, respectively. Among non-reformed countries, El Salvador (12 per cent), Guatemala (11.3 per cent) and Nicaragua (11.1 per cent) registered the highest growth rates. Table 9.1

Effects of privatization on main line deployment

Category

Privatized countries Non-privatized Source:

4.2

Annual growth (%)

Annual growth (%)

prior to privatization 5.7 1981–1989 5.1

after privatization 13.5 1990–1997 8.7

ITU data, calculated by Gutierrez and Berg (2000).

Effects of Reforms

While Table 9.1 shows a generally positive effect of privatization on improving telecommunications infrastructure, various studies on the interrelationship between reforms and results are not conclusive (Gutierrez and Berg (op. cit)). For example, some researchers found that market competition helped increase telephone penetration but not the privatization efforts. Others show contradictory results that privatization increased network expansion while competition did not. Privatization sometimes had positive effects on telephone deployment but only in those countries with relatively high income (above per capita GNP of US$1500). Therefore, improving telecommunications and Internet access is a matter of combined efforts in regulatory reforms, privatization as well as other positive developments in national income and political and economic environment. In their study of 19 Latin American studies, Gutierrez and Berg presented a ranking of explanatory variables in the order of their effects on improving telephone service (Table 9.2). Variables with the largest impact were population density, GDP per capita and cellular subscription rates, in that order. Telephone expansion was more likely when an area is more densely populated, has higher disposable income and is affected by competition such as cellular communications. These are geographic and economic variables. Regulatory frameworks and market liberalization variables also have significant effects, but to a lesser

Kagami 02 chap 06

2/15/04

2:54 PM

196

Page 196

Country/area studies

degree than the former. In any case, Latin American countries are experiencing substantial improvements in access to basic telecommunications networks whether due to market reforms or improving economic conditions. Table 9.2

Statistical explanatory power on teledensity

Aspects

Standard deviation

Population density GDP Cellular subscription Government type Regulatory framework Economic freedom Source:

2.89 1.45 1.36 0.78 0.95 0.21

Increase in teledensity (%) 13.10 6.57 6.13 3.53 4.30 0.93

Gutierrez and Berg (2000).

Other measurements of effective reforms are seen in the level of competition in telecommunications service markets and in growing broadband services. For example, consumers in Argentina by 2000 had four competitors for domestic and international long-distance service from only one in 1995 (Yomal (2000)). In addition to two privatized companies (Telecom Argentina and Telefónica Argentina), new entrants such as AT&T, Bell South-Movicom, and CTI-GTE are offering new services in pager, cellular and Internet markets. Fierce competition in the long distance service market is evident in the flat or diminishing levels of revenues during the period of 1996–2000, at just over US$10 billion (Pyramid Research (2001)). Mexico, Argentina, and Peru each had more than 20 operators in the long distance market in 2000. 4.3

Broadband Internet Access

For dial-up and broadband Internet access, both improvements in basic infrastructure (such as telephone lines, DSL networks and cable penetration) and market competition must occur simultaneously. Most Latin American countries are pushing toward market liberalization in the hope that competitive players will invest in basic infrastructure. However, most users are from upper to middle income groups who are located mostly in urban areas. Infrastructure investment is geographically limited and heavy competition is focused on a small group of affluent users who already have an Internet connection. Argentina is best poised for broadband expansion. With 5.2 million cable households, Argentina is the only Latin American country with more than a 50 per cent cable penetration rate (Figure 9.12). The Strategis Group

Kagami 02 chap 06

2/15/04

2:54 PM

Page 197

Internet and telecommunications outlook in Latin America

197

(http://www.strategisgroup.com) survey indicated that more than half of Latin American revenues from cable modems come from Argentina. Colombia (39.8 per cent), Mexico (16.2 per cent) and Brazil (7.2 per cent) have low penetration rates but represent substantial markets in numbers, 3.3 million, 2.8 million and 2.8 million households, respectively. Combined with expanding DSL offers by international entrants, consumers can expect growing opportunities for broadband Internet access. Nevertheless, sustained investment to upgrade networks and offer such services are necessary beforehand, for which great improvements are not expected. Deteriorating economic conditions also mean the prospect of full deployment of 2.5G wireless is in doubt and the 3G licensing timetable is being postponed. 60 50 40 %

30 20 10

Source:

Brazil

Peru

Venezuela

Mexico

Chile

Puerto Rico

Colombia

Uruguay

Argentina

0

Deutsche Bank (2000).

Figure 9.12

Cable penetration rates, 2000

The DSL revolution being experienced in Asia, especially in Korea, is influenced by market reforms as well as a fierce competition among multiple service providers. Simply liberalizing and opening up the telecommunications market will not guarantee improvements in Internet access and the availability of advanced telecommunications services to the general population. The most serious problem in the Latin American telecommunications market is the cannibalization of the existing market by new entrants without significantly enlarging the number of people who are connected and can afford to subscribe

Kagami 02 chap 06

198

2/15/04

2:54 PM

Page 198

Country/area studies

to new service offerings. Although recent reform efforts have stimulated better than average improvement in Internet access, market liberalization alone will not be enough to sustain such growth or to extend Internet access to the universal service level envisioned by policy-makers.

5.

EXPANDING INTERNET ACCESS IN LATIN AMERICA

Many developing nations, and especially disadvantaged groups in those countries, are still struggling to access traditional telephone services. Telephone lines in many low income countries are less than two per 100 inhabitants, compared to 70 for the US and an average of 39 for European countries, according to International Telecommunication Union (ITU) data for 2000. While the Telecommunications Act of 1996 in the US represented a shift in universal service policy from telephone service to advanced telecommunications services, many Latin American countries face the daunting task of improving basic telecommunications infrastructure while at the same time attempting to expand the level of access to the Internet. Universal service objectives in developed countries are concerned with ‘not leaving someone behind’ in the digital revolution. In contrast, many Latin American countries find themselves tackling the age-old problem of extending basic telecommunications services to the majority of their population. Some of the new technologies indeed help latecomers. For example, with satellite and cellular networks, countries are no longer dependent on fixed line networks that can be very expensive to connect to sparsely populated areas. Distance is now a lesser barrier to universal service than before. Declining costs also enable countries to accelerate infrastructure deployment in urban as well as rural areas. These and other socio-economic conditions such as market liberalization and global competition paint a very positive scenario under which Internet, wireless, satellite and associated telecommunications technologies will enable developing countries to catch up with developed nations faster than they thought they ever could. For example, Internet access centres and telecentres can be established in any remote area with a satellite uplink, providing voice as well as data services for a large number of inhabitants. Nevertheless, most Internet access modes require fixed lines such as plain old telephone networks, or coaxial cables. Thus, an insufficient development in basic telecommunications infrastructure continues to be a critical obstacle in expanding Internet access in those countries. Wireless communications, despite lofty expectations, are stumbling to come into the market. High investment costs associated with expensive spectrum licensing also tend to preclude any possibility that wireless providers will offer affordable services to low income populations. Under these circumstances, universal service objectives in the

Kagami 02 chap 06

2/15/04

2:54 PM

Page 199

Internet and telecommunications outlook in Latin America

199

Internet age have not changed significantly from those for conventional telephone access. Policies to expand access to telephone service and the Internet can be evaluated by three broad measures: availability, accessibility and affordability. Availability refers to the network’s penetration into households. Accessibility initiatives focus on providing access points such as public telephones and community Internet access sites. Affordability is measured in terms of price or cost of connection weighted by the level of income. There are policies and programmes designed to improve each measure, and we investigate these in turn below. 5.1

Availability through Privatization or Competition

The impact of new technologies such as cellular phones on network accessibility is apparent from Figure 9.13 that shows the number of telephone lines per 100 inhabitants in selected countries. In Chile, for example, there were 22.4 cell phone lines per 100 inhabitants in addition to 22.1 fixed telephone lines. Inclusion of cell phone lines improves telephone availability in Latin America significantly, compared with Figure 9.10 that considered only fixedline telephones. 50 45 Per 100 inhabitants

40 35 30 25 20 15 10 5 0 Argentina

Brazil

Chile Cell phone

Source:

International Telecommunication Union.

Figure 9.13

Telephone availability, 2000

Mexico Phone

Peru

Venezuela

Kagami 02 chap 06

200

2/15/04

2:54 PM

Page 200

Country/area studies

Nevertheless, Fuentes-Bautista (2001) pointed out that the best performers, Chile and Argentina, have no flexible universal service objectives within their national telecommunications reform agenda. Both Chile and Argentina began market liberalization early, in 1988 and 1990 respectively, and competition is significant in the telecommunications sector. In other countries, the privatization process was aided by grants of monopoly privileges to private companies in order to make the sales more palatable. But even with explicit universal access clauses, these countries are finding it difficult to inject any incentives to improve availability. Privatization does not necessarily amount to market competition. 5.2

Expanding Accessibility

Accessibility measures usually consider whether there are geographically usable access points for any given population. Public telephones and community technology centres for Internet access are designed to increase accessibility. Figure 9.14 shows that Brazil, Peru, Mexico and Venezuela, which were relatively low in availability measures, successfully expanded public phone lines following telecommunications reforms. Reforms are characterized by the privatization, partially or completely, of state-owned telecommunications companies or the introduction of competition in local or long distance services. In Figure 9.14, the year of reform or liberalization is marked with a rectangular box (Argentina 1990, Brazil 1995, Chile 1988, Mexico 1990, Peru 1994, Venezuela 1991). In most cases except Chile, reforms appeared to expand public accessibility. Chile has by far the best availability (Figure 9.13) due to a competitive market environment. Chile’s inability to increase accessibility highlights the difficulty in balancing private initiatives with universal service policy goals. Like public phones, public access to the Internet is also augmented by community technology centres and public Internet access sites. Mostly funded by governments, community technology centres are designed as a community support facility, such as Peru’s Red Científica Peruana, Chile’s Centro Comunitario Internet El Encuentro, Colombia’s Neighborhood Information Units, and Brazil’s Telecentres. Compared to the US where community access sites are primarily used by the low income population, Latin America’s public access sites are used by middle income as well as low income groups (Fuentes and Straubhaar (2001)). Users’ age and education levels also differ significantly. In Peru, for example, most telecentre users are under 30, while US users are more evenly distributed by age groups. High school and college graduates dominated in Peru, while the largest group of users in the US are those who didn’t finish high school. Clearly, community centres in Latin America are meeting the unfulfilled demands of its population, while US sites serve primarily its universal service objectives.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 201

Internet and telecommunications outlook in Latin America

201

Payphones for 1000 population

3.5 3 2.5 2 1.5 1 0.5 0 1990

1991

1992 1993

1994

Argentina Mexico Source:

1995

1996 1997

1998

Brazil Chile Peru Venezuela

International Telecommunication Union.

Figure 9.14

Telephone accessibility, 1990–1998

A different kind of public Internet access sites are commercial Internet cafés and cybercafés. In Mexico, there were 286 registered cybercafés in 1998. These public sites can offer a very inexpensive opportunity to access the Internet. For example, ‘PC-bangs’, Korea’s cybercafés and game rooms, charge between US$1 and US$2 an hour and they are open 24 hours a day. There were over 21 000 PC-bangs in Korea at the end of 2000, greatly expanding Internet access to beyond home PC users. The majority of Korean PC-bang franchises are connected to the Internet at 512Kbps or T1. Although PC-bang users are young adults who primarily play interactive games, such Internet access sites offer ready and easy access to the Internet on virtually every city’s streets, villages, near train and bus stations and in resort areas. Operating mainly as a for-profit business, PC-bangs have greatly increased Internet access in a relatively short time. 5.3

Costs and Affordability

During the previous decade, market reforms and privatization efforts in Latin American countries greatly improved public access to telecommunications networks. Nevertheless, they lag far behind the US, European and some Asian countries, affording less than 50 per cent of the population with a basic telephone

Kagami 02 chap 06

2/15/04

2:54 PM

Page 202

202

Country/area studies

service. At the same time, some telephone operators claim that there is not enough demand for such service. According to a World Bank estimate (1995), 91.1 per cent of telephone demand was met in Mexico. This measure is based on the average waiting time for a new telephone line (one month for Mexico). This peculiar aspect is explained as ‘depressed demand’ by Fuentes-Bautista (op. cit.). In other words, those who are in the high income group already have access to the telephone or have cellular phones, while those in the low income bracket have no demand due to high costs. Lack of competition will increase prices beyond the reach of many. The emphasis on privatization instead of market competition has resulted in high costs, low quality and discouraging levels of service. The giant Telmex, for example, has improved Mexican telecommunications infrastructure somewhat, but it fails to offer reliable, quality services at reasonable prices. The cost of a monthly telephone subscription in Latin America far outstrips the per capita income level. In Peru, for example, the monthly residential subscription rate was seven times greater than monthly per capita income in 1998 (Figure 9.15). One third of annual income would have to be spent on telephone fees in Chile and Mexico. Rates for cellular phone subscription are at least 8 7.0

7 6 5 4.0

3.9

4 3 2

2.4 1.9

1.7

1 0 Argentina Figures: Source:

Brazil

Chile

Mexico

Peru

Monthly residential phone subscription/monthly per capita income. International Telecommunication Union.

Figure 9.15

Telephone affordability, 1998

Venezuela

Kagami 02 chap 06

2/15/04

2:54 PM

Page 203

Internet and telecommunications outlook in Latin America

203

twice that of a fixed line telephone. At such high prices, only a limited portion of their population can afford any telecommunications service. In order to access the Internet, consumers must pay for a telephone connection as well as subscription fees for ISPs. Therefore, any effort to expand Internet access must include both rate restructuring in local telephone services and similar policies targeted at Internet subscription and interconnection fees. With metered billing for a local connection and ISP connection charges that often levy long distance fees, many Internet users must pay an enormous amount in fees for a minimum level of Internet access. Some innovative schemes have been used to introduce different pricing regimes for traditional voice traffic and Internet data traffic. Argentina, in particular, experimented with a special long distance area code for dial-up users to encourage Internet access. Its effect will be discussed in more detail in the next chapter. The important aspect of this experiment was the realization that simple privatization did not consider fundamental differences between conventional phone networks and Internet access. Even though they might be transferred through the same physical networks, usage patterns differ widely and rates and tariffs must be altered accordingly in order to stimulate Internet usage. When income levels are comparatively low, small variations in access modes and charges may provide additional incentives for those who are denied Internet service despite their willingness to participate.

6.

OUTLOOK FOR FUTURE INTERNET GROWTH

The basic premise that pushed for privatization and market competition in the telecommunications industry was that these measures would improve investment in basic telecommunications infrastructure and access to telecommunications services, including the Internet, for the masses. The vigour toward privatization and competition was one constant policy variable in most Latin American countries during the past ten years. Nevertheless, the result is mixed. While there certainly appear to be some positive developments in telephone and Internet access measures, the growth trend or current momentum does not seem to be sufficient to bring such services to the majority of their population. Are privatization and competition the answer to gaining universal access? The fundamental roadblock seems to be the level of income that is unable to support a wider diffusion of telecommunications technologies. Privatization initiatives, while removing government controls, leave pricing to private companies at a level such that the majority of the population cannot afford basic service. While they were designed to promote competition, which was expected to lower prices, the reality is that market forces seldom satisfy such objectives as universal service. Companies are eager to offer services to those who can

Kagami 02 chap 06

2/15/04

2:54 PM

204

Page 204

Country/area studies

afford them while market conditions often discourage them from investing in remote, underserved areas. More importantly, deregulation may not be the answer in promoting wider telephone and Internet access. While others investigated the difference between privatization and competition, Guillén and Suarez (2001), based on their study of 141 countries, argued that they both had little impact on increasing the number of Internet hosts and users. Rather, an index of predictable policy-making – as an indicator that a business environment is unlikely to change without warning – is found to be most significant, after such usual variables as income levels and telephone lines, in explaining the growth in Internet hosts and users. A predictable and stable policy-making environment is a necessary condition to encourage investment and entrepreneurship. Although Argentina is one of the leading Internet hotbeds in Latin America, its current status pales in comparison with Ireland and Singapore. For example, Argentina has a relatively liberal telecommunications policy but somewhat unfavourable environment for entrepreneurship conditions. The Argentine government intervened to bring down leased line prices in 1997 – the creation of special area code (0610) for Internet usage. But small Internet startups were gobbled up by large, vertically integrated players who control the telecommunications market and are heavily financed by foreign interests such as Stet/France Telecom and Telefónica of Spain. Singapore’s telecommunications policies are relatively interventionist in that service operators and ISPs are tightly controlled by the government. But regardless of the policy framework, both Ireland and Singapore are far ahead of Argentina because of their favourable business environment. This result signals a very difficult problem for most Latin American countries in their drive to expand Internet access. The hurdles they need to overcome are not simply those of privatization or opening up their markets to competition. The overall income level, economic policies, and political stability all contribute to a success in the Internet age. Market initiatives, which are often made ineffective by monopolized private companies, may not bring about substantial gains that enable the majority of their population to participate in the digital revolution.

REFERENCES Business News America (2001), ‘Incumbents Increase Control of ISP Accounts’, 9 July. Deutsche Bank (2000), ‘Equity Research: Latin American Media’, http://www. growthmarkets.com/pdf/GM_LatMedia.pdf eMarketer (2001), ‘The Web in Latin America: A Status Report’, http://www. latinoutdoor.com/Speakers/Presentations/Noah_Elkin.pdf

Kagami 02 chap 06

2/15/04

2:54 PM

Page 205

Internet and telecommunications outlook in Latin America

205

Fuentes, M. and J. Straubhaar (2001), Improving Public Internet Access in Brazil: Moving Beyond Connectivity, Telecommunications and Information Policy Institute, University of Texas at Austin. Fuentes-Bautista, M. (2001), ‘Universal Service in Times of Reform: Affordability and Accessibility of Telecommunication Services in Latin America’, 29th Research Conference on Communications, Information and Internet Policy, September. Guillén, M. and S. Suarez (2001), ‘Developing the Internet: Entrepreneurship and Public Policy in Ireland, Singapore, Argentina, and Spain’, Telecommunications Policy, Vol. 25, pp. 349–71. Gutierrez, L. and S. Berg (2000), ‘Telecommunications Liberalization and Regulatory Governance: Lessons from Latin America’, Telecommunications Policy, Vol. 24, pp. 865–84. International Telecommunication Union (ITU) (2001), World Telecommunication Indicators. Jamison, M. (2000), ‘Strategies for Latin America in the Global e-Economy’, http://bear.cba.ufl.edu/centers/purc/primary/jamison/mjLABCpaper.pdf Pyramid Research (2001a), B2B Portals and Marketplaces: e-Business Strategies in the Productive Sector. Pyramid Research (2001b), Benchmark Research Report: The Death of Long Distance in Latin America, Cambridge, MA. Telegeography, Inc. (2001), Packet Geography 2002, Washington, DC. US Department of Commerce, International Trade Administration (US DOC, ITA) (2000), Export IT Latin America. World Bank (1995), ‘Performance Indicators for Telecommunications Sector’, http://www.worldbank.org/html/opr/pmi/telecom/telecom0.html. Yomal, R. (2000), ‘Telecommunications: Now Playing with a Full Deck’, Comments Argentina, July, pp. 14–15. Zona Latina (2001), ‘Why Latin Americans Buy (or not Buy) over the Internet’, available online at http://www.zonalatina.com/Zldata181.htm

Kagami 02 chap 06

2/15/04

2:54 PM

Page 206

10. Policies for Internet access: cases of Mexico and Argentina Soon-Yong Choi 1.

INTRODUCTION

Mexico and Argentina, besides Brazil, are the largest economies in Latin America, with populations of 99 million and 37 million, respectively, and gross domestic products of US$484 billion and US$283 billion, respectively, in 2000. In terms of Internet population, these countries are also ranked within the top 20 in the world. Nevertheless, Hong Kong with a population of just under seven million has more Internet population than either of these countries, according to the 2001 Nielsen/Net Ratings Global Internet Index. Typical of other Latin American countries, both Argentina and Mexico have a 10 per cent or lower rate of Internet penetration while leading Internet nations surpass the 50 per cent rate. Certainly, this disparity is in part due to their comparatively low GDP and per capita income levels. The inadequacy in basic telecommunications infrastructure is another main factor that contributes to their low rates of Internet penetration. Associated problems of high costs for telephone/Internet access, limited network service options, and uneven income distribution all contribute to their low performance. Nevertheless, several policy initiatives were put in place during the last ten years, mainly the privatization of state-owned telephone monopolies and the introduction of competitive, open telecommunications policy frameworks. As a result, substantial progress has been reported in terms of telephone and network availability, as discussed in the previous chapter. However, case studies of Mexico and Argentina will offer some lessons on the overall effectiveness of these initiatives and further understanding of their prospects for full participation in the digital revolution in the coming years. Mexico represents in many ways a typical case of a Latin American political, economic and regulatory environment where a large privatized company, Telmex, dominates the telecommunications industry. Argentina, offering a more pro-competitive market, has instituted several innovative strategies, such as a special area code for Internet dial-up access. Argentina’s experience was 206

Kagami 02 chap 06

2/15/04

2:54 PM

Page 207

Policies for Internet access: Mexico and Argentina

207

one of the fastest growths among all Latin American countries. This case study is intended to focus on those factors that made it a success story and evaluate them for policy effectiveness in a long-term perspective.

2.

TELECOMMUNICATIONS IN MEXICO

2.1

Teléfonos de México

Teléfonos de México (Telmex), the giant monopoly, traces its history back to 1903 when Sweden’s Ericsson founded Mexeric in cooperation with Mexican interests. In 1947, Mexeric became Telmex along with other investment groups including the government of Mexico. The government’s share in the enterprise remained at 48 per cent until 1972 when Echeverria’s government took majority control of the company. As a state-owned monopoly, its performance was more focused on generating a steady flow of income rather than improving the network infrastructure and its service level. Privatization took place in 1990 when a consortium of Grupo Carso, BellSouth and France Telecom obtained 51 per cent of Telmex stock. In return for local monopoly status, Telmex was required to achieve a 12 per cent increase in main telephone lines annually through 1994 and to expand payphone access to various underserved communities. By 1996, the legal limit for local ownership was lifted and the government sold its share of Telmex stock to international investors. Thus, 1996 marks the complete privatization of the government-owned telephone industry. The long distance service market became competitive in early 1997. Although local telephone service markets were organized as a duopoly by allowing a competitor, Telmex maintains a strong, monopoly position in most markets. The mobile cellular market was also established as regional duopolies, for which Telmex is a main player through its subsidiary, Telcel, and although there are seven major regional cellular players, Telcel is the only player that provides nationwide coverage. Unlike the US, cellular charges are paid by callers, called ‘calling party pays’ (CPP) policies that are used in many Asian and European countries. Cable television subscription is low, at below 20 per cent, far behind Argentina, Uruguay and Colombia. Although the cable TV market is fragmented, the largest operator, Cablevision, is owned by Televisa, a network operator, and Telmex. High speed Internet access through cable modem faces the stiff challenge of first bringing the coaxial/fibre optic network to a larger part of the population. The government of Mexico now controls telecommunications policy through an independent regulatory body, Comisión Federal de Telecommunicaciones (COFETEL), established in 1996 as part of privatization initiatives. COFETEL

Kagami 02 chap 06

2/15/04

2:54 PM

Page 208

208

Country/area studies

oversees various regulatory activities such as registration of tariffs and equipment certification as well as spectrum auctions for new communications services. 2.2

Industry Overview

Mexico’s telecommunications market generated revenue of US$13.7 billion in 2000, second only to Brazil’s US$24.9 billion. The largest market segment is voice service over fixed telephone lines, followed by mobile voice services (22 per cent) (Figure 10.1). Internet service revenues accounted for 8 per cent of total market revenues. In 2000, the number of cell phone lines per 100 inhabitants exceeded that of fixed telephone lines (Figure 10.2), and is expected to grow even more in the coming years. Pyramid Research projects that mobile service revenues will account for 28 per cent of the US$30 billion total by 2005. The next most active growth area is Internet service, projected to be 20 per cent of the total. Figure 10.2 also shows the effect of Telmex privatization during the 1990s. Till 1994, Telmex was required to fulfil Mexico’s universal service objectives by increasing main lines annually. Once that obligation expired, the figure shows a clear slowdown, even a dip in 1996, in the penetration rate. Most Internet 8% Data comm 6% Pay TV 6% Telephone 58%

Mobile 22%

Source:

Pyramid Research (2001).

Figure 10.1

Communications service revenue shares, 2000

Kagami 02 chap 06

2/15/04

2:54 PM

Page 209

Policies for Internet access: Mexico and Argentina

209

30

Per 100 inhabitants

25 20 15 10

Telephone Source:

Cell

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

0

1987

5

Total

ITU (2000).

Figure 10.2

Telephone penetration rates, Mexico, 1987–2000

subsequent growth occurred mainly through the growth in mobile line subscriptions, and to a lesser degree through increased deployment of payphones. Nevertheless, the combined fixed and mobile telephone lines represent 26.7 phone lines per 100 inhabitants, behind Chile, Argentina and Brazil. Telmex also plays a leading role in Internet services, being the largest ISP. Incumbent telcos in Latin America controlled 23 per cent of the market in 1999, extending their grip to over 40 per cent by the end of 2000 (The Yankee Group (2001)). With Telmex’s monopoly over local exchange customers, the Internet access market’s consolidation into a few providers will continue. For Internet traffic, there are few alternative connection schemes. Mexico’s international backbone is limited to a 1Gbps connection from Mexico, DF, to Dallas/Fort Worth NAP. Regional distribution goes through three regional hubs, Guadalajara, Puebla and Monterrey, with separate lines running from Monterrey to Chihuahua and Mexicali, and from Puebla to Merida in Yucatán. From the Latin American region, the major international line is from São Paulo, Brazil to Miami (and from Miami to Madrid), but there have been few regional interconnections within the Latin American regions. This means that communications from Mexico to Brazil are often relayed through US network access points, adding considerable costs to Latin American Internet users.

Kagami 02 chap 06

2/15/04

2:54 PM

210

Page 210

Country/area studies

In 1998, Mexican consumers paid one of the highest rates for access to the Internet. An OECD survey (1999a) showed that the largest Mexican ISP, Telmex Internet Directo charged US$24.78, the highest among all OECD countries, for three hours of access per month. Since it had a flat-rate plan, the relative amount decreased when one considered 20 or more hours of access. Telmex Internet Directo’s pricing structure was significantly skewed toward charging low for telephone service and levying a hefty sum for Internet access. Telephone charges accounted for only 18.6 per cent of the total basket, compared to a 56.5 per cent average for OECD countries. Figure 10.3, showing a comparison of the total Internet access basket (telephone charges plus ISP Internet access fees), is reproduced here from the earlier chapter for reference. 80 70 60 US$

50 40 30 20

ISP

Germany

Switzerland

Greece

Japan

Mexico

Korea

UK

US

Turkey

Canada

0

Finland

10

Telephone

Based on the largest ISP in each country, for 20 hours online access a month. In US$ purchasing power parity. Source:

OECD (1999a).

Figure 10.3 Comparison of total Internet access charges, off-peak rates, 1998 By 2000, the situation had been reversed. According to a new OECD survey (2001), Mexico now ranks as one of the countries with the lowest Internet access costs (Figure 10.4). Shown in purchasing power parity dollars, the total cost in Mexico was US$37.4 compared to an OECD average of US$66.4. Some 66

Kagami 02 chap 06

2/15/04

2:54 PM

Page 211

Policies for Internet access: Mexico and Argentina

211

US Turkey

Fixed telephone

Canada

Telephone usage

Mexico

ISP

New Zealand Australia Korea Finland Iceland Italy Denmark Japan Germany Greece Sweden France UK Norway Switzerland EU OECD Austria Ireland Spain Portugal Belgium Netherlands Luxembourg Poland Hungary Czech Republic 0.0 Source:

50.0

100.0

150.0

200.0

OECD (2001a).

Figure 10.4 Comparison of total Internet access charges, 40 hours at peak rates, 2000

Kagami 02 chap 06

212

2/15/04

2:54 PM

Page 212

Country/area studies

per cent of the charges were local telephone fixed charges. ISP charges have been reduced greatly while telephone charges are no longer levied per usage. Notably, most low cost countries do not include usage-based telephone charges. This certainly shows a complete reversal in the way consumers are charged for telephone and Internet access in Mexico.

3.

MEXICAN TELECOMMUNICATIONS REFORMS

3.1

Market Liberalization and Regulatory Experience

By January 1999, the OECD’s survey showed that all local, trunk and international telecommunications markets in Mexico were competitive, with ten, fourteen and seven players in each market segment, respectively. The market share of new entrants in the international market share jumped from zero to 32 per cent in one year following liberalization in 1996. The analogue cellular market was organized as a duopoly. Mexico’s new regulatory law enacted in 1995 allowed competition in the entire telecommunications market, including local, long distance, cellular, cable and satellite communications. The unexpectedly large gains by competitors were alarming to Telmex, which leveraged its market position in commercial negotiations with competitors to set interconnection charges, so that regulatory bodies faced the first challenge in managing the new competitive, marketoriented regulatory environment, where the incumbent Telmex still controlled local exchange services as well as long-distance and mobile services. The regulatory agency, Secretaría de Communicaciones y Transportes (SCT), despite its initial reluctance to intervene, had to settle the interconnection disputes between Telmex and seven long distance providers, among which Avantel (Banamex and MCI) and Alestra (Alfa Telecom and AT&T) represented major US interests. The SCT was not prepared to face the fact that even when markets are liberalized, the government’s role continues in mediating disputes among the players. COFETEL, created in 1998, managed to resolve it quickly, by replacing a 58 per cent surcharge or settlement fee with a fixed fee of US$422 million. While Telmex leveraged its monopoly position, Avantel and Alestra put pressure on through NAFTA and WTO agreements. But Mexico’s professed goal of letting market negotiations resolve any disputes meant that the regulatory authorities could play only a minor role as an arbiter. This tended to strengthen the position of the incumbent monopolist such as Telmex. Similar tariff disputes arose in regard to cellular telephony, where Telmex used high interconnection charges as an anti-competitive tool.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 213

Policies for Internet access: Mexico and Argentina

3.2

213

Indicators of Success

Perhaps the best indicators of telecommunication reforms are the growth trends in telephone infrastructure investment and penetration rates and the measure of telephone affordability. From the initial privatization in 1990 until 1994, Telmex was required to invest in telephone lines. The effect is seen in the steady growth in the number of main lines per 100 households (Figure 10.5). 60

Per 100 households

50 40 30 20

Mexico

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

0

1987

10

World

World figures (not to scale) only shows the growth trend in total main lines. Source:

ITU (2000).

Figure 10.5

Main telephone lines per 100 households, 1987–1998

However, after the market was completely liberalized and investment decisions were left entirely to market forces, the rate growth had been stagnant and even negative. As shown in Figure 10.2, most increases in telephone access occurred via mobile phone subscriptions since 1996. If fixed and mobile lines were simply added, as we have done in Figure 10.2, Mexican telephone lines grew steadily. If, however, we consider some duplication of subscribers between fixed and cell phones, the net effect would be a negative growth in telephone penetration. Considering the high costs of cellular phones, it is highly likely that cell phone subscribers were indeed also subscribers of fixed telephone lines.

Kagami 02 chap 06

2/15/04

2:54 PM

214

Page 214

Country/area studies

During the late 1990s, it was apparent that high income, high margin services were expanded while basic telephone services suffered from both insignificant investments in infrastructure and continuing increases in costs. Monthly telephone subscription fees were 1.8 times greater than monthly per capita income in 1990 (Figure 10.6). 16

Mth charges/mth income

14 12 10 8 6 4

Fixed Source:

1998

1997

1996

1995

1994

1993

1992

1991

1990

0

1989

2

Mobile

ITU (2000).

Figure 10.6

Changes in telephone affordability in Mexico, 1989–1998

This steadily increased to four times income level by 1998. At the same time, market competition in high end cellular services brought down cell phone prices substantially, from over ten times monthly income at the beginning to a low of seven times by 1998. If telecommunication reforms were primarily targeted to increase investments in basic infrastructure, expand access, and lower prices, privatization and competition have hardly been adequate to meet those purposes. 3.3

Evaluation

The OECD’s evaluation (1999b) of Mexican reforms summarized five benefits for Mexico through: • Reducing prices and increasing quality for Mexican consumers; • increasing competitiveness and productivity of the industry;

Kagami 02 chap 06

2/15/04

2:54 PM

Page 215

Policies for Internet access: Mexico and Argentina

215

• Promoting innovation and new technologies; • Increasing the adaptability of the Mexican economy; • Establishing efficient regulatory institutions and methods. However, on most points, our previous discussion illustrates differing evaluations of the reforms’ effects. Prices and affordability for consumers have not improved, if not worsened. Although new technologies such as cellular communications grew rapidly, its beneficiaries seemed to be limited to those who already had access to telephone service. Local Telmex struggled to survive in the face of foreign-financed competitors. Instead of improving cost structure and productivity, Telmex mostly relied on leveraging its monopoly status to survive. The SCT and COFETEL found that there were limits in mediating disputes when commercial agreements largely replaced published rate and tariff schedules. Finally, whether the Mexican economy became more agile and adaptive to economic shocks remains to be seen. In addition to Telmex, Mexico’s other privatized giant, Satelites Mexicanos (Satmex) also faced stiff competition from the US and European companies that were entering the market in late 2001 (Los Angeles Times (2001)). The entry by PanAmSat, mostly owned by Hughes Electronics Corp., SES Global based in Luxembourg (buying GE Americom) and Televisa signals an opportunity to expand advanced telecommunication services such as wireless. But opening up the market to competition does not automatically improve services. Although Telmex is privatized, it still behaves very inefficiently as it did under government ownership. At the same time, foreign competitors, who may focus only on high revenue consumers, may actually retard universal service goals. The rationale for market liberalization stems from the belief that market forces will make local companies more competitive, which will ultimately benefit consumers. But the devil is in the details.

4.

INTERNET AND E-COMMERCE IN MEXICO

As in other Latin American countries, business-to-business (B2B) transactions dominate Mexico’s e-commerce. According to one estimate, B2B accounted for 70 per cent of all online transactions in Mexico in 2000. Major multinational manufacturers are at the forefront of B2B applications. For example, Cemex, the largest cement-maker based in Monterrey, began a US$50 million project on B2B e-commerce. The Mexican government for its part enacted a law in 2000 that legally equates electronic contracts and transactions with offline counterparts, augmenting a more favourable legal environment for e-business. While B2B e-commerce revenues are estimated to be in the billions of dollars,

Kagami 02 chap 06

2/15/04

2:54 PM

Page 216

216

Country/area studies

business-to-consumer (B2C), or the online consumer retailing market is relatively small. Mexico is the second largest e-commerce market in Latin America (Figure 10.7). Overall in Latin America, the largest B2C market segment is the online auction (33 per cent), followed by computer hardware and software (12 per cent), financial services (11 per cent), and entertainment (10 per cent) (Curry et al. (2001)). US$ million Others 104

Argentina 85

Brazil 300

Mexico 91 Source:

Curry et al. (2001).

Figure 10.7

B2C, retail e-commerce revenues, 2000

Most Internet users access websites via dial-up services through an ISP, in school or at work. Telmex, at 56 per cent, dominates the ISP market. Most traffic is directed to major portal sites that are mostly operated by groups based in Mexico, the US and Argentina (Table 10.1). T1msn is a new joint venture between Telmex and Microsoft; and Todito is owned by TV Azteca and Esmas by Televisa. Terra and El Sitio operate in many Latin American countries. Major B2B sites include Cemex (www.cemexsuppliers.com) which is owned by the large cement manufacturer based in Monterrey, which also operates an advertising and information site (www.cemex.com). Similar to most portal sites, Mexican, Latin American and US-based Spanish portal sites also incorporate retail sales on their sites, but there are a number of sites specializing in online retailing and other services (Table 10.2). Although

Kagami 02 chap 06

2/15/04

2:54 PM

Page 217

Policies for Internet access: Mexico and Argentina

Table 10.1

Major portal sites in Mexico

Name

URL

Yahoo! StarMedia Yupi El Sitio Todito Esmas Infosel AOL Latin America T1msn Terra Umbral Mexis Adnet Quepasa

mx.yahoo.com www.starmedia.com www.yupi.com.mx www.elsitio.com.mx www.todito.com www.esmas.com.mx www.infosel.com www.americaonline.com.mx www.t1msn.com.mx www.terra.com.mx www.umbral.com www.mexis.to2.com www.adnet.com.mx www.quepasa.com

Table 10.2

217

Users

7.6m 2.8m 250k 60k 1.3m 250k

60k

Primary location US US US Argentina Mexico Mexico Mexico US/Mexico US/Mexico Mexico/L.A. Mexico Mexico US

Major online retail sites in Mexico

Name

URL

Location Products

Agugutata.com Comprasdirectas Compuprice Decompras Despagar Eshop.com Estafeta Floresflowernet Laborum Liverpool Mercadolibre Notimex Notivisa Pl@zabanamex Roberts.com Rshopping.com Sanborn’s Submarino Ticketbus.com VirtualPlaza ZonaFinanciera

www.agugutata.com www.comprasdirectas.com www.compuprice.com.mx www.decompras.com www.despagar.com ws3.tecnofin.com.mx www.estafeta.com.mx www.floresflowernet.com.mx www.laborum.com www.Liverpool.com.mx www.mercadolibre.com www.notimex.com.mx www.notivisa.com www.plazabanamex.com.mx www.Roberts.com.mx www.rshopping.com.mx www.sanborns.com.mx www.submarino.com.mx www.ticketbus.com.mx www.virtualplaza.com.mx www.nonafinaciera.com

Mexico Mexico Mexico US US Mexico Mexico Mexico Chile Mexico US Mexico Mexico Mexico Mexico US Mexico Mexico Mexico Mexico US

Baby products Retail Computer Retail Travel Electronics Shipping Flowers, gifts Jobs Dept. store Auction News News Retail Men’s clothing Retail Dept. store Books, CDs Bus ticket Mall Financial

Kagami 02 chap 06

218

2/15/04

2:54 PM

Page 218

Country/area studies

there are no specific data, many Mexican shoppers use the Internet to buy products from the US. As a result, many Spanish language sites in the US cater to Mexican e-commerce from their US sites as well as through establishing a Mexican web presence. Future prospects for e-commerce in Mexico will ultimately depend on the continuing growth in Internet access. However, low telephone penetration, high Internet access costs, low PC ownership and slow broadband expansion will become significant barriers to e-commerce growth. Fundamentally, local telephone access charges can add up to a prohibitive amount since local calls are metered. Even with new free ISPs (Gratis1 and Terra), local connection is a serious barrier to e-commerce growth. Other structural problems include consumers’ concerns about payment security. Also, the low usage of credit cards among Mexicans prevents many of them from participating in online commerce. Finally, most websites on the Internet are not in Spanish, limiting availability and accessibility, although most Internet users in Mexico are more or less proficient in English, to the level that they prefer US-based, English language sites for more reliable information and products.

5.

TELECOMMUNICATIONS IN ARGENTINA

After Brazil and Mexico, Argentina has the third largest Internet population in Latin America, with 2.5 million users in 2000. Although these three countries are similar in many aspects of telecommunications development, Internet usage and reform policies, Argentina is unique in some aspects. Argentina has the highest per capita income among all major Latin American countries (US$7731 in 2000). Per capita IT spending at US$68.7 is larger than that of Chile or Brazil, although only slightly, and far greater than that of Mexico. A particularly significant fact is its high cable penetration rate, which, at 57 per cent, is comparable to the US and other developed countries. Argentina’s six million cable subscribers account for almost half of Latin America’s cable users. In this respect, Argentina is well positioned to lead the broadband transition in Latin America. Although initial expectations were that its extensive cable networks would extend basic telephone access, it is more likely that its high cable penetration will impact future Internet growth very positively. The state-owned monopoly, Empresa Nacional de Telecommunicaciones (ENTel) was privatized in 1990, broken into two regional monopolies, each with a share of the profitable Buenos Aires market. Telefónica de España (Telefónica) owns the majority stock of the southern regional operating company. The northern regional company, Telecom Argentina (Telecom) is majority owned by France Telecom and STET of Italy. As in Mexico’s reform efforts, both Telefónica and Telecom were required to invest in expanding the

Kagami 02 chap 06

2/15/04

2:54 PM

Page 219

Policies for Internet access: Mexico and Argentina

219

basic telephone infrastructure in exchange for a local monopoly. By 1998, both companies had fully converted their lines into digital networks, while adding more lines and introducing DSL technologies. Since 1999, Argentina’s wireless providers, Movicom (lead by BellSouth) and CTI (GTE-led consortium) were also allowed to offer fixed line telephone services by the Comisión Nacional de Comunicaciones (CNC), a quasi-independent regulatory body created in 1996. In 1997, Argentina instituted major initiatives to revamp its telecommunications policy and to enhance the lagging Internet industry. Since ENTel’s privatization, telephone access grew steadily, second only to Chile in 2000. But compared to Chile, Mexico and Brazil, Argentina was far behind in several measures of Internet development. For example, Internet penetration rates in 1999 were led by Chile (1.35 per cent) while Argentina was least successful at 0.8 per cent. By 2001, Argentina was leading all Latin American countries in terms of Internet penetration (Figure 10.8). 12 10 8 % 6 4 2 0 Brazil

Mexico 1999

Argentina 2001

Source: Latin American Information Center, ‘Trends in Latin American networking (http://www.Lanic.utexas.edu/project/tilan/), and Nielsin/Netrating (2001).

Figure 10.8

Internet penetration, 1999–2001

Changes in the number of Internet hosts show the even greater progress that Argentina was able to achieve in a short period (Figures 10.9 and 10.10). In 1993, the number of hosts was far below other Latin American countries. By

Kagami 02 chap 06

2/15/04

2:54 PM

220

Page 220

Country/area studies

1.2 1 0.8 0.6 0.4 0.2 0 Argentina Source:

Venezuela

Brazil

Mexico

Chile

Petrazzini and Guerrero (2000).

Figure 10.9

Number of Internet hosts per 10 000, 1993

45 40 35 30 25 20 15 10 5 0 Argentina Source:

Venezuela

Brazil

Mexico

Petrazzini and Guerrero (2000).

Figure 10.10

Number of Internet hosts per 10 000, 1999

Chile

Kagami 02 chap 06

2/15/04

2:54 PM

Page 221

Policies for Internet access: Mexico and Argentina

221

1999, as Figure 10.10 shows, Argentina’s website density far outstripped that of Brazil or Mexico. It is of great interest to investigate what factors or policy measures enabled it to greatly improve Internet access during the late 1990s. Certainly, we can isolate several key initiatives implemented in 1997 when it was declared, through Presidential Decree 554/97, that Internet access was a matter of national interest. We discuss these initiatives in more detail in the next section.

6.

INITIATIVES FOR INTERNET GROWTH: ARGENTINA

If privatization and the introduction of competitive forces into the telecommunications market were the primary reasons for Argentina’s rapid growth, similarly spectacular phenomena might have been expected in most other countries since such market reforms were shared by other Latin American countries. To a certain degree, the peculiarity of Argentina’s experience points to the fact that privatized, liberalized market forces were not the only important factors. Rather, Argentina instituted several, some unorthodox, measures that include regulatory interventions regarding local access tariffs, an innovative calling plan, and a concerted push to expand community Internet access programmes. 6.1

Regulatory Intervention

In June 1997, the Presidential Decree declared that the Internet was a matter of national interest, promulgating universal service mandates that aimed at providing citizens with access to modern multimedia applications with reasonable tariffs and quality. During the public hearing phase that was intended to garner public input toward a set of strategies, local ISPs focused on the problem of high fees for leased lines where there was little competition. Through a decree, Argentina instituted price regulation on leased lines, lowering it as much as 45 per cent. The market for ISPs grew at an accelerated pace, from 35 at the end of 1995 to 136 in 1997 and to 170 in 1999. Nevertheless, Argentina’s ISP market is dominated by a few large players, including Arnet (Telecom), Advance (Telefónica), Impsat (Grupo Pescarmona) and Ciudad Digital (Grupo Clarín). The first two are owned by two incumbent telephone operators, and these top four ISPs together were reported to control more than 80 per cent of the market by 1999. However, the reduction in leased line fees also encouraged ISPs in underserved areas outside of Buenos Aires. Prior to 1997, the capital commanded 70 per cent of total ISPs. After the price reform, 50 per cent of ISPs operated in the interior.

Kagami 02 chap 06

222

6.2

2/15/04

2:54 PM

Page 222

Country/area studies

Network Efficiency Measures: NAP and 0610

A major problem for Internet access in most Latin American countries is the lack of regional and interregional interconnection. Without such connections and peering arrangements, most international connections had to go through the US. Alternatives were a 64Kbps satellite link provided by Imsat in 1996 or using a somewhat unreliable dedicated link between ISPs. Most Argentine ISPs were charged a higher amount of fees for international connection when they were direct clients of US and European network access points (NAPs). For example, instead of paying for interconnection fees only, they were charged for the whole circuit plus the access rate at the connection point. But in 1998, Cabase NAP opened to remedy the situation. Future efforts will be centreed around establishing regional NAPs that connect Argentina with Chile, Brazil and other Latin American countries’ NAPs. At the local network level, dial-up access to the Internet was expensive mainly because Internet traffic was charged the same way voice traffic was charged. Since telephone service was metered, long hours of Internet connection could lead to an enormous sum in fees. Furthermore, when local ISP numbers were not offered, Internet calls had to be made as long distance calls. To reduce the financial burden and explore a differential pricing scheme that differentiated voice from data traffic, Argentina instituted a special dialing area code (0610). By dialling 0610 to connect to an ISP, consumers were guaranteed special pricing. Under the 0610 plan, calls were metered differently than voice connection. For example, a variation of the 0610 tariff scheme would charge the first 15 minutes at normal telephone rates, subject also to time of day differentiation. The next 15 minutes would be free, and from the 31st until the 60th minute a normal fee schedule would be applied, with the following 15 minutes without a charge. On average, such plans can save up to 45 per cent in local access fees for the first hour of usage. One result of the 0610 experiment was the proliferation of free ISPs. According to a recent Cabase survey, almost 37 per cent of all dial-up accounts belonged to free ISPs in 2001. Free providers, instead of using banner advertisements or other marketing gimmicks, earn revenues from telephone operators. In reality, part of the payments to telcos include Internet access fees, albeit at a discounted rate. The popularity of free ISPs in Argentina is growing and the major players, Uyuyuy and AlternativaGratis, have dropped the 0610 plan, relying instead exclusively on revenue-sharing arrangements with telcos. There are no concrete statistical data available to evaluate whether the introduction of the special area code did or did not encourage more active Internet access. But it would be reasonable to assume that such a special rate arrangement did lower the cost of Internet usage substantially for many users.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 223

Policies for Internet access: Mexico and Argentina

6.3

223

Community Technological Centres

The third component of Argentina’s Internet initiative was part of universal access objectives through community telecentres. Through ‘Argentin@ Internet.todos’, the government initiated a project to establish 1000 Centros Tecnológicos Communitarios (CTC) in 1998. Although such projects are by no means unique to Argentina (see http://www.tele-centros.org), the pace and scope of CTC is much wider than most community access projects in other Latin American countries. CTC also expands Internet access through public libraries by providing a special telephone line for the Internet and up to US$100 of support for access calls. In 1999, Argentina also began an ambitious project to provide each citizen with a free e-mail address through post offices equipped with PCs throughout the country.

Source:

Figure 10.11 6.4

www.tele-centros.org.

A community access centre in Peru

Evaluations

The net effect of these initiatives partially explains the rapid growth in Internet hosts, ISP services and Internet usage in Argentina. The immediate effects of new rate policies in 1997 resulted in substantial decreases in local access and leased line charges (Table 10.3). In addition, cost reductions and the 0160 plan also affected the way its population access the Internet. Over a third of dial-up Internet accounts are associated with the so-called free ISPs. Although charges for Internet access are subsumed in local phone access payments, the new rate plan enables consumers to benefit from substantially lower overall costs or a near-flat-rate schedule. Nevertheless, Argentina enjoys one of the highest levels of income within the region, which renders a favourable development environment to begin with. Over a third of its population is concentrated in or near Buenos Aires, making it easy to raise telecommunications service penetration. New rate plans, however, encouraged more ISP presence and Internet services outside of the capital area.

Kagami 02 chap 06

2/15/04

2:54 PM

224

Page 224

Country/area studies

Table 10.3

Prices changes due to rate intervention, 1997

Intervention Local peak time Local regular Local discount 64Kbps leased line 2Mbps leased line Source:

Before

After

Change (%)

1.35 1.01 1.01 8905/mth 114 096/mth

0.64 0.64 0.42 4793/mth 63 182/mth

–53 –37 –58 –46 –45

Petrazzini and Guerrero (2000).

The question remains whether improvements in telecommunications infrastructure will have a significant effect on the long-term growth of Internet usage. Even when there is a favourable network environment, the ultimate user preference will depend on the availability and type of online content and activities. Infrastructure building activities must be followed by development efforts in increasing and diversifying usable contents, e-commerce sites, and new applications. To make advances in content areas, the next stage of efforts must focus on providing legal and commercial environments that are conducive to consumers and businesses alike. One critical shortcoming is the lack of a concerted effort to improve broadband capacity and connection. Argentina has the unique opportunity to deploy broadband faster than any other Latin American country due to its high cable penetration rate. In the multimedia-rich Internet era, accessing usable content will require high bandwidth. To sustain past rapid growth, Argentina will need new initiatives that target the changing Internet environment.

7.

LESSONS LEARNED

The three major economies in Latin America, Brazil, Mexico and Argentina, have experienced significant improvements in telecommunications infrastructure and high growth in Internet users during the past few years. Nevertheless, a striking feature in these countries is the low Internet penetration rate: while other Internet leaders are surpassing or approaching the 50 per cent level of penetration, Brazil’s rate remains at 7 per cent and Mexico’s at 3.4 per cent. For Argentina, due to several effective initiatives, 10 per cent of its population now access the Internet. Despite privatization efforts and market competition that were expected to lower telecommunications costs, Internet access is limited to those who are in the top income level.

Kagami 02 chap 06

2/15/04

2:54 PM

Page 225

Policies for Internet access: Mexico and Argentina

225

Market liberalization had a positive impact on expanding the basic telecommunications infrastructure and lowering the costs somewhat, as discussed in this chapter. Nevertheless, market forces alone are inadequate to induce companies to invest in traditionally underserved areas. Competitive forces also tend to encourage providers to focus on high margin groups, neglecting the disadvantaged and low income population. Surveys show that the basket price for local access and Internet connection in these countries is comparable to or much higher than those in developed countries. The basic telephone service alone costs several times more that the average income, ranging from two times in Argentina to four times in Mexico to seven times in Peru. Despite market reforms, such measures did not improve, except for some signs of increasing accessibility in terms of higher numbers of main lines. Then, the depressed income will continue to be the most critical obstacle to further improvement in telecommunications access and Internet usage. Given economic problems and income distribution in Latin America, past rapid growth in telecommunications services and the Internet may have been a sign that those with available disposable income were finally offered the service they desired. In other words, upper and middle income households are now connected to the Internet but there is little prospect for sustained penetration beyond these segments in the future, regardless of popular projections by market analysts that paint a very rosy scenario. At the same time, a few developed countries have also experienced a very slow pace of Internet growth. For example, Germany and Japan which had a penetration rate below 20 per cent in 1999 (Figure 10.12), gained substantially by 2001, to 34 per cent and 37 per cent, respectively. France, on the other hand, remains at 19 per cent. Therefore, a low penetration rate is not a great concern by itself. To achieve the level of universal service, governments must make investments in underserved and disadvantaged areas. Community access centres, although they have become more prevalent, are limited in number and reach. However, a side effect of privatization is the inability to collect revenues that can be used for infrastructure investment. Once telecommunications policies are delegated to markets and private enterprises, universal service objectives become subject to fickle market incentives. Instituting rate regulations and universal service fees is increasingly difficult. As an alternative, many Latin American countries are trying to leverage revenues from wireless licences and spectrum auctions for universal service initiatives. But the prospect for a widespread deployment of wireless and mobile services is very low at this time. Revenues from spectrum auctions and licensing, once expected to boost government revenues, have evaporated as service providers struggle to find market potential. Once the relatively small number of households in the upper and lower income brackets is fully served,

Kagami 02 chap 06

2/15/04

2:54 PM

226

Page 226

Country/area studies

45 Canada

Sweden

Internet penetration (%)

40 35 Australia

30

Denmark Finland

USA Norway

25 20

UK

15

France

Japan

Germany

10 5 China

0 0

5000

10000 15000 20000 25000 30000 35000 40000 GDP per capita

Source: European Telework Online, statistics relating to information technology in EU, http://www.eto.org.uk/eustats

Figure 10.12

Internet penetration and GDP, 1999

there is little prospect that Latin American countries will continue to expand Internet access to the level of US, European and a few Asian markets.

REFERENCES Curry, J., O. Contreras and M. Kenney (2001), ‘The Internet and E-Commerce Development in Mexico’, BRIE Working Paper 144, Berkeley Roundtable on the International Economy. International Telecommunication Union (ITU) (2000), Americas Telecommunication Indicators. Los Angeles Times (2001), ‘Free Trade May Be Costly for Mexican Satellite Company’, 23 August. Nielsen/Netratings (2001), Global Internet Index. OECD (1999a), Communications Outlook 1999, OECD, Paris. OECD (1999b), The OECD Review on Regulatory Reform in Mexico, OECD, Paris. OECD (2001), Communications Outlook 2001, OECD, Paris. Petrazzini, B. and A. Guerrero (2000), ‘Promoting Internet Development: The Case of Argentina’, Telecommunications Policy, Vol. 24, pp. 89–112. Pyramid Research (2001), Communications Markets in Mexico: Blurring Barriers and Escalating Competition, Mexico Country Report, Cambridge, MA. The Yankee Group (2001), ‘Incumbents Increase Control of ISP Accounts’, Global News Wire, 9 July.

Kagami 03 chap 11

2/15/04

2:54 PM

PART II

Challenging issues

Page 227

Kagami 03 chap 11

2/15/04

2:54 PM

Page 228

Kagami 03 chap 11

2/15/04

2:54 PM

Page 229

11. Tipping, standardization and convergence: catch-up and failure in Japan’s standards strategy Mitsuhiro Kagami 1.

INTRODUCTION

Technical standards* are basic for using goods. If one person uses miles while another uses kilometres to describe speed/distance on a car speedometer, neither can necessarily understand it. Of course if a conversion factor exists, both can calculate the speed/distance. In this case the conversion is 1 mile = 1.6km. By the same token, if the electrical frequency is different, home appliances cannot be used in such different cases as 50 cycles or 60 cycles. In this case, too, a converting switch exists to change cycles. These simple examples show that technical specifications or standards are necessary for products that stipulate technical characteristics. For, if you do not know them, like foreign languages, you cannot use them (or communicate with others). Here again, however, if conversion factors (translators) exist, both can be converted. Market size is another factor that comes into play when consumers choose a specific product. If people like a product for its features: design, form, handling, usefulness, economy and popularity, then sales increase, and eventually the product will dominate the market. A slight change in the function or character of a product’s features however may produce a change in consumer response to the said product and result in falling sales and a loss of market dominance. This is called ‘tipping’. One of the most notable examples is the VHS vs. Betamax rivalry for the home videotape market. When a specific product dominates a market, standards attributable to the said product are called ‘de facto standards’. VHS became the de facto standard in the above example. On the other hand, when international organizations or governments determine specific standards for public safety, the environment, and so on, these are called ‘de jure standards’. Norms and standards determined by the Japanese Industrial Standards (JIS) and the International Organization for Standardization (ISO) are examples of de jure standards. 229

Kagami 03 chap 11

2/15/04

2:54 PM

230

Page 230

Challenging issues

Networks are common features of telecommunications, transportation (airlines, railways, and roads), electricity, mail services and online businesses. Network economy has the characteristic that the more users participate in the network, the more convenient the network becomes for the user. This is called ‘network externalities’. Users may prefer a specific software or machine in the Internet world and if the network externality works, a particular product or specific software may dominate the cyber market. This is called ‘locked-in effects’. It is probable over time that the network economy will create de facto standards due to the given externalities. It is said that convergence takes place as telecommunications and information technologies advance. Digital TV may combine with personal computers (PCs) because both utilize digital technologies. Therefore, if the same technical standards are deployed for different purpose products, both can be converted for same purpose usage. Thus, convergence issues are deeply related to standards. This chapter describes the relationship between standards and consumer tastes. The first section explains tipping and de facto standards. The second section treats de jure standards. Japan’s defeats in standards wars are illustrated in the third section. The fourth section again explains consumer preferences and the selection of goods based on demand or human factors. Convergence issues are also briefly examined in the fifth section and finally concluding remarks follow.

2.

TIPPING AND DE FACTO STANDARDS

2.1

VHS vs. Betamax

A famous example of tipping was the market war between SONY and Japan Victor/Matsushita regarding home videocassette recorders (VCRs) (Ida (2001)). Both companies tried to introduce videotapes with a 0.5 inch width and two hours of recording time. SONY first introduced a videotape called Betamax (0.5 inch with one-hour playtime) with high resolution pictures in 1975. One year later, Victor marketed a videotape called VHS (0.5 inch with two hours playtime). The Victor/Matsushita alliance used heavy advertising to promote its VHS product. In response to the alliance’s offensive SONY introduced a Betamax with a two-hour playtime in 1977, but the Victor/Matsushita alliance already had the upper hand and had gained a 50 per cent market share by 1978, and after that their share continued to expand. Finally SONY decided to produce VHS tapes in 1988, ending the videotape war. The victory of VHS can be explained by the following factors: (i) SONY concentrated on providing the highest quality picture possible while Victor/Matsushita did not (the latter was satisfied with near-perfect but generally

Kagami 03 chap 11

2/15/04

2:54 PM

Page 231

Catch-up and failure in Japan’s standards strategy

231

acceptable picture quality levels); (ii) SONY initially sold Betamax as a superior picture product but in doing so sacrificed playtime; (iii) Victor/Matsushita on the other hand relied on extended playtime over picture quality and used their strong combined marketing power through their sales networks to get the message across to the general public. This is a good example of de facto standards. However, some lessons can be drawn from this example. First, quality is an important factor in selling products but it is only one factor. Consumers may prefer tapes with longer playtime or other advantages over quality. Above average quality may be acceptable. Second, advertising and sales networks are necessary in order to sell products. Third, those who bought Betamax (tapes and VCRs) had to eventually change to the VHS system (loss for consumers). Fourth, no converter existed between the two systems so consumers who bought Betamax were unable to convert and store old footage. Fifth, SONY faced enormous revenue losses for initial R&D, marketing costs, and finally production costs in changing from Betamax to VHS. 2.2

Forum Standards

The result of the above ‘war’ brought about consortium or group methods in the production of new products among manufacturers in order to avoid such enormous losses. Thus, the so-called ‘forum standards’ were born. This is the third way to set up standards in addition to de facto and de jure standards. Manufacturers usually organize a consortium or association to form common standards for a new product. Membership is open (entry and withdrawal are free). A member can use other members’ patents if the member also offers its patent. In this case, the time taken to formalize standards is greatly reduced and thus less initial cost is incurred. In a sense, forum standards are a kind of strategic alliance. A good example of this was seen in the development of digital versatile disks (DVD) by a consortium (later the DVD forum in 1997).1 2.3

Voluntary Standards

Between de facto and de jure standards, there is another type of standard. That is called ‘voluntary standard’ (or industry-based standard). How is a voluntary standard determined in Japan? Let us take the standard on terrestrial digital TV as an example. The Ministry of Posts and Telecommunications (now called the Ministry of Public Management, Home Affairs, Posts and Telecommunications (MPHPT) after the government reform of 2001) has the Council on Telecommunication Technologies that is composed of equipment manufacturers, academics and engineers. The Council submitted a technical report on the said topic to the Ministry in May 1999 and two ministerial ordinances based on the

Kagami 03 chap 11

232

2/15/04

2:54 PM

Page 232

Challenging issues

technical report were stipulated in November 1999. These ordinances usually give only directions and frameworks. The details are discussed and determined by producer associations such as the Association of Radio Industries and Businesses (ARIB).2 ARIB formulated the standard on the transmission method of terrestrial digital TV in May 2001 as its standard, ARIB STD-B31. However, these standards are not compulsory.

3.

DE JURE STANDARDS

There are two types of de jure standards: one by public or government institutions; the other by international organizations. 3.1

De Jure Standards by Governments

Such items as infrastructure and public goods have de jure standards. Public standards are necessary for private goods when they are related to safety, the environment, and public health. Those standards are also called ‘mandatory standards’. Moreover, stated technical standards are convenient to all users within a country so usually domestic standards are stipulated for private goods by government institutions. In the case of Japan, the most famous body is the Japanese Industrial Standards (JIS). JIS was established in 1949. JIS’s role is now determined and managed by the Japanese Industrial Standards Committee (JISC) under one unit of the Ministry of Economy, Trade and Industry (formerly the Agency of Industrial Science and Technology) since 1952. JISC’s task is summarized as follows: (i) establishment and maintenance of JIS; (ii) administration of accreditation and certification; (iii) participation in and contribution to ISO activities; and (iv) development of measurement standards and technical infrastructure for standardization. JISC has 9461 national committee members and covers 8364 JIS standards (see Figure 11.1). Nowadays, the borderline between de jure and de facto or voluntary standards is not clear because of the very rapid technological changes. A previous standard may be outdated immediately when completely new products appear. This short product cycle, which is particularly visible in the electronics and computer industries, will change the system of public standardization. From a competition point of view a single standard hinders new developments, whereas multiple standards enhance competition among manufacturers to obtain de facto standards. However, this also depends on the given characteristics of products and markets. It is said that one-to-many correspondence, such as radios and TVs (one station with many receivers), may require de jure standards since it takes time and money to transform to new ones.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 233

Catch-up and failure in Japan’s standards strategy

233

JISC Chairman

JIS (8364) Standardization

Standards Council

PASC

Divisional Council National Committee

ISO/IEC

Certification and Accreditation

JIS Mark JNLA ILAC/APLAC

Members of JISC (9461)

Technical Regulation, Standards and Conformity Assessment Policy Unit,

Technical Infrastructure and Measurement Standards

Ministry of Economy, Trade and Industry (METI)

JSA (Japanese Standards Association) Source:

JISC, http://www.jisc.org

Figure 11.1 3.2

Sales of JIS and ISO/IEC Standards

Japanese industrial standards committee (JISC) system

Standards by International Organizations

It is logical that standards for resources or goods for international use should be regulated by international organizations. Such organizations as the International Telecommunication Union (ITU), the International Electrotechnical Commission (IEC) and the International Organization for Standardization (ISO) are examples. The ITU regulates IT-related standards and the IEC does the same for electronics-related standards while the ISO covers standards on goods and services including systems that are internationally traded. Take the ISO for example. It was established in 1947 in Geneva and currently its members represent 140 countries. ISO has 187 technical committees with 572 subcommittees. Under the subcommittees 2063 working groups are tackling standards for goods and services (see Figure 11.2). Each country has one vote. In the case of Japan, JISC is the representative body. Up to now, over 13 000 standards have been stipulated, from the simple screw to business models (or systems). Well known examples include the ISO 9000 series for industrial quality and the ISO 14 000 series for environmental considerations. It is said that a shepherd (secretariat) country in a Technical Committee can sometimes pass its own technical standards as global standards. This, of course,

Kagami 03 chap 11

2/15/04

2:54 PM

Page 234

234

Challenging issues

Policy development committees: • Committee on conformity assessment (CASCO) • Committee on consumer policy (COPOLCO) • Committee on developing country matters (DEVCO)

GENERAL ASSEMBLY (140 countries) • Principal officers • Delegates of: – Member bodies – Correspondent members – Subscriber members

Council standing committees: • Finance • Strategies

Council

Central Secretariat Ad hoc advisory groups

Source:

Technical Management Board Committee on reference materials (REMCO) Technical advisory groups Technical committees (187)

ISO, http://www.iso.org

Figure 11.2

ISO structure

can lead to a form of hegemony, and the shepherd country can gain unfair advantages from such global standards (this point is addressed later). Therefore, the rules of the game to determine standards should be transparent and equal for all member countries. Some crucial decisions have also been made by the World Trade Organization (WTO) relating to global standards. The WTO’s Agreement on Technical Barriers to Trade (TBT) became effective in 1996, which determined that domestic standards sometimes work as trade barriers. The basic premise here was that ISO standards are preferable to any other standards. Since then each country has had to apply ISO standards or revise versions of domestic standards in order to comply. For instance, in the case of Japan, 2000 out of around 8300 JIS standards had to be revised and adjusted to comply with ISO standards between 1995 and 1998 (Jou (2001)).

4.

JAPANESE DEFEATS IN THE STANDARDS WAR

It is commonly accepted that ISO’s standards are strongly influenced by European standards owing in great part to the very origin of the organization.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 235

Catch-up and failure in Japan’s standards strategy

235

The European Committee for Standardization (CEN), established in 1961, has around 3000 standards. Nineteen European countries are members of CEN and its norms are automatically treated as ISO standards (by the Vienna Agreement). European countries, especially the UK, think standards are economic weapons in international negotiations. Therefore, let us say that a country has a strategy to globally promote its own products, in such a case it is obviously advantageous to let its own standards be global ones. On the other hand, Japan had not shown much interest in global power games until quite recently. As a ‘catch-up’ country, Japan had always followed or assimilated to the environment of the host country in order to sell its products. Japan made an effort to search for and understand foreign standards and exported products with such standards. For example, Japan manufactured and exported industrial products under UK standards to Singapore and Malaysia during the 1950s and 1960s and under US standards to the US during the 1970s and 1980s (for example, left-hand drive cars). However, Japan never thought of nor attempted to export Japanese standards. Japanese manufacturers are capable and skilled enough to produce products to any standard. However this has proven to be a weak point. In the present discussion on standards for terrestrial digital TV, Japanese TV set manufacturers are said to be reluctant to introduce the Japanese standard. Standards do not matter to them because their priority is to sell their products, as many as possible, whatever the technical standard. Meanwhile, China has now developed its own policy. It has embarked on a course to introduce its own standard for terrestrial digital TV (as explained later). This begs the point: How can Japan win without a clearly defined strategy? Two such international standards failures for Japan are illustrated here. Notorious failures for Japan were the ISO 9000 and ISO 14 000 series (Jou (op. cit.)). Japan did not take the initiative in the ISO discussions though Japan was a leader in these fields. 4.1

ISO 9000 Series

The ISO set up a Technical Committee (known as TC176) to discuss quality management and quality assurance. Canada was named to be the shepherd country and the first meeting was held at Ottawa in 1980. The major standards on ISO 9000 series were determined based on the British Standards (BS5750) in 1987. The Japanese government did not send any mission to the first meeting. Yet, up to date quality management was Japan’s pride at that time as exemplified by quality control (QC) and the Deming Prize (Kagami (1995)), but Japan did not take any initiative in the discussion. It came as a bolt from the blue to Japan when European clients asked to see ISO 9000 certificates for products made in Japan during the 1990s. Since then Japanese manufacturers and exporters have had to obtain these certificates.

Kagami 03 chap 11

2/15/04

2:54 PM

236

Page 236

Challenging issues

Moreover Japan’s need for acquisition of this series increased after 1997, when the government decided to require such certification as a condition of tender for government procurements. 4.2

ISO 14 000 Series

Following an environment summit held in Rio de Janeiro in 1992 the United Nations Committee on Environment and Development (UNCED) asked ISO to formulate system standards on the environment. The ISO established a Technical Committee (TC207), and the first meeting was held in Toronto in 1993. Five major standards on ISO 14 000 series were stipulated in 1996. The core of these standards it was said, were promoted by the UK, based on its own standard (BS7750). Here again, Japan failed to take any positive initiative in the discussion although Japan had extensive experience in environmental management. It is ironic that Japan’s certificate acquisition number on this ISO series is number one in the world. This passive attitude shown by Japan in the arena of world discussion certainly undermined its position in global trade and investment, resulting in a worsening of its reputation internationally.

5.

TECHNOLOGY AND SATISFACTION

The relationship between technology and consumer satisfaction is here analysed from a different angle. Why do some products remain after alternative products appear? Substitutes are not always substitutes but sometimes act as complements or coexist. The evolution of products depends on technological advance and changes in consumers’ preferences. This is why tipping takes place. Several cases of coexistence and disappearance are illustrated as follows: 5.1

Coexistence

5.1.1 Letter vs. telephone Writing letters is an old means of communication. People write letters to each other to communicate what they want to tell or know, for example, how they are getting along, their daily affairs, love affairs, commercial transactions, and official uses. During the Edo era a professional group of hikyaku carried letters on foot. It took six days between Tokyo and Kyoto. Under the Meiji government official mail services began in 1871 first between Tokyo-Kyoto-Osaka. The introduction of telephone services in 1890, first between Tokyo and Yokohama, started 14 years after the invention of the telephone by Graham Bell in 1876. Telephone messages have several features such as (i) verbal communications,

Kagami 03 chap 11

2/15/04

2:54 PM

Page 237

Catch-up and failure in Japan’s standards strategy

237

(ii) transmission is far quicker than mail, and (iii) no preparation is necessary such as in writing letters. Both above methods of communication continue to exist today because of their own strengths. Letters, especially, are superior as a means of expressing personal affection and the human touch. Moreover, business and official letters are often important due to authentication and verification of signatures or official seals/emblems. Even facsimile or electronic letters are no substitute for written letters so far. The telephone is quick but is usually followed by a letter of confirmation in business practice. (This may be improved if electronic signature and identification systems are developed in e-business.) 5.1.2 Radio vs. television NHK (Japan Broadcasting Corporation, a public broadcasting organization and formerly called Tokyo Broadcasting Station) first broadcast radio programmes in 1925 in Japan. NHK was then established as a special corporation in 1950. Private radio broadcasting stations started operating in 1951 and NHK started FM broadcasting in 1957. On the other hand, NHK started experimental black and white TV broadcasting in 1950 and began TV programmes in the Tokyo metropolitan area in 1953. NHK colour TV broadcasting started in 1960. In developing countries, for example, Brazil started colour TV programmes in 1972. The Football World Cup held in Argentina in 1978 boosted sales of colour TVs in Brazil. The introduction of TVs did not wipe out radio receivers because there are many radio lovers such as long distance drivers, students working late at night and music fans, particularly, of high fidelity sound. Local radio stations also play an important role for community information. Here again, coexistence can be observed because people have their own taste for both services. In Japan terrestrial TV enterprises now number 129, AM radio operators 48 and FM operators 194 (Ministry of Public Management, Home Affairs, Posts and Telecommunications (2001)). Technological changes, however, kill some products. Functions and price are the main reasons for their failure. 5.2

Disappearance

5.2.1 Typewriter vs. word processors/PCs The first typewriter machine made by E. Remington & Sons went on sale in 1874. Since then typewriters have worked hard for almost 100 years. People mastered strangely arranged keyboards (qwerty)3 and how to type became a must-skill for office workers. However, the appearance of the word processor, especially since the 1980s, drastically changed the situation. Word processors or PCs have editing and memory functions. This is the key element explaining

Kagami 03 chap 11

238

2/15/04

2:54 PM

Page 238

Challenging issues

why typewriter machines have been almost completely forgotten. Only the keyboard function remains as standard in the word processor/PC as a remnant of the old technology. 5.2.2 Mainframe computer vs. PCs Another example can be illustrated in computers. People were surprised to see the initial IBM computer systems. A solemn mainframe computer with card and magnetic tape reader machines enshrined in an air-conditioned room. The systems seemed to be almighty, virtual magic. Then came the PC. At first slow and expensive, PCs have gradually become powerful and cheap. They can easily be connected to create LAN systems, and once such networks are formed, PCs can function on equal terms with the old mainframe system. In addition PCs can be bought and mastered to an acceptable extent by individuals and used at home as well as work due to their relatively inexpensive purchase price and easy application systems. Now large mainframe computers only exist in the fields of large and high speed calculation such as meteorology, astronomy, and quantum physics. IBM’s history of glory, collapse and recovery shows even a giant can fail if market conditions (people’s demand) and technological innovation are not well matched. 5.3

Consumer’s Preference

To summarize, good products remain even when there are alternatives because people’s tastes differ while bad products disappear due to limited functions and cost considerations. Here people’s tastes and personal inclinations are of importance because individual participation becomes very crucial in the Internet world. That is why the change from the batch computer system to the PC is symbolic. It highlighted the end of centralized information systems and the birth of decentralized network information systems reflecting the changing attitudes and perceptions of contemporary society. High-tech products are sometimes too high-tech. People do not want to deal with overly complex equipment or too sophisticated contents. PC manuals are still difficult to read. Multifunctional equipment with a high price tag all too often does not correspond with the wishes of consumers who want simple functions. Sometimes IT machines seem to exist for themselves, not for users (in this respect, see, for example, Naemura (1995)). By the same token government-engaged projects sometimes forget what the public want. Over-investment or investment under misguided demand will cause disaster later. For example, in Japan NTT heavily invested in integrated service digital network (ISDN) but the service was basically ignored by consumers due to its slowness compared to new technologies such as DSL (Digital Subscriber Line) or ADSL (Asymmetric Digital Subscriber Line) in

Kagami 03 chap 11

2/15/04

2:54 PM

Page 239

Catch-up and failure in Japan’s standards strategy

239

the broadband diffusion. Another failure in Japan is NHK’s efforts to invent analogue-type high definition TV. Key factors here are technological changes, timing, and interpreting changes in user demand and taste.

6.

CONVERGENCE OF TECHNOLOGY

The introduction of terrestrial digital TV broadcasting is in a heated state of discussion in several countries. Here the pros and cons of digital TV are explained. Especially, the convergence of TV and PC is discussed. 6.1

Analogue vs. Digital TV

The invention of the digital TV method (DigiCipher) in the US in 1990 was a great step forward. There was a long dispute in the US during the 1990s regarding high definition TV (HDTV) or advanced TV (ATV). Digital TV finally won out against the analogue-type in the Federal Communications Commission (FCC) discussion because of the invention of DigiCipher. NHK developed HDTV along analogue lines but failed to be selected in this discussion (Hamano and Hattori (2001)). It is said that as a point of national strategy the US disliked the idea of Japanese dominance in the analogue-type HDTV and thus made efforts to find new standards, that is, digital TV methods. Digital TV has the following merits over analogue TV: (i) (ii)

Clear and beautiful pictures; Two-way communications are possible (users can ask for and get information through TVs); (iii) Efficient use of frequency4 (multi-channels, no ghost, mobile telecommunications for vacant spectrum, and so on); (iv) TV receivers can be used as a PC due to digitization.5 On the other hand, digital TV has several difficulties to overcome as follows: (i)

Digital TV sets are very expensive and analogue receivers need adapters called set-top boxes (STB); (ii) Broadcasting stations need to change equipment for digital TV transmission (enormous investments are required); (iii) Demarcation of satellite broadcasting and CATV stations; (iv) Content issues (uncertainty about digital content, while consumers question the need for HDTV as analogue content continues to be popular).

Kagami 03 chap 11

240

2/15/04

2:54 PM

Page 240

Challenging issues

In the UK, terrestrial digital TV started in September 1998 and satellite digital broadcasting in October 1998. The UK plans to transfer to digital TV by 2010. In the US, digital services started through satellite in June 1994 and through terrestrial stations in November 1998. Meanwhile the FCC decided that all private TV stations must begin digital broadcasting by 2002. So far the progress of digital TV is slow. The accumulated purchase of digital TV sets in the US is barely over one million as compared with yearly purchases of analogue TVs of 25 million sets. One of the reasons for the slow development of digital TV is that CATV’s share in the US is as high as 68 per cent, but CATV stations have not transmitted digital programmes instead of conventional analogue programmes. It is now being discussed as to whether CATV stations must carry digital programmes (a must-carry regulation). From the consumer’s point of view, the content versus cost argument of digital TV is not very attractive owing to the very expensive price tag of digital TV sets. In a sense consumers are now in a wait-and-see mode. 6.2

Development of Digital TV in Japan

Satellite digital broadcasting began in June 1996 under the Communication Satellite (CS).6 At present two stations are functioning: Sky-perfecTV and Wire Broad Networks. NHK started digital services under the Broadcasting Satellite (BS) in December 2000. Now BS digital services include eight TV companies (three channels for NHK plus one for each of the seven private companies), ten private FM radio companies and nine private data service companies. In addition a new satellite was fixed at the East longitude 110° in 2000 to operate CS/BS digital services from early 2002. CATV started digital services in July 1998 and began retransmission of BS digital programmes from December 2000. With respect to terrestrial digital TV, services for three areas (Tokyo, Nagoya and Osaka) were due to start in 2003 and other areas in 2006. All analoguetype broadcasting is planned to end by 2011 (Amendment of the Electric Wave Act in July 2001). In the meantime CATV has to be fully digitized by 2010. This may create confusion among consumers. The reasons are as follows: (i) the diffusion of digital TV sets is slow like in the US. As of April 2001 BS digital TV sets shipped accumulated to 203 thousand and digital tuners to 402 thousand sets (Ministry of Public Management, Home Affairs, Posts and Telecommunications (op. cit.)). Prices for this equipment are high7 and content not so attractive (such as baseball and golf); (ii) TV stations cannot prepare for digital services within such a short time in terms of money8 and programming; (iii) the end-date of analogue services is too soon. Digital services may not cover all areas (universal service); (iv) it is thought that frequency allocation is under pressure from mobile phone companies or other sectors which want to

Kagami 03 chap 11

2/15/04

2:54 PM

Page 241

Catch-up and failure in Japan’s standards strategy

241

expand their services. People do not want to buy high priced digital TVs in order to help consumer electronic companies nor cellular phone companies. 6.3

Unknown Factors: Interest Groups, Technology, and Demand

When we think of the future of digital TV, there exist unknown factors. First, there are severe conflicts of interest between interest groups such as CATV groups, satellite TV groups, other broadcasting stations, the computer industry, TV manufacturers, mobile phone companies, MNCs, and new entrants who are interested in multimedia and entertainment.9 Long-lasting discussions in the US on digital TV show that the dispute has not yet been resolved. Second, technology will produce cheaper methods of TV transmission. For example, a flying object in a fixed orbit at 20 000 metres may work as a relay station. Actually, Honda and Cranfield University (UK), are developing such an airplane (called ‘space bird’).10 They estimate that the cost will be onehundredth of that of a satellite launch. Moreover, as we have seen so often in recent years, new technologies may come up in the TV-related field not only to reduce costs quickly but also to change concepts fundamentally. Third, a reasonably large number of consumers may remain satisfied with their present analogue TVs for quite a while to come and digital TVs may not diffuse as originally planned. Convergence of TV sets and PCs may not occur simply because people prefer to keep separate various functions such as PC use for private e-mail contacts while wide-screen digital TV sets may become popular for comfortable living room family relaxation. People’s tastes and preferences are difficult to predict. 6.4

Digital TV Standards

It is understood that convergence takes place when the same technical standards are adopted by two different products, such as TVs and PCs in the above case. Liquid crystal displays (LCDs) can now receive TV programmes and LCDs are also used for PCs. Therefore, digital technology can soon attain convergence of TVs and PCs if standards are integrated. At present there are four types of digital TV standards (see Table 11.1). NHK mainly developed the Integrated Services Digital Broadcasting-Terrestrial (ISDB-T) method, based on the US (ATSC) and European (DVB-T) methods. That is why ISDB-T is more flexible than the previous two methods. It is said, however, that the Chinese method (DMB-T) is better than the Japanese method in terms of mobile reception and data reception. China has a clear strategy that technical dependence is not permissible. China wants to develop its own standards for third-generation cellular phones, too, called TD-SCDMA against W-CDMA (Japan-Europe) and cdma2000 (US).

242

19.4 No TV No

Source: EE Times, 29 January 2001.

Data rate in Mbps Mobile reception Power consumption Burst data reception

ATSC (Advanced Television Systems Committee) Single carrier (8-VSB)

US

Japan

DVB-T ISDB-T (Digital Video (Integrated Services Digital Broadcasting – Terrestrial) Broadcasting – Terrestrial) Multi carrier Multi carrier (OFDM = Orthogonal (OFDM) Frequency Division Multiplex) 5 to 32 4.9 to 23.2 Some Good TV TV No No

Europe

5.4 to 32 Best Port TV & PDAs Yes

DMB-T (Digital Multimedia Broadcasting – Terrestrial) Multi carrier & spread spectrum

China

2:54 PM

Modulation

Aspect

Digital TV standards

2/15/04

Table 11.1

Kagami 03 chap 11 Page 242

Kagami 03 chap 11

2/15/04

2:54 PM

Page 243

Catch-up and failure in Japan’s standards strategy

243

Brazil announced that it would start terrestrial digital TV broadcasting from 2002. NHK is eager to introduce the ISDB-T method because if Brazil adopts the system, there is an expectation that it will spread to other Mercosur countries and later South America as a whole. However, Japanese TV makers are said to be reluctant to push the idea. Standards do not matter for them to sell digital TV sets since they can correspond to any type of standard. Because Europeans are also enthusiastic to introduce their method, competition seems very tough.

7.

CONCLUDING REMARKS

Tipping takes place when consumer preferences change, especially changes in function, price and technical standards. Tipping may be slow if a converter exists between two different standards. Technical standards have basically three categories: de facto standards, forum standards and de jure standards. De facto standards are the result of competition. Consumers decide the popularity of products. Thus, tipping is related to de facto standards. Moreover, network externalities that work in the present online networks produce the one-got-all phenomenon. Group efforts by manufacturers bring about forum standards that can avoid losses and risks that come from severe competition through de facto standards. De jure standards come from public or international organizations to regulate and harmonize public and international concerns. We should utilize these standards according to consumer satisfaction, market conditions, product characteristics, product cycles, social as well as global systems. Determination of new standards should be free from the egoism of advanced countries or multinational corporations (MNCs) because they have a tendency to reign over markets. Particularly, international standards should be transparent and neutral.11 Japan has always followed foreign standards in order to export its products. Japan has never taken initiatives in standard discussions up until quite recently. Japan was forced to adopt European or US standards as global ones on several important occasions. Since Japan’s technical levels and management skills are satisfactory as compared with those of other advanced countries, Japan should contribute more to international standard formations.12 The IT revolution has brought about tremendous impacts on our society (Kagami (2001)). These impacts are continuing both positively as well as negatively. Digitization of TV broadcasting networks is expected to make another major push because such TVs can be utilized as PCs: interactive communications, the Internet, games, T-business, and e-mail functions. Convergence of telecommunications and broadcasting requires fundamental adjustments in infrastructure,

Kagami 03 chap 11

2/15/04

244

2:54 PM

Page 244

Challenging issues

hardware, software, and content. Moreover, strategic alliances of firms including MNCs will take place and legal structures will also need to change. Individual participation and people’s demands, tastes or human touch (handwritten, voice and face-to-face contacts) are the key to this transformation. Strong initiatives from governments or large firms will often fail since supplyside approaches always ignore changes in human satisfaction or future necessities. Aggressive supply plans for new products will be balanced with shifting demand. Moreover, technological changes are very quick. Development plans for top-of-the-line, state-of-the-art products such as digital HDTVs should be cautious, especially when huge income disparities exist in a society. Convergence in IT-related products means that technical standards merge in different-purpose products. This will create tremendous welfare for mankind. However, consumers’ choice does not change if the new products do not respond to their preferences or tastes.

NOTES *

1.

2.

3. 4.

5. 6.

7.

ISO defines standards as follows: Standards are documented agreements containing technical specifications or other precise criteria to be used consistently as rules, guidelines, or definitions of characteristics, to ensure that materials, products, processes and services are fit to their purpose. In order to produce DVD video for movies and DVD for PCs the forum worked well but standards for DVD for other purposes failed to integrate due to participants’ different interests such as DVD-RAM. At present, three different types of DVD-RAM standards exist. ARIB is designated by MPHPT as the Centre for Promotion of Efficient Use of the Radio Spectrum under the terms of Article 102–17 of the Radio Law of Japan. ARIB and its committees conduct studies, research and development, and standardization activities on radio systems to encourage high density use of the radio spectrum. ARIB also cooperates with similar standards groups around the world in seeking to harmonize standards for radio systems (http://www.arib.or.jp/). The QWERTY keyboard is also said to be another example of tipping against the DVORAK keyboard (Katz and Shapiro (1994)). One digital bandwidth can transmit three to four channels of standard TV programmes due to the compressed technology of digitization. Also digitization can solve the ghost and noise problems usual on analogue frequency caused by interference by electric waves. If analogue TV broadcasts cease, the frequencies can be utilized for other uses such as mobile phones. An auction system for the allocation of bandwidth for wireless services is a new topic of conversation (Nakamura and Agata (2001)). If digital TV uses the progressive method in scanning lines, it is more compatible with computer algorithm than interlaced scanning. The World Radio Communication Conference 1977 allocated eight channels on the orbit passing at the East longitude 110° for Japan. Now Japan has two BS satellites (BSAT-1a and BSAT-2a at 110°) and four CS satellites (JCSAT-4A at 124°, JCSAT-3 at 128°, SUPERBIRDC at 144°, and JCSAT-2 at 154°). ITU has recently allowed Japan to have another satellite at 110° and it was launched in October, 2000 for CS uses. But due to the same longitude positioning as BS it can be used for CS/BS digital broadcasting. Initially the retail price of a digital TV set was more than US$10 000 and now it costs around US$3500 (but still high compared with standard colour TVs at around US$800).

Kagami 03 chap 11

2/15/04

2:54 PM

Page 245

Catch-up and failure in Japan’s standards strategy

245

8. It is said that the necessary investment in digitization for all analogue stations is estimated at around US$8.3 billion. Moreover, there is another problem relating to the frequency allocation. Digital TVs use the UHF band but some Japanese analogue TVs are also using part of this UHF band so that a rearrangement of bandwidth is necessary for simultaneous broadcasting of analogue and digital TVs (so-called analogue-to-analogue conversion problem). This investment was initially estimated to cost US$606 million but recent estimations put the cost at over US$1.6 billion due to their being more affected receivers than previously expected (Nihon Keizai Shimbun, 21 November 2001). 9. In the UK, satellite B-Sky-B (digital) TV has reached 5.5 million households and CATV reached around three million since its start in 1998 while terrestrial digital TV (ITV Digital) was sluggish at 1.2 million as of September 2001. The success of B-Sky-B is attributable to the distribution of free antenna and free STBs and a gamble on the lure of interactive services (Nihon Keizai Shimbun, 26 November 2001). 10. NASA succeeded in establishing a height record of 29 400 metres for an unmanned airplane powered by solar battery on 13 August 2001 (called ‘helios’) (Nihon Keizai Shimbun, 9 August and 14 August 2001). 11. Sometimes standards formation is done by public subscription for fairness. A new standard for encryption algorithms (called the Advanced Encryption Standard or AES) for US government use was offered by public subscription and a method invented by two Belgians was chosen in 1997, called ‘Rijndael’ (Nihon Keizai Shimbun, 31 December 2001). 12. Japan is said to be advanced in the area of DVDs, cellular phones, game software, ceramics, nanotechnology and biotechnology.

REFERENCES Hamano, Yasuki and Katsura Hattori (2001), Japan-US War on Digital TV (in Japanese), Ascii Co., Tokyo, (translation of Joel Brinkley, Defining Vision: The Battle for the Future of Television (1997)). Ida, Takanori (2001), Network Economics (in Japanese), Nippon Hyoronsha, Tokyo, pp. 116–20. Jou, Yoshihiko (2001), ‘De Jure Standards in the Construction Industry’ (in Japanese), in Fukutaro Watanabe and Toru Nakakita (eds), Strategies and Formation of Global Standards: An Analysis of De Jure Standards (in Japanese), Japan Institute for International Affairs, Tokyo, pp. 160–74. Kagami, Mitsuhiro (1995), The Voice of East Asia: Development Implications for Latin America, Institute of Developing Economies, Tokyo. Kagami, Mitsuhiro (2001), ‘IT Revolution and its Meaning to Society’, in Mitsuhiro Kagami and Masatsugu Tsuji (eds), The ‘IT’ Revolution and Developing Countries: Late-comer Advantage?, Institute of Developing Economies (JETRO), Tokyo. Katz, Michael and Carl Shapiro (1994), ‘Systems Competition and Network Effects’, Journal of Economic Perspectives, Vol. 8, No. 2, Spring, pp. 93–115. Ministry of Public Management, Home Affairs, Posts and Telecommunications (2001), White Paper on Information and Telecommunications, Gyosei, Tokyo. Naemura, Kenji (1995), ‘User Involvement in the Life Cycles of IT and Telecommunication Standards’, in Richard Hawkins, Robin Mansell and Jim Skea (eds), Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Edward Elgar, Aldershot, pp. 93–102. Nakamura, Kiyoshi and Koichiro Agata (eds) (2001), Convergence of Telecommunications and Broadcasting in Japan, United Kingdom and Germany, Curzon Press, Richmond, Surrey, p. 149.

Kagami 03 chap 11

246

2/15/04

2:54 PM

Page 246

Challenging issues

Nihon Keizai Shimbun (Japan Economic Daily), ‘A Relay Station in the Sky at 20,000 m’, 9 August and ‘NASA’s Unmanned Airplane Established the Height Record’, 14 August 2001. ——, ‘Jamming of Terrestrial Digital TVs’, 21 November 2001. ——, ‘B-Sky-B’s Run Away in British Digital Broadcasting’, 26 November 2001. ——, ‘Selection of Cipher Technologies: Japan’s Return Match in Europe’, 31 December 2001. http://www.arib.or.jp http://www.cenorm.be http://www.iso.org http://www.jisc.org http://www.wto.org

Kagami 03 chap 11

2/15/04

2:54 PM

Page 247

12. Is the Japanese press a dinosaur in the 21st century?: the IT revolution and newspapers in Japan Kojiro Shiraishi 1.

INTRODUCTION

After 20 months of research, the Japan Newspaper Publishers & Editors Association (JNPEA) released a study on the future of newspapers in February 1998. The project was motivated by the uncertainty spreading through the Japanese media, a result of the explosive growth of the Internet, foreign media moguls’ entry into Japan,1 newly launched digital broadcasting services, rapidly developing technologies in editing and printing, and the shrinking ad market caused by Japan’s economic downturn. Some media critics liken the Japanese press to a dinosaur. These observers think highly of digital media, such as the Internet and satellite broadcasting. They often say that electronic newspapers distributed through the Internet, satellites or both will replace conventional ink-on-paper. They also point to ecological concerns, saying that newspaper publishing is eating up the world’s forests. The Association’s study, ‘Newspapers Take On The Digital Information Age: Can Journalism Survive?’ came to four main conclusions. First, that it is becoming more important for editors and publishers to make every effort to heighten the reliability of newspapers in the multimedia age. Second, that we should maintain the Resale Price Maintenance System2 in order to protect our public role as a medium of expression and reporting. Third, that we should try to aggressively take part in electronic and electric wave media that merge telecommunications and broadcasting services, and in which non-media enterprises are ready to participate. And finally, that we should unite to solve such imminent problems, instead of spending excessive time and energy on competitive battles to expand our readership. The Japanese press has been forced to deal with drastic changes initiated by quickly advancing digital technologies since the late 1990s, even before ‘IT 247

Kagami 03 chap 11

248

2/15/04

2:54 PM

Page 248

Challenging issues

revolution’ became a buzzword in Japan. With the burst of the IT bubble economy in the United States, Japanese business leaders are likely to think twice before letting the IT revolution into their offices, envisioning a huge investment to build up a digital network that ultimately yields poor returns. Again and again, they have the same nightmare: digitized offices equipped with networked computers defy the promise of a ‘paperless office’ and instead produce mountains of printed documents. They doubt whether the IT revolution can truly enhance productivity. But even with such doubts about its economic benefits, major industries in Japan cannot afford to ignore the IT revolution. Japan’s press is no exception. Tough competition among Japanese national and local newspapers has forced publishers to digitize the processes of editing, printing and shipping. At the same time, pressure from overseas has led them to expand their news distribution services from conventional ink-on-paper to such digital media as the Internet, cell phones and satellite data broadcasting. Stakes are high with such newly launched digital services because of their lower profitability. Japanese newspaper publishers worry that Japanese youth, who are potential subscribers to newspapers, will be addicted to digital gadgets and pay little attention to a print culture. This chapter analyses the Japanese press in the age of the IT revolution, focusing on one of the leading national dailies, The Yomiuri,3 where the author is employed. Section 2 describes the characteristics of Japan’s press and its current situation. Section 3 outlines newspapers’ business ties with TV stations in Japan, and Section 4 refers to their news distribution services via digital media.

2.

NEWSPAPERS IN JAPAN

The first modern, printed newspaper in Japan appeared in the late 19th century, after the Meiji Restoration of 1868 led Japan to open its doors to Western culture. In feudal times, especially in the Edo period under the Tokugawa Shogunate, ordinary people loved to read gossip written on pieces of wood named ‘kawara-ban’ and sold for a penny or two on the street. The Yomiuri, one of the oldest Japanese newspapers, takes its name as well as its origins from kawara-ban. ‘Yomiuri’ literally means a newspaper with its headlines read (yomi) and sold (uri) on the street. The Yomiuri will celebrate its 130th anniversary in 2004. Today, Japan’s press is comprised of 122 daily newspapers printing 72 million copies every day across the nation. This compares to 56 million copies in the United States and 42 million copies in China (Table 12.1). Given that Japan has a population of about 126 million people in 43 million households,

Kagami 03 chap 11

2/15/04

2:54 PM

Page 249

The IT revolution and newspapers in Japan

249

this works out to more than one newspaper per household. At the root of this world-leading readership are Japan’s almost 100 per cent literacy rate, economic prosperity and a coast-to-coast home delivery system.4 Table 12.1 Aspect

Newspaper circulation by country Japan

US China India Germany Russia

Total circulation (million) 72.41 56.18 42.47 Per 1000 people 577 201 36

26.54 28

24.85 303

22.80 141

Source: World Press Trends, a report issued by the World Association of Newspapers for 1998.

Once a month, Japan has a press holiday, on which neither a morning nor evening edition is printed. These holidays were established to give delivery personnel a break from their demanding schedule. Yomiuri Shimbun prints The Yomiuri in Japanese and The Daily Yomiuri in English. The Yomiuri has both morning and evening editions printed at 22 domestic printing sites and five overseas printing sites. A total of 10.2 million copies are printed of the morning edition and four million of the evening edition, excelling over competing newspapers The Asahi, The Mainichi, The Nihon Keizai and The Sankei. The Guinness Book of World Records lists The Yomiuri as the world’s biggest commercial newspaper. A copy of the morning edition costs 130 yen and the evening edition 50 yen (including consumption tax). The monthly subscription fee for both editions is 3925 yen, again including tax. The Yomiuri is one of five major dailies categorized as a general interest national daily newspaper that covers almost all of Japan, as opposed to local newspapers that are based in an individual prefecture (Japan consists of 47 administrative entities called prefectures). National dailies are most widely read in big cities like Tokyo and its suburbs, Kanagawa, Saitama and Chiba. In the Tokyo metropolitan area, for example, where more than 60 per cent of The Yomiuri’s 10.2 million readers live, it is read in one of every 2.9 households, a diffusion rate of about 35 per cent (Table 12.2).5 In smaller cities, however, local newspapers dominate the market. Japanese newspapers have been expanding their business in dog-eat-dog competition both with each other and with magazines, television and radio. Circulation has risen year by year, but recently seems to have reached a plateau. Subscription fee revenue accounts for almost 50 per cent of overall revenue, with advertisement accounting for about 40 per cent. TV broadcasting has already excelled over print media in advertisement sales (Table 12.3).

Kagami 03 chap 11

2/15/04

2:54 PM

250

Page 250

Challenging issues

Table 12.2 Circulation of ‘The Yomiuri’ (copies of morning edition) and diffusion rate (per cent in parenthesis) by prefecture Prefecture Tokyo Kanagawa Hokkaido Kyoto Hiroshima Fukuoka

Local daily The Yomiuri The Asahi The Nihon Keizai 284 779 (5.2) 236 282 (6.9) 1 229 121 (50.4) 425 973 (42.0) 660 612 (58.4) 632 614 (32.5)

1 555 848 (28.3) 1 094 317 (32.0) 260 670 (10.7) 189 367 (18.7) 159 923 (14.1) 410 201 (21.1)

1 271 773 (23.1) 992 418 (29.0) 163 841 (6.7) 199 366 (19.7) 138 932 (12.3) 340 619 (17.5)

586 019 (10.7) 280 978 (8.2) 59 775 (2.5) 61 022 (6.0) 61 123 (5.4) 86 858 (4.5)

Source: Diffusion Rate, January–June 2001, published by Japan Audit Bureau of Circulation Association (JABCA).

Table 12.3 Daily newspapers’ circulation, diffusion rate and subscription rate Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Circulation Copies Copies per Monthly Subscription (unit: 1000) per household 1000 persons rate (Yen) 72 536 71 690 72 043 71 924 72 047 72 705 72 699 72 410 72 218 71 896

1.24 1.22 1.22 1.20 1.19 1.19 1.18 1.16 1.15 1.13

589 580 581 578 578 582 580 576 573 570

3190 3650 3650 3850 3850 3850 3850 3925 3925 3925

Source: The Japanese Press 2001, JNPEA.

What worries Japanese newspaper publishers most is how long they will be able to enjoy the large circulations that now support their business. They fear that coming generations will spend more time Net-surfing or playing electronic

Kagami 03 chap 11

2/15/04

2:54 PM

Page 251

The IT revolution and newspapers in Japan

251

games than reading printed materials. For example, Japanese readers used to have both morning and evening editions delivered to their homes every day. With lifestyles changing rapidly – husbands working until late in the evening and housewives taking part- or full-time jobs – there is often no one home to read an evening newspaper. Subsequently, the circulation of evening editions has been going down. The Sankei, a major national daily, announced in its morning edition for 8 November 2001 that it would stop printing its evening edition in April 2002, due to its readers’ growing indifference toward evening newspapers. Another reason was that the publisher has found it difficult to make ends meet if it keeps printing two editions. Other national dailies, for the time being, are not following suit. Japan’s current recession makes it hard for newspapers to gather enough ads to fill their evening editions, and it is becoming a heavy burden for sales agents to employ newspaper boys to distribute the evening edition.

3.

NEWSPAPERS AND TELEVISION

Japanese newspapers have strong ties with private TV stations (with the exception of the government-sponsored broadcaster Nippon Hoso Kyokai (NHK), known in English as Japan Broadcasting Corporation). The Yomiuri is affiliated with Nippon TV Network (NTV), Channel 4; The Asahi with Asahi National Broadcasting Co., Ltd., Channel 10; The Mainichi with Tokyo Broadcasting System, Inc. (TBS), Channel 6; The Nihon Keizai with TV Tokyo, Channel 12; and The Sankei with Fuji TV, Channel 8. There are similar ties between local newspapers and local TV stations. In the early days of TV broadcasting after World War II, the Japanese press acquired influence over TV companies by providing them with money and personnel. Newspapers and their sister TV stations in general maintained cooperative relations. For example, The Yomiuri provided NTV with a news programme compiled from its morning and evening editions for three decades, to help strengthen NTV’s news content. When it comes to networks for gathering news across the country, major national dailies are superior to leading TV stations. All in all, newspaper reporters provide more quality and quantity than their TV counterparts. The Yomiuri took advantage of cross-media reporting, airing its own news programme and reaching potential readers through the TV screen. Other media groups also adopted these tactics and began to air similar programmes. In addition to TV news programmes, The Yomiuri has provided news services for cable TV networks, radio stations and electric billboards (on the street and on bullet trains) since the 1980s. In the late 1990s, Japanese national dailies launched websites in keeping with the trend in the US news media. In 1999

Kagami 03 chap 11

2/15/04

252

2:54 PM

Page 252

Challenging issues

they began distributing news on the cell phone Internet connection services initiated by NTT-DoCoMo, a digital mobile phone company. In the meantime, the Japanese press has diversified its business by entering satellite broadcasting services in cooperation with electronic manufacturers and trading companies. According to a recent survey conducted by JNPEA, 89 newspapers and news agencies have opened websites to distribute news materials. A total of 41 news organizations circulate breaking news through cable TV networks and mobile phones. Twelve major newspapers and news agencies are involved in digital satellite broadcasting services,6 and several newspapers have begun testing readers’ response to electronic newspapers and electronic flyers.

4.

OUTLOOK OF NEW BUSINESSES

4.1

Internet

Yomiuri Shimbun Co. Ltd is a media conglomerate encompassing a variety of business entities, including a professional baseball team (The Tokyo Yomiuri Giants), a travel agency and an orchestra. The flagship of Yomiuri Group is, without doubt, The Yomiuri. Member companies are supposed to secure and help its core business – newspaper publishing. The Asahi, The Mainichi and The Nihon Keizei are also based on their own convoy system as comprehensive media-related business groups. By diversifying the front lines of their business, news organizations aim to catch consumers’ attention and motivate them to start, or continue, subscribing to newspapers. Japanese newspapers frequently host concerts and exhibitions of fine art to underline their contribution to cultural activities. When necessary, they also ask their readers to donate money to aid domestic and foreign victims of natural disasters and terrorism. Primarily motivated by the rapid launch of websites among the US media, The Yomiuri launched its own Internet news services in June 1995, as a way to multiply its news distribution channels. Yomiuri On Line (YOL) carries breaking news and editorials both in Japanese and English. In the early 1990s, just after the bubble economy burst, Japan was criticized for being faceless or invisible, meaning that it was not clear who governs in Japanese politics – lawmakers, bureaucrats or business leaders. Editors of The Yomiuri came to see the Internet as a way of reaching the world, beyond time zones and international borders, to help Japan be recognized as it is. Besides giving foreigners easy access to today’s Japan, the Internet also helps Japanese nationals living overseas know what’s going on back home. It prints

Kagami 03 chap 11

2/15/04

2:54 PM

Page 253

The IT revolution and newspapers in Japan

253

overseas editions of its Japanese-language paper in New York, Los Angeles, London, Bangkok and Hong Kong. A major concern for newspapers is how to recover the expenditures made to set up Internet businesses. The US media originally expected three streams of revenue: subscription, advertisement, and transaction. Before long, however, they found it difficult to charge users for access, because premium sites lost visitors and were subsequently less attractive to advertisers. Internet users had a strong belief that information on the World Wide Web should be freely exchanged. To them, the Internet was free TV broadcasting funded by commercial film revenue, and they did not want to provide their e-mail addresses and personal data to obtain free access to news sites. The Internet ad market was too small to feed news sites, and commissions from selling and buying goods on their sites was almost nil. Given this state of affairs, The Yomiuri had little hope of making a profit from Internet business. The editors in charge of Internet services thought it would take a few years just to break even, which turned out to be a slightly optimistic forecast. People in the sales department did not rule out a kind of cannibalism, in which certain readers might shift to free Internet services and undermine their business. Ad people, too, feared that their market would shift from ink-on-paper to the Internet. Ultimately, however, The Yomiuri decided to give it a try and jumped onto the bandwagon of Internet news distribution services. YOL kicked off as a rudimentary website in mid-June 1995. It was updated only twice daily, in keeping with the deadlines for the morning and evening editions of The Yomiuri. In the embryonic days of Japan’s Internet age, there were no exact access figures. According to YOL’s analysis, only 8961 pages were viewed in the first seven days after its launch. In an article announcing the startup of YOL, the 28 July 1995 issue of The Yomiuri estimated that a total of about five million computers were connected to the Internet in 150 countries, with an estimated 50 million users. As of 1997, 11 million Japanese people used the Internet and 2.8 million households were connected to it, making for a diffusion rate of 6.4 per cent. Its ad market was about ¥6 billion, one-thirteenth as much as in the United States. Over time, the number of Internet users has exploded worldwide, thanks to improved connections and better performance from both hardware and software. As of December 2000, there were 47 million Internet users in Japan, according to a survey conducted by the Ministry of Public Management, Home Affairs and Posts and Telecommunications.7 In September 2001, YOL’s monthly page views reached a record high of 90 million due to coverage of the terrorist attacks in the United States. The Yomiuri has revised its homepage almost every year to improve its content and meet requests from ad departments. Today, it offers ten times as much content as the original version.

Kagami 03 chap 11

254

2/15/04

2:54 PM

Page 254

Challenging issues

The Yomiuri puts up about 200 stories on its website every day, updating the site around the clock. Around 80 per cent of the website content is taken from The Yomiuri’s morning and evening editions, with the remainder being original product written for YOL. Stories are taken verbatim from the newspaper, and visitors to the site can read these stories at no charge. One special programme titled Internet Broadcasting uses streaming technology to transmit moving pictures and sound on the Web. Internet users can enjoy news services similar to TV. This service is still being offered on an experimental basis, to determine its business possibilities in the coming age when telecommunications and broadcasting will be fused together. YOL concentrates on offering web magazines that provide original content, independent from The Yomiuri. Its Web magazines are aimed at specific interest groups; Ote-Komachi, for example, is meant for career women in the business centre of Tokyo, while @cars is for young adults who love automobiles. Yomiuri Shimbun’s Media Strategy Bureau is in charge of Internet operations and wireless news distribution services. In 1995–96, the number of Internet staff was increased by five times, to the current strength of about 60 employees, including temporary workers. Most of the reporters and editors were transferred from the newsroom, while computer engineers were brought in from the newspaper production bureau. Yomiuri Shimbun’s Internet services are still in the red, but the situation has been improving somewhat. In fiscal year 2000, the revenue from banner ads on YOL more than doubled to 465 million yen, up from 200 million yen in the previous year. Expenditures in fiscal year 2000 totalled 877 million yen, 60 per cent of which was allotted for employees’ salaries. YOL’s editors probably expected their Internet services to move into the black in fiscal 2001, but the ad market failed to grow as anticipated. It may have contracted to the level of the previous year; at the very least, it has remained flat. In fiscal year 2001, they would see even worse cost-performance from Internet news services. A few US and European newspapers were forced to reduce their digital workforces because of difficulties turning a profit. The Yomiuri has not been driven this far, but has had to make every cost-cutting measure possible. Among national dailies, The Nihon Keizai is the only newspaper that makes a profit from its online business. Nihon Keizai’s website supervisor Tomomi Tsubota attributes the success of Nikkei Net to the fact that the newspaper, which is comparable to The Wall Street Journal in the United States, deals with economic and financial news critical for businesspeople. Subsequently, it receives an enormous amount of access,8 which leads to greater ad revenues than those achieved by general interest national dailies such as The Yomiuri and The Asahi. Japanese newspapers are trying to charge for their Internet news services, but they have yet to find a way to make this feasible.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 255

The IT revolution and newspapers in Japan

255

‘Internet users have the mistaken idea that information on the Web is free’, warns Ikuo Nishioka, former vice president of Intel Japan.9 ‘We should create a new culture in the Internet Age, one in which recipients of valuable information pay an appropriate fee to content providers.’ 4.2

Cell Phone News Service

Japanese media are hoping that their news distribution services on digital cell phones will make up for their losses on the Web. Since 1999, NTT-DoCoMo’s Internet mode services (i-mode) have made it possible for subscribers to access websites specially designed for digital cell phones. As a result, cell phones have turned from a conventional device for voice transmission into an innovative mobile channel for information and entertainment. Subscribers to i-mode services cannot only exchange e-mail but also read breaking news, reserve tickets, and buy newly released pop songs. Young Japanese are heavy users of cell phones, and regard their gadgets as a sort of personal digital assistant powered by telephone technology. ‘Cell phones are an indispensable tool for college students these days,’ says Tatsuo Inamasu, a sociology professor at Hosei University in Tokyo. ‘For example, when a senior gets a job interview, he or she is asked to leave their cell phone number for further contacts. Until recently, you had to give a home telephone number.’10 Mobile phone subscribers now outnumber solid line subscribers. A total of 72 million people were counted as cell phone users in November 2001, meaning that more than one in two Japanese carries a cell phone. Competing mobile telecommunications companies J-Phone and AU have tried to catch up with NTT-DoCoMo and expand the market by offering new services. As a result, one-third of Japan’s cell phone users are now believed to subscribe to i-mode or similar services. The Yomiuri accepted an offer from NTT-DoCoMo to be one of its initial content providers in 1999. Ever since i-mode was launched, The Yomiuri has provided breaking news abridged for the service, along with information on sports, entertainment and financial markets. The Yomiuri also cooperated with its sister newspaper The Sport Hochi to beef up its coverage of sports and entertainment. At their peak, during Japan’s professional baseball season, The Yomiuri and its collaborators gathered a total of 280 000 subscribers on three mobile telephone companies’ services. Each subscriber pays 200 yen per month to access pages of The Yomiuri, in addition to basic charges for voice and data transfer. Nine per cent of these fees go to the mobile phone company that operates the services. Most people subscribe to more than one service, ultimately paying more than 10 000 yen monthly.

Kagami 03 chap 11

2/15/04

2:54 PM

256

Page 256

Challenging issues

The Yomiuri is reaping a great harvest from the secondhand use of news stories originally edited for its own website. Only five or six Yomiuri staff members are involved in cell phone news services, but it is expected to yield about 100 million yen in profits each year. Table 12.4 Subscribers to ‘The Yomiuri’ site on cell phone companies’ Internet services Year

i-mode, NTT-DoCoMo

April 99 April 00 April 01 June 01 Source:

4.3

1462 42 083 189 527 215 379

J-Sky, J-Phone

No services available No services available 35 089 39 007

EZ-Web, AU

Total

No services available 1462 No services available 42 083 26 654 251 270 31 312 285 698

Media Strategy Bureau, The Yomiuri.

Digital Satellite Broadcasting

The year 2000 was a watershed for Japan’s broadcasting industry. Major broadcasters formed new business alliances with non-media enterprises to enter into high definition broadcasting services, also known as digital hi-vision. Among them were NTV backed by The Yomiuri, TV Asahi tied up with The Asahi and TV Tokyo supported by The Nihon Keizai. These key terrestrial wave-broadcasting stations did not rush into new services. They were motivated partly by self-defence, needing to protect themselves from new participants from outside the broadcasting business. They were also dubious about a government policy to digitize broadcasting to keep up with the IT revolution. Manufacturers of electrical appliances, on the other hand, supported the policy of digitization, in the hope of cashing in on a mass market for digital TV sets or set-top boxes to catch digital broadcasting satellite services. The Japanese TV market has already reached saturation point, with each household in Japan containing on average more than one television set. Prices for conventional TV sets have been going down. Japan’s electronic industry, therefore, has a strong desire to revitalize the domestic market by introducing digital broadcasting services. However, broadcasters have yet to air whole programmes in high definition, and against the expectations of those who, like it or not, got involved in satellite broadcasting services, Japanese consumers declined to buy new TVs only to watch brief clips in this high quality format. The Japanese government’s original forecast for the spread of digital satellite broadcasting was ‘ten million households in 1000 days’. As of October 2001, however, 11 months after its

Kagami 03 chap 11

2/15/04

2:54 PM

Page 257

The IT revolution and newspapers in Japan

257

launch, digital satellite broadcasting had only spread to 800 000 households. According to an interim financial report, all major broadcasters suffered huge losses in 2000, between ¥4.4 billion and ¥1.1 billion each. The Yomiuri entered a new field of digital satellite broadcasting called independent data broadcasting, in which only text and sound data are transmitted. It is a simplified version of Internet satellite broadcasting, which can be displayed on a TV screen. Nippon Data Broadcasting (NDB – Channel 940) mainly provides news, weather forecasts and market information, almost the same as the content of YOL. With few viewers and little revenue from ads and subscription charges, NDB is in the red and not expected to make a profit for several years to come. Apart from data broadcasting, The Yomiuri provides news commentary programmes for six hours a day on a communications satellite broadcasting channel, from the spring of 2002. Japanese communications satellites, originally for information exchanges among private and public organizations, were partly opened to broadcasting in the late 1980s. A new communications satellite launched in 2000 is allotted mainly for digital broadcasting, and The Yomiuri formed an alliance with NTV and other partners to establish CS Nippon, a broadcasting service company using this satellite.

5.

CONCLUSIONS

Takao Suzuki, director of the JNPEA’s ad department, emphasized the Internet’s impact on newspapers when he summed up a story from the 17 July 1999 edition of The Economist. According to Suzuki’s summary, the story ‘made an ominous prediction that the Internet would deliver a direct blow to advertising and editing, the lifeline of newspapers, and that if proper measures were not taken, newspapers would be driven out of the market within a few years.’11 Suzuki’s opinion is widely shared by Japanese newspaper publishers and editors. A sense of fear toward new media prompted investments to improve their editing and printing systems. For instance, in 1997 The Yomiuri established an Asynchronous Transfer Mode (ATM) network that enables high speed, high capacity data transmission between its head and branch offices and printing sites, in preparation for cross-media news distribution services. Regarding cross-media news services, Naoshi Hashimoto, a senior researcher in JNPEA’s technology department, reports on Media General, an American news syndicate based in Tampa, Florida. Its news centre combines newspaper, broadcasting and Web services, requiring its reporters to do stories for different media and appear on TV programmes.12 In response to such diversified information channels, The Yomiuri started operation of its own General Systems Centre in May 1997. Its core functions are a materials management system,

Kagami 03 chap 11

2/15/04

258

2:54 PM

Page 258

Challenging issues

general database, new advertising system and new business administration system. The main feature of the materials management system is efficiency in processing colour photos. This task once required a specialized operator, but now any reporter can do it. With its automatic image-processing function, even photos submitted right before deadline can be included in the paper. Tsuneo Watanabe, president, CEO and editor-in-chief of The Yomiuri and president of JNPEA, says in a company brochure13 that in the 21st century, information technology would surely develop at a startling pace. ‘While aiming to be a comprehensive communication media group that leads the way toward the future, the Yomiuri group is also determined to step up its efforts to advance and invest in electronic media,’ Watanabe said. He does not view the relationship between newspaper and electronic media as a ‘zero-sum game’. He also cited that ‘Besides their strong capability to cover news, newspapers also excel in areas such as ease of reading, document preservation, provision of information on which readers can base their value judgments, in-depth commentary and transportability … the more media diversify, the more prominent the role of newspapers will become.’ In November 2000, The Yomiuri launched a new department called the News Distribution Centre to efficiently disseminate breaking news to the Internet, satellite broadcasting and cell phones. The Yomiuri plans to have every reporter equipped with a digital video camera within a few years, so as to take both moving pictures and still photos. It has also started negotiating with its sister broadcaster NTV to build up the Video Centre to share news materials in the near future. All in all, Japanese reporters and editors have begun to change their ‘newspaper deadlines first’ policy and become more cooperative toward news distribution services for electronic and electric wave media. Robert Fiddler, of the Media Laboratory of the Knight-Ridder Group once said, ‘There are many people who believe that newspapers are dinosaurs and that they are going to become the road kill on the Information Super Highway in the not-too distant future. We believe exactly the opposite. We believe that newspapers can in fact evolve into a new form of media that blends the old familiar aspects of newspapers with the new technologies that are emerging. So that you have the ability to read, browse and scan, and at the same time being able to interact with the newspapers, to interact with advertisers through your newspapers in ways that are not possible through media today.’14 Like it or not, the Japanese press must change in keeping with the IT revolution.

NOTES 1. Media mogul Rupert Murdoch surprised the Japanese media in 1996 by purchasing stock in TV Asahi, one of Japan’s five leading TV stations.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 259

The IT revolution and newspapers in Japan

259

2. JNPEA has strongly opposed attempts by the Fair Trade Commission (FTC) to abolish the Resale Price Maintenance System, which has been permitted for newspapers because they are highly cultural products dealing with speech. JNPEA contends that if the principle of ‘the same paper, the same price everywhere in Japan’ were thrown away, it ‘would plunge the newspaper industry into a cutthroat underselling war’. 3. From now on, Japanese newspapers that have the word ‘Shimbun’ (newspaper) in their title will be abbreviated to The Yomiuri, and so on. 4. According to Data Book on Japanese Newspapers 2001, published by JNPEA, 93.5 per cent of newspapers in Japan are sent directly to homes or offices through news agents. 5. The Yomiuri Shimbun 2000/2001, published by The Yomiuri in June 2000, p. 15. 6. Data Book on Japanese Newspapers 2001, published by JNPEA, p. 32. 7. 2001 White Paper: Information and Communications in Japan, by The Ministry of Public Management, Home Affairs, and Posts and Telecommunications, available at www.soumu.go.jp 8. Japan Printer, March 2001, p. 14. 9. Statements at the Yomiuri IT Forum, held by The Yomiuri in Tokyo on 18 December 2001. 10. Ibid. 11. The Japanese Press 2000, published by JNPEA, p. 22. 12. Japan Printer, March 2001, p. 3. 13. The Yomiuri Shimbun 2000/2001, op. cit., p. 1. 14. Statements in video footage titled ‘Tablet Newspaper’, produced by Knight-Ridder.

REFERENCES Advertisement Bureau, The Yomiuri (2001), Media and Advertisement (in Japanese), The Yomiuri, Tokyo. Gilder, George (1992), Life After Television, W.W. Norton & Company, New York. Katsura, Keiich (1990), Japanese Newspapers Today (in Japanese), Iwanami-shoten, Tokyo. Kusano, Atsushi (2000), How To Deal With TV Reporting (in Japanese), PHP Institute, Tokyo. Muro, Kenji and Akio Nakamata (2000), What Has Happened to Reading, Dainipponinsatsu, Tokyo. Sasaki, Hiroshi (2001), SONY’s Strategy for Broadband (in Japanese), Nihonjitsugyoushuppan, Tokyo. Shimano, Isao (2000), Broadcast (in Japanese), Jitsumukyoiku-shuppan, Tokyo. Smith, Anthony (1979), The Newspaper: An International History, Thames and Hudson Ltd, London. Takahashi, Toru (2000), Beyond the Internet Revolution (in Japanese), KK-Bestsellers, Tokyo. http://www.yomiuri.co.jp (website of The Yomiuri).

Kagami 03 chap 11

2/15/04

2:54 PM

Page 260

13. PKI solutions for trusted e-commerce: survey of the de facto standard competition in PKI industries Atsuho Maeda 1.

INTRODUCTION

In this section, we focus on a new field of technology called ‘electronic certification’, which is expected to be an effective tool for ensuring the safety of commercial transactions (including international trade) in the midst of the increase in e-commerce on the open networks through the Internet. We will discuss the activities of the new enterprises related to the de facto standard at the industry level, and analyse their strategies for making inroads into Asian markets and their impact. We have been shifting our transaction basis from a norm of ‘face-to-face commercial transactions’, which involved only the trusted participants, into a business (including trade) format of ‘open electronic commerce’ conducted via the Internet. The question of how to ensure mutual trust between those involved in the transaction has become an important issue. Along with the growth of the Internet as the basis of business affairs, ‘the positive impact’ is that conventional business practices of exclusive associations have been shaken off, and business opportunities have expanded. On the other hand, ‘the negative impact’ is that it is also necessary to be aware of the unlimited increase in the danger of illegal access to the data for transactions conducted on the Internet, including ‘wiretapping,’ ‘tampering’ and ‘impersonation’. Transactions that are made on an open network are between multiple, unspecified users (individuals, corporations, government agencies), unlike the dealings within a business hierarchy upon guaranteed mutual trust. For these open network transactions, the issue of the reliability of the other party is an extremely difficult one to find the best solutions for. The mutual trust system for business transactions is divided into the following two areas: 260

Kagami 03 chap 11

2/15/04

2:54 PM

Page 261

PKI solutions for trusted e-commerce

261

(i)

Credit rating (evaluation of the quality as a business partner and its solvency); (ii) Authentication (verification that the communicating party is actually who they claim to be, as well as the integrity of the transferred contents). The first of these, (i) credit rating, has been commonly used by businesses and financial institutions, and is the important point that requires the most attention for a commercial transaction. It will remain difficult to handle electronically, even in the future. For item (ii) authentication, as there has been a shift to the Internet as the basis for commerce, this has been an emerging issue, particularly since the latter half of the 1990s. There continues to be lively development by the rising system solution enterprises aiming to offer a trusted communication technology platform. Some of the systems established so far are a digital signature (verification using an appended digital ID) as a means of verifying the true identity of another party, and a certificate authority (CA) to verify registration of the communication partner. These are both based on technologies of encrypted data transfer by the Public Key method. For this version, these are generally called PKI (Public Key Infrastructure), as Figure 13.1 shows. The purpose of this section is to clarify the characteristics of the de facto standard strategies of the main vendors associated with PKI (hereafter, called PKI vendors). Hence, with regard to the credit rating, it will not be discussed here.

RA STEP 1

Digital ID

IA Certificate Authority

STEP 3

INTERNET

STEP 6

STEP 2

Sender A STEP 5 STEP 4

Figure 13.1

Crypto graph

+

Digital ID

Recipient B

Public Key for Sender A

Public Key for Recipient B

Private Key for Sender A

Private Key for Recipient B

Image of Public Key Infrastructure

Kagami 03 chap 11

262

2/15/04

2:54 PM

Page 262

Challenging issues

The PKI forming the basis for (i) authentication originally developed from technology that was used by the military. In order to ensure the security of top secret and highly sensitive information, state-of-the-art element techniques were adopted. Among these, encryption technology has been known since the period of the ancient Roman Empire as an effective and convenient tool. Since the late 1980s, international society has been facing the rapid demise of the Cold War structure. The increasing military competition between superpowers based on nuclear weapons has abated. As a result, the security menace has shifted from being certain nations that possessed special weapons to being anonymous individuals with an excellent understanding of advanced Internet technology. In a sense, nowadays, the authentication infrastructure has the same significance as the national security systems during the Cold War. However, it is not necessarily required to prevent society from using the electronic certification systems based on strong encryption technology. Society’s awareness of issues such as (1) expansion of the use of the Internet, (2) spread of e-commerce (B2B, B2C, and so on), (3) legal framework to deal with the digitalization of business transactions, and (4) security problems related to e-commerce has progressed rapidly. As a result, at the beginning of the 21st century, the following major changes have begun in the environment surrounding encryption technology: • Relaxation of encryption technology export control regulations by the US Department of Commerce (January 2000): The DOC announced that, in accordance with President Clinton’s policy on regulation relaxation, the export of any cryptographic commodity to an individual or commercial enterprise outside the US would be allowed without government approval, except for some special cases. • Expiration of the RSA method encryption technology patent (basic PKI patent) (September 2000): The exclusive right of use within US held by RSA Security Inc. for 17 years regarding the patent on the RSA method of public key encryption expired in September 2000, and the algorithm entered into the public domain. As this shows, ‘the year 2000’ was a significant turning point for the commercial application of encryption technology, particularly in the US, with rapid progress in openness and liberalization. It is impossible to measure the significance of relaxation of the technological entry barriers and the expansion of the possibility of international deployment of authentication technology (see the huge R&D cost in PKI industry, irrespective of profit level, in Figure 13.2). Especially significant is the lapsing of the RSA patent, which will eliminate the licensing fees that PKI vendors have been required to pay in order to use the PKI basic patent (encryption technology). Reduction in licensing fees →

Kagami 03 chap 11

2/15/04

2:54 PM

Page 263

PKI solutions for trusted e-commerce

263

Drop in PKI-related costs → Increase in the number of PKI users. In line with the economic effect of network externalities, it is expected that there will be global expansion of PKI. 100 From CY1998 to CY2000 Results

P/(L) per revenue (%)

50

RSA

Entrust

0 VeriSign –50 –100

Baltimore –150 0

5

10

15 20 25 R&D cost per revenue

30

35

Source: US Securities and Exchange Commission: http://www.verisign.com/corporate/investor/index.html; http://www.rsasecurity.com/; http://www.entrust.com/investor/financials.htm; http://www.baltimore.com/investors/index.asp.

Figure 13.2

Profit/loss, R&D per revenue of major PKI vendors

On the other hand, from the perspective of PKI business, in response to PKI market expansion, various kit products will be offered, and product differentiation will progress (for example, development of products for mobile communications, and so forth). As a result, intense competition among businesses to become the de facto standard in the system solution field has begun. The analysis described below primarily focuses on the main PKI vendors.

2.

ANALYSIS OF MAJOR PKI VENDORS

At present, the major enterprises known throughout the world, and expected to develop internationally in the field of electronic authentication (including certificate authority supporting systems) are;

Kagami 03 chap 11

2/15/04

2:54 PM

264

A. B. C. D.

Page 264

Challenging issues

RSA Security Inc. (US); the licensor of the RSA patent described above; VeriSign Inc. (US); the largest company, spun off from RSA; Entrust Inc. (Canada); an off-shoot of Nortel Networks Ltd.; Baltimore Technologies plc (Ireland); EU standard.

Each of these companies has their own strategy. However, what is common is that they all have strategies for government, industry and business, pursuing aggressive sales with the aim of establishing the de facto standard. ‘The division of the world’ in PKI business has been started by these four players. A, B and C are struggling for supremacy in North America, while D commands the largest share of the European market. The question of Asia will be discussed in detail in a later section. This section deals with the specifics of the strategies to become the de facto standard by studying the global strategies of each of these companies. See Figures 13.3 and 13.4 for comparisons of revenue per capita and of quick ratios, respectively.

400

CY2000

1000 US$/person

354

291

300

200

Revenue per capita

180

US

133

US 100

Canada

EU

0 VeriSign 2291

RSA 1093

Entrust 1108

Baltimore 1185

Number of employees Source: US Securities and Exchange Commission: http://www.verisign.com/corporate/investor/index.html; http://www.rsasecurity.com/; http://www.entrust.com/investor/financials.htm; http://www.baltimore.com/investors/index.asp.

Figure 13.3

Revenue per capita by major PKI vendors

Kagami 03 chap 11

2/15/04

2:54 PM

Page 265

PKI solutions for trusted e-commerce

A.

265

RSA Security Inc.

Corporate strategy The strategic theme for the company is ‘shaking off from its patent dependency and establishment of supremacy through its technical advances other than patent’. Since September of 1983, when the patent for ‘Public Key Encryption and Decoding Algorithm’1 was granted, the company has taken advantage of its patent, which continues to be the effective standard in the huge US market. Most of the rival PKI vendors sell PKI systems that use the RSA method, and a portion of their profits have been paid as licensing fees to RSA. RSA has aggressively defended these patent rights, including suits claiming patent infringement. However, as explained in an earlier section, this patent lapsed in September 2000, hence the situation has changed. During the time that the patent was in effect, PKI peripheral technology, such as access authorization systems and so on, that the company established, was predominant. In order to continue this dominance, these products and PKI products are being integrated as a market expansion strategy. Business model Unlike other companies, the certification system is not set according to a specific business model; instead, RSA emphasizes the interoperability with systems from other PKI vendors. The global standards SecurIDTM (a two-factor access authorization system, with an 80 per cent share in the US market) and BSAFETM (an encryption generation toolkit, 800 million licences sold worldwide as of the end of 2000), and the highly compatible certificate authority supporting system KEONTM were developed for corporations performing their own PKI structure, and sales expanded rapidly (Worldwide 1999: US$5M → 2000: US$20M). However, they do not issue certificates as a certificate authority. The sales operation of RSA is based on direct sales. Particularly for developing new markets outside the US, there is a great deal of importance placed on the sales channels, like resellers and distributors, and they are more cautious than other companies about establishing branch offices. For this reason, entry to the market outside the US has been delayed. B.

VeriSign Inc.

Corporate strategy The objective of this company, as stated by the company President & CEO, Stratton Sclavos, is ‘From a security services company into the Internet’s most

Kagami 03 chap 11

2/15/04

2:54 PM

266

Page 266

Challenging issues

Quick assets ÷ Current liabilities × 100% 900 800 700 600

M&A

Stock Sales

VeriSign

RSA

% 500

Stable

M&A

400 300 200 100 0 1999

Entrust

Baltimore

2000

Source: US Securities and Exchange Commission: http://www.verisign.com/corporate/investor/index.html; http://www.rsasecurity.com/; http://www.entrust.com/investor/financials.htm; http://www.baltimore.com/investors/index.asp.

Figure 13.4

Quick ratio comparison of major PKI vendors

trusted utility’. In keeping with this, company strategies are to establish a solid business foundation as a pillar of the US and to embrace customers and powerful industries through M&A; as a result, VeriSign’s quick ratio2 shows its weakened financial balance as shown in Figure 13.4. Customers include public agencies like the US Internal Revenue Service (IRS) as well as major businesses like Bank of America, Hewlett-Packard, Kodak, and Texas Instruments. It is known as the ‘world’s largest electronic certificate (digital ID) issuing organization’, with more than 500 000 (total as of March 2001) electronic certificates for websites issued. The company is succeeding in getting a network of multinational businesses based in the US using its company’s certification networks. VeriSign’s strong domain is that it also supports companies in issuing electronic certificates on their own. Customers are not required to build the difficult-to-manage certificate authority infrastructure; they are able to completely outsource all the work prior to the actual issue of electronic certificate, to VeriSign.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 267

PKI solutions for trusted e-commerce

267

Business model This company’s PKI product is a certificate authority supporting system called VeriSign OnSiteTM. The features, as mentioned above, are the high level certificate processing functions that allow total outsourcing to meet the customer requirements. It has passed the system audit by the American Institute of Certified Public Accountants (AICPA), known for strict on-site audits, and VeriSign also issues electronic certificates as a certificate authority. With regard to the composition of the authentication systems, this company has adopted its own unique strategy. A worldwide certification service network, called VTN (VeriSign Trust Network) has been established, and the company is succeeding in promoting the VeriSign brand as a certificate authority through associated companies throughout the world. In other words, the strategy is to build systems connecting global sales networks based on the certificate structure, centred at the US headquarters. It is a ‘hierarchical authentication structure’, but a defining difference is that the highest level in the authentication hierarchy is not a public Root CA (that is, an authority that is recognized by a public sector, and so forth), but a private business. The advantage is that it is possible to expand the communication network to any entity or individual in the world, based on the assumption of ‘trust in the closed community of VTN’. As a result, the company’s clients include government agencies with strict security requirements, including the FBI (Federal Bureau of Investigation) and IRS (Internal Revenue Service), as well as the NRC (Nuclear Regulatory Commission). One of the problems that can be pointed out is the risk of all authentication functions being in the private sector. There is likely to be particular resistance to this among users outside the US, since the highest authority in the system is a US company. It is not possible to get away from this trade-off in this case. In this regard, there is a great deal of attention being paid to public safety assurance, and VeriSign obtained a certification of qualification of common criteria EAL4 (Evaluation Assurance Level 4) in December 1999, recognized as an international standard of the ISO (International Organisation for Standardization) (ISO/IEC15408). The sales system of VeriSign is primarily website transactions bolstered by its strong brand identity. As mentioned earlier, in the overseas markets its subsidiaries are the private sales network called VTN. C.

Entrust Inc.

Corporate strategy Started as the electronic certification business department of Nortel Networks Inc., the Canadian communication equipment manufacturer (as of the end of December 2000, Nortel held 26 per cent of the Entrust shares). This company’s strategy is to build an authentication network focused on large clients in North

Kagami 03 chap 11

268

2/15/04

2:54 PM

Page 268

Challenging issues

America, and to enhance the direct sales system aimed at powerful customers. Among its strategic customers, in addition to the Ontario provincial government and the Canadian federal government, are many relatively large government agencies and companies, such as the US Department of Defense, NASA, NSA (National Security Agency), SWIFT (Society for Worldwide Interbank Financial Telecommunication S.C.), J.P. Morgan Chase & Co. This company has increased clients on the basis of certificate compatibility with major certification systems for specific government and industry users. There is a striking tendency to form blocks of the same type of customers, such as the federal government and securities industry. In July 2001, the company successfully closed a deal for the largest order since its founding (value US$18 million) for the Secure Channel Project (part of a digital government plan being conducted by the Canadian government). Since May 1999, Entrust began operating an electronic certificate issuing business with the platform ‘Entrust.net’. However, the operation is fundamentally different from the outsourced certificate authority offered by VeriSign. Entrust.net assumes the function of the certificate authority is operated by the enterprise itself (in-house). As a result, Entrust’s customers are mainly large enterprises that are capable of operating as a certificate authority on their own. Business model The company’s main PKI product is the certificate authority supporting system Entrust/PKITM. As mentioned above, the presumed customers are enterprises running their own certificate authority, and this defines the business model. In other words, it inevitably adopts an operation style that is adapted to big business, since its customers are large enterprises who can run a certificate authority themselves within their own organization. For example, compatible certification is built into the authentication system for each client enterprise, forming an enormous certificate network. There is an assumption of a distributed certification system that has a ‘cross-certification structure’. The remarkable difference from VeriSign is that the network is expanded on the basis of the certification systems of the client enterprises, not by absorbing the customers into the company’s own network. In order to implement this original business model, Entrust has a strong tendency to enhance the direct sales system focusing on large institutional customers. D.

Baltimore Technologies plc

Corporate strategy This company is headquartered in the UK. As clearly seen from the shipments in 1999 (80 per cent to European countries), its primary markets are in the EU. The main customers include EU bodies, the British Ministry of Defence, and

Kagami 03 chap 11

2/15/04

2:54 PM

Page 269

PKI solutions for trusted e-commerce

269

Deutsche Telecom. However, since severe international competition has started in the electronic certificate market, it has become necessary to develop bases of operation abroad, as Figure 13.5 indicates. To accomplish this, the company began a thorough M&A strategy in 2000, moving into markets outside of Europe, such as the US, Canada and Japan. The company’s acquisition of Security Domain Pty Ltd in March 1998 marked the start of its entry into Australia. Later, it began accessing markets in areas with different cultural backgrounds, like the US and Asia. Excluding Japan and Korea, it has conducted business through wholly owned subsidiary companies. Furthermore, the company has become a major PKI vendor, forming a global network and establishing a cooperative relationship with Identrus LLC, which performs Root CA for financial institutions. Identrus is affiliated with 47 financial institutions worldwide (as of July 2001), including ABN AMRO Bank, N.V., ING Group, Citi Group, and the HSBC Group. This company has been designated the SET certificate authority by VISA and MasterCard International, starting with Asia and the Pacific region. It is characteristic of the company to focus on certificates for the financial sector. Business model First, this company’s product strategy is to provide certificate authority supporting systems as well as certificate authority hosting services. It does not conduct any electronic certificate issuing business (except in Australia). From the perspective of certificate system structure, this company adopts a hierarchical structure. This means the Root CA (supreme level certificate authority) is at the top of a pyramid format authentication system. The upper level CA issues the electronic certificates to the lower level CA. The Root CA is typically established for each field of industry, ensuring industry-wide PKI. Naturally, if the secret key of the Root CA is compromised, the entire PKI as a whole collapses, so there is an extremely high demand for security. Specifically, Identrus (this company’s customer) is a typical Root CA, issuing electronic certificates for financial institutions. Identrus has introduced the Baltimore Technologies CA supporting system, and issues an electronic certificate meeting global shared specifications for the member financial institutions. For the businesses with which these financial institutions have dealings, electronic certificates with the Identrus electronic signature (digital ID) form an impressive credit guarantee system for e-commerce, including accounts settlement. Identrus is not merely in the certificate business, it is also a comprehensive credit rating business. In the past, direct sales were the main format of the operations, but there is a tendency to focus on indirect sales through its partners for advancing into markets.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 270

270

Challenging issues

CY1999

CY2000

US 22%

Others 21%

Asia 22%

EU 44%

EU 79%

UK Ireland Rest of EU

Africa Australia Asia N/S America

Source: US Securities and Exchange Commission: http://www.verisign.com/corporate/investor/index.html; http://www.rsasecurity.com/; http://www.entrust.com/investor/financials.htm; http://www.baltimore.com/investors/index.asp.

Figure 13.5

3.

Consolidated revenue of Baltimore Technologies plc by region

STRATEGY FOR ASIAN MARKET

In this section, we will review how these PKI vendors enter into the Asian PKI market, especially from the strategic point of view by country. 3.1

Republic of Singapore

VeriSign Inc. In the past, VeriSign supported the Singapore market from MSC Trustgate.com (Malaysian related company). In response to the expansion of the electronic certificate market, in August 2001 a strategic alliance, including capital investment (less than 50 per cent), was started with a private Singapore certificate authority, TrustAsia Inc. (established April 2001). The company also received investment funding from the US, Singapore, and Japan (Nippon Venture Capital Corp.), and the range of operations is expanding through ASEAN and into China. For this reason, the centre of operations for Southern

Kagami 03 chap 11

2/15/04

2:54 PM

Page 271

PKI solutions for trusted e-commerce

271

Asia is in Singapore, and a base for Northern Asia operations is being established in Shanghai, China. Unlike the VTN member companies so far, TrustAsia has its own R&D facilities (encryption technology research in Shanghai). In addition, there is a unique business strategy, such as moving into regions where there are already VTN affiliated companies, like Malaysia and China. Baltimore Technologies plc This company concluded a procurement agreement for its PKI supporting system (UniCERT) with ID.Safe in December 1999. ID.Safe has been the official electronic certificate provider for the e-ASEAN working group since November 2000, using Baltimore’s technology. It also has a business tie-in with Singapore Network Services (SNS) Pte Ltd, a trading system enterprise in Singapore. Since January 1989, it has also been performing certificate services for TradeNet, an electronic trading system. 3.2

Malaysia

The first licensed certificate authority in Malaysia was DIGICERT Sdn Bhd. In 1998, POS Malaysia Bhd was established with funding from MIMOS Berhad, a government science and technology research organization, and the network business GITN Sdn Bhd. iD2 Technologies (Swedish company, funded by CISCO, Ericsson, Reuters, and others) is a PKI vendor for DIGICERT. After DIGICERT, the second licensed certificate authority in Malaysia, MSC Cybersign International Sdn Bhd (renamed MSC Trustgate.com Sdn Bhd) was established as an agency of the Multimedia Development Corporation (MDC) which holds an 80 per cent share. In June 2000, MSC Trustgate.com concluded a strategic alliance (including capitalization) with VeriSign and the Malaysian communications business, Lityan Holdings Berhad. In July 2000, it was licensed as a certificate authority by the CCA (Controller of CA). MSC Trustgate.com participated in the e-ASEAN Task Force conference held in Kuala Lumpur in August 2000, advocating the establishment of an international electronic certificate system making use of the ASEAN framework (integration of cross-certification and the legal framework) and the establishment of a ‘Central e-ASEAN Certificate Authority’. 3.3

Hong Kong (Special Administrative Region)

VeriSign Inc. In order to conform to the system of regulations and launch an aggressive offense on the rapidly growing Hong Kong market, VeriSign established HiTRUST.COM (HK), a Hong Kong subsidiary of HiTRUST.COM Inc.

Kagami 03 chap 11

2/15/04

272

2:54 PM

Page 272

Challenging issues

(Taiwanese company) as part of the VTN (December 2000). At that time, in addition to Hong Kong, major cities in China were added to the area of operation for HiTRUST.COM (HK), including Hebei, Jiangsu, Guangdong, Beijing, Shanghai, and Guangzhou. Baltimore Technologies plc In October 1999, this company established a base of operations into the Asian market in Hong Kong, opening a branch office. It was the first large PKI vendor to move into Hong Kong. In the Asian market, this company is a strong PKI builder for the trading sector, successfully marketing to the trading and financial EDI system operations groups in various Asian countries, including Hong Kong’s Tradelink, SNS in Singapore, and Trade-Van in Taiwan. The business organizations headed by these system operation entities conform to this company’s PKI system concept of a hierarchical structure. Since January 1997, Tradelink has been issuing electronic certificates for trade-related businesses as a voluntary certificate authority. Baltimore announced a business alliance with Hong Kong Post in January 2001. The purpose of the alliance is to offer this company’s WTLS (Wireless Transport Layer Security) technology. As a result, the Hong Kong Post has decided to adopt Baltimore as the electronic certificate system for mobile communications devices, allowing Baltimore’s product to become the de facto standard in Hong Kong. 3.4

Japan

RSA Security Inc. In November 1996, a wholly owned subsidiary (Japan RSA Ltd K.K.) was established. In May 1998, Security Dynamics Ltd K.K. was established as a wholly owned subsidiary of Security Dynamics Technologies, Inc. In November 1999, these two subsidiaries merged to form the current RSA Security Japan Ltd. In order to strengthen its operations organization in Japan, an agreement was reached with Fujitsu in August 1999 to form a sales alliance. The company is well known for the interoperability of its systems with the systems of other vendors, and it is expected to participate in Japan, which is likely to become crowded with multiple system vendors at national and regional government levels. VeriSign Inc. VeriSign Japan K.K. was established in February 1996, and has come to lead the market as a private certificate authority. In March 1997, the PKI Processing Centre was opened and operations for Japan’s first certificate authority started. As a result of a business strategy of sales alliances, this company’s electronic

Kagami 03 chap 11

2/15/04

2:54 PM

Page 273

PKI solutions for trusted e-commerce

273

certificate system is incorporated in system products, like SCM and VPN, sold by system integrators such as NTT Data, NEC, Toshiba and Nihon Unisys. VeriSign Japan is the first overseas corporation of the VeriSign Group. With regard to capitalization, in addition to the 51 per cent from VeriSign headquarters, investors include Mitsui Sumitomo Bank, Toshiba, NEC, Sony, and others. In December 2000, there was a capital increase through a thirdparty allocation of shares, adding investors like Tokio Marine and Nomura Securities, bringing the total number of investor companies to 33. When the capital increase through a third-party allocation of shares was conducted, VeriSign headquarters increased its stake to 69 per cent. Entrust Inc. This company’s entry into the Japan market was in June 1998 with SECOM Co. Ltd as its sales distributor in Japan. In December 1998, Entrust Japan Co. Ltd was established. However, Entrust Inc. provided only a little more than 10 per cent of the funding, with SECOM, the largest stockholder at 46 per cent, making Entrust Japan Co. Ltd a joint venture between 16 companies, including NTT Data, SONY, Tokyo Mitsubishi Bank, Mitsui Sumitomo Bank, Nomura Securities, and Sumitomo Electric. As Entrust, the business operations focus mainly on supplying PKI products and consulting, with the bulk of the operations handled by NTT Data and SECOM TrustNet. In recognition of its penetration into the Japanese market, a capital increase through a third-party allocation of shares was held in April 2002 at the same time as Entrust Inc. announced its intention to increase its share to 37 per cent (SECOM 38 per cent). In order to support digital government in Japan, which assumes certificate compatibility through Bridge CA, this company released the electronic certificate system ‘Entrust PKI e-Government Edition’ in April 2001. Baltimore Technologies plc The earliest progress in Japan by the four major PKI vendors was the recognition of NSJ Corp. (established August 1995) as a sales distributor by Baltimore in June 1998. In March 2000, Baltimore acquired 73 per cent of NSJ Corp., and in May 2000 changed the name to Baltimore Technologies Japan Co. Ltd (buy-out completed in July 2000). As a result, Baltimore has come to have a 45 per cent share3 of the Asian PKI market, including the rapidly developing Japanese market. Baltimore Japan has funding from 14 companies, including NTT DoCoMo, Nomura Securities, Tokyo Electricity, and the Japan Research Institute. In May 2001, an alliance was announced with SOK, a company that had started operations as a voluntary certificate authority, and Baltimore Japan decided to build a PKI structure to issue SOK brand electronic certificates. Through this, it began its pursuit of SECOM which had already begun operating

Kagami 03 chap 11

2/15/04

2:54 PM

274

Page 274

Challenging issues

in the network security business as a voluntary CA through its alliance with Entrust. In addition, Teikoku Databank, which is affiliated with this company, in September 2001, was recognized as the accredited certificate operator for the electronic bidding system for the Ministry of Land, Infrastructure and Transport (MLIT), and the vendor of this system is Baltimore Technologies. Teikoku Databank has amassed results for 100 years performing corporate credit services, and can combine electronic certification and credit rating administration. VeriSign, Entrust, and Baltimore have formed an electronic application promotion consortium targeting the digital government infrastructure for Japan; and, in cooperation with companies like Netmarks Inc., are realizing the world’s first certificate compatibility for systems from different PKI vendors. As a result, when PKI users perform electronic applications or electronic bidding, it will be possible to generally support each vendor’s certificate authority supporting system. 3.5

Republic of Korea

VeriSign Inc. This company is developing its certificate business by agreeing to cooperate in a strategic alliance (announced September 1999) with Korea Electronic Certification Authority, Inc. (KECA, established March 1999, Korea’s first certificate authority), operating as a certificate authority under the umbrella of KECA, and establishing CrossCert Inc. (established March 1999 as a separate company from the Korean PKI vendor, Syntech) as the base of VTN for Korea. In addition, in August 2001, an application was submitted to the MIC for designation as a licensed CA. Customers include large businesses such as POSCO, Samsung Electronics, Dongbu Insurance, and Hanhwa Securities. In June 2001, investments were also made in iTrusChina Co. Ltd, a VTN member certificate authority in China. Entrust Inc. This company established Entrust.net Korea Ltd in December 1998 as a CA for Korea. In August 2000, a business tie-up with Nattrak, a Korean Internet security company was made for electronic certificate business operations supporting Wireless Application Protocol (WAP), beginning the first certification business for mobile communications in the Korean market. Baltimore Technologies plc The SET (Secure Electronic Transactions) solutions vendor BARA e-Business and Communications Co. Ltd (established in April 1997) changed its name to BARA Baltimore Technologies Korea Co. Ltd and was designated as the sales agency for the Korean market (September 2000).

Kagami 03 chap 11

2/15/04

2:54 PM

Page 275

PKI solutions for trusted e-commerce

3.6

275

People’s Republic of China

VeriSign Inc. Through HiTRUST.COM(HK), a member of VTN in Hong Kong, VeriSign is aiming to expand the network into China. In December 2000, the CA processing centre was established in Hong Kong and began offering certificate services for users in Hebei, Jiangsu, Guangdong, Beijing, Shanghai, and Guangzhou, as well as Hong Kong. In April 2001, HiTRUST Inc. was established as a Chinese subsidiary of HiTRUST.COM Inc. Furthermore, in September 2000, iTruschina Co. Ltd was established, and began operations as a VTN affiliate following approval from the government in April 2001. Entrust Inc. In May 2000, this company opened a branch office in Beijing, and has been proposing independent CA construction systems for the People’s Bank of China (PBOC). The China Financial Certification Authority (CFCA) established as a joint venture of 13 commercial banks under the auspices of the PBOC, adopted the Entrust system in August 2001, and is conducting certificate services for bank users and those making security transactions. Baltimore Technologies plc In March 2000, this company and its business partner in Korea, BARA eBusiness and Communications Co. Ltd, established a PKI service company called Beijing Ether Electronics Group Co. Ltd, as a joint venture with the Engineering Research Centre for Information Security Technology (ERCIST), an agency of the Chinese Academy of Science and Technology (CAS). In September 2000, the company formed a cooperation and business alliance with Neusoft Corporation, the largest software developer in China, reaching an agreement on PKI educational support for China.

4.

FUTURE PROSPECTS

So far, we have looked at the business strategies of the major European and North American vendors that are proceeding to divide up world PKI markets, and analysed the most recent circumstances of their progress into Asian PKI markets. As indicated by the statement4 ‘Strategic globalization of a business accelerates the new hegemony of competitiveness through the creation of new operation strategies to respond to global markets, such as international transfers of business know-how, and optimisation of operation resources for the new region,’ it should not be surprising that the PKI vendors from advanced countries

Kagami 03 chap 11

276

2/15/04

2:54 PM

Page 276

Challenging issues

in Europe and North America are aiming at geographic expansion in Asia, including Japan. So, what are the developments we can expect in the future in the PKI market (including Asian markets)? The following is a case study of the likely major trends in the PKI industry. 4.1

Scenario 1: Establishment of the Open Global Standard for PKI

As mentioned in the previous section, since the dawning of the PKI market in 2000, the Asian market has been aggressively pursued by major European and North American vendors. Many of the public sector CAs in Asia rely on solutions provided by these western vendors. In the private sector as well, the main attitude from the Internet society seems to be ‘Introduce a world-class level electronic certificate foundation from a leading PKI vendor, and participate in a global certification network,’ minimizing the power of nations or governments. VeriSign’s VTN and Entrust’s ‘cross-recognition type’ network of CAs and Baltimore’s ‘hierarchical type’ network of CAs are probably designed for this kind of international society. This tendency is especially notable in areas with a strong dependence on trade with Europe and North America, and relatively smaller domestic markets, such as Singapore, Malaysia, Hong Kong, and the Philippines. In such areas, the authentication and certificate technology of the western PKI vendors is actively introduced, with the priority on achieving smooth trade transactions with ‘certification tools’ that conform to western standards. The following are some concrete examples for the above ‘Scenario 1’. Establishment of an Asian PKI Forum PKI vendors are not only competing to establish themselves as a de facto standard. They are also beginning to build cooperative relationships in order to develop interoperability and compatibility of PKI. In the US, with the most developed ecommerce industry, a ‘PKI Forum’ was established in December 1999. Major PKI vendors, including RSA, Entrust, Baltimore, IBM, and Microsoft, as original founders, participate with the intent to expand the number of users through an open strategy of technology standards. Later, VeriSign also joined. In Asia and the Pacific regions too, there has been a spurt of international activity to promote and spread PKI. The Japan Promotional Association for Asia PKI Forum (APKI-J) was set up in Japan in December 2000, mainly by Hitachi, NEC, Fujitsu, Mitsui and Toyota. In June 2001, the first joint meeting was held in Tokyo with attendance from the following eight regions: Australia, China, Hong Kong, Japan, Korea, Malaysia, Singapore and Taiwan. And it was agreed to establish an ‘Asia PKI Forum’ for building interoperable PKI in Asia. In Korea as well, activities have begun, such as the Korea PKI Forum (established March 2001). However, when the Korea PKI Forum sponsored a

Kagami 03 chap 11

2/15/04

2:54 PM

Page 277

PKI solutions for trusted e-commerce

277

conference in Seoul on building a shared Root CA for Asia in April 2001, in addition it laid an emphasis on ‘the threat to the Asian market from multinational PKI vendors based in the US and Canada’.5 International ‘cross-recognition’ of PKI in Asia Some examples of bilateral cross-recognition in Asia have been started already. At the specific industry level, some international schemes for interconnection of PKI have been launched. At the fifth Conference (held in Beijing, August 2001) of the Pan-Asian e-commerce Alliance (PAA), which is a private sector group on trade EDI in Asia, formed by representatives from Hong Kong, Singapore, Taiwan, China and Korea (Japan and Malaysia joined later), it was announced that field testing for PKI interoperability would commence between Asian nations by early 2002. As mentioned above, the trade sector is a very important area for Singapore and Hong Kong, and others, and internationalization is a higher priority than regional individuality in this field. Even to establish an effective trade system, it is urgent to establish a shared certification and authentication infrastructure. It is believed that the motivation for integration of the individual electronic certificate systems is from those fields in which there is a high priority on interaction. For reference, Tradelink of Hong Kong, a leader in PAA, as well as SNS in Singapore, and Trade-Van in Taiwan are all Baltimore users. For the financial sector, the importance of interoperable PKI is high too. Hence the similar industry-based PKI project called ‘Identrus’ is in progress as stated in Section 2. From Asia, participants include Bank of Tokyo–Mitsubishi, Mizuho Corporate Bank, Sumitomo Mitsui Banking Corporation, UFJ Bank, the Korea Exchange Bank, and ChoHung Bank. Identrus could take the initiative to integrate PKI of member banks including their worldwide client networks. Identrus is also a Baltimore user. Reorganization of PKI under one vendor As mentioned above, the global electronic certificate market is divided up by four major PKI vendors. However, this framework is not at all stable. As stated by Alex Van Someren, the CEO of NCipher Corporation Ltd, ‘It is difficult to participate in a PKI market controlled by VeriSign.’ In extreme cases, the market can be looked at in terms of the opposition between VeriSign’s network and the alliance of other PKI vendors. Along with such movements, a simple structure will arise under a strong ‘hierarchy’. 4.2

Scenario 2: Revival of Regionalism in PKI

However, it cannot be denied that there is a risk regarding certification and authentication as mere tools. At the beginning of this chapter, authentication was

Kagami 03 chap 11

278

2/15/04

2:54 PM

Page 278

Challenging issues

defined as ‘verification that the communicating party is actually who it claims to be, as well as verification of registration and the integrity of the transferred content’, but it is not possible to simply determine whether electronic certification can be handled by this kind of telecommunication technology, even in the future. As indicated,6 a CA could expand its business to include services which accumulate information such as financial status, ability to settle accounts, and transaction history of those involved in the transaction. If it is possible to clearly demonstrate appropriate evaluation criteria for the trade partner to the e-commerce user on demand, this should make an enormous contribution to the growth of e-commerce. From this point of view, we should seriously consider the risk of specific technology being controlled by European and North American businesses. There is a deep-seated worry, particularly with regard to US companies, in many Asian countries and it is considered entirely natural for government policy to eliminate monopolies of a business sector controlled by foreign capital. In fact, for the electronic certificate industry in Korea, the solutions from foreign PKI vendors are offered for the certificate authorities under the close scrutiny of the government. As discussed later, there is research underway on Chinese characters (Kanji)-based electronic certificate services for China, Hong Kong, and Taiwan. In Japan as well, large Japanese electronics companies, like NEC, Hitachi, and Fujitsu are setting up a joint venture as the first accredited CA in Japan (JCSI). These examples show it is important to pay careful attention to the trust and credit structure that reflects national cultural backgrounds. Some concrete examples for the above ‘Scenario 2’ are shown below. Regionalism in the US (after the 9-11 terrorist attacks on the security leader) We have no doubt about the position of the US as the leader in the security field including the PKI industry. However, after the terrorist attacks on the US on 11 September 2001, there have been large tremors in previous ideas about security. There is a sudden shift in public opinion in the US toward strengthening the government’s ability to monitor activities and perform criminal investigations beyond the security of businesses and individuals. There has been a reversal on the point of security, from ‘protecting individual privacy (information)’ to ‘ensuring safety at the national level’. The anti-terrorism act, the ‘Uniting and Strengthening America (USA) Act’ enacted on 26 October 2001 was proposed by Republicans John Kyl (senator from Arizona) and Orrin Hatch (senator from Utah). This act legalizes investigative techniques, such as interception of communications on the Internet by the relevant government authorities for the purpose of countering terrorist activity. Furthermore, the Republican senator from New Hampshire, Judd Gregg, stated that there should

Kagami 03 chap 11

2/15/04

2:54 PM

Page 279

PKI solutions for trusted e-commerce

279

be an international system of cooperation between national governments and PKI vendors, and that there should be cooperation to break encryption when necessary. There is a high possibility that terrorists use encryption technology to communicate with each other, and it has been pointed out that ‘online terrorism’ is another danger, following the attacks and the anthrax mailings. In such circumstances, there is probably no choice but to impose some degree of limitation on privacy rights on the Internet. However, measures which rock the foundation of PKI, such as worldwide prohibition of the use of encryption technology and the Private Key registration (deposit) system, are beyond the authority of countries. There is a limit to the public control of Private Keys, and the risk of information control by the government is excessively high. Incidentally, Senator Gregg also insisted that ‘PKI vendors have “influence strong enough to determine the fate of nations”, so they, as corporate citizens, should cooperate in the development of decryption technology by the US government’. The US government is destined to work toward restoring confidence in the security system, even if it means disregarding the protection of personal information. US regionalism may destroy a security network that has been built at the worldwide level. Regionalism in China As mentioned in Section 2 on VeriSign’s business strategy, there is a risk in having all authentication functions in the private sector. It is likely that the Chinese government will be among those with the strongest aversion to this situation. Moreover, for an American company, the Chinese government will certainly want to eliminate such influence as much as possible. The Shanghai Electronic Certificate Authority Centre Co. Ltd (SHECA) mentioned in Section 3 on Hong Kong PKI , is working with Hong Kong on a project to build a Chinese character (Kanji)-based electronic certificate system. There is a strong tendency for China to erect hurdles for foreign PKI vendors based on some sense of sovereignty. In addition, the goal in the future seems to be to build an authentication and certification mechanism that is optimized for the business customs and regulatory system within China. Furthermore, in China, regionalism is definite even at the level of domestic CAs; and the construction of a certificate infrastructure has been initiated at the local city administration level, such as Beijing, Shanghai, Sichuan, and so on, under its strong ‘hierarchy’. The above scenarios offer a completely different perspective. But these are the realities to be confronted in the early 21st century. What is important for PKI will be how the balance of power between government and business enterprises (major vendors) is kept. Especially, how to prevent egoism of leading nations is significant to the development of an open security standard for global networking.

Kagami 03 chap 11

2/15/04

280

5.

2:54 PM

Page 280

Challenging issues

CLOSING REMARKS

For social efficiency and convenience, apparently, we shall see ‘Network globalization’, apart from the conventional hierarchy, in setting up the new information infrastructure. However, it will take an enormous time to establish an infrastructure agreeable to anyone. Particularly, it will be difficult to build up an international infrastructure related to social credibility. So far, we have reviewed PKI as a business tool for authentication in Internet communication, with each vendor struggling to be the global standard, pushing each business model to the market as the best solution. On the way to the establishment of the best infrastructure as a final solution, it will be required to test more business tools through de facto standard competition in the private sectors. For a more sophisticated standard, business enterprises will continue to invent new technologies, such as PKI in combination with biometrics using unique characteristics (fingerprints, iris, and so forth). In this regard, the construction of original business models will be important as the great social contrivance. Whatever it may be, the technical effect will be proven in the future. On the other hand, for the public sector, much wider perspectives are necessary, especially with respect to a trust system, such as PKI, that could be said to be a national issue. The trust framework of each society will be determined in accordance with the status of each society respectively. It is not a simple matter of business tools, but of complicated political issues. Before selecting the trust system of society, we have to determine the national policy. Should the government have a decentralized or centralized structure? Should the society accept the international standard or not? If the trust system is set before the social consensus is reached on these points, such a society shall be ‘distrustful’.

NOTES 1. US Patent No. 4,405,829: Generally, called ‘RSA Algorithm Patent’ on Cryptographic Communications System and Method. This patent is used for creating a set of Public Keys and Private Keys which are important factors for encryption under PKI. Effective from September 1983 with a duration of 17 years. 2. Quick ratio: A (financial) stringent test that indicates if a firm has enough short-term assets to cover its current liabilities, also called ‘acid test ratio’. Calculated at (Cash + Marketable securities + Accounts receivable + Other short-term investments, without Inventory) ÷ Current liabilities × 100%. From the viewpoint of business administration theory, a company with a ratio more than 100 per cent can pay its current liabilities immediately, thus should be regarded as favourable. 3. Asian PKI market share: Comment by Patrick Hugh Holahan, Executive Vice President Marketing, Baltimore Technologies plc in PKI seminar on 7 July 2000 in London, England. 4. Advantage of strategic globalization: Keegan, W.J. (1999), Global Marketing Management, Prentice-Hall, Englewood Cliffs, NJ.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 281

PKI solutions for trusted e-commerce

281

5. US and Canadian initiatives in Asian markets: Conference presentation by Korea PKI Forum on 17 April 2001 in Seoul, Korea. 6. Business expansion of certificate authority: Miyawaki, K. and Kato, K. (2001), Electronic Certification Will Change Japan, Japan Productivity Centre for Socio-Economic Development, June.

REFERENCES Adams, Carlisle and Steve Lloyd (2001), Understanding Public-Key Infrastructure: Concepts, Standards, and Development Considerations, Macmillan Technical Publishing, Basingstoke. Discussion Paper (2000), ‘Current issues on e-Commerce Legal Framework’, Jurist, No. 1183, Yuhikaku Publishing Co. Ltd, 15 August. Davies, Ross and Toshiyuki Yahagi (2001), Retail Investment in Asia Pacific: Local Responses and Public Policy Issues, Oxford Institute of Retail Management, Templeton College, University of Oxford, January. Miyagawa, Shoko and Juichiro Yamazaki (1998), ‘“Trust” and “Reputation” in the Internet Society’, Hitotsubashi University Business Review, Vol. 46, No. 2, Toyo Keizai, Inc., November. UNIZON Corp. (2001), Framework of the Network Security – Chart Guidance, D. Art Corp., April.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 282

14. The interconnection and pricing of the Internet Takanori Ida and Masashi Ueda 1.

INTRODUCTION

One of the current problems which economics is facing is how to analyse the rapidly developing Internet. In the 1990s, the Internet became a driving force for the growth of the world economy, and the theory of ‘the New Economy’ was quite popular. However, as the new century began, the growth of IT industries slowed down, and an end to the IT bubble was whispered. At any rate, it is true that the spread of the Internet and the development of IT industries have significantly changed the structure of the world economy; for example, the 1997 US Department of Commerce report on ‘The Emerging Digital Economy’ discussed various aspects of the digital economy such as ecommerce and, more generally, the introduction of computers and related technology in the workplace. The purpose of this chapter is to consider to what extent we can analyse this rapidly evolving digital economy from the viewpoints of economics. Needless to say, it is beyond the ability of one person to analyse the Internet as a whole, but a number of studies have been conducted to this end over the last decade. Accordingly, this chapter will comprehensively survey studies made so far and, on the basis of these foundations, try to provide a model framework to analyse the industrial structure of the Internet. This chapter roughly consists of two parts. The first part goes from Section 2 to Section 4, where conventional studies are summarized, while the second part contains Section 5 and Section 6, which will correspond to the original contribution of this chapter. Section 2 will discuss what the Internet is and explain the history, the difference from telecommunications, and the convergence of services. Section 3 will investigate the pricing of the Internet and make its problems clear, whereas Section 4 will examine the industrial structure of the Internet, especially focusing on the vertical features and interconnection. Section 5 will theoretically analyse the topology of the Internet with game theory, where a components model will be used to integrate the one-way model 282

Kagami 03 chap 11

2/15/04

2:54 PM

Page 283

The interconnection and pricing of the Intenet

283

and the two-way model. Section 6 develops a model analysis to compare price cap regulation and unbundling. Section 7 will give further discussion.

2.

WHAT IS THE INTERNET?

2.1

The History of the Internet

The Internet is the worldwide networking of networks connected via a common communications protocol, namely TCP/IP, which provides a common language for interoperation between computers that use a variety of local protocols.1 The essence of the Internet can be summarized in the following two points: first, the Internet is a set of protocols to allow access by various host computers in a fully distributed way; second, the Internet is the software and system agreements that allow disparate hardware and software to talk across disparate networks.2 As is well known, the origin of the Internet dates back to the period of the ‘Cold War’. In the late 1960s, the Advanced Research Projects Administration (ARPA), a division of the US Defense Department, developed the ARPAnet to link universities and high-tech defence contractors for military reasons. TCP/IP was then adopted to provide a standard protocol for ARPAnet communications. In the mid-1980s, the National Science Foundation (NSF) established the NSFNET to connect its supercomputers and to provide other general services. The NSFNET also adopted TCP/IP and provided a high speed backbone for the emerging Internet.3 Let us here turn to the development of the Internet in Japan. The JUNET (Japan UNIX/University Network), a joint research project of three universities (University of Tokyo, University of Keio, and Tokyo Institute of Technology) was established in 1984 and connected to the American CSNET (Computer Science Network) in 1986. The Project of WIDE (Widely Integrated Distributed Environment) started in 1988 so that academic networks and commercial networks could interconnect. At that time, the network connecting three universities and ten private companies began operating. This is now called the start of the Internet in Japan. There are two commonly held misconceptions about the Internet even today. The first is that the development of the Internet was historically paid for by the American government. In fact, the National Science Foundation paid for less than 10 per cent of the Internet until 1994, and Internet infrastructure and software development were mainly paid for by private companies and organizations. The second is that the Internet is free. It is true that the marginal cost for some Internet traffic may be zero due to statistical sharing, but social costs, such as congestion costs, are quite significant.4

Kagami 03 chap 11

2/15/04

2:54 PM

Page 284

284

Challenging issues

2.2

The Difference between Telecommunications and the Internet

It will be useful, at this moment, to explain the difference between telecommunications and the Internet. The critical characteristics of the architecture of telephone networks can be summarized as follows: first, the control is highly centralized and integrated; second, the hierarchical interconnection is based on a central office switch to conserve bandwidth at the trunk side; third, the basic transport is the twisted-pair copper wire to the end-user while optical fibre is used in the loop and to large scale users; fourth, the interface is the telephone handset and recently has expanded to include the use of a data set.5 How is the Internet different from the telephone networks explained above? In fact, most backbone and regional network traffic moves over leased phone lines; therefore the Internet may be the same as telecommunications as a basic technology. However, there is a fundamental distinction between how the lines are used by the Internet and the telephone. The Internet is connectionless packetswitched whereas the telephone is circuit-switched.6 Telephone networks use circuit-switching where an end-to-end circuit must be set up before the call begins. A fixed share of network resources is reserved for the call, and no other call can use those resources until the original connection is closed. One advantage of circuit-switching is to assure QoS (quality of services) such as guaranteed maximum delay. On the other hand, the Internet uses packet-switching where the data stream from a computer is broken up into packets of about 200 bytes. One advantage of packet-switching is to permit a statistical sharing on the communication lines on a first come, first served basis. If the network becomes overloaded, however, packets are delayed or dropped.7 We will discuss this problem of the current Internet in the next section. 2.3

The Convergence of Services via the Internet

The development of the Internet has provoked the collision and convergence of services previously regarded as separate ones (computers and data communications, content and conduit, equipment and services), and this development of innovative services unavoidably requires new business models, new network architectures, and new policy frameworks.8 Especially, Internet telephony will be the most prominent application at the nexus of the convergence of the traditional public switched telephone network (PSTN) and the emerging global information infrastructure based on the Internet protocol (IP).9 It should be noted here that there are various meanings of Internet telephony. In a narrow sense, it represents just the first-generation systems for making telephone calls over the Internet (called Internet protocol telephony), or at best the second-generation systems such as packet telephony, voice over

Kagami 03 chap 11

2/15/04

2:54 PM

Page 285

The interconnection and pricing of the Intenet

285

IP (VoIP), voice over the network (VON), or voice conferring. However, the latest concept should be defined from a wider viewpoint. Internet telephony is currently defined as the transformation of the Internet from an application on the public telephone network to a general communications infrastructure capable of supporting telephone services as well as other multimedia applications. It is important that all of the multimedia applications can be supported over the IPderived protocols running on diverse physical media (such as CATV, telephone, wireless, and others).10 The latest Internet telephony applications can be classified into three types. The first is the type that uses the Internet only to provide POTS (plain old telephone services) between existing telephone equipment. An example is the bypassing of international call or local loop services. The second is the type that interoperates between the existing telephone and Internet networks and provides communications between users with either computers or existing telephone sets as end nodes. Examples are computer telephone and voice access to the Web. The third is the type that uses Internet-attached computers to provide some form of human communications across the packet-switched Internet. Here examples are more general moving-picture services such as teleconferencing and teleworking.11 As a matter of fact, we will have to satisfy the following five conditions for Internet telephony to be put to practical use.12 These are: (1) the guarantee of QoS; (2) an appropriate pricing scheme for enhanced services; (3) the improvement of services; (4) the realization of always-on connectivity, and (5) the arrival of the ubiquitous society. Nonetheless, it will take a while to satisfy all of these conditions. Therefore, we will investigate the current problems of Internet pricing in the next section.

3.

PRICING THE INTERNET AND ITS PROBLEMS

3.1

Pricing of the Internet

There are in general three main elements of network costs: (1) the cost of connecting to the network; (2) the cost of providing additional network capacity; and (3) the social cost of congestion.13 Therefore, when we think of the pricing scheme of the Internet, we have to cover the three elements of costs mentioned above. At present, there are three representative schemes of Internet pricing.14 The first is flat-rate pricing, in which Internet users pay a fee to connect but are not billed for each bit sent. For example, users may pay for a Tier 1 link regardless of how many bits they send or receive. The second is usage-sensitive pricing, in which users pay a portion of their Internet bill for a connection and a possibly varying portion for each bit they send or receive. The marginal

Kagami 03 chap 11

286

2/15/04

2:54 PM

Page 286

Challenging issues

monetary cost of sending or receiving another bit is non-zero during certain time periods such as peak hours. The third is transaction-based pricing which is different from usage-sensitive pricing in that prices are determined not by the number of bits but by the characteristics of the transaction. Flat-rate pricing for access might be the simplest and most effective solution to increasingly complicated interconnection.15 In fact, it can be said from a historical point of view that flat-rate pricing has worked successfully in the infant-age of the Internet. We can list some advantages of flat-rate pricing16: (1) predictable fees reduce risk for users and to some extent for providers as well; (2) flat fees encourage usage, which increases customer satisfaction and does not hurt the provider (if marginal costs are zero); and (3) flat fees avoid the (possibly considerable) administrative costs of tracking, allocating, and billing for usage. 3.2

Problems of Internet Pricing

The conspicuous characteristics of the information society, such as competitive environment, complex mixture of services, order-of-magnitude range of carriage speeds, and the blurring of administrative boundaries, seem to provoke ‘the dilemma of Internet pricing’, that is how to price distributed computers’ interconnection.17 Since interconnecting distributed computers is all the Internet really does, the dilemma is deeply rooted in the essence of the Internet. One of the problems that Internet technology faces is how to guarantee QoS for various demands of users. Simple flat-rate pricing cannot finely differentiate QoS depending on the various demands. In general, QoS can be classified into four categories:18 (1) there is no lower bound but an upper bound; (2) there is no bound above or below but congestion on peak periods; (3) it remains near the promised average at all times; and (4) there is a full reservation of resources, as well as a constant and guaranteed rate. It has recently become desirable to establish explicit mechanisms to allow users to specify different needs, with the presumption that they should be differentially priced.19 One example of such mechanisms is multi-part tariff, which comes from the basic idea that the charge to users should be based on some observable characteristics of users. The characteristics to be considered are, for example, as follows.20 (1) Access: whether the user is in fact connected to the system. (2) Capacity: the maximum rate at which a user can move information through the system regardless of whether the user actually uses the capacity. (3) Usage: a charge for the actual quantity of information sent through the system. (4) Priority: a charge for displacing other users in the event of congestion. There is no doubt that the most serious problem of Internet pricing is how to control congestion. This is because a user’s incremental packet imposes costs on other users in the form of delay or even dropped packets when the network

Kagami 03 chap 11

2/15/04

2:54 PM

Page 287

The interconnection and pricing of the Intenet

287

is at over capacity. To internalize the social cost of congestion, a real time auction mechanism called ‘smart market’ has been proposed by economists.21 Packets should be prioritized based on the value that a user puts on getting the packet through quickly. To do this, the user assigns to packets a bid measuring willingness to pay for immediate servicing. On congested routes, packets are prioritized based on bids. If all congestion revenues are reinvested in the expansion of capacity, then the capacity will be expanded to the socially desirable point where its marginal value is equal to its marginal cost. Furthermore, other economists suggest that a settlement policy, like in telecommunications, is necessary for the Internet. One may argue that traffic flows are so sufficiently symmetric that a settlements policy is not required. In fact, the current Internet has not adopted any settlements on the basis of billand-keep. However, resource usage may not always be symmetric, and therefore the opportunities to free ride on capacity investments by other network providers are increasing.22

4.

INDUSTRIAL STRUCTURE OF THE INTERNET AND ITS INTERCONNECTION PROBLEM

4.1

The Industrial Structure of the Internet

The Internet is a network of networks and the global aggregate of various, large or small, local networks. Internet structure can be summarized simply as three levels:23 the bottom level local access, the regional or mid-level conveyance, and the backbone level interconnection. There are two principal types of Internet providers. One is ISP (Internet service provider), which provides access services for end-users at a retail level, while the other is IBP (Internet backbone provider), which provides transmit services for ISP at a wholesale level. Therefore, an ISP has to connect to an IBP so that it connects to another ISP. The IBP interconnects other IBPs at network access points (NAPs).24 There are a wide array of ISPs ranging from a mom-and-pop operation offering services to a small number of subscribers in a specific location to a national operator that offers access across the whole country.25 Since the cost of becoming a local ISP is very low, competition among ISPs is extremely intense, and there exist a significant number of entries and exits of ISPs. Besides, the IBP that provides backbone services at the wholesale level can be considered as an upstream supplier from the viewpoint of an ISP. Most ISPs have been integrated forward into an IBP.26 Figure 14.1 summarizes the discussion above. The industrial structure of the Internet can be summarized as follows when we focus on the vertical perspectives.27 (1) Incentives to vertically integrate on

Kagami 03 chap 11

2/15/04

2:54 PM

Page 288

288

Challenging issues

The X IBP Network

The X ISP Network

The Y IBP Network

The Y ISP Network

The X Network Users

The Y Network Users

The X Area

The Y Area

Figure 14.1

The Internet topology

the Internet are still strong. (2) The viability of the wholesale market is enhanced. (3) Downstream integration is more likely than upstream. (4) Facilities-based providers will integrate into Internet services. To understand these features of the industrial structure of the Internet, it seems helpful to use the game-theoretical approach of industrial organization theory. 4.2

Japanese Internet Providers

Here we will briefly comment on the current situation of Japanese Internet providers.28 First, let us begin with types of Internet providers. Most Internet providers used to be personal computer communication companies that did not have their own telecommunications facilities. However, telecommunications companies that have their own telecommunications facilities have recently increased their market share. Figure 14.2 illustrates the market share of Japanese Internet subscribers according to the type of provider each one is from 1998 to 2000.29 We can understand that telecommunications-type providers have captured a significant proportion of the customers of personal computer communication companies. This is because price competition among the Internet providers has become intense and conventional personal computer communications companies have lost price competitiveness against the large scale and vertically integrated telecommunications companies.30 In addition, the breakdown of the subscribers’ share of telecommunicationstype providers is shown in Figure 14.3. Here we can see that the market share of the NTT group is high enough to be almost equal to that of all other common carriers. This is because the brand image of the biggest telephone company in the world has a positive effect even as an Internet provider. Furthermore, the

Kagami 03 chap 11

2/15/04

2:54 PM

Page 289

The interconnection and pricing of the Intenet

289

80 70 60 50 % 40 30 20 10 0 1998

1999 2000 Year The total of telecommunications corporations The total of personal computer communications

Source: The White Paper of the Internet (2001) Impress Co., Japan.

Figure 14.2 The number of Japanese Internet subscribers according to the types of providers 40 35 30 25 % 20 15 10 5 0

1998

1999 2000 Year The NTT group Other telecommunications Foreign capital telecommunications

Source: The White Paper of the Internet (2001) Impress Co., Japan.

Figure 14.3 The number of Japanese Internet subscribers to telecommunications corporations

Kagami 03 chap 11

2/15/04

2:54 PM

290

Page 290

Challenging issues

foreign capital providers have been gradually increasing the number of subscribers. This tendency will continue in the future because Japan Telecom merged with a big British cellular phone company, Vodaphone. Furthermore, broadband services such as ADSL and FTTH have been operating since 2000, and a competitive-power gap between a company with telecommunications facilities and one without continues to widen. The number of subscribers to ADSL services in September 2001 is shown in Figure 14.4. The NTT group has about 50–60 per cent of the market share of ADSL services and has a dominant position in this area. It follows from the above evidence that the Japanese Internet industry is still highly oligopolistic and vertical integrators have the competitive edge. Others 8% Japan Telecom 5%

NTT-East 34%

E-Access 11%

Yahoo!BB 16% NTT-West 26% Source:

Multimedia Research Institute (2001), ‘M&D Report 2001’.

Figure 14.4

The market share of ADSL providers, September 2001

5.

THE MODEL ANALYSIS OF THE INTERNET

5.1

The One-Way Model and the Two-Way Model

It will be helpful to explain the key elements, such as the one-way model of connection, the two-way model of connection, and the components model, on which this analysis depends. The one-way model of connection is the network structure where one company needs access to another but the reverse does not

Kagami 03 chap 11

2/15/04

2:54 PM

Page 291

The interconnection and pricing of the Intenet

291

hold.31 One example is: some ISPs provide access services to end-users in a retail market and at the same time serve as IBPs that provide transit services to ISPs in a wholesale market, while other ISPs that do not serve as IBPs have to buy transit services from IBPs to provide final services to customers. The two-way model of connection demonstrates that customers calling each other belong to two different local networks and each carrier must buy termination access from the other network.32 One example is: IBPs have to interconnect at a point of interconnection so that end-users on each network can exchange information.33 It will be useful here to adopt a ‘components model’ in order to consider the one-way model and the two-way model systems respectively. It was Matutes and Regibeau (1988) and Economides (1989) who introduced the components model. The components model analyses how complementary components are combined to produce a valuable system.34 For example, a computer itself will not be of any use to a consumer without a complementary monitor. That is to say, the computer system is composed of a computer and monitor. 5.2

The Basic Setup of the Model

This subsection will explain the basic setup of the model. We assume at this moment that there are two levels of ‘components’, A and B. For example, component A means an access service from ISP to end-users while component B means a transit service between IBP and ISP. It is also assumed that each component has two types, namely A1/A2 and B1/B2. We call the combination of components ‘system’. Systems are A1B1, A1B2, A2B1, and A2B2. Defining the price of components Ai and Bj are Pi and Qj respectively, the price of system AiBj, Sij, can be represented as the sum of the prices of components. Sij = Pi + Qj; i, j = 1, 2

(1)

Let us assume linear demand functions of systems AiBj, Dij, for the sake of simplicity, following previous research into components model. The own effect of price to demand is represented by coefficient b, the cross-effects are represented by coefficients c, d, and e. Accordingly the demand functions of systems are given as follows. D11 = a – bS11 + cS12 + dS21 + eS22 D12 = a – bS12 + cS11 + dS22 + eS21 D21 = a – bS21 + cS22 + dS11 + eS12 D22 = a – bS22 + cS21 + dS12 + eS11

(2)

Kagami 03 chap 11

2/15/04

2:54 PM

Page 292

292

Challenging issues

The demand functions of components are given as follows. For example, the demand of component A1, DA1, is represented as the sum of the demands of systems A1B1 and A1B2, D11 and D12. DA1 = D11 + D12 DA2 = D21 + D22 DB1 = D11 + D21 DB2 = D12 + D22

(3)

At this moment, we need further assumptions to make the analysis simple without affecting the overall analysis. • • • • 5.3

The constant cross-effects of prices: c = d = e. The non-negative equilibrium prices and quantities: b > 3c > 0. The zero marginal costs of production. Furthermore, all parameters are normalized by b: b = 1. The One-Way Model

The one-way model (abbreviated as 1w) demonstrates the case where two types of providers exist: one is a dominant company that monopolizes one level of components and the other is a partial entrant. The dominant company provides A1/B1/B2 whereas the partial entrant does only A2. The possible interpretation is that the dominant company is an IBP that integrates an ISP and the partial entrant is a disintegrated ISP. Customers can choose component A from two providers but component B from only the dominant provider. Therefore, there are four kinds of systems as final products, A1B1, A1B2, A2B1, and A2B2; firm The one-way

The two-way

B1, B2

B1, B2

A1

Figure 14.5

A1

A1

The components model of interconnection

A1

Kagami 03 chap 11

2/15/04

2:54 PM

Page 293

The interconnection and pricing of the Intenet

293

1 sets the price of components A1/B1/B2, and firm 2 sets the price of component A2. The left of Figure 14.5 illustrates the one-way structure as explained above. The companies’ profit functions Π1 and Π2 are given as follows. Π1 = P1 DA1 + Q1 DB1 + Q2 DB2 Π2 = P2DA2

(4)

The first-order conditions of profit maximization can be obtained by differentiating profit functions Π1 and Π2 with respect to prices P1/Q1/Q2 and P2. The equilibrium prices and quantities are shown in Table 14.1. 5.4

The Two-Way Model

Next, we move on to explain the two-way model (abbreviated as 2w). The twoway model means that there is no dominant carrier and that firm 1 provides A1/B1 while firm 2 provides A2/B2. For example, it can be thought that component A is a transit service between IBP and ISP and component B is the interconnection service between IBPs. Each component has two types, A1/A2 and B1/B2. Component B is compatible with two providers, and B1 can be combined by A2 as well as A1 to produce the systems A1B1 and A2B1 respectively. The same thing can be said of B2. This means two-way connection in this context. We have here four kinds of systems as final products, A1B1, A1B2, A2B1, and A2B2. Firm 1 sets the price of A1/B1 and firm 2 sets the price of A2/B2. The right of Figure 14.5 illustrates the two-way structure as explained above. The companies’ profit functions Π1 and Π2 are given as follows. Π1 = P1 DA1 + Q1 DB1 Π2 = P2 DA2 + Q2 DB2

(5)

The first-order conditions of profit maximization can be obtained by differentiating profit functions Π1 and Π2 with respect to prices P1/Q1 and Q2/P2. The equilibrium prices and quantities are shown in Table 14.2. 5.5

The Results of the Analysis of the One-Way and Two-Way Model

Now let us investigate the results of the one-way and two-way model. The main conclusions obtained from Tables 14.1 and 14.2 can be summarized as the following lemmas and proposition.35 Lemma 1.1 (The components prices of 1w): P11w < P21w, Q11w = Q21w Lemma 1.2 (The components prices of 2w): P12w = P22w = Q12w = Q22w

ΣΣD1w =

Total demand :

a( −5 + 3c) 3( −1 + c)

294

P12 w =

2w D11 =

ΣΣD2 w =

Component prices

System demands

Total demand :

2w D12 =

a( −3 + 5c) −7 + 17c 4 a( −3 + 5c) −7 + 17c

P22 w =

2a 7 − 17c

a( −3 + c) 6( −1 + c)

a( −3 + 5c) −7 + 17c

2a 7 − 17c

1w D12 =

a( −3 + c) 6( −1 + c)

a 3 − 3c

2w D21 =

Q12 w =

1w D21 =

Q11w =

a( −3 + 5c) −7 + 17c

2a 7 − 17c

a 3

a 3 − 12c + 9c 2

2w D22 =

Q22 w =

1w Q22 =

Q21w =

a( −3 + 5c) −7 + 17c

2a 7 − 17c

a 3

a 3 − 12c + 9c 2

2:54 PM

The result of the two-way model (2w)

1w D11 =

System demands

P21w =

a 6 − 6c

2/15/04

Table 14.2

P11w =

The result of the one-way model (1w)

Component prices

Table 14.1

Kagami 03 chap 11 Page 294

Kagami 03 chap 11

2/15/04

2:54 PM

Page 295

The interconnection and pricing of the Intenet

295

Lemma 1.3 (The comparison of components prices between 1w and 2w): P11w < P12w, P21w (<) P22w if c (>)0.091, Q11w = Q21w > Q12w = Q22w Lemma 1.1 examines the component prices of the one-way model. Since the price of A1 that a dominant provider sets is lower than that of A2 that an entrant sets, the system demand of the dominant provider is larger than that of the entrant. This result demonstrates the advantage of the dominant provider because it can set the price of vertically integrated components taking system demands into consideration. Lemma 1.2 examines the component prices of the two-way model. Since two symmetric vertical integrators compete, the prices of all components that they set are identical. Therefore, all prices of systems are also identical, and system demands are symmetric. This is because neither has an advantage in competition. Lemma 1.3 compares the component prices of the one-way and the two-way model. In the one-way model, the price of component B is relatively higher because competition does not work at this level. By contrast, in the two-way model, the price of component B is relatively lower because competition works at this level conversely. These effects of component B exceed those of another component A. Proposition 1 (The comparison of total demands between 1w and 2w): ΣΣD1w < ΣΣD2w Proposition 1 compares the social welfare between the one-way and twoway model.36 Since the total demand of the two-way model is higher than that of the one-way model, the social welfare of the two-way model is higher than that of the one-way model. Why is the two-way model more socially desirable than the one-way model? There are two reasons for this. One is the internalization of the vertical externalities. Vertical integrators can set prices considering the demands of systems instead of components. In other words, the two-way model can avoid the problem of double margins that increase the prices of systems and decrease system demands and the social welfare. The other is the action of the horizontal substitution effect. That is to say, since there is no company that monopolizes components, the price competition in terms of systems works effectively and system demands are increased.

6.

DEVELOPMENT OF THE ONE-WAY AND TWO-WAY MODEL

6.1

The Regulatory Model of the One-Way Model

In this section we will compare price cap regulation and unbundling to regulate the one-way model that we know to be socially inferior to the two-way model.

Kagami 03 chap 11

296

2/15/04

2:54 PM

Page 296

Challenging issues

Proposition 1 stated that the one-way model is socially inferior to the twoway model. Thus, we will examine how to regulate the one-way model and improve its social welfare. There are two possible methods of regulation: price cap regulation and unbundling. Price cap regulation: The government regulates the upper limit of the price of component B that a dominant company monopolizes. Unbundling: The government disintegrates component B from the dominant company and determines its price. 6.2

The Price-Cap Model of the One-Way Connection

The price-cap model of the one-way connection (abbreviated as 1w-pc) demonstrates what happens when the government regulates the upper limits of prices of components B1, B2 up to f, whereas a dominant company provides A1/B1/B2 and an entrant provides A2. Systems as final products are A1B1, A1B2, A2B1, and A2B2. Firm 1 sets the price of component A1 and firm 2 sets the price of component A2. We follow the previous section with regard to the basic setup of the model. Profit functions, Π1 and Π2, are given as follows. Π1 = P1DA1 + fDB1 + fDB2 Π2 = P2DA2

(6)

The first-order conditions of profit maximization can be obtained by differentiating profit functions Π1 and Π2 with respect to prices P1 and P2. The equilibrium prices and quantities are shown in Table 14.3. 6.3

The Unbundling Model of the One-Way Connection

The unbundling model of the one-way connection (abbreviated as 1w–ub) demonstrates what occurs when the government disintegrates the component B1 and B2 from a dominant company and determines the prices while the dominant company provides A1 and an entrant provides A2. Systems as a final product are A1B1, A1B2, A2B1, and A2B2. Firm 1 sets the price of component A1 and firm 2 sets the price of component A2. We employ the same type of model as outlined in the previous section. Profit functions Π1 and Π2 are given as follows. Π1 = P1DA1 Π2 = P2DA2

(7)

Kagami 03 chap 11

2/15/04

2:54 PM

Page 297

The interconnection and pricing of the Intenet

297

The first-order conditions of profit maximization can be obtained by differentiating profit functions Π1 and Π2 with respect to prices P1 and P2. The equilibrium prices and quantities are set out in Table 14.4. 6.4

The Result of the Analysis of the Regulation Model of the One-Way Connection

Now let us investigate the result of the regulation of the one-way model. The main conclusions obtained from Tables 14.3 and 14.4 can be summarized as the following lemmas and proposition. Lemma 2.1 (The component prices of 1w-pc): P11w–pc < P21w–pc Lemma 2.2 (The component prices of 1w-ub): P11w–b = P21w–ub Lemma 2.3 (The comparison of the component prices between 1w–pc and 1w–ub): P11w–pc P11w–ub, P21w–pc P21w–ub Lemma 2.1 examines the component prices in the case of price-cap regulation. Under the price-cap regulation, since the price of A1 that a dominant company sets is lower than that of A2 that an entrant company sets, the component demand of the dominant provider is larger than that of the entrant. Lemma 2.2 examines the component prices in the case of unbundling. Since two symmetric vertical disintegrators compete, the prices of all components that they set are identical. Therefore, all prices of systems are also identical, and system demands are symmetric. Lemma 2.3 compares the component prices in the case of price cap regulation and unbundling. In both cases, the prices of B1 and B2 are regulated up to f. The prices of A1 and A2 are lower in the pricecap regulation than those in the unbundling. This conclusion may be counter-intuitive because unbundling is a stricter regulation for the dominant company than the price cap regulation. Proposition 2 (The comparison of total demands between 1w–pc and 1w–ub): ΣΣD1w–pcΣΣD1w–ub Proposition 2 compares the social welfare between the price-cap regulation and the unbundling. Since the total demand in the case of price-cap regulation is larger than that in the case of unbundling, the social welfare in the case of price-cap regulation is higher than that of unbundling. Why is the price-cap regulation socially superior to unbundling? This is because the price-cap regulation can avoid the problem of double margins as Proposition 1 has stated. Under the price-cap regulation, the dominant company can determine the price

298

Total demand :

System demands

Component prices

Table 14.4

2 a( −1 + c) + (1 − 2c − 3c 2 ) f −1 + 2c

ΣΣD1w − ub =

2( −1 + c)( a + ( −1 + 3c) f ) −1 + 2c

( −1 + c)( a + ( −1 + 3c) f ) −2 + 4c

a − f + 3cf 2 − 4c

( −1 + c)( a + ( −1 + 3c) f ) −2 + 4c

P21w − ub = 1w − ub D12 =

a − f + 3cf 2 − 4c

1w − ub D11 =

P11w − ub =

The result of the one-way model with unbundling (1w–ub)

ΣΣD1w − pc =

1w − pc D12 =

a + ( −1 + 2c + 3c 2 ) f 2 − 4c 1w − pc D21 =

Q11w − pc = f

( −1 + c)( a + ( −1 + 3c) f ) −2 + 4c

1w − ub D21 =

Q11w − ub = f

a( −1 + c) + c(1 − 2c − 3c 2 ) f ( −1 + c)( a + ( −1 + 2c + 3c 2 ) f ) −2 + 4c −2 + 4c

P21w − pc =

( −1 + c)( a + ( −1 + 3c) f ) −2 + 4c

1w − ub D22 =

Q21w − ub = f

( −1 + c)( a + ( −1 + 2c + 3c 2 ) f ) −2 + 4c

1w − pc D22 =

Q21w − pc = f

2:54 PM

Total demand :

1w − pc D11 =

a + ( −2 + 7c − 3c 2 ) f 2 − 4c

a( −1 + c) + c(1 − 2c − 3c 2 ) f −2 + 4c

P11w − pc =

The result of the one-way model with price cap regulation (1w–pc)

2/15/04

System demands

Component prices

Table 14.3

Kagami 03 chap 11 Page 298

Kagami 03 chap 11

2/15/04

2:54 PM

Page 299

The interconnection and pricing of the Intenet

299

of component A1 to maximize the total profit of the systems even though it is not allowed to determine the prices of components B1 and B2. It is interesting to note that the price-cap regulation of bottleneck or essential facilities is more socially desirable than unbundling, so that the divesture of a dominant company is not always socially preferable. Corollary of Proposition 2 (The equivalence of 1w–pc and 1w–ub under the marginal-cost regulation): P11w–pc = P11w–ub, P21w–pc = P21w–ub, ΣΣD1w–pc = ΣΣD1w–ub if f = 0 The corollary compares the price-cap regulation and unbundling when the government lets the prices of components B1 and B2 be marginal costs, namely f = 0. In this case, the equilibrium prices and equilibrium quantities in the case of price-cap regulation are exactly equivalent to those in the case of unbundling. However, since the company needs to recover the fixed cost and to secure the appropriate profit, the marginal-cost regulation is in fact difficult to implement, therefore the price-cap regulation and unbundling will be different from each other when the prices are higher than the marginal cost.

7.

FURTHER DISCUSSION

This chapter has discussed to what extent economics could investigate the evolving digital economy, mainly the Internet. First, we have comprehensively surveyed previous research and summarized the economic problems of the Internet: how the Internet is different from telecommunications; what kind of problems the Internet is facing; and what the industrial structure of the Internet is like. After discussing these problems in detail, we have moved on to model analysis from the viewpoint of game-theoretical industrial organization: the analysis of the one-way model made clear that the two-way model is socially preferable to the one-way connection: furthermore, the analysis of the one-way model was developed so that price-cap regulation is socially preferable to unbundling. In conclusion, we think that it is sufficiently possible to economically analyse the Internet and obtain some deep insights. However, the problem of how we should utilize these academic insights in actual economic policy-making will remain. One economist has pointed out that regulation should be more akin to safety and consumer protection than pricing and cost recovery.37 Another has stated that the distinction between regulated ‘telecommunications’ and unregulated ‘information services’ is at the centre of the US’s 1996 Telecommunications Act, but this distinction was rooted in the conventional telephone network; it does not work in the IP world, therefore we needed to develop new ways of reconciling old telephone regu-

Kagami 03 chap 11

2/15/04

300

2:54 PM

Page 300

Challenging issues

lations with new IP networks.38 Seemingly, the 1996 Telecommunications Act had the unintended and unfortunate consequence of increasing the division between the telecommunications ‘haves and have-nots’.39 All of this seems really right. It will be more difficult to answer the political questions proposed above and design an institutional blueprint in the future than to summarize past events and to theoretically explain the current situation. This will be the common task imposed on all economists.

NOTES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

See MacKie-Mason and Varian (1997), p. 28. See McGarty and McKnight (2001), p. 47. See MacKie-Mason and Varian, op. cit., p. 28. See McKnight and Bailey, op. cit., p. 4. See McGarty and McKnight, op. cit., p. 51. See MacKie-Mason and Varian, op. cit., p. 33. See MacKie-Mason and Varian op. cit., p. 33. See McKnight et al. (2001b), p. 4. See Lehr (2001), p. 93. See McKnight et al. (2001b), p. 3. See Clark (2001), p. 18. See Clark, op. cit., p. 23. See MacKie-Mason and Varian, op. cit., p. 39. See McKnight and Bailey (1997b), p. 13. See Anania and Solomon (1997), p. 93. See Clark (1997), p. 230. See Anania and Solomon, op. cit., p. 93. See Sharifi (2001), p. 309. See Clark (1997), p. 248. See Crawford (1997), p. 387. See MacKie-Mason and Varian, op. cit., p. 46. See MacKie-Mason and Varian, op. cit., p. 58. See MacKie-Mason and Varian, op. cit., p. 30. See McKnight and Leida (2001), p. 195. See Lehr, op. cit., p. 103. See Lehr, op. cit., p. 106. See Lehr, op. cit., p. 109. Furthermore, it will be useful to note two current pieces of research of the industrial structure in broadband services: first, Faulhaber and Hogendorn (2000) analysed the future industrial structure of broadband infrastructure and found that oligopolistic competition is likely to emerge for demand levels, thus approaching that of the CATV market; second, Rubinfeld and Singer (2001) analysed the economic incentives of vertical merger of the Internet content provider and the cable system operator to engage in two distinct vertical foreclosure strategies, namely conduit discrimination and content discrimination. 28. The figures used here are derived from ‘The White Paper of the Internet’ by the Impress Co. However, we have to be careful about the result because the figures are largely based on the method by which the data are extracted. 29. It should be noted that since the investigation admits multiple answers the sum of market share exceeds 100 per cent. 30. One reason why the market share of telephone companies has increased rapidly in recent years has been an indirect or direct effect of the sudden spread of Internet services through cellular phones, and in particular the success of NTT DoCoMo’s ‘i-mode’.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 301

The interconnection and pricing of the Intenet

301

31. See Laffont and Tirole (2000), p. 6. Also see the following for details of the one-way model: Armstrong et al. (1996), Baumol and Sidak (1994), Laffont and Tirole (1994). 32. See Laffont and Tirole (2000), p. 8. Also see the following for details of the two-way model: Armstrong (1998), Giovannetti (2001), Laffont et al. (1998). 33. In the model analysis below, we will extend the definition of the two-way interconnection such that one customer can choose a component of services from multiple companies. 34. See Shy (2001), p. 36. Also see the following for details of the component model: Economides and Salop (1992), Economides (1996), Matutes and Regibeau (1992). 35. We can draw all lemmas and propositions by direct calculation. However, since we have to solve the equation of higher degree under the parameter conditions, it is quite difficult to figure out by hand; we used here MATHEMATICA. The results will be available on request. 36. Since prices are higher than marginal costs, outputs are considered socially too few. Accordingly, we can regard the total demand of systems as a sort of index indicating social welfare. It is thought that the more the total demand, the higher the social welfare. 37. See Clark (2001), p. 40. 38. See Weinberg (2001), p. 352. 39. See Hoffman and Novak (2000), p. 256.

REFERENCES Anania, L. and R.J. Solomon (1997), Flat – The Minimalist Prices, in McKnight and Bailey, pp. 91–118. Armstrong, M. (1998), ‘Interconnection in Telecommunications’, Economic Journal, Vol 108, pp. 545–64. Armstrong, M., C. Doyle and J. Vickers (1996), ‘The Access Pricing Problem: A Synthesis’, The Journal of Industrial Economics, Vol. 44(2), pp. 131–50. Baumol, J.B. and J.G. Sidak (1994), Toward Competition in Local Telephony, MIT Press, Cambridge. Brynjolfsson, E. and B. Kahin (2000), Understanding the Digital Economy, MIT Press, Cambridge, MA. Clark, D.D. (1997), Internet Cost Allocation and Pricing, in McKnight and Bailey (1997a), pp. 215–52. Clark, D.D. (2001), A Taxonomy of Internet Telephony Applications, in McKnight et al. (2001a), pp. 17–42. Crawford, D.W. (1997), Internet Services: A Market for Bandwidth or Communication? in McKnight and Bailey (1997a), pp. 380–400. Economides, N. (1989), ‘Desirability of Compatibility in the Absence of Network Externalities’, American Economic Review, Vol. 79, 1165–81. Economides, N. (1996), ‘The Economics of Networks’, International Journal of Industrial Organization, Vol. 14, pp. 673–99. Economides, N. and S.C. Salop (1992), ‘Competition and Integration among Complements, and Network Market Structure’, Journal of Industrial Economics, Vol. 40(1), pp. 105–23. Faulhaber, G.R. and C. Hogendorn (2000), ‘The Market Structure of Broadband Telecommunications’, Journal of Industrial Economics, Vol. 48, pp. 305–29. Giovannetti, E. (2002), ‘Interconnection, Differentiation and Bottlenecks in the Internet’, Information Economics and Policy, Vol. 14(3), pp. 385–404. Hoffman, D.L. and T.P. Novak (2001), ‘The Growing Digital Divide: Implications for an Open Research Agenda’, in Brynjolfsson and Kahin (2000), pp. 245–60.

Kagami 03 chap 11

302

2/15/04

2:54 PM

Page 302

Challenging issues

Laffont, J.-J. and J. Tirole (1994), ‘Access Pricing and Competition’, European Economic Review, Vol. 38(9), pp. 1673–710. Laffont, J.-J. and J. Tirole (2000), Competition in Telecommunications, MIT Press, Cambridge. Laffont, J.-J., P. Ray and J. Tirole (1998), ‘Network Competition: Overview and Nondiscriminatory Pricing’, Rand Journal of Economics, Vol. 29(1) pp. 1–37. Lehr, W. (2001), ‘Vertical Integration, Industry Structure, and Internet Telephony’, in McKnight et al. (2001a), pp. 93–124. MacKie-Mason, J.K. and H.R.Varian (1997), ‘Economic FAQs about the Internet’, in McKnight and Bailey (1997a), pp. 27–62. McGarty, T.P and L.W. McKnight (2001), ‘Virtually Global Telcos: International Internet Telephony Architectures’, in McKnight et al. (2001a), pp. 43–91. McKnight, L.W. and J.P. Bailey (1997a), Internet Economics, MIT Press, Cambridge, MA. McKnight, L.W. and J.P. Bailey (1997b), ‘An Introduction to Internet Economics’, in McKnight and Bailey (1997a), pp. 3–24. McKnight, L.W. and B. Leida (2001), ‘Internet Telephony Service Providers’, in McKnight et al. (2001a), pp. 193–215. McKnight, L.W., W. Lehr and D.D. Clark (2001a), Internet Telephony, MIT Press, Cambridge, MA. McKnight, L.W., W. Lehr and D.D. Clark (2001b), ‘An Introduction to Internet Telephony’, in McKnight et al. (2001a), pp. 1–13. Matutes, C. and P. Regibeau (1988), ‘Mix and Match: Product Compatibility without Network Externalities’, Rand Journal of Economics, Vol. 19, pp. 221–34. Matutes, C. and P. Regibeau (1992), ‘Compatibility and Bundling of Complementary Goods in a Duopoly’, Journal of Industrial Economics, Vol. 40(1), pp. 37–54. Rubinfeld, D.L. and H.J. Singer (2001), ‘Vertical Foreclosure in Broadband Access?’, Journal of Industrial Economics, Vol. 49, pp. 299–318. Sharifi, H. (2001), ‘Internet Telephony Carrier Strategies’, in McKnight et al. (2001a), pp. 303–23. Shy, O. (2001), The Economics of Network Industries, Cambridge University Press, Cambridge. Weinberg, J. (2001), ‘Internet Telephony Regulation’, in McKnight et al. (2001a), pp. 325–66.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 303

15. Conclusion Mitsuhiro Kagami, Masatsugu Tsuji and Emanuele Giovannetti Deregulation and liberalization during the 1990s brought about competition in the telecommunications sector. Relaxation of market ownership restrictions and the lowering of entry barriers resulted in cross-market entry and made many combined business activities possible. Coupled with computer and wireless technologies, especially, digitization produced convergence among telecommunications, computer, and broadcasting industries. As a result, telephony (fixed and mobile), CATV, satellite, computer, and even home appliances have been connected through the Internet. Such services as voice, packet- or circuitswitched data, movies/videos, images (photos), TV, and e-business can be enjoyed by Internet users. Particularly, broadband access services have been expanding and they will be the mainstream in the future network society. From our joint study we can draw some lessons from the rapidly changing state of the IT revolution, especially from the developing countries’ point of view. Negative aspects are: (i) digital divide and universal service; (ii) monopoly and hegemony; and (iii) demand considerations, while the positive sides are: (iv) leapfrogging the industrialization process; and (v) broadband expansion.

1.

DIGITAL DIVIDE AND UNIVERSAL SERVICE

We noticed that disparities in Internet access among socio-economic groups are growing in all countries. It is observed that higher income groups can better utilize the Internet than other income groups. Education and ethnicity also affect Internet diffusion. In particular, large cities are unevenly distributed in Internet use and the urban poor have very limited access. In addition, the gap between rural and urban areas is also expanding. An even more serious issue from our point of view is the disparity among developing countries. It is clearly detected that Latin America as a region lags far behind East Asia in terms of Internet penetration. According to the International Telecommunication Union (ITU), Chile, the leading regional country in Internet use (1155 303

Kagami 03 chap 11

304

2/15/04

2:54 PM

Page 304

Challenging issues

per ten thousand inhabitants in 2000) is less than Malaysia (1590), the sixth ranked (after Korea, Japan, Hong Kong, Singapore, and Taiwan) in East Asia. Why is Internet penetration in Latin America delayed? There are mainly four reasons though the situation varies country by country. First, there exist enormous income disparities that cause low penetration. Low income groups cannot have Internet access even in advanced countries, so this tendency is strengthened in Latin America. Moreover, as broadband access develops, these deficiencies will further intensify. Second, the lack of basic infrastructure is crucial. Telephone networks are bases for Internet connection. DSL/ADSL also use these telephone lines. Prompt and secure mail delivery services are very important for e-business for distributing ordered online products. Stable electricity supply is also essential for computers and IT-related equipment. It is said that basic infrastructure and services remain weak in Latin America. Third, privatization of state-owned enterprises (SOEs) did not necessarily bring about efficient Internet diffusion. During the 1980s and 1990s state-owned telephone companies were privatized in the region but these incumbent companies still have monopolistic power and interconnection charges were not lowered as had been initially thought. Fourth, Latin America lacks interconnection between major network access points within the region. More concerted investment efforts for interconnection are required. Regarding universal service or universal access, more efforts should be made to increase public/private access centres to underserved and disadvantaged groups. Although some countries in Latin America have their own plans to establish telecentres and discount Internet rates, examples from other countries such as ‘E-rate grants’ and the Technology Opportunity Program (TOP) in the US, ‘telecottages’ in Hungary, the ‘Village Road Programme’ in Estonia, and ‘PC-bangs’ in Korea are useful.

2.

MONOPOLY AND HEGEMONY

Generally speaking, new entrants in Internet access businesses have two options: to build new networks or to lease lines from incumbent carriers. As mentioned above, incumbent telecommunications companies have monopolistic power due to their huge sunk cost and influential stakeholders such as the government and powerful politicians, so access prices failed to decline as expected in developing countries (they have gradually decreased due to competition and new entrants such as cable and satellite TV as compared with ten years ago, for example). However, an even more complicated problem has emerged regarding interconnection charges.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 305

Conclusion

305

Simply put, Internet connection has three tiers of groups: final users, Internet service providers (ISPs), and Internet backbone providers (IBPs). In this geographical classification, price structure is basically divided into two: retail prices which users pay for ISPs (prices for the ‘last mile’), and wholesale prices which ISPs pay for IBPs (sometimes called ‘upstream’ connectivity prices). IBPs are generally global conglomerates such as AT&T and Sprint. Because of technical reasons it is difficult to trace packet-data information routes. Moreover, confidential contracts among carriers make wholesale price structures ambiguous. Recent trends of mergers and acquisitions (M&As) among large IBPs have made this problem more complex. Potential anti-competitive activities should be monitored. In this volume tracing the information routes (number of hops) was examined and these trials will help with efforts to understand prices in upstream connectivity. Standards and intellectual property rights are another issue. International standards have often been used as a global strategy, or hegemony, to sell own products by multinational corporations (MNCs) or even governments of advanced countries. In the high-tech world, innovations become very complex and difficult to understand, backed by enormous R&D investment. Recent examples include third-generation cellular phones, digital TV methods, and encryption for online networks. Some measures are necessary to control these hegemonic movements and to bring transparency in decision-making of de jure standards. In this context, it is worth mentioning that the US government offered the advanced encryption standard (AES) for e-government use for public subscription in 1997. By the same token, industrial ownership or intellectual property rights have also sometimes been misused for latecomer developing countries. Recent WTO decisions to produce generic HIV drugs at a lower cost than the original royalty payment is a good example of a situation where modifications to the present system can be carried out if necessary.

3.

DEMAND CONSIDERATION

High-tech products are sometimes too high-tech. People do not want to deal with overly complex equipment or contents that are too sophisticated. Multifunctional equipment with a high price tag all too often does not correspond with the wishes of consumers who want simple functions. It is very important to read consumers’ preferences and people’s real demands, something that suppliers or producers often forget. Sometimes IT machines seem to exist for themselves, not for users. Japan’s failures illustrate this point. NTT invested heavily in ISDN but users ignored it because of its slow download time. NHK’s invention of analoguetype high definition TV was also left on the shelf after digital TV appeared.

Kagami 03 chap 11

306

2/15/04

2:54 PM

Page 306

Challenging issues

Both giant companies invested heavily in the wrong products and their investments were never recovered. Governments also fail in discerning real demand and allocate limited budgets for bad projects. Too many over-the-top IT plans such as the Multimedia Super Corridor in Malaysia have been promoted far outstripping people’s demands. In this regard, hasty and expensive terrestrial digital TV projects in Japan and Brazil seem to be under scrutiny from the demand point of view.

4.

LEAPFROGGING INDUSTRIALIZATION

It is interesting that the growth and dissemination of mobile phones in developing countries that generally have very limited traditional telephone copper-line networks have been so rapid. Basically in this situation the wireless or cellular phone networks leapfrog the fixed line system. This sort of leapfrogging (‘digital jump’) can be observed in other high-tech industries such as IT. Analysis in this volume of the Indian software industry suggests that certain developing countries may follow a different, not the traditional path, to successful industrialization. It concludes that Indian domestic firms may follow an alternative route in overcoming the difficulties of promotion and join the high value club after comparing profiles of three types of software entities in India: foreign firms; domestic firms with foreign subsidiaries; and domestic firms. The software sector has two distinct characteristics. One advantage is its absence of backward linkages, or what we call an ‘island of competitiveness’. The other is its small initial investment due to knowledge intensity. Due to these characteristics we can set up software businesses anywhere. However, a pre-condition may be relatively high education levels in English, mathematics and computer sciences. Moreover, location – a place where knowledge is agglomerated such as proximity to universities or national research centres, or where satellite links are easily facilitated – is also important. We have to carefully watch the Indian case further and hope that some developing countries can use the software sector as leverage for development.

5.

BROADBAND EXPANSION

Basic trends can be observed in the IT revolution in terms of structural changes. That is: (a) from fixed to mobile telephone; (b) from telephony to the Internet; and (c) from narrowband to broadband. Particularly, the growth of high speed Internet or broadband access (DSL/ADSL, CATV, FTTH, and FWA) is unprecedented. Demand is expected to increase further, if access prices go

Kagami 03 chap 11

2/15/04

2:54 PM

Page 307

Conclusion

307

down and content quality improves. Therefore, for further broadband expansion the following points must be addressed. First, new entry and competition should be promoted to reduce access prices. Second, distribution of quality contents needs to be promoted. The arrival of easily accessed broadband can act as a ‘big bang’ to get away from the present IT recession. In the near future multimedia convergence will take place and TVs and PCs will have the same function in the digital age. Third-generation mobile phones can be used as a high speed information terminal communicating voice, data, and pictures. Portable PCs with wireless LAN cards can work as small intelligent offices if handy wireless LAN boxes are facilitated in nearby shops like restaurants and coffee stands in towns. We hope that technological progress will find astonishing breakthroughs for enriching our lives at lower cost. However, as we progress into this ‘Brave New World’ we should not forget about the disadvantaged groups of people in developing countries and in pockets of developed countries who are excluded from benefiting from such achievements.

Kagami 03 chap 11

2/15/04

2:54 PM

Page 308

Kagami 03 chap 11

2/15/04

2:54 PM

Page 309

Index 0610 plan 222 @cars 254 Abbate, J. 179 AboveNet 42, 49 accelerators 118–23 contrastive effects on e-commerce 123–9 access charges 304 Argentina 223, 224 CEE 164–5 Mexico 210–12 accessibility 1, 199, 200–201 activities and strategies 102, 104–6 ADI (Thailand) 146 adoption of new technologies 64 ADSL (Asymmetric Digital Subscriber Line) 3–4, 115, 238–9, 290 Advance 221 advanced encryption standard (AES) 305 Advanced Research Projects Administration (ARPA) (US) 283 Advanced Television Systems Committee (ATSC) 241, 242 affordability 1, 199, 201–3, 214 Africa one 57 age digital divide and 144 firms 101, 103–4 Albania 161–9, 184 Alestra 212 AlternivaGratis 222 Amazon.com 3, 17, 18, 153 American Institute of Certified Public Accountants (AICPA) 267 analogue TV 239–40 ANS 49 antitrust 36, 47 analysis for backbone market 48–51 AOL Latin America 88

application areas 102, 104–6 Argentina 8, 9, 206–7, 218–25 broadband 67, 196–7 e-commerce 192 expanding Internet access 200–203 initiatives for Internet growth 221–4 Internet penetration 64, 65, 185–91, 219–21 telecommunications 196, 204, 218–21 telecommunications imports from US 87–9 Armstrong, M. 39 Arnet 221 Arora, A. 93, 98, 99, 100, 102, 105 ARPAnet 283 Asahi, The 249, 250, 251, 252, 256 Asahi National Broadcasting Company 251, 256 Asia 136, 142 Asian PKI market 270–75 international cross-recognition of PKI 277 see also under individual countries Asia-Pacific region 66 Asia PKI Forum 276–7 Asian Institute of Technology (AIT) 137 Association of Radio Industries and Businesses (ARIB) 232, 244 Association of the Thai Computer Industry (ATCI) 138–40 Asundi, J.M. 102 Asynchronous Transfer Mode (ATM) network 257 AT&T 49, 50, 84 average revenue per minute 74, 75 deregulation 70, 71–2 share of long distance service revenues 73, 74 AT&T Microelectronics (Thailand) 146

309

Kagami 03 chap 11

2/15/04

2:54 PM

Page 310

310

Index

ATMs (automatic transactions machines) 18 AU 255, 256 authentication 261–2, 277–8 see also electronic certification availability 1, 199–200 Avantel 212 B FLET’S 25 B2B e-commerce CEE 169–70 Korea 123, 127–9 Latin America 192–4 Mexico 215 B2C e-commerce CEE 169 Korea 122, 123, 125–6 Latin America 192–3 Mexico 216 Thailand 141 backbone 5, 35–61, 189, 190 antitrust analysis for the backbone market 48–51 European Internet backbones 39–41, 42–4 exploring through cyber-geography 51–3 South Africa 53–8 see also Internet backbone providers (IBPs) Baltimore Technologies plc 10–11, 263, 264, 266, 268–70, 276 Asian market 271, 272, 273–4, 274, 275 Band-X 44–7 bandwidth 190–91 leasing costs 41–6 Banerjee, A.V. 99 banking, Internet 178–9 BARA Baltimore Technologies Korea Company 274 Barnes and Noble 153 Beatty, J. 84 Beijing Ether Electronics Group Company 275 Bell Canada International 88 Bell South 88 Berg, S. 194, 195–6 Betamax 229, 230–31 Bhattacharjee, A. 94

Bolivia 195 Bray, H. 35 Brazil 243, 306 broadband 67 e-commerce 192 Internet access 64, 65, 185–91, 224 expanding 198–203 telecommunications 194, 197, 208 telecommunications imports from US 87–9 broadband 1, 3–4, 65–7, 224 expansion 306–7 Japan see Japanese broadband strategy Korea 4, 66, 115 Latin America 196–8 US 77–8, 79 see also under individual forms of access broadcasting 27 Broadcasting Satellite (BS) 240 BSAFE 265 BT Ignite 42 Bulgaria 161–9 business areas 102, 104–6 business models 265–9 Cabase NAP 222 cable modems see CATV cable penetration rates 68–9, 196–7, 218 cable TV 207, 251–2 Cable & Wireless 42, 49, 50, 52, 53 Canon 146 Carrier1 41, 42 Carterfone 72 CATV 3 Japan 20–21, 22, 23, 24, 24–5, 26, 27, 240 Korea 115 US 66–7, 77 Cave, M. 48 cell phones see mobile telephony cellular subscription rates 195–6 Cemex 215, 216 Central e-ASEAN Certificate Authority 271 Central and Eastern Europe (CEE) 7–8, 160–84 attracting investment 168–9 competition policies 167–8 Estonia see Estonia

Kagami 03 chap 11

2/15/04

2:54 PM

Page 311

Index pricing policies 164–7 range vs depth of access 161–4 transactions, services and rural initiatives 169–72 Centre for Democracy and Technology 168, 170 Centros Tecnológicos Communitarios (CTCs) (Argentina) 223 CERFnet 52, 53 certificate authority (CA) 10, 261 certification 120–1 electronic see electronic certification chaebols 127, 128 Chandra, P. 102 charges 20, 21–2 Chile 192, 194, 197, 199, 200, 201, 202 Internet access 185, 187–91 China 235, 241, 242, 275, 279 China Financial Certification Authority (CFCA) 275 Cho, Chang-Eun 122 CINX (Cape Town Internet Exchange) 53, 56 circuit-switching 284 Citicorp Overseas Ltd 97 Ciudad Digital 221 Claffy, K. 52 clusters 95 CMM 102, 105 coexistence 236–7 Cohendet, P. 94, 95, 100 COLT Telecom 42 combini banks 18 Comisión Federal de Telecommunicaciones (COFETEL) (Mexico) 207–8, 212, 215 Comisión Nacional de Comunicaciones (CNC) (Argentina) 219 commoditization 46 Communications Authority of Thailand (CAT) 137, 149–50, 154 communications satellites 240, 257 community health and welfare management 28 community technology centres 200, 223 competition broadband in US 77–8 deregulation see deregulation of telecommunications

311

international and firm level responses 94–6 local loop 74–7, 119–20 network competition and Japanese broadband strategy 24–6 and privatization in Latin America 199–200, 203–4 competition policy CEE 167–8 Internet upstream connectivity and 5, 35–61 competitive local exchange carriers (CLECs) 75–7 component prices 293–5, 297–9 components model 11, 291–9 Compuserve 49 computer viruses 129–32 Concert 42 congestion costs 38–9, 285, 286–7 connection models 11, 291–9 connectivity 2 Internet upstream connectivity and competition policy 5, 35–61 structure of the Internet 37–8, 137–8, 287–8 consortia 231 consumer electronics 150 consumer preferences 305–6 consumer satisfaction 236–9 content providers 37 convergence of services via the Internet 284–5 of technology 230, 239–43, 307 Cooperative Association for Internet Data Analysis (CAIDA) 51 co-production 92–3, 96 corporate strategies 265–9 Council on Telecommunication Technologies (Japan) 231 Covisint 128 credit cards 169, 170 credit rating 261 Cremer, J. 47 CrossCert Inc. 274 CS Nippon 257 CSNET (Computer Science Network) 283 CTI 219 culture 129 cyber crime 129–32

Kagami 03 chap 11

2/15/04

2:54 PM

Page 312

312 Cyber Crime Investigation Squad (Korea) 131 cyber-geography 36, 51–3 Cyber Korea 21, 120 Cyber Terror Response Centre (Korea) 131, 132 cyber terrorism 130–32 Czech Republic 161–9, 184 DACOM 123 Data-General 146 data path 51–2 D’Costa, A.P. 99, 103 de facto standards competition in PKI industries 10–11, 260–81 Japan’s standards strategy 9, 229, 230–32, 232 de jure standards 9, 229, 232–4 Deltathree.com 117 demand 305–6 depressed 202 digital TV 241 depth of access 161–4 deregulation of telecommunications Estonia 172–4 Korea 118–20 Latin America 194–6 Mexico 212 Thailand 154–5 US 71–4 Deutsche Telekom 42 developing countries 2 and consumer electronics production 150 dial-up charges 79–81 diffusion accelerators 118–23 rapid in Korea 115–18 diffusion rate of the Internet 135–6 Digex 49 DIGICERT 271 DigiCipher 239 digital divide 2, 62 CEE 170–71 Japan 24, 33 Thailand 141–5 and universal service 303–4 US 81–7 digital jump 306

Index Digital Multimedia Broadcasting-Terrestrial (DMB-T) 241, 242 digital satellite broadcasting 240, 252, 256–7 digital signature 10, 261 digital TV 235, 239–43, 306 digital versatile disk (DVD) 231, 244 Digital Video Broadcasting-Terrestrial (DVB-T) 241, 242 disappearance of technologies 237–8 distance learning 29–30, 170 international 30–2 division of labour, international 6, 92–113 Dodge, M. 51 domain name restrictions 164–5 domestic companies with foreign subsidiaries 98–109, 110 without foreign subsidiaries 98–109 DSL (Digital Subscriber Line) 3–4, 197 Japan 19–20, 21–2, 23, 24, 238–9 US 66–7, 77 Duflo, E. 99 Dymond, A. 171 e-ASEAN Task Force conference 271 e-BAY 17 e-commerce 17–18 CEE 169–70 difficulties in Internet shopping 125–6 Japan 32 Korea 123–9 Latin America 192–4 Mexico 192, 215–18 PKI solutions 10–11, 260–81 Thailand 141, 151 E-incubator 170 e-Japan project 22–6 E-rate grants 85, 86 e-toy 17 EasyNet 43 economic growth 92 Economides, N. 75, 291 edge pricing 39 education distance learning 29–32, 170 Thailand 147–9, 156–7 US 82, 83 EEBone 174 Eesti Telefon (ETC) 172–3, 178

Kagami 03 chap 11

2/15/04

2:54 PM

Page 313

Index eEurope initiative 39–40 El Salvador 195 El Sitio 216, 217 electronic application promotion consortium 274 electronic certification 10–11, 260–81 analysis of major PKI vendors 263–70 future prospects 275–9 strategy for Asian market 270–75 Electronics and Telecommunications Research Institute (ETRI) (Korea) 121 employees, number of 101, 103–4 employment 136, 152–3 Empresa Nacional de Telecommunicaciones (ENTel) 218 encryption technology 262, 279 see also electronic certification Energis 43 Entrust Inc. 10–11, 263, 264, 266, 267–8, 276 Asian market 273, 274, 275 Entrust.net 268 Entrust/PKI 268 Entrust PKI e-Government Edition 273 environmental standards 236 EO Mega-air 26 EO Mega-fibre 26 Epoch 49 equity capital 101, 103–4 Estonia 7–8, 161–9, 172–80, 184 banking 178–9 Digital Signature Act 2000 174, 179 harmonization 174 liberalization 172–4 national promotion of IT 174–7 public-private partnerships and FDI 178 Estonia Education and Research Network (EENet) 174, 175 Estonian Communications Board 174 Estonian Information Technology Foundation 177 Estonian Mobile Telephone 178 EuroISPA 40 Europe DVB-T 241, 242 Internet backbones 39–41, 42–4 European Commission 48–51 European Committee for Standardization (CEN) 235

313

European Union, accession to 167–8, 168–9 Exodus 49 export intensity 102–3, 106–9 export markets 102, 104–6 export processing zones 97 exports India 101, 102–3, 103–4, 106–9 Thailand 146 US telecommunications 87–9 factor inputs 101, 103–4 Federal Communications Commission (FCC) (US) 71–2 Fernandes, R.J. 93 fibre glut 41 see also FTTI; optical fibres Fiddler, R. 258 Finland 178 firm capabilities 94–6 firm size 101, 103–4, 144–5 fixed penetration 163–4 flat-rate pricing 38–9, 285–6 foreign companies 98–109 see also multinational corporations foreign direct investment (FDI) 92–3 CEE 168–9, 178 India 97 Thailand 145–7, 150–1 foreign subsidiaries 98–109, 110 forum standards 9, 231 France 225 France Telecom 43 free ISPs 222 free riding problem 38 FTTH (fibre to the home) 1, 3–4 Japan 19, 20, 21–2, 23, 24, 24–5 Fuentes, M. 200 Fuentes-Bautista, M. 200, 202 Fuji TV 251 Fujikura 146 Fujita, M. 95 Fujitsu 146, 272 FWA (fixed wireless access) 3–4 Japan 21, 22, 23, 24, 24–5, 25–6 GDP Internet penetration and 226 Latin America 195–6 gender 144

Kagami 03 chap 11

2/15/04

2:54 PM

Page 314

314 Germany 225 Ghemawat, P. 110 Giovannetti, E. 39 GITN 271 Global Crossing 40, 41, 43, 57 globalization 92–6 international competition and firm level responses 94–6 government role de jure standards 232–3 Korea 120–22 Thailand 156–8 Gregg, J. 278–9 GridNet 49 group consciousness 122 GTE 49, 50 GTS 43 Guatemala 195 Guerrero, A. 224 Guillén, M. 204 Gutierrez, L. 194, 195–6 Guyana 195 hacking 129–32 Hamano, Y. 239 Hanaro Telecom 119–20 Hansapank 179 harmonization 174 Hashimoto, N. 257 Hatch, O. 278 Hattori, K. 239 health care 27–9 Heeks, R. 97, 99, 103 hegemony 304–5 Hewlett-Packard (HP) 97 Hicks, D.M. 93 high definition TV (HDTV) 239 high tech crimes 129–32 high technology markets 94 Hindustan Computers Ltd (HCL) 97 Hitachi 145 HiTRUST.COM (HK) 271–2, 275 HiTRUST Inc. 275 Hobbes, R.Z. 38 homogeneity 122 Hong Kong 206, 271–2 Hong Kong Post 272 hops (routers) 46, 51 household income 77, 79, 82, 83, 144 human capital 101, 103–4

Index human resource development 155–8 Human Resource Development Project for Vietnam 30–32 Hungary 8, 161–9, 170, 171–2, 180, 184 Hush-a-Phone 71–2 i-mode (Internet mode services) 255, 256 iBASIS 117 IBM 146, 193, 238 ID.Safe 271 Ida, T. 230 Identrus 269, 277 ILO 141–2, 153 immature technology 18 Impsat 221, 222 Inamasu, T. 255 income disparities 304 GDP see GDP household 77, 79, 83, 84, 144 incubators for entrepreneurs 97–8 independent data broadcasting 257 India 6, 92–113, 306 typology of firms 98–109 USAID project 156 Industrial Revolution 135 industrialization, leapfrogging 6, 306 industry-based (voluntary) standards 231–2, 232 information processing technology 135 information policy 7–8, 160–84 information technology (IT) IT revolution 1, 135, 247–8 quick changes in 1–2 recession 2–4, 16–17 spending per capita in Latin America 187, 189 Information Technology College Estonia 177 information transmission technology 135 Infosys 97 infrastructure 1, 304 Internet infrastructure and Internet access in US and the Americas 64–70 Korea and network infrastructure 120 telecommunications Latin America 187–91 Thailand 144, 145, 149–50

Kagami 03 chap 11

2/15/04

2:54 PM

Page 315

Index telephones infrastructure investment in Mexico 213 Latin America 187, 188, 190 InNET 49 inputs 101, 103–4 integrated circuits (ICs) 145–6 Integrated Services Digital BroadcastingTerrestrial (ISDB-T) 241, 242, 243 intellectual property rights 305 interconnection 304 and pricing of the Internet 11, 282–302 interest groups 241 Intermedia Communications 49 international cooperation 131–2 international cross-recognition of PKI 277 international distance learning project 30–32 international division of labour 6, 92–113 International Electrotechnical Commission (IEC) 233 International Standardization Organization (ISO) 102, 105, 229, 233, 234 9000 series 235–6 14 000 series 236 international organizations 233–4 International Telecommunication Union (ITU) 233, 303–4 Internet 1, 252–5 bubble 15, 16–18 connectivity see connectivity convergence of services via 284–5 difference from telecommunications 284 diffusion rate 135–6 e-commerce see e-commerce history of 283 interconnection and pricing 11, 282–302 Japanese press and 252–5 Latin America see Latin America model analysis 290–99 social issues caused by diffusion of 129–32 structure 37–8, 137–8, 287–8 Internet access Asian countries 142

315

digital divide see digital divide Estonia 174–7 Internet infrastucture and in US and Americas 64–70 Korea 115–18 Latin America see Latin America policies for in Mexico and Argentina see Argentina; Mexico range vs depth in CEE 161–4 Thailand 138–41, 143 US 81–2 universal service 85–7 Internet addiction 129 Internet backbone providers (IBPs) 2, 287–8, 305 Internet upstream connectivity and competition policy 5, 35–61 models of connection 11, 290–99 Internet banking 178–9 Internet Broadcasting 254 Internet cafés 69–70, 122, 201 Internet exchange points (IXPs) 37, 38 Internet penetration rates 62, 63, 64–5, 140, 185, 186, 225, 226 Japan 253 Korea 115, 116 Latin America 185, 186, 219–21, 224, 303–4 Internet pricing 79–81 and interconnection 11, 282–302 Japan 20, 21 Mexico 210–12 policies in CEE 164–7 problems 38–9, 286–7, 304–5 Thailand 138, 139 upstream connectivity and competition policy 5, 35–61 Internet service providers (ISPs) 2, 5, 37, 38, 155, 287–8, 305 Argentina 221, 222 connection models 11, 291–9 Estonia 173–4 European 40 free 222 Internet pricing 39, 79–81, 203, 210–12, 305 Japan 287–8 small and free peering 48 Internet Service Providers Association (ISPA) (South Africa) 53

Kagami 03 chap 11

316

2/15/04

2:54 PM

Page 316

Index

Internet Solutions (IS) 54, 56, 57 Internet telephony 284–5 Internet Thailand (INETTH) 137 Interoute 41, 43 Interpark 123 Interpol 132 investigation framework for cyber crime 131–2 Ireland 204 ISC 62 ISDN (Integrated Services Digital Network) 28, 29–30, 30–32, 238 IT see information technology ITCX 117 iTrusChina Company 275 iTS Communications 26 Iyer, K. 102 J-Phone 255, 256 Jamison, M. 194 Japan 278 assistance for Thailand 155–6 attacks on Japanese organizations by Korean hackers 131–2 demand and product failures 305–6 development of digital TV 240–41 electronic certification 272–4 Internet growth 225, 283 Internet providers 288–90 Japanese MNCs 145–6 Japan Newspaper Publishers and Editors Association (JNPEA) 247, 252, 259 Japan Promotional Association for Asia PKI Forum (APKI-J) 276 Japan Telecom 290 Japanese broadband strategy 4–5, 15–34 current state of broadband 4, 18–22 estimation of future broadband access 22–4 lessons learned from Net bubble 16–18 National Broadband Network Initiative 22–6 possible applications 26–32 recent broadband development 24–6 Japanese Industrial Standards (JIS) 229, 232, 234 Japanese Industrial Standards Committee (JISC) 232, 233 Japanese press 10, 247–59 cell phone news service 252, 255–6

composition 248–51 digital satellite broadcasting 252, 256–7 Internet 252–5 and TV 251–2 Japanese standards strategy 9–10, 229–46 defeats in standards war 234–6 Jensen, M. 55 JINX (Johannesburg Internet Exchange) 53, 56 job consciousness 123 jobs relocation 153 joint ventures 100 Jou, Y. 234 JUNET (Japan UNIX/University Network) 283 Jupiter Communications 192 K-Opticom 25–6 Kagami, M. 235, 243 Kansai Electric Company 26 Kattuman, P. 94, 102 kawara-ban 248 Kende, M. 40 KEON 265 Korea, South 6, 114–34 accelerators 118–23 Act on Digital Signature 1999 123 Basic Act on Electronic Commerce 1999 123 broadband 4, 66, 115 e-commerce 123–9 electronic certification 274, 278 PC-bangs 69–70, 122, 201 rapid diffusion of new technologies 115–18 social issues caused by diffusion of the Internet 129–32 Korea Electronic Certification Authority 274 Korea Institute of Industrial Technology Evaluation and Planning (ITEP) 121–2 Korea Network Information Centre (KRNIC) 118, 119, 122 Korea PKI Forum 276–7 Korea Telecom (KT) 119–20 Korean backbone computer network project 120

Kagami 03 chap 11

2/15/04

2:54 PM

Page 317

Index

317

Korean information infrastructure (KII) 120 Korean National Police Agency (KNPA) 130–32 KPN Qwest 40, 41, 43 KSC Comnet 137 Kula Tree (Village Road) programme 174–5 Kyl, J. 278

local loop competition 74–7, 119–20 local access charges 79–81 pricing in CEE 164–5 local newspapers 249, 250 locked-in effects 230 long-run growth rate 102–3, 106–9 Lotte department store 123 Loxinfo Company 137

labour 147–9, 152–3 Laffont, J.-J. 39 Lakha, S. 93 language 129 large firms 144–5 Lateef, A. 93 Latin America 2, 5–6, 8, 64–5, 67–8, 185–205, 225–6 broadband 196–8 current status of Internet development 185–91 e-commerce 192–4 expanding Internet access 198–203 Internet penetration rates 185, 186, 219–21, 224, 303–4 outlook for future Internet growth 203–4 telecommunications reform 194–6 US telecommunications exports to 87–9 see also under individual countries Latin Internet portal 88 Latvia 161–9, 179, 184 Lawrie, M. 53 leapfrogging industrialization 6, 306 learning 6, 110 see also education Lee, Nae-Chan 118 letters vs telephone 236–7 Level 3 40, 41, 43 LG Electronics 127 LG Telecom 117 liberalization see deregulation of telecommunications Lie, Han-Young 118 linkages 37–8 Linux 2 Lithuania 161–9, 184 Lityan Holdings 271 local area networks (LANs) 24, 27, 238

M-Web 55 Mackie-Mason, J.K. 39 mainframe computers 238 Mainichi, The 249, 251, 252 Malaysia 271, 306 Malmberg, A. 94, 95, 101 mandatory standards 232 Marchand, R. 84 market failure 24 market liberalization see deregulation of telecommunications market shares backbone market 49–50 Indian software industry 102–3, 106–9 Masel, G. 53 Maskell, P. 94, 95, 101 Mason, R. 39 Matsushita Corporation 145 Matutes, C. 291 McDaniel, J. 167 MCI-WorldCom 56 rejection of proposed merger with Sprint 47, 48–51 Media General 257 Media Strategy Bureau 254 mentality 122–3, 128 merger, proposed between MCIWorldCom and Sprint 47, 48–51 Metrix Interlink 49 Mexeric 207 see also Telmex Mexico 9, 206–18, 224–5 broadband 197 e-commerce 192, 215–18 Internet access 64, 65, 67, 185–91 expanding 198–203 telecommunications 207–12 reforms 212–15 US telecommunications exports to 87–9

Kagami 03 chap 11

318

2/15/04

2:54 PM

Page 318

Index

MIC 117 Microlink 175 Microsoft 88, 97 Microwave Communications Inc. (MCI) 72 see also MCI-WorldCom Minebea 146 Mitsubishi 145 mobile Internet 117 penetration in CEE 162, 163–4 mobile telephony 1, 306, 307 CEE 165, 167 cellular subscription rates 195–6 Latin America 199 news services in Japan 252, 255–6 model analysis of Internet 11, 290–99 basic setup 291–2 one-way model 11, 290–91, 292–3, 293–5, 295–9 regulatory model 295–9 results of analysis 293–5 two-way model 11, 291, 293–5 monopolies 165, 166, 304–5 Motorola 97 Movicom 219 MSC Cybersign International 271 MSC Trustgate.com 270, 271 Multimedia Super Corridor 306 multinational corporations (MNCs) India 97, 98–109 investment in Thailand 145–7 multi-part tariff 286 Murray, B. 171–2 Naemura, K. 238 NASSCOM 98 National Broadband Network Initiative (Japan) 22–6 National Communications Committee (NCC) (Thailand) 150 National Electronic and Computer Technology Centre (NECTEC) (Thailand) 137, 141 National Science Foundation (NSF) (US) 283 National Information Infrastructure (NII) (US) 85 National Semiconductor 146

National Telecommunications and Information Administration (NTIA) (US) 85–7 National Thai 145 nationalization 194 Net bubble 15, 16–18 Net2Phone 117 network access points (NAPs) 37, 38, 190, 191, 222, 287 network competition 24–6 network economy 230 network efficiency measures 222 network externalities 230 Neusoft Corporation 275 ‘New Economy’ 15, 282 newspapers circulation by country 248–9 Japanese see Japanese press NHK (Nippon Hoso Kyokai) 237, 239, 240, 243, 251, 305–6 Nicaragua 195 Nidec 147 Nielsen Net Ratings 62, 63, 64, 65 Nihon Keizai, The 249, 250, 251, 252, 254, 256 Nikkei Net 254 Nippon Data Broadcasting (NDB) 257 Nippon TV Network (NTV) 251, 256, 258 Nishioka, I. 255 NL Net 49 Nordea 179 Nortel Networks Inc. 267 Novell 97 NSFNET 283 NSJ Corporation 273 NTT 19, 25, 288, 289, 290, 305–6 NTT-DoCoMo 252, 255, 256 NUA 62 OECD 210–12, 214–15 Oestmann, S. 171 Oftel 41, 47–8 ‘One Product for One Tamboon’ project 156 one-way model of connection 11, 290–91, 292–3, 293–5 regulation 295–9 online retail sites 216–18 open global standard for PKI 276–7

Kagami 03 chap 11

2/15/04

2:54 PM

Page 319

Index optical fibres 1, 19, 25, 33, 40 see also FTTH Oracle 97 Ote-Komachi 254 output 102–3, 106–9 outsourcing 92–3, 96 overestimation of scale and pace 17 packet loss 46 packet-switching 284 Pan-Asian e-commerce Alliance (PAA) 277 PanAmSat 215 Pappalardo, D. 49, 50 pari pari tendency 122, 126 Patibandla, M. 102, 110 Patni Computer Systems 97 payment procedures 125–6 PC-bangs 69–70, 122, 201 peering 40, 48, 58 penetration rates cable 68–9, 196–7, 218 Internet see Internet penetration rates networks in CEE 161–4 telephones see telephones personal computers (PCs) 1 convergence with TVs 241, 307 disappearing technologies 237–8 ownership rates 68, 81–2, 138–40, 142–3, 191 personal information, leakage of 125, 126, 129 personal life 129 Peru 195, 200–201, 202 Petrazzini, B. 224 piracy 129 PKI Processing Centre 272 Poland 161–9, 180, 184 political stability 204 population density 77, 78, 195–6 portals 37, 216, 217 POS Malaysia 271 Posco 127 Post and Telegraph Department (PTD) (Thailand) 149 press see Japanese press; newspapers price-cap regulation 296, 297–9 pricing, Internet see Internet pricing private key registration (deposit) system 279

319

privatization 304 Latin America 194–5, 199–200, 203–204, 207, 218–19 Thailand 150 productivity 102–3, 106–9 products 102, 104–6 projects 102, 104–6 PSINet 43 public Internet access sites 69–70, 122 Estonia 174, 175, 176, 177 Latin America 200–201, 223 public key infrastructure (PKI) 1–2, 10–11, 260–81 analysis of major PKI vendors 263–70 future prospects 275–9 open global standard 276–7 regionalism 277–9 strategy for Asian market 270–75 public networks 24 public-private partnerships 178 Pyramid Research 192, 194, 196, 208 quality IP quality index 46–7 management 235–6 of services 286 standards in Indian software 102, 104–6 quick ratios 266, 280 race 82, 83 radio 237 Rakbankerd.com 151 range vs depth of access 161–4 recession IT 2–4, 16–17 Japan 15 Regibeau, P. 291 regional Bell operating companies (RBOCs) 72, 74–5 regional cooperation 155–6 regionalism 277–9 regulation 212, 221 model of connections 295–9 see also deregulation of telecommunications Reichl, P. 38 reputation 101 Resale Price Maintenance System (RPMS) 247, 259

Kagami 03 chap 11

320

2/15/04

2:54 PM

Page 320

Index

research and development (R&D) 262, 263 retail prices 39, 79–81, 203, 210–12, 305 retail sites, online 216–18 revenues Indian software 101, 102–3, 103–4, 106–9 per capita for PKI vendors 264 robustness of existing systems 17–18 Rogers, E.M. 64 Romania 161–9, 184 Root, C.A. 269 round trip time (RTT) 46, 51 routers (hops) 46, 51 Royce, W.W. 95 RSA Security Inc. 10–11, 262, 263, 264, 265, 266 Algorithm Patent 262, 265, 280 Asian market 272 rural initiatives 170–72 rural-urban divide 82, 83 SAFE 56 SAIX 54, 56, 57 Samnuk-Rakbankerd Foundation 151 Sampo Pank 179 Samsung Electronics 127 Sankei, The 249, 251 Sanyo 145 SAT-3/WASC 56 Satelites Mexicanos (Satmex) 215 Satellite Data Networks (SDN) 53–5, 57–8 satellite digital broadcasting 240, 252, 256–7 satellite networks 198 SBC Communications 88 Schenker, S. 39 Sclavos, S. 265–6 Seagate Technology 146 SECOM 273, 273–4 Secretaría de Communicaciones y Transportes (SCT) 212, 215 Secure Channel Project 268 SecurID 265 security 126, 179, 278–9 Security Dynamics Ltd K.K. 272 Serome 117–18 services Indian software 102, 104–6 trade in 70

SES Global 215 Setouchi Town tele-education system 29–30 settlements policy 287 Shanghai Electronic Certificate Authority Centre (SHECA) 279 Siemens 97 Signetics 146 Siil, I. 165 Singapore 157–8, 204, 270–71 Singapore Network Services (SNS) 271 size, firm 101, 103–4, 144–5 skilled labour 152 Slovakia 170, 184 Slovenia 161–9 passim, 170, 172, 184 small and medium enterprises (SMEs) Korea 128 Thailand 144–5, 151–2, 156 smart market 287 society 129 software 6, 92–113, 306 waterfall model of development 95–6 SOK 273 SONY 230–31 Sony Pictures 26 Soros Foundation for an Open Society 162 South Africa 53–8 development of backbone industry 53–7 telecommunications policy 57–8 ‘space bird’ 241 special dialling area code (0610) 222 specialization 102, 104–6 spectrum auctions 225 Sport Hochi, The 255 Sprint 52, 53 rejection of proposed merger with MCI-WorldCom 47, 48–51 standardization 9–10, 229–46 standards 2, 9–10, 229–46, 305 convergence of technology 239–43 de facto see de facto standards de jure 9, 229, 232–4 technology and satisfaction 236–9 Stigler, J. 71 stock prices 16–17 strategies and activities of Indian software firms 102, 104–6 corporate 265–9

Kagami 03 chap 11

2/15/04

2:54 PM

Page 321

Index Straubhaar, J. 200 streaming 117, 254 Suarez, S. 204 Sutton, J. 95, 110 Suzuki, T. 257 Sweden 178 system demands 293–5, 297–9 systems, robustness of existing 17–18 T1msn 216, 217 tacit knowledge 95 targeted degradation 47 Tauzin-Dingell bill 66–7 Tata Consultancy Services (TCS) 96–7 taxation 168 Technical Barriers to Trade (TBT) 234 technical standards see standards technological achievement index 161 technology adoption of 64 and consumer satisfaction 236–9 convergence of 230, 239–43, 307 evaluation 120–21 immature 18 Technology Opportunities Program (TOP) (US) 85–7 Teikoku Databank 274 Telcel 207 tele-health system 28 tele-home-care system 27–8 telecare 27–9 telecentres 165, 171–2 Telecom Argentina 218–19 Telecommunication Business Act (Korea) 119 telecommunications 8, 225 Argentina 196, 204, 218–21 competition policies in CEE 167–8 deregulation see deregulation of telecommunications differences between Internet and 284 exports from US to Latin America 87–9 infrastructure see infrastructure Mexico see Mexico monopolies in CEE 165, 166 reforms and Internet usage in US and the Americas 70–74 telecommunications policy 57–8 telecottages 171–2

321

Telefónica de España (Telefónica) 218–19 Teleglobe 44 Telegraph Organization of Thailand (TOT) 149–50 Telenor 44 telephones coexistence with letters 236–7 costs of Internet access in Mexico 210–12 expanding Internet access in Latin America 198–203 infrastructure see infrastructure local see local loop mobile see mobile telephony penetration rates 68 Mexico 208–9, 213–14 Thailand 144, 145 subscription rates in Latin America 202–3, 214 universal service in US 84–5 Televisa 207, 215 television (TV) 1 coexistence with radio 237 convergence with PCs 241, 307 digital 235, 239–43, 306 Japanese newspapers and 251–2 Telia 44 Telkom 53, 56, 57 Telmex (Teléfonos de México) 9, 88, 202, 206, 207–8, 208–9, 212, 215, 216 Telmex Internet Directo 210 Terra 216, 217 terrestrial digital TV 235, 240, 306 terrorism 278–9 cyber terrorism 130–32 test of the hypothetical monopolist 41 Texas Instruments (TI) 97 Thai Social/Scientific, Academic and Research Network (ThaiSARN) 137 Thailand 7, 135–59 current IT status 137–50 impact on employment 152–3 impact on firms 150–52 Internet 137–8 IT consumer 138–45 IT producer 145–50 policy issues 153–8 Telecommunications Master Plan 149–50

Kagami 03 chap 11

2/15/04

2:54 PM

Page 322

322 360networks 40, 41, 42 throughput 46 Tiger Leap programme 176–7 tipping 9, 229, 230–32 Tirole, J. 39 Todito 216, 217 Tokyo Broadcasting System (TBS) 251 Toshiba 145 total demand 294, 295, 297–9 Traceroute 46, 51, 61 trade 70 Tradelink 272 TradeNet 271 training programme 157–8 transaction-based pricing 286 transit arrangements 40, 58 transmission control protocol/internet protocol (TCP/IP) 37, 58, 283 TRON 2, 12 trust 260–2, 280 see also public key infrastructure TrustAsia Inc. 270, 271 Tsubota, T. 254 Tsuji, M. 19, 33 TV Tokyo 251, 256 two-way model of connection 11, 291, 293–5 typewriters 237–8 UBS Warburg 120, 122 Ühispank 179 unbundling 296–7, 297–9 unemployment 148 UniCERT 271 Unicom-Pipex 49 Uninet 53 United Communication Industry (UCOM) 151 United Kingdom (UK) 240 United Nations, Human Development Report 161 United Nations Committee on Environment and Development (UNCED) 236 United Nations Development Programme (UNDP) 156 United States (US) 5–6, 53, 62–91 American firms’ investment in Thailand 145–6 Communications Act 1934 71, 73

Index competition in local loop and broadband service 74–8 costs of access 79–80 Department of Commerce (DOC) 262, 282 Department of Justice 49, 51 digital divide 81–7 digital TV 240, 241, 242 Framework for Universal Service 84–5 HDTV 239 internet access 64–7 IT recession 2–3, 16–17 National Information Initiative (NII) 24 ‘New Economy’ 15 outsourcing 93 regionalism 278–9 Telecommunications Act 1996 5, 70, 73–4, 74, 84–5, 198, 299–300 telecommunications exports 87–9 telecommunications reforms 70–74 Uniting and Strengthening America (USA) Act 2001 278 United States Agency for International Development (USAID) 171 universal service digital divide and 303–4 US framework for 84–5 extending to Internet access 85–7 Universal Service Administrative Company 85 upstream connectivity prices 305 competition policy and 5, 35–61 urban-rural divide 82, 83 Uruguay Round 118 usage-sensitive pricing 38–9, 285–6 UUNET 44, 48, 49, 52, 53 UUNET Africa 55–6 UUNET South Africa 54, 55–6, 57–8 Uyuyuy 222 Van Sommeren, A. 277 Varian, H.R. 39 Venezuela 200, 201 venture businesses 121–2 Verisign Inc. 10–11, 263, 264, 265–7, 276 Asian market 270–71, 271, 271–2, 272–3, 274, 275

Kagami 03 chap 11

2/15/04

2:54 PM

Page 323

Index Verisign OnSite 267 VHS 229, 230–31 Victor/Matsushita 230–31 Video Centre 258 videocassette recorders (VCRs) 229, 230–31 Vietnam, Human Resource Development Project for 30–32 village enterprises 151 Village Road (Kula Tree) programme 174–5 viruses, computer 129–32 Vodaphone 290 Voice-over-IP (VoIP) 117–18 Volkswagen 193 voluntary standards 231–2, 232 VTN (VeriSign Trust Network) 267, 271, 272, 276 WAP 117 Watanabe, T. 258 waterfall model 95–6 web magazines 254 Web2Phone 118 webcasting 117 wholesale prices 305 Internet upstream connectivity and competition policy 5, 35–61 WIDE (Widely Integrated Distributed Environment) 283 wide area networks (WANs) 24

323

Williams Communications 40, 49 Wilson, K.G. 71 Wipro 97 wireless licences 225 word processors 237–8 workforce 147–9, 152–3 World Bank 156, 202 World Trade Organization (WTO) 70, 118, 305 Basic Telecommunications Agreement 119, 155 TBT 234 WorldCom 40, 44, 56 see also MCI-WorldCom Yahoo BB 19, 21 Yahoo Latin America 88 Yankee Group, The 209 Yomal, R. 196 Yomiuri, The 10, 248, 249, 250, 256, 257 cell phone news service 255–6 General Systems Centre 257–8 Internet 252–4 News Distribution Centre 258 and TV 251 Yomiuri On Line (YOL) 252–4, 257 Yomiuri Shimbun Company 252 Media Strategy Bureau 254 Yusen Broadnetworks 19, 25 Zona Latina 192–3