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 1171 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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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

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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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 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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


7

8

9

10


Market share, 2000–01 8.00

Density

6.00 4.00 2.00 0.00

0

Foreign cos.

1

2

3


4

5

6


Kagami 01 chap 01

2/15/04

2:55 PM

Page 109


109

Age (in years), 2001

Density

0.12 0.08 0.04 0.00

0

Foreign cos.

5

10

15


20

25


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


200


Productivity (Rs.mn./Employee) 0.40

Density

0.30 0.20 0.10 0.00

0

Foreign cos.

1

2

3


4

5

6

7


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


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


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


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


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

2/15/04

2:54 PM

Page 117


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


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


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


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


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


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


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

2/15/04

2:54 PM

Page 125


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


(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


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


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


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


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. ** 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

2/15/04

2:54 PM

Page 295


295

Lemma 1.3 (The comparison of components prices between 1w and 2w): P11w < P12w, P21w ()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


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

2/15/04

2:54 PM

Page 299


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


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

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

Developing Countries and the WTO: Policy Approaches

Innovation Policy: A Guide for Developing Countries

Trade Policy in Developing Countries

Trade Policy in Developing Countries

Macroeconomics for Developing Countries

The Digital Divide

Transport and Developing Countries

Globalization and the Developing Countries

Cyberspace Divide: Equality, Agency and Policy in the Information Society

Technology Transfer for Renewable Energy: Overcoming Barriers in Developing Countries

Information Technology, Innovation System and Trade Regime in Developing Countries: India and the ASEAN

The Political Economy of Taxation: Lessons from Developing Countries

WTO Law and Developing Countries

Public Information Technology: Policy and Management Issues

Gender and Computers: Understanding the Digital Divide

Bridging the Class Divide: And Other Lessons for Grassroots Organizing

Digital Bridges: Developing Countries in the Knowledge Economy

Gender and Computers: Understanding the Digital Divide

Soil conservation in developing countries: project and policy intervention

Marine biotechnology and developing countries

Digital Bridges: Developing Countries in the Knowledge Economy

Industrial Policy in Developing Countries - Overview and lessons from seven country cases

Minimum Wages and Social Policy: Lessons from Developing Countries (Directions in Development)

Developing Countries in the WTO

Cases on Information Technology: Lessons Learned (Cases on Information Technology Series) (Cases on Information Technology Series)

Economics of Developing Countries

Regionalism Among Developing Countries

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

Developing Countries and the WTO: Policy Approaches

Innovation Policy: A Guide for Developing Countries

Trade Policy in Developing Countries

Trade Policy in Developing Countries

Macroeconomics for Developing Countries

The Digital Divide

Transport and Developing Countries

Globalization and the Developing Countries

Cyberspace Divide: Equality, Agency and Policy in the Information Society

Technology Transfer for Renewable Energy: Overcoming Barriers in Developing Countries

Information Technology, Innovation System and Trade Regime in Developing Countries: India and the ASEAN

The Political Economy of Taxation: Lessons from Developing Countries

WTO Law and Developing Countries

Public Information Technology: Policy and Management Issues

Gender and Computers: Understanding the Digital Divide

Bridging the Class Divide: And Other Lessons for Grassroots Organizing

Digital Bridges: Developing Countries in the Knowledge Economy

Gender and Computers: Understanding the Digital Divide

Soil conservation in developing countries: project and policy intervention

Marine biotechnology and developing countries

Digital Bridges: Developing Countries in the Knowledge Economy

Industrial Policy in Developing Countries - Overview and lessons from seven country cases

Minimum Wages and Social Policy: Lessons from Developing Countries (Directions in Development)

Developing Countries in the WTO

Cases on Information Technology: Lessons Learned (Cases on Information Technology Series) (Cases on Information Technology Series)

Economics of Developing Countries

Regionalism Among Developing Countries

Recommend Documents