WORLD BANK TECHNICAL PAPER NO. 383 Work in progress W TP383 for public discussion Nov . lIq% Technology Institutions and Policies I h/wir Rol, i, I)/'e /-l,opip lh( /i(iol(/,gi ,al (Cpa//ilit i11 IJ(flsot,' -$u1 E1UFIEnU 0 - Hrk'ill~~~~~~~~~~~il I~~ . Hrr i /,,'()S WithiTu Ra/p/ id/( ,~I/'i RECENT WORLD BANK TECHNICAL PAPERS No. 310 Elder and Cooley, editors, Sustainable Settlement and Development of the Onchocerciasis Control Programme Area: Proceedings of a Ministerial Meeting No. 311 Webster, Riopelle and Chidzero, World Bank Lendingfor Small Enterprises 1989-1993 No. 312 Benoit, Project Finance at the World Bank: An Overview of Policies and Instruments No. 313 Kapur, Airport Infrastructure: The Emerging Role of the Private Sector No. 314 Valdes and Schaeffer in collaboration with Ramos, Surveillance of Agricultural Price and Trade Policies: A Handbookfor Ecuador No. 316 Schware and Kimberley, Information Technology and National Trade Facilitation: Making the Most of Global Trade No. 317 Schware and Kimberley, Information Technology and National Tradle Facilitation: Guide to Best Practice No. 318 Taylor, Boukambou, Dahniya, Ouayogode, Ayling, Abdi Noor, and Toure, Strengthening National Agricul- tural Research Systems in the Humid and Sub-humid Zones of West and Central Africa: A Frameworkfor Action No. 320 Srivastava, Lambert, and Vietmeyer, Medicinal Plants: An Expanding Role in Development No. 321. Srivastava, Smith, and Fomo, Biodiversity and Agriculture: Implicationsfor Conservation and Development No. 322 Peters, The Ecology and Management of Non-Timber Forest Resources No. 323 Pannier, editor, Corporate Governance of Public Enterprises in Transitional Economies No. 324 Cabraal, Cosgrove-Davies, and Schaeffer, Best Practicesfor Photovoltaic Household Electrification Programs No. 325 Bacon, Besant-Jones, and Heidarian, Estimating Construction Costs and Schedules: Experience with Power Generation Projects in Developing Countries No. 326 Colletta, Balachander, and Liang, The Condition of Young Children in Sub-Saharan Africa: The Convergence of Health, Nutrition, and Early Education No. 327 Vald6s and Schaeffer in collaboration with Martin, Surveillance of Agricultural Price and Trade Policies: A Handbookfor Paraguay No. 328 De Geyndt, Social Development and Absolute Poverty in Asia and Latin America No. 329 Mohan, editor, Bibliography of Publications: Technical Department, Africa Region, July 1987 to April 1996 No. 330 Echeverria, Trigo, and Byerlee, Institutional Change and Effective Financing of Agricultural Research in Latin America No. 331 Sharma, Damhaug, Gilgan-Hunt, Grey, Okaru, and Rothberg, African Water Resources: Challenges and Opportunitiesfor Sustainable Development No. 332 Pohl, Djankov, and Anderson, Restructuring Large Industrial Firnis in Central and Eastern Europe: An Empirical Analysis No. 333 Jha, Ranson, and Bobadilla, Measuring the Burden of Disease and the Cost-Effectiveness of Health Interventions: A Case Study in Guinea No. 334 Mosse and Sontheimer, Performance Monitoring Indicators Handbook No. 335 Kirmani and Le Moigne, Fostering Riparian Cooperation in International River Basins: The World Bank at Its Best in Development Diplomacy No. 336 Francis, with Akinwumi, Ngwu, Nkom, Odihi, Olomajeye, Oku.nmadewa, and Shehu, State, Community, and Local Development in Nigeria No. 337 Kerf and Smith, Privatizing Africa's Infrastructure: Promise and Change No. 338 Young, Measuring Economic Benefitsfor Water Investments and Policies No. 339 Andrews and Rashid, The Financing of Pension Systems in Central and Eastern Europe: An Overview of Major Trends and Their Determinants, 1990-1993 No. 340 Rutkowski, Changes in the Wage Structure during Economic Transition in Central and Eastern Europe No. 341 Goldstein, Preker, Adeyi, and Chellaraj, Trends in Health Status, Services, and Finance: The Transition in Central and Eastern Europe, Volume I No. 342 Webster and Fidler, editors, Le secteur informel et les institutions de microfinancement en Afrique de l'Ouest No. 343 Kottelat and Whitten, Freshwater Biodiversity in Asia, with Special Reference to Fish No. 344 Klugman and Schieber with Heleniak and Hon, A Survey of Health Reform in Central Asia (List continues on the inside back cover) WORLD BANK TECHNICAL PAPER NO. 383 Finance, Private Sector, and Infrastructure Network Technology Institutions and Policies Their Role in Developing Technological Capability in Industry Melvin Goldman Henry Ergas with Eric Ralph and Greg Felker The World Bank Washington, D.C. Copyright © 1997 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing November 1997 Technical Papers are published to communicate the results of the Bank's work to the development community with the least possible delay. The typescript of this paper therefore has not been prepared in accordance with the proce- dures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data in- cluded in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denomninations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such bound- aries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissem- ination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which con- tains an alphabetical title list with full ordering information. The latest edition is available free of charge from the Dis- tribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France. Cover illustration by Christopher Chapman and design by Catherine Kocak. ISSN: 0253-7494 Melvin Goldman is senior technology development specialist for Asia in the World Bank. Henry Ergas is a visiting professor at the University of Auckland in New Zealand and an advisor to the Australian Competition and Consumer Commission. Eric Ralph is assistant professor at the Graduate Telecommunications Center at the George Washington University in Washington, D.C. Greg Felker is completing his Ph.D. at the Woodrow Wilson School at Princeton Uni- versity. Library of Congress Cataloging-in-Publication Data Technology institutions and policies: their role in developing technological capability in industry / Melvin Goldman . .. [et al.]. p. cm. - (World Bank technical paper; no. 383) Includes bibliographical references (p.). ISBN 0-8213-4052-2 1.Technology and state-Developing countries. 2. Technology and state. 3. Research, Industrial-Developing countries. 4. Research, Industrial. 5. Technological innovations-Economic aspects- Developing countries. 6. Technological innovations-Economic aspects. I. Goldman, Melvin, 1945- . II. Series. T49.5.T4446 1997 338'.064'091724-dc2l 97-33551 CIP? (COINIENTS FOREWORD ....................................................... iv ACKNOWL]EDGM]ENTFS .v..........................................................................v ABSTRACT ....................................................... vui EXECUTI[VE SUMMAlRY. .......................................................1. CHAPTER ii: APPROACH OF TH]E STU[DY ........................................................7 The Natuire and ][mportance of Techinological Capability .....................7 Focus of the Study ............... . .. 8 Alppiroac]h: How the Study Was Carried Out ...............................................9 CHAPT]ER 2: SURiVEY OV]ERV](EW AND RESULTS ........................................ 12 TI Characteristics ....................................................... 12 Firmn C]haracteristics .............................................................. 13 ]Firmnst External Technlology Sourcing ......................................................... 14 ]Firms' Demand for Technical Services ........................................................ 16 CHAPT]ER 3: ]FIRMS' PATTMERNS O]F USE ............................................................. 18 I[mntroduiction .............................................................. 18 Firms an(d Learning ........................................................................ 19 I][nternral Calpabilities: The Firmts Involvement in R&D ............ ........... 20 Looking Outside the Firm: The Relationship between Ilinteirnal and External Technological Resources ................................... 21 Patteirns of External lU[se: Customers and Suppliers ............................... 22 Govermnment Policies. Services versus Financial Assistance ............. 24 Policy ][mplications .............................................................. 24 CHAPT]ER 4: MA]K]ING TECHNOLOGY INSTITUTIONS WORK ............. 27 ][ntroductioin ................................... 27 W]hat Constitu[tes Good Performance ................................... 27 How T]I Managers Can Improve TI Performance ..................................... 28 Variations by Sectoir and Technology . ........................................................ .o Cooperatives.31 CHA]PT]ER ,: GOVERNMENT RKOILE IN REFORMING TECHNOLOGY INFRASTRUCTURE .............................................................. 34 Goverinment Role in Technology Development ............... ..................... 34 T]he I[rmlpoirtance of a Diffu[sion Orientation .............................................. 34 Diffutsion foir Simall Firms ............................................ ......................... 35 Reforming or Rkestructuriing TX[s ................................................................... 36 ILessons for Governrment Policy and the TI System ............. ................. 39 REFERENCES ...................................................................................................................41 iii ]FOR]EV(O)]RLD) S uccessful industrialization depends on tech fective technology polices? What sorts of technology nological development. Continuous im supporting institutions (TIs) are appropriate, and what provements in technology enable firms, indus- roles can they play? Although there have been company, tries, and economies to enhance productivity and build industry, and country-level case-studies, there has been competitiveness in new areas. Advances in the economic a lack of comparative analysis taking account of varia- theory of innovation, together with accumulated expe- tion across industries with different core technologies, rience of policy makers and practitioners, have shed light industrial structures, and public institutions. To begin on the process of industrial technology development to fill this gap, the Asia Technical Department designed (ITD) and the potential role of supporting policies and and led a multi-country, multi-sector study titled "Insti- institutions. Government's ability to drive accumula- tutional and Policy Priorities for Industrial Technology tion of technological capabilities by firms and industries Development". The project assembled research teams through "public goods" investments in basic scientific in eight economies-China, India, Japan, Korea, Taiwan, research is now seen as limited, and important efforts Mexico, Canada and Hungary-to study how finms in are being mounted to reform isolated and ineffective six industrial sectors have expanded their technological public research establishments. It is also recognized that know-how, obtained support from external institutions, industrial firms must invest in acquiring technological markets, and policy-induced sources, and interacted knowledge and mastery, drawing on stocks of technolo- with TIs. The study involved both historical research gies from abroad and from external sources in their own drawing on secondary sources and the administration countries. Beyond individual efforts in technological of surveys (through interviews and by mail) of firms and learning, firns are supported in important ways by their TIs using the same questionnaires in each location. Its enviromnent including the policies and economic insti- results provide a rich base of quantitative and qualita- tutions that shape firms' incentives to invest in techno- tive information on the policy and institutional aspects logical learning, and provide complementary assets- of ITD. technical skill, information, technical and financial sup- port-to make such investments more productive. Carl Dahlnian Translating these generalities into policy making in Director-World Development Report developing countries remains a challenge. What are ef- July 1997 iv AVCIKN(O)WV]L)E]DGM]ENTS ] F his study required a massive effort from enormous energy and speed. Shekhar Chaudhuri also many quarters. I have tried to note the willingly dedicated himself to providing a host of new many who have provided support, substan- materials. tive as well as logistic. I have surely omitted a few and offer my deepest apologies. Among the advisers, Lewis Branscomb, Professor and former Director of Science and Technology at John There are a number of people whose contributions Kennedy School of Harvard University, Narayanan far exceeded their commitments. Henry Ergas (initially Vaghul, Chairman of ICICI Limited, India, and Fumio an adviser) and Eric Ralph contributed to the analytical Nishikawa, previously Executive Vice President of Toray, framework and analysis. Henry wrote most of Chapter all provided valuable support, as did Charles Davis of 3 and Eric contributed portions. Without Eric's dogged International Development and Research Council of approach to organizing the multi-country data and use Canada. In the Bank, Carl Dahlman and Ashoka Mody of proper econometric tools, the project would be on provided useful feedback. Without Khalid Siraj's sup- more tenuous grounds. Henry's brilliance and willing- port, patience and guidance, as well as that of Carl ness to spend so much time paid off with rich results. Dahlman in the earlier stages, the work would never Greg Felker sacrificed time from his dissertation by pre- have been completed. The idea of a shorter policy-ori- paring parts of Chapter 2, serving as a critic, and pro- ented, non-jargon paper was Dan Ritchie's, and he de- viding enormous support for the related sector studies serves credit for facilitating this study. to be published. Anthony Ody provided support and criticism far beyond any peer reviewer with whom I have Substantial financial support was provided by the dealt. His comments and skepticism led frequently to International Development Research Council of Canada, useful insights and to plugging important holes. Edito- the Overseas Economic Cooperation Development Fund, rial assistance was provided by Bennett Minton. With- the Government of Japan through various trust funds, out him the results would have been even more unap- in addition to the World Bank, and the Industrial Credit proachable. and Investment Corporation of India. Many of the country team leaders took time from For preparing and testing the questionnaires, assis- busy schedules to revise this work or to provide more tance was provided by Shakuntala Gunaratne, Helen information on case studies or data problems. Profes- Garcia and by team leaders and Morris Teubal. Particu- sors Xu of China and Kim of Korea were very helpful as larly helpful were Professors Okada, San Gee and Rejean were Professors Wang and Ms. Lu of China, Professors Landry. Word processing support was provided by Dixit and Ramnarayan of India, Inzelt, Gyorgy and Gladys Stevens, Maria de Vera and Merced Doroteo, but Havas of Hungary, Yoon of Korea, San Gee of Taiwan, it was the dedication over many years in organizing and and Kojima, Watanabe, Maeda and Ohmori of Japan.All word-processing of Alloysius Ocheni that is particularly responded to requests for additional sector studies for noteworthy and without which we would not be at this publication. point. Yoshitaka Okada provided feedback, support and The team leaders and principal researchers con- ideas, beyond all reasonable expectations, and all with sisted of: Canada: Rajean Landry (Project Leader); v China: Xu Zhao-xiang and Wang Huijiong (Co- Charles Park and Se Joon Yoon; Mexico, Kurt Unger Leaders), Tang Siguo and Lei Wu; Hungary: Annamaria (Project Leader), Javier Jasso, Victor Paredes, and Arturo Inzelt (Project Leader), Katalin Gyorgy, Attila Havas and Torres; lTaiwan: San Gee (Project Leader) and Cheng Judith Mosoni-Fried; India: Shekhar Chaudhuri (Project Ying-chen. Many assisted each of the teams and we also Leader), .M. R. Dixit, S. Ramnarayan, C. B. Dasgupta, acknowledge their assistance. Naresh Chotai, Arijit Skikdar and Anil Yadav; Japan: Yoshitaka Okada (Project Leader), S. Watanabe, Y. Melvin Goldman Maeda, Sakura Kojima, M. Wada, T. Ohmori and N. Washington, D.C. Tanaka; Korea: Kee Young Kim (Project Leader), Sejo Oh, July 1997 vi AIB,STRA(CT ] F he importance of developing and dissemi pan, TIs can meet the needs of SSI in traditional sectors nating technology for industrial develop where they dominate. Financial and other incentives fa- ment and competitivenes is widely recog- vor large enterprises even more heavily. nized. Companies have to build technological capabili- ties themselves but also need to source technology out- The most successful TIs are those which know their side the firm from customers and suppliers to technol- clients, understand clients' needs, upgrade their knowl- ogy institutions (TIs). This study of six industries in eight edge regularly and maintain incentives related to their economies examined the role in building industrial tech- mission. They complement other sources of technology nological capability played by outside sources particu- usually helping in the absorption, adaptation, improve- larly technology institutions (TIs) such as research and ment or development of technology only in part with standards organizations and by directed policies such the rest done by the firm, perhaps with imported knowl- as tax incentives. It concluded that TIs have made a pro- edge or with ideas of customers or suppliers. TIs cannot found difference on industrial technology development substitute for a good business environment, a stable where there are sufficient imperatives such as a com- petitive environment or manufacturing for an overseas Governments can stimulate the building of TIs, de- quality-driven market pushing firms to improve their velop TIs and programs supporting SSI and ensure that technology. TIs were particularly valuable for economies diffusion of known technologies is emphasized. But they and in sectors where their role is to diffuse technology should not run TIs. Where TIs do not support industry, to help catch up technologically. they should be restructured to do so, modifying leader- ship, financial systems, structures and incentives. Gov- TIs and financial incentives have proved less ad- errnments need also to ensure an adequate flow and stock equate in helping technologically less sophisticated and of high level technical manpower to staff TIs as well as smaller finns. Most countries have neglected traditional private firms. Financial incentives need to be targeted small-scale industry (SSI) and many TIs are not appro- to market deficiencies, such as the use of TIs by SSIs and priate for communicating with enterprises lacking tech- avoid substituting for what firms would have done in nical capability. If properly targeted, however, as in Ja- any case. vii ]EX]EC1U[TI[V]E SUU[MMAJY What Is the Study about? economy of the six sectors studied (three each for Hun- gary and Canada). The sectors were selected to include This paper summarizes the findings of a study a range of characteristics related to science and technol- that aimed to determine the key characteristics of tech- ogy intensity, industrial structure, economies of scale, nology support institutions (TIs) and technology poli- young versus mature industries, and kinds of custom- cies which support more rapid growth of technological ers. Study teams in each economy used the same instru- capability in industry. These characteristics were ex- ments to survey firms (about eighteen interviews in each pected to vary depending on: the nature of the industry sector plus a random mailed survey) and TIs (up to ten (or sector), including the industrial structure; how rap- interviewed in each sector), and also analyzed the evo- idly technology is changing and the degree to which it lution of the sectors, particularly as regards technologi- is science-based; the country endowment, including the cal and institutional development. These analyses were quality of training and its educational and institutional complemented by case studies of interactions between culture; and the characteristics of firms and their capa- firms and TIs. bilities, as embodied for example in technical manpower, facilities and management attitudes. TIs' characteristics What Do Firms Seek? that were expected to be important were how TIs are organized and operate, including their internal incen- The surveys confirm an intensive use of outside tive structure, management background and modus support for technological improvement, and the range operandi, and approach to services and income genera- of these sources is varied. Users consider them tion. The study attempted to determine the efficacy and complementary to other external, as well as internal, relative importance of various technology policies, in- sources rather than as substitutes. If a firm has its own centive measures, TIs and other sources of technologi- technical resources, personnel or an R&D department, cal know-how. it is far more likely to use outside technological sources, particularly public and private institutions. Larger firms This paper also examines how various firms in are more apt to use TIs than small firms. Over 80 per- different sectors and countries improve their technology cent of firms with an in-house lab and more than 350 to increase productivity and product quality and develop employees use at least one public TI compared to 31 per- new products and processes. It seeks to understand the cent of firms with fewer than 50 employees that do not roles of different participants in the innovation process, have an in-house R&D lab. (See figure 1.) including firms' engineering and R&D departments, cus- tomers, suppliers, foreign licensors, and private and Customers, followed by suppliers, are the most public technology institutions. used external sources of technology, but the intensity of use of publicly supported Tls is more surprising. When Six economies--Japax., The Republic of Korea, the various kinds of public TIs are lumped together, more The Peoples' Republic of China, India, The United Mexi- firms have used at least one public TI than either a cus- can States and Taiwan (China)*--were the subject of full tomer or supplier as a source of technology. There are studies. In Hungary and Canada study was more lim- major country differences, however; intensive use is ited in scope. Although each economy is at a different made of Ts in China, Japan, India and Korea. In Canada, stage of development, all have sizable industries in each least use is made of TIs even after controlling for the Hereafter, these economies will be referred to in a shortened form, e.g. The Peoples Republic of China will be referred to as China. 1 2 much smaller Canadian average firm size. In part, this used according to TIs as well--this is true even when variation reflects differing historical approaches toward considering only Tls that carry out R&D. The "demand" learning and independence-what might be described as for diffusion of known technology is not surprising. In cultural differences, but it also relates to the supply of so many industries technological know-how is so diverse services. China and Japan have each built many tech- and specialized that individual firms cannotpossibly be nology centers, helping to explain high usage. current in all fields (unless they have a large technology support group of their own and are well structured to The overwhelming demand by industrial firms use it--which is all too rare). Most are aware of only a is for services related to what might be called diffusion, small fraction of the possibilities for innovation using that is, the transfer and application of known technol- the existing stock of technological know-how. The small- ogy. Firms most frequently use services related to infor- est firms in mature industries have little capability to mation, standards and testing, problem solving (and identify their technical needs or questions, let alone trouble shooting) and technology training. And even search for answers. when they use R&D services of TIs, they tend to want answers to particular questions, rather than the devel- Some needs can be satisfied through normal opment of entirely new technologies. The study un- market channels or, as in Japan, through keiretsu -intri- earthed few examples where firms purchased off-the- cate private sector channels of cooperation--or through shelf, self-contained technologies developed by Tls. This cooperative associations that may include independent is in stark contrast to the claims and publicly projected firms across keiretsu lines. But even in Japan, where such image of many TIs, particularly R&D labs. private mechanisms are extensive, and where private research firms and consultants also have recently devel- Not only do firms say they use diffusion ser- oped, government still supports many areas where the vices most, but those services are also the most frequently private sector does not. While less important and per- Figure 1. Use of Public Tis by Firm-s Izo and In-house lab I department 90 . 80 F. 70- 601 200 ID 20 Sma1 (1-50) Mudk,i (51-300) Laq. (351k) EWithout in-house lwiWith in-house lia Akk rubrb .1, rUSb.gi, edoohgy 005*, hDsJ S& G000- 1f Institutions and Policiesfor Industrial Technology Development 3 vasive in another advanced country, Canada, external and investing in developing capabilities in anticipation support generally, and from TIs in particular, is impor- of future needs. tant in technology improvement. Guidelines for management will depend on the The study does not suggest an ideal model for domain of the TI--its market and the role it wants to play. cooperative inter-firm relationships, but it does empha- The mix of activities to meet clients' current demand and size the contribution such interaction can make to tech- investment in building capacity for their future need, nology upgrading as well as ways that it might be en- for example, will vary by the type of service organiza- couraged. Business practices that spur cooperation be- tion-standards and testing laboratory, training institu- tween supplier and customer, and policies that encour- tion, R&D institute or multipurpose center; the nature age companies to learn from their domestic and over- of the sector and clients--whether its clientele is princi- seas customers and suppliers (and more indirectly to ex- pally large-scale or small; and their degree of techno- port), clearly enhance a country's accumulation of tech- logical sophistication. The education and experience re- nological capability. Firm interaction also can be fos- quired of key technical staff depends on the clients and tered through TIs. A number of cooperatively formed the services. TIs in Japan and one in India illustrate the potential for sharing technology as well as the delicate nature and Staff quality must be technically excellent for the importance of timing in any cooperative effort. Here TI's tasks, with a level of expertise well ahead of its cli- too, government canbe a catalyst. But these efforts work ents. But expertise is not enough. TI staff must interact more effectively in certain sectors--textiles for one--that with industry regularly to know their needs and how to have many medium-scale firms and whose technologi- communicate with firm staff. Furthermore, the TI should cal needs are not so advanced or closely tied to develop- organize its incentives systems--particularly criteria for ing science. In science-based sectors like polymers, sev- staff advancements (promotions, degree of autonomy, eral publicly supported research labs in India have pro- increasing responsibility and increased compensation)- vided industry more effective support than have the to encourage individual staff to work for the goals of industry-run cooperatives in Japan. the organization, but also to ensure cooperation and teamwork. Characteristics of Effective TIs Many other characteristics contribute to perfor- mance: systems for confidentiality, signals as well as Most of the elements that go into a well-perform- incentives to staff, and varying approaches to bringing ing TI are not especially different from those in a well- in new ideas and blood. Whether a TI is owned by gov- managed firm. An effective TI knows its market, under- ernment or the private sector (or whether the CEO is a stands its capabilities and limitations, matches the quali- world-renowned researcher or an industrialist) are not ties and skills of its staff to the needs of the market, un- important determinants of performance. That is not to derstands where the market is going and can adjust to say that the CEO is unimportant or that government has it, and balances advertising its know-how and guarding no role; in fact, both are critical. Government normally its clients' confidence. A TI must identify and study its must contribute to the institutional setup, and ensure market and know its clients intirnately, including their that industry supports an infrastructure that helps raise current desires and likely future needs. This implies that the technological capability of many firms, not just a few TIs need to interact with their clients, know their prob- leaders. Similarly, the leadership must set the tone of lems, and identify what they need. They must then de- the TI, get the incentives right and emphasize the val- termine what they need to learn and seek to achieve the ues of a service orientation and technical excellence. A right balance between responding to customer demand technical and industrial background is likely to help a 4 CEO set the tone; TIs having a CEO with industry expe- rience have on average nearly 20 percent more indus- trial revenue. is in building basic technological and industrial skills and with liberalization it is reaping the benefits. Government's earlier promotion of education and indus- Incentives and Policies for Technology trial development--the building and operation of small Development and large-scale factories in the private and public sec- tor--provided the experience and developed the skills Although the broader policy framework was not needed for today's rapidly growing and increasingly ef- the study's principal focus, the survey results and the ficient industrial sector. All successful examples have sector studies shed light on what constituted effective had to pay a temporary price of inefficiency in order to policies. First, where technology incentives--such as learn. technology (or R&D) tax incentives, loans, subsidies, protection--are not specifically directed at small-scale firms, they have tended to be more concentrated and Government's Role in Technology Development regressive by firm size than technical assistance by TIs. One-third as many firms as use a TI use any incentive. As with its role in industrial development gen- And these findings apply across different types of in- erally, government's role with respect to TIs is, first and centives and countries. foremost, to create an appropriate environment for tech- nology acquisition and development, and to make avail- Historically, government interventions to guide able the right incentives. That means that a local indus- or control technology acquisition and to license or limit try (regardless of ownership) needs to operate in a com- production capacity have had mixed results. Japanese petitive environment domestically and/or abroad to en- foreign exchange restrictions and controls on foreign sure concern for product quality and range and efficiency technology acquisitions in the two-and-a-half decades of production. In turn, that creates demand for TI ser- after the Second World War, pushed firms to cooperate vices. for licensing, diffusing and developing technology In some sectors, notably casting, whose technology The appropriate environment includes an edu- changed slowly and which had the skills to absorb tech- cation system that produces high quality technical man- nology, the effect was dramatic. In polymers, whose tech- power and technology infrastructure with equipment nology is more science-based and changes rapidly, Japa- and people that can support industry. Government's role nese government policies to constrain licenses, diffusion is to pro-mote both, but not to manage them. Where and capacity led to inefficiency and relatively slow tech- needed Tls do not exist, government might stimulate in- nology development. Indian government policies across dustry, an educational institution, or a private group to the industrial sector from the late 1960s to the 1980s simi- form them, or start them itself. Government can offer larly retarded many sectors. More positive, but still the right incentives and financial assistance to stimu- mixed results, occurred in Korea. But what distinguished late growth and service to industry. Korea and Japan from India was the intense competi- tion, both domestically and for export markets, among Where an infrastructure exists but is not provid- Japanese and Korean firms, while in India, until recently, ing support, government is frequently part of the prob- exports took a back seat and competition was discour- lem: in particular, the way it finances TIs and its fre- aged through capacity licensing and excessive protec- quently close control over TIs' activities. There are suc- tion. Many firms in all three countries, however, also ben- cessful examples of institutional transformation high- efited from the policy of induced learning, resulting from lighted in Chapter 5 (Boxes 7, 8 and 9). Common ele- the need to be innovative and to reverse-engineer in or- ments in successful turnarounds are "hard budget con- der to compete. Where India appears to have succeeded straints", leadership as well as a clear vision of the institu- Institutions and Policiesfor Industrial Technology Development 5 tion's role and procedures, including incentives and pro- A contrasting approach, which covers a smaller grams for industry. percentage of the SSI population, is the Taiwanese pro- ductivity center. The center is an SSI model of a TI that An important role for government relates to develops technological expertise in a generic area with small-scale industry (SSI), which often receives almost wide applicability to a range of SSIs. The Taiwan center's no support. Only Japan maintains a network of TIs dedi- expertise is industrial engineering and automation. Its cated to SSIs, and these services are subsidized. It is a services are well regarded, but they are used by only a mistake to tack on support for SSIs to general TIs whose few SSIs. This model is followed somewhat in Korea principal clienteles are larger more sophisticated firms. and is being adapted and spread in parts of China. The SSIs have different needs and lack the technical person- center's services are heavily subsidized and the TI is nel to communicate with sophisticated technical staff of wholly Government-owned. This model too can be most TIs (hi-tech firms are different). TIs that focus on adapted for different environments by reducing the level SSIs need to expose the companies to the benefits of of subsidy and the role of government in management. change and generate a demand for technology improve- ment, as well as resolve problems brought to them. Conclusion Two approaches for supporting SSIs seem to be effective. The Japanese approach is decentralized and Technology policy and institutions are impor- directed at geographic clusters of industries. TIs are sup- tant tools to help countries industrialize. But there are ported by prefectural (county) government, and their no simple solutions for targeting such policies, and some primary purpose is to support local SSIs. TI staff work of the easiest to create, such as R&D and tax incentives, with their clients in many ways and most work relates may be the least useful. Ultimately, there is no substi- to diffusion. The clustering of related industries in Ja- tute for independently functioning TIs, which respond pan permits regional institutional specialization in par- to industrial demand and look to their clients rather than ticular sectors. This approach has wider replicability Government for their normal operating expenses. TIs since clustering is common in many places, although per- which perform well for the economy are technical and haps the model may need adaptation to function with a social demonstrations of good management practices. reduced Government role in countries where public in- They know what their industrial clients demand and lead stitutions are weak. them to what they will need. (C]H[A]P'TER iL APPR(O)ACIH[ O)F T]H[]E STlUIDY This report examines the use that firms in a group of developed and developing economies make of O cope with breakdowns, unexpected outages, external sources of technical support in product and pro- deficiencies in layout and scheduling, and unex- cess innovation. It pays particular attention to the role plained changes in product or process performance; in the innovation process of technology support institu- and tions (TIs)--universities and technical colleges, private consultants and contract R&D providers, as well as pub- O adjust to process and product improvements lic bodies such as central and local government labora- when these occur. tories, industrial extension services, and productivity centers. The ability to master these skills is the central component of industrialization. The payoffs include the capacity to implement existing processes and to adjust The Nature and Importance of Technological to the changes in supply and demand that are the hall- Capability mark of dynamic, open economies. Yet the extent of this ability varies greatly among countries, with some--such The acquisition of technological capability is as the fast growth economies of East Asia--rapidly up- essential for industrial growth. "Technological capabil- grading their skills, while many others fail to implement ity" is not synonymous with the ability to develop en- even the simplest industrial processes. Those who can tire technologies, the "breakthroughs" that play so vis- thrive grow faster under the impetus of "learning by ible a role in advanced economies. Rather, it is first and doing," while the rest fall farther behind. Strengthen- foremost the ability to learn, to master progressively the ing the ability to learn is consequently a central chal- many elements involved in producing industrial goods. lenge for public policy. For almost all developing countries, this entails absorb- ing technologies that are in widespread use elsewhere. Economists have traditionally viewed firms' in- Absorption is neither easy nor passive. Considerable novation as depending mainly on their internal capa- skill is required, for example to: bilities and efforts, a view reflected in the emphasis in - conventional analyses on in-house R&D, and the relega- O identify the technologies that should be used, tion of external support to secondary status. The diffi- and design and build the facilities in which they are culties of contracting for technology have been viewed embodied; as limiting firms' reliance on outside for technical sup- port, and external reliance has been seen as only a step 4 adapt facilities, processes and products to the toward full internalization of R&D. Policymakers and peculiarities of local raw materials and other inputs, industrialists have long challenged this view and econo- factor prices, and market demand; mists now have begun to do so. Three elements are im- portant in this respect. O train personnel; First, current analyses of the innovation process O ensure that plant and equipment are main- place great emphasis on spillovers--on the interdepen- tained and that quality standards are enforced; dence of firms' innovation efforts. These effects occur 7 8 through a range of means, including observation by each These newer views of the innovation process, firm of the innovation efforts of other firms, the publica- though s(een in the academic literature, have not sub- tion of research, conferences and colloquia, the transfer stantially affected public policy. This partly reflects the of skills and knowledge between customers and suppli- limits of the work carried out to date. The analytical ers, and the movement of staff among firms. Though it framework is often poorly specified and incomplete; little is conventional to view imitation as a relatively easy if empirical research has been done, and even less has been not costless process, these spillovers do not occur auto- done to translate the perception that "institutions mat- matically; rather, they require purposive action by the ter" into institutional design. It may also reflect the domi- firms involved, most notably investment in the people nance of policy makers and advisors more concerned, and systems needed to learn. Firms' capacities to exploit out of fiscal necessity or ideology, with cutting back Gov- these investments clearly affect their technological per- ernment involvement. formance. Empirical studies of the characteristics of effec- Second, the extent and pattern of these spillovers tive TIs and policies have followed one of three courses. is increasingly viewed as depending not solely on the A common approach is analyzing TI cases in one or more efforts of firms themselves but also on their institutional countries with or without an overall institutional frame- context. Recent analyses of "national systems" of sci- work (see Arnold, Bessant et al., 1993, and IDRC, 1993). ence and technology (see para. 1.10) highlight the role Another approach is to analyze the experience of a group of social institutions in determining the capacity to in- of TIs in one country. This is commonly carried out on novate. Some of these institutions, such as national re- behalf of policy makers eager to restructure or reduce gov- search laboratories and standards-setting bodies, pro- ermient contributions to the technology infrastructure mote the development and diffusion of technology; oth- (notably commission reports in England and India such ers, such as educational institutions and the institutional as Lord Bessborough, 1972, and Abid Hussain et al., 1986). infrastructure of capital, labor and product markets, may The third approach is to examine technology policies and influence the ability and incentive to innovate, just as institutions as an element within the overall science and they shape other aspects of economic behavior. technology system of a country or countries (Nelson, 1992, and Dahlman, 1989,1990). While many of these provided Third, it is increasingly recognized that exter- insights arLd hypotheses for this study, they do not have a nal sources of technological capability may be especially common approach with an underlying framework to be important in industrializing economies. Being further tested across a range of countries. Few have dealt with from the technological frontier, firms in these economies the organizations and incentives of institutions, and none can most readily learn by observing the behavior of their looked at sectoral differences. One study (Goldman, 1994), more advanced competitors and adapting it. They can which examtined four sectors in three European countries, also draw on the expertise of suppliers and customers, served as a pilot for this study. (A range of works con- who can help identify best practice. Finally, innovation sulted for the study is in the bibliography. The list is not requires the mobilization of a range of assets, many at exclusive and some materials undoubtedly and unknow- least partially external to the firm (for example, people, ingly have been omitted.) sales and distribution networks, and suppliers). Pro- ducers in developing economies are most likely to be constrained by shortages of these complementary assets, Focus of the Study and henc'e to depend for their long-term technological performance on public and private action to increase the The study's aim was to determine key charac- availability of those assets. teristics of TIs and technology policies that support growth of technological capability by industry. These Institutions and Policiesfor Industrial Technology Development 9 characteristics were expected to depend on: 4 Korea has had an outward-looking industrial- + the nature of the industry (or sector) includ- ization strategy for more than three decades. While ing the industrial structure, how rapidly tech- encouraging firms to compete on export markets, nology is changing and the extent to which it is the government for many years promoted a highly science-based; concentrated industrial structure, based on a small number of large, diversified conglomerates. 4 country endowments, including the quality of training and education and its institutional C0i Hungary was an early Eastern European culture; and industrializer, but from the Second World War until recently it was centrally planned. It was again the 4 the firm and its capability as embodied, for earliest of Eastern European countries to initiate example, in its technical manpower and facili- market reforms, the effect of which was traumatic, ties and management attitudes. since firms no longer had guaranteed markets in Eastern Europe and had to compete with sophisti- TI characteristics that were expected to be impor- cated products from the West. Only certain indus- tant included TI's organization and operations, man- tries survived the adjustment, including a few that agement background and mode of operations, and have increased exports. approach to services and income generation. The study attempted to determine the efficacy and im- 4 Mexico long followed import substitution poli- portance of various technology policies and incen- cies similar to those of many large developing coun- tives, as well as types of TIs. tries. The 1982 crisis led to adjustment policies that have resulted in industrial readjustments. Its prox- imity to the US has a complex effect on institutional Approach: How the Study Was Carried Out as well as industrial development. Most research on the innovation process has 4 Taiwan has long had an outward approach to examined firms in industrial market economies, industrialization. Though large, often publicly particularly the United States and the United King- owned firms have been significant in heavy indus- dom, but the questions here require a broader view. tries, small to very small enterprises have accounted This paper therefore examines a sample of econo- for most industrial output. Many are family-owned mies--India, China, Taiwan, Korea, Japan, Hungary, and many also generally highly export-oriented. Mexico and Canada--spanning a range of levels of development. They also have had differing ap- 4 In Japan, as in Korea, conglomerates are impor- proaches to industrial development: tant, though they are generally looser networks of affiliated enterprises, and they draw on small and 4 China has been undergoing a rapid transi- medium-sized secondary suppliers. Their flexibil- tion from a centrally planned economy to one ity has been central to the adjustment capability of in which market mechanisms are dominant. an economy that is intensely competitive and trade- oriented. 4 Although now in the process of liberaliza- tion, India's economy has long been subject to 4 Finally, Canada, though it has had substantial extensive regulation and protection, condoning tariff protection for many years and restricted for- inefficiency and leaving large parts of industry eign investment, is integrated into the North Ameri- inward-looking. can market. Firms have achieved productivity lev- 10 els that are very high by world standards. An important caveat. All of these countries al- 4 A desk analysis, reviewing the evolution of ready have a substantial industrial sector that reflects technology, industry structure and technology poli- different stages of industrial development. Historical cies and institutions overall and in each of the sec- analysis of sectoral and institutional development has tors. Interviews were conducted with industry as- been carried out in most of the countries, which along sociations, policymakers and academics, to provide with statistical and case analysis helps explain their evo- background on industrial, trade and technology lution as well as outcomes. This study does not, how- policy. ever, pretend to address the case of less advanced de- veloping economies with only a nascent industrial sector. 4- Interviews of a sample of about eighteen firms in each sector representing a cross-section of the in- This study makes use of common survey instru- dustry in terms of firm size, technological dynamism ments in the countries and across the same sectors in and Location. The interviews aimed to understand each. Countries were selected to ensure a range of policy how firms improve products and processes and the environments and industrial structures, as well as varia- conbtibution of various actors and policies includ- tion in achieving rapid technological and industry de- ing (and particularly) TIs. velopment. The existence of a sizable industrial sector was important to permit a big enough sample of firm 4 Interviews with up to ten relevant TIs in each interviews for statistical comparison. sector in each country to understand their services, structure (including incentives), staffing, funding, Sectors were selected to ensure significant pro- and perceived roles. duction and a range of firms in each country. The com- bination of sectors also includes industries built on dif- 4 A questionnaire survey mailed to randomly se- ferent sorts of technologies--some mature, others emerg- lected firms to check the interview results (on a non- ing. They have variety in the following characteristics: purposive and larger sample) so as to expand the science or technology based, structure, economies of sample to permit analysis of a wider range of char- scale, young and dynamic versus mature, and typical acteristics. customer (the pubLic, oligopsony). 4 Finally, case studies were done in most sectors The six sectors are: foundries, textiles, auto parts, of selected technology transactions between TIs and machine tools, software and polymers. All six were stud- firms to understand the process of technological sup- ied in Japan, Korea, China, India, Mexico and Taiwan; port and the relationships of participants in a tech- only three sectors were studied in Canada and Hungary. nology support endeavor. In addition to these, pharmaceuticals, electronics, and textile machinery were studied in some economies to Analysis was carried out at the country level and provide complementary insights. Pharmaceuticals were on a cross-national basis. The country reports are supple- surveyed in India and Hungary, textile machinery in Ja- mented by an analysis of the data combined from the pan, and electronics in Korea and Taiwan. various countries, which permitted hypotheses to be tested statistically, helping assure the validity of the re- Country (economy) teams all followed the same sults. For example, the quantity and importance of de- approach and used identical pretested interview and mand for varying kinds of technological services is ex- mail questionnaires designed by Bank staff and refined amined, riot merely for different kinds of firms of vary- with team leaders based on pilot testing. Each team con- ing technological capability, but also in the same sector ducted the following activities: across coumtries at different stages of technological de- Instithtiotns and Policiesfor Industrial Technology Development 11 velopment. case studies were used to deepen understanding and illustrate conclusions. We have approached the information from The results of the eight country studies are presented three perspectives. The country and sector reports, in Chapters 2 to 4. Chapter 2 provides an overview of the including historical studies, help identify trends and quantitative results, and Chapter 3 concentrates on what issues to be analyzed statistically. Second, statistical firms seek and also explains firms' use of internal and ex- analyses test hypotheses about TI performance and ternal sources of technology (including TIs) as well as of firm demand. The 167 TI, and 704 firm interviews, technology incentives and policies and discusses country and 2049 completed firm questionnaires were the and sectoraldifferences. Chapter4illustratesbycaseand basis for the analysis. Nearly all firm data cited in analysis the characteristics of successful technology insti- this report are from the mail survey. To give indica- tutions and policies. An attempt is made to identify what tions of the findings, results of regression analysis constitutes good performance and to advise how to are presented throughout this paper. The reader achieve it. Also discussed are the roles of various actors should view these as indicative. Any particular nu- in building an effective institutional and policy environ- merical estimate should be understood as the mid- ment for technology development. Chapter 5 attempts to point of a likely range of outcomes. Unless other- present a vision of a successful technology policy and in- wise indicated, such estimates are significant at the frastructure for a middle-income developing country. It 95 percent confidence level. Similarly, when a char- presents a summary of the role of government and on acteristic is cited as an indication of effectiveness of the need for and approach to a diffusion and small-scale a TI or a predictor of a firm's action, this is also sig- industry orientation. It also offers cases and suggestions nificant at the 95 per cent confidence level' . Finally, on how to transform TIs that use excessive government resources and have little effect on the economy. Examples of successful transformation illustrate the possibilities. 1 Cross-sectional analysis, which forms the basis of the statistical work of this study, is fraught with difficulties. First, the complex underlying structure of the relationships to be examined are generally unknown. While the models that formed the basis of the analysis were built on the best of available information, it would be serendipitous if they should coincide with the relationships examined. Second, even if accurate models could be developed, relations in cross-sectional data are extremely complex and involve many variables. While data availability generally limits the number of variables that may be considered, thereby leading to model mis-specification. For these reasons, the results produced in the paper have been subjected to a number of robustness tests, including comparisons among different models, as well as application of different statistical techniques to the same data and model. The following rules have, however, been generally followed. When the variable to be explained was qualita- tive, but scaled - e.g. degree of competition from 0 (none) to five (high) - ordered logits or probits have been used; when the explanatory variable involved discrete outcomes (ownership structure) multinomial regressions (logit and probit) were used; when the explanatory variable was censored, but otherwise quantitive, a maximum likelihood tobit estimator was used. CIHEA]PT]F]E]R 2, SUIV]KVEY OVERVIEW AND IRESUILTS TI Characteristics The survey' of TIs covers inter- views conducted of 167 institutions us- ing identical questionnaires (Table 1).2 Figure 2 Reliance on Industry Contract Revenues and TI Seventy-two percent of TIs were orga- Dynamism by TO Type nized mostly along sectoral lines rather than by services, technical discipline, or 3.5 50 interdisciplinary expertise. Ownership 45 structure varies greatly, with public TIs 3 40 predominant in China, followed by Ko- 2.5 35 ¢ rea and India. Japan has the largest share 0 2 of private TIs. Japan and China have 25 huge numbers of TIs, so that the sample . 20 there represents a small percentage of the 1 __ relevant TI population, while in Korea 0.5 10 a_ and Mexico almost all relevant TIs were visited. Private for-prol¶i G*vaenrnmer NofrPrrad & coops Universdije TI-typo The share of operating budget revenues derived from contracts with in- Source: Global survey results. dustry is an important measure of TI per- formance, and it varied significantly by country. Japa- had the hlighest average reliance on contract revenues, nese TIs had the lowest reliance on contract revenues. in part because of decade-old measures to reduce un- However, the variance was high, suggesting that a strong tied budgetary support to TIs. Figure 2 groups Tls into subsidy element in technology services need not frus- ownership categories to compare their reliance on in- trate incentives for TIs to work with industry. China dustrial contracts and the interviewers' composite esti- Table 1 Sample TI Characteristics by Country China Mexico Korea Japan India Number in sample 46 17 16 62 26 Mean year of founding 1972 1981 1979 1953 1965 Percent of sample Tis that are sector-specific 78 53 56 79 65 Mean staff 416 88 419 92 439 Mean percent technical staff 46 66 42 71 33 Mean percent of contract revenues (1992) 57 41 56 23 39 Source: Global survey result. 1 Unless otherwise stated, all tables, graphics, and data cited result from the analysis of the survey in the various economies and are cited as Global survey results. 2 Data for Tis is not presented for Canada, Hungary and Taiwan, because it is not available in a comparable form. 12 13 mate of the TI dynamism, scored on a scale of one to linked TIs on average rank highest. There are variations five. The dynamism score was based on the TIs' extent from these averages including many dynamic private and ways of keeping abreast of and building capability TIs, particularly in Japan, while many government or in new technologies and incorporating environmental university TIs have high shares of earnings from clients. and industrial needs into their work. As expected, pri- The TI characteristics that lead to successful performance vate for-profit TIs generate the highest percentage of rev- are treated in Chapter 4. These results do suggest that the ownership of the TI is far less impor- tant to performance than might be as- Figure 3 Reliance on Industry Contract Revenues and sued Furter,othene to genrae re- TI Dynamism by Client Size sumed. Further,theneedtogeneraterev- enue can encourage Tls to be industry-ori- T4.. - 35 ented, but is not necessary for promoting 3'5 ~ ~ e9 . - 30 an industry orientation. -25 Figure 3 shows the results found when TIs are grouped by size of client firm (num- E, 2- * ber of employees). About 20 percent of _ _ -1S TIs reported small firms as their primary. s- 10 t~~~~~~~~~~~~~ clients, and a simidlar number served me- 1 -lo diurn-sized firms, in contrast to the 60 per- 0.5 - 5 cent of surveyed TIs whose primary clients were large firms. This result corresponds Small (1-50) Medium (51-300) Large (301+) to a basic finding of the surveys of firms, Size of Firm Clientele namely that large firms tend to use TIs more extensively than do small and me- Source: Global survey results. dium-sized ones. As expected, the share of contract revenues in the budgets of TIs enues from industrial contracts and universities the low- serving large firms was twice that of TIs serving small est. This indicator, however, correlates poorly with the firms. TIs serving medium-sized firms, however, were dynamic score, in which government- and university- judged most dynamic. Table 2 Firm Mailed Survey: Sample Size by Sector and Country Sector China Canada Korea Japan India Taiwan Total (China) Polymers 35 39 29 25 6 22 156 Machine tools 36 19 35 44 10 26 170 Auto parts 35 16 34 64 35 94 278 Textiles 35 20 33 35 25 62 210 Foundry 36 13 35 51 8 98 241 Textile machinery 0 2 58 27 0 8 95 Software 35 4 0 42 27 117 225 Pharmaceutical 0 19 0 1 21 16 57 Electronics 0 52 42 2 0 187 282 Other 0 214 0 31 0 89 335 Total 21 2 398 266 322 132 719 2049 Source: Global survey results. 14 Firm Characteristics Figure 4 Mean Number of Employees In the mailed survey of 2,049 6000 T firms (Table 2, below), Taiwan had the largest sample size followed by Canada, so00 and India had the smallest. In five of 5 0 the six sectors (polymer, machine tool, auto parts, textiles and foundry), ques- 4000 tionnaire responses were fairly evenly Ns 0. distributed across countries. The aver- E age age of firns reflects the timing of 0 30000 each country's industrialization, as well E as the pace of firm entry and exit. 2000 t Japan's firms are on average the oldest, followed by China, Canada, and India. 1 i - The youngest samples of firms were in 1000 hX Korea and Taiwan. Textile firms have KeCaiea the oldest mean age among individual 0 - - Japan sectors, while software and electronics Textile Polymer Mach Tool Autopart Foundry Taiw Sector firms are the youngest. A basic indica- tor of industrial structure is provided by Taiwan average firm-size as measured by num- C h in a Canada Korea Japan (China) ber of employees. China's firms are the Textile 267 1 52 153 502 428 Polymer 2636 65 1 237 3437 292 largest by far, registering the importance Machine tool 1773 23 838 249 1 63 of large state-owned enterprises in the Auto parts 1817 307 131 721 385 country's industrial structure as well as Foundry 712 164 201 406 240 in the sample's coverage, and perhaps Source. Global survey results. China's low ratio of labor to other costs and over staffing compared to other countries in this Japan's firms have the lowest average R&D intensity of sample. Taiwan's small average firm size reflects the the countiies surveyed, although this partly reflects the predominance of small and medium-scale firms in its large sample and high response rate in low R&D-inten- industrial structure. Japan and India have a mix reflect- sity sectors like foundries and auto parts. As expected, ing their industrial structure. Figure 4 and its corre- software was the most R&D-intensive sector, and poly- sponding table present a picture of the average size firm mers also reported relatively high R&D ratios. Techno- by country and sector. logically rnature sectors such as foundries and textiles have low R&D intensity, as do auto parts, where firms R&rD intensity (R&D expenditures as a percent- rely more heavily on technology from principal buyers. age of sales) is a measure of firm investment in techno- logical innovation. Canada's sample had the highest average R&D intensity (8.1 percent). Korea and Taiwan Firms' External Technology Sourcing also were above 5 percent. These high rates reflect the technology-intensive nature of sectoral coverage, and Firms' use of TIs and technology services high- perhaps selection effects--more firms conducting R&D lights several important issues. Firms were asked about may have chosen to respond to the questionnaire. their interaction with a comprehensive range of technol- Institutions and Policiesfor Industrial Technology Development 15 ogy sources, including public and private TIs as well as differences in size distribution of firms, as well as the formal and informal sources of technology support. An level of sophistication and rate of change of their respec- important issue is the degree to which firms rely on pub- tive core technologies. Figure 5 illustrates TI use for four lic or noncommercial institutions, in contrast to private sectors--polymers, auto parts, foundries, and software contractual sources of technology support or links with as well as the total sample. Firms in the science-based their major customers and suppliers. This question was polymers sector reported comparable and high use of addressed by grouping the technology sources into four various technology sources, while auto parts firms cited categories: (a) public TIs, including national institutes long-term customers or suppliers far more often than and industry or research associations; (b) private con- private TIs or in-house research units. In the techno- tractual sources, such as foreign licensers and for-profit logically mature foundry sector, 80 percent of firms used contract laboratories; (c) technical assistance from long- a public TI, surprisingly the highest percentage of the term suppliers and customers; and (d) firms' internal three sectors. R&D or technology departments. When we separate public TIs into national TI, Striking was the magnitude of firms' use of for- local and regional TIs, and universities, and also con- mal public or quasi-public technology institutes. More sider suppliers and customers separately, the order of than 70 percent had used a public TI at least once, while use frequency changes. Among individual sources, (i.e., customers and suppliers were cited by fewer firms, and when not grouping categories of technological sources) in-house R&D units by only half of the sample. The re- long-term customers was the most often-cited source of sult held true across all countries except Canada, with technology support (53 percent of firms), closely fol- the highest percentage of firms using public TIs reported lowed by suppliers and in-house R&D departments. by China, India, and Japan. Similarly, public TIs were Foreign investors and licensers were among those used cited by more firms than any other source of technology by the fewest number of firms, along with academic and in every sector except pharmaceuticals (where use of research associations. National TIs, private contract labs, foreign investors/licensers made private contractual local TIs and consulting firms each were used by slightly sources the most cited source). Use of technology sources more than a third of the firms. Important differences in varied considerably across sectors, reflecting perhaps the national TI systems are apparent. For example, national Figure 5 Firms' Use of Technology Source by Sector [Data excludes Taiwan (China)] 8 s- 'a 60 50 .240 E 20 "- 00 0 0 t 0 Polymers Autoparts Foundries Software Total 1 External Public a External Private * Informal Private E In-house Source: Global survey results. 16 in the same rough order, Figure 6 Firms Use of Technology Source by Sector but the difference be- [Data excludes Taiwan (China)] tween sources is rela- 70T tively small. >60± One lesson 50 - that emerged from the 0140 study's interviews and historical comparisons g 30 is that while public TIs E 20 ~ U l * * * U are important, diverse IiL T l * L L L institutions are required 1 10 o l ] 1 _ l + l + ] < to serve industries that vary in their structures Polymers Auto parts Foundries Software Total and in the nature of their CO Foreign Sources C Long-term Customers * National Tls O Local & Reg. nT core technologies. Source: Global survey results. = _ ~~~~~~~~~~~~~~~~~Figure 1 in the Executive Summary (p. TIs and foreign investors are important to Korean firms, 2) illustrates the frequency of use of public TIs by firm but less so in Japan, where local/regional TIs and in- size. Larger firms tend to use TIs more extensively be- dustry associations play a greater role. Firms in India cause of their greater internal resources. More signifi- also use industry associations as well as private consult- cant than size is the presence of internal labs or techni- ing firms. cal departnents: firms with internal resources were more likely to seek outside help, from public institutions in The pattern of use of individual technology particular. Moreover, small firms with internal labora- sources was roughly similar across sectors with long- tories were twice as likely to use public TIs as those with- term suppliers and customers again most heavily used. out internal laboratories. These results suggest the dif- But there were differences by sector. In-house R&D labs ficulty in reaching and serving small firns, particularly were least important in mature sectors such as found- those without internal capabilities. ries and customer-driven sectors like auto parts, but they were more prominent in innovation-intensive sectors like polymers and software. Figure 6 compares the use of Firms' Demandfor Technical Services foreign investors and licensers, long-term customers, na- tional TIs and regional/local TIs in four sectors (plus Fimrs used basic technological services - infor- the total sample). While long-term customers were the mation, education and training, and standards and test- most frequently cited in each, use of foreign investors ing - more often than advanced services requiring more and licensers ranged from almost half in polymers to creativity, such as contract R&D or technical problem- only 10 percent in foundries. Polymer firms reported solving. 'Ihis result was true for all countries, though using national and local Tls for technology support al- the differeLnce was largest in India and Japan and small- most as frequently as they use long-term customers, est in China (where state-owned enterprises reported fre- whereas relatively fewer software firms relied on public quent use of R&D services from group laboratories). institutions. Of the firms who use different sources, the Similarly, all sectors reported more frequent use of basic ratings of importance for technological improvement are services than advanced ones. Institutions and Policiesfor Industrial Technology Development 17 The use of individual ser- Figure 7 Importance of TI Services by Sector vices varied across countries, sec-4. ........ tors, and sizes of firm. The most- 4.1 often cited service in Japan was education and training; in Korea it 4 was information services; and in * 3.9 0) India it was standards and testing. uz 3.8 : il§1 Contract R&D was least cited in m virtually all countries. Individual 3.7 - sectors differed in the importance 3.6 M i attached to their use of technologi- cal services, as illustrated by Fig- 0 3 _a1 ure 7. While most sectors reported ME3. information and standards/testing 3 as most frequently used, problem MM solving and information were most 3.2 Polymers Autoparts Software Total important when firms rated ser-ll la Information o Education & Training M Problem Solving 0 Contract R&D n Standards vices for their effect on technologi- cal improvement. Polymer firms Source: Global Survey Results. attached greater importance to problem-solving and trouble shooting than to other ser- lar across firm size. However, as Figure 8 ilustrates, the vices and also attached greater importance and used con- gap between use of basic services -information and tract R&D services more than firms in other sectors. Thi standards/testing-and contract R&D was relatively reflects the sector's science basis and rapidly changing sman technology. The pattern of service-use was roughly simi- ler among large fis. In sum, basic services that dif- fuse known technology are the most Figure 8. Use of TI Services by Firm-Size critical element of TIs' support of [Data excludes Taiwan (China)] firms' technological learning efforts. 90 | When firms look for external technol- 80- ogy support, it is most often in thie X 70 4, v i !1'' ........... !i.!,! E''''''' --!''<-'<'-'''-'' * - |form of information services, stan- M 60 dards/testing, and education and 50 training. Larger firms and firms in science-based or high-technology M .- Wn-- B - th t -:-Eii3 7- v j j -s. I sectors make greater use of contract M 30 fu | iSovi | iit - t | l ! . | R&D and technical problem-solving 2 0 i i N v v miii g iil-- |but continue to source basic services 1~~~~~ even more frequently. However, 0 o E | firms that use more complex services Small (1-50) Medium (51-350) Large (+350) like problem-solving tend to value || Information S Training S Problem Solving Z Contract R&D Standards | them more highly. Source: Global survey results. (CIH[APTMIE]R. 3- FIfRMS' PN]FT]ERNS (OF 'UESE Introduction firm's in-house capabilities, which must be exploited to access external sources of technological informa- This chapter summarizes the results of the firm tion. Without internal technological strength a firm's surveys and draws implications for policy recommen- ability to utilize outside sources is severely handi- dations. One major conclusion is that, to be effective, capped. This model is strongly supported by the sur- TIs and governments must develop a market-by-mar- vey data, detailed in the chapter's remaining sections. ket focus to provide a wide range of firms with techno- Section C begins within the firm, looking at what logical help. For external support to be of help, it must drives R&D. It finds R&D expenditures to be most be matched to the diverse needs of firms. A second strongly associated with large technologically ad- major conclusion is characterized by the adage: to those vanced enterprises with high levels of training and who have, more is given. A weakness in government, an export orientation. It also seems that firms which university, and industry technology support efforts is see therrLselves as more technologically advanced at that they tend to end up developing technology with home thim abroad are more likely to engage in R&D. larger firms and those that have in-house expertise, These predictors of R&D expenditure turn out to be even if their objectives are broader. This is both ineffi- also strongly related to a firm's use of external tech- cient--program benefits could be increased by shifting nological sources. some expenditures to services and industries that re- ceive little attention, such as diffusion of available tech- Section D begins the analysis of external nology in traditional sectors like textiles and foundries sources by highlighting the largely complementary re- --and inequitable, as most resources are claimed by a lationship between use of in-house and outside sources. narrow set of firms. That is, a firm's internal resources are critical to its abil- ity to access external sources. This relationship is stron- Analysis of survey results also suggests that in- gest for medium-sized firms, because to some extent novative firms look beyond domestic markets. For ex- larger firms appear to substitute internal sources for ample, R&D expenditures tend to be higher among external. The Chapter continues examining the pattern firms with a greater share of exports. Firms that are of external use, starting with those sources with regu- aware of a technological gap with the rest of the world lar nontechnical contact with the firm (customers and are more likely to be engaged in R&D, and are more suppliers) and then moving to more specialized and likely to seek external advice. India is the exception distant sources (foreign licensers, private consultants, that might prove the rule. Its firms' poorer R&D per- contract R&D facilities, and government TIs). Use again formance in the survey might be explained by their tra- is found to be positively related to the firm's size, its ditional insulation from international competition, and internal capabilities, export orientation, and the differ- the failure of export oriented Indian firms to seek do- ence between the firm's technological rank in domes- mestic external information sources might stem from tic and world markets. Broader government programs- skepticism as to the value of TIs "insulated" from the -grants, tax concessions, technology loans, tariffs, etc.- world industrial and technology market. -are considered in Section F. The chief beneficiaries of these are large companies, especially those owned by Section B offers a model of how firms learn. the government. The chapter concludes with policy In particular, it argues that learning depends on a implications in Section G. 18 Institutions and Policiesfor Industrial Technology Development 19 Firms and Learning R&D do not arise and external sources, such as suppli- ers of capital goods or purchasers of final output, may Fizns learn from each other. Not only is con- already be active in the market. At the same time, the scious imitation widespread, but often technical staff of firms' internal technical skills are likely to be weaker, competing firms assist one another. Firms also "buy" and their ability to increase skills more constrained by technical informnation and skill bundled with other goods shortages of capital and of skilled staff than firms in the and services-be it through inputs (relations with sup- industrialized world. External sources may then allow pliers) or outputs (relations with customers). And firms some reaping of economies of scale and scope--which purchase technical knowledge as such, through contracts can be considerable in areas such as instrumentation, with consultants, public research institutions, and train- technical documentation and training--as well as easing ing centers, and they participate in cooperative projects bottlenecks of financial resources and trained person- involving technical advance, ranging from the setting nel. of standards to collaborative research. Evidence of "learning from each other" was found in all countries, These considerations suggest an interdepen- sectors, and types of firms surveyed. dence of internal and external sources (as Box 1 illus- trates) in the innovation process. External sources may These external sources for innovation may be provide stimulus and support, but at least some inter- even more important in developing economies, where nal technical resources are needed to assimilate skills and firms mostly seek older technology. As a result, com- capabilities from outside. A three-stage approach may plex transactional problems involved in out-sourcing assess the extent and nature of this interdependence: Box I Korean Auto Parts Company Develops Technological Capability Korea's largest auto parts company, Mando Machinery Corportation, manufactures motors, electrical and electronic parts, air-conditioning, brakes, and machine tools. It employs 6,500 including 410 technical personnel, and since the mid-1980s has built a large research organization. Formed in 1962, Mando developed much of its technological capability through licensing and reverse engineering. As technology became more sophisticated and foreign companies became reluctant to license, the company built up its technological capability, concentrating on sophisticated value-added products and materials and engineering applications in nine specialized centers . The company uses external sources, particularly TIs, extensively including the well regarded Korea University where it holds two-day workshops four times a year to increase communication and technical knowledge among its technical staff. It has worked with many universities and the Korea Institute for Science and Technology (KIST) to develop components and has participated in multi-company, government- coordinated R&D projects. In 1988, realizing that foreign companies would not readily license recently developed sophisticated technology related to electronically controlled suspension (ECS), the company decided to build an ECS system itself. The technology is difficult to master without capability in many technical areas, induding vehicle dynamics, sensor technology, electronics and mechanical engineering. Instead of developing all capabilities in-house, it obtained support from KIST. With commitment and push from the principal customer, Hyundai (which initially was skeptical about Mando's capability to develop the technology), by 1994 the company's ECS was being installed in Hyundai automobiles. in 1995, Mando produced 30,000 units. Source: Based on the survey work of Se Joon Yoon and Snag Pyo Kim of Yonsei University, Korea. 20 identifying first the characteristics that appear to shape T'he ratio of exports to sales is also positively firms' internal technological capabilities; then assess- related to R&D intensity. Participation in world markets ing the influence of these capabilities on the use of ex- not only provides competitive discipline, but also en- ternal sources of technological support; and finally ex- hances opportunities to learn and secure returns from amining the determinants of the sources and services innovations. that require some external support. Low investment rates in R&D are observed in firms facing especially intense competition in their home Internal Capabilities: The Firm's Involvement market, possibly reflecting a reduced ability to appro- in R&D priate gains from innovation. The ratio of R&D expenditure to sales3 varies Corporate strategy. Especially important are the among countries and sectors. As is typical of cross- firm's policies for human resource development. Train- sectional analyses, much of the variation does not ap- ing and R&D intensities appear to be strongly pear to be systematically related to the variables cov- complem(entary. This is presumably because exploiting ered by the survey. However, where there is a pattern, the results of R&D requires skilled people. three sets of factors seem important in explaining R&D expenditures: context, market position, and corporate Given these results, what might explain the ap- strategy. parent over-investment in R&D in Korea and the corre- sponding under-investment in India? Two other rela- tionships emerge from the analysis. First, the Indian Context. While the firm's size does not seem to firms seera to have faced less domestic competition than influence its R&D intensity, the sector in which it oper- their counterparts, particularly in Japan and Korea, and ates does. In technologically mature industries offering domestic market-share leadership seems to have shel- fewer opportunities for technical advance, such as found- tered Indian firms from competition more than is the ries, automobile parts and textiles, firms tend to invest case in the other countries.4 Japanese firms, almost re- less in R&D than their other characteristics would sug- gardless of their market share, rank themselves as highly gest. Country factors are also relevant. Firms in India exposed to competition. In Korea firms with high do- appear to "under-invest" in R&D, while those in Korea mestic market shares report greater competition than do invest more heavily than predicted by other character- small firmis. But for Indian firms the reported intensity istics. of competition falls as domestic market share rises. Market position. The closer firms consider Second, in India lower domestic competition themselves to be to the frontier of best practice, the seems to correspond to a weaker orientation to export- greater the R&D expenditure. ing. While Indian firms appear to have significantly lower export propensities overall, export shares seem R&D is also higher among firms that rate them- especially low among firms with a high share of the do- selves as technologically more advanced when compared mestic market (though this effect weakens at the top end to domestic rivals than they do when compared with of the market share distribution). In contrast, newer rivals in foreign markets. Such a firm might be better firms in India seem to have higher export propensities positioned and motivated than its domestic competition than would be expected on the basis of their other char- to take on international best practice. acteristics. 3 Broadly similar results were obtained using other indicators for internal technological capabilities, such as the share of technical staff and whether or not a firm has an internal R&D department. 4 Surveys were carried out in 1 993 and 1 994. The competitive environment in India has changed significantly since then. Institutions and Policiesfor Industrial Technology Development 21 Together, these findings suggest that the Indian ing outlays to sales (and which are consequently under-investment in R&D may reflect the degree to most likely to be able to draw on in-house skills and which dominant firms have survived without facing the resources) are less likely to draw on external tech- challenges of either domestic competition or export nological institutions. markets. The contrast between Korea and India is espe- cially striking in this respect: Korea's high market-share 4 This last relationship, however, does not hold for me- firms, which are relatively active exporters, invest espe- dium sized firms. Those medium sized firms with cially heavily in R&D, while a similarly placed Indian high training outlays are especially likely to use ex- firm underinvests in R&D. This reinforces the sugges- temal technological support. In other words, for larger tion that an exposure to international competition can firms internal skills can substitute for external affect a firm's R&D expenditure. sources, but among medium firms reliance on exter- nal support complements internal skill development. Looking Outside the Firn: The Relationship While these patterns hold across the full sample Between Internal and External Technological of countries, there are also striking differences. Three Resources deserve note. The survey results indicate that firms view their In India, the firms most likely to use external internal technical departments as relatively low-cost, help in product and process innovation have high shares easy to access, and good at maintaining confidentiality. of the domestic market. External sources seem less at- Still they make substantial use of external sources in de- tractive than they are elsewhere to smaller and medium- veloping products and processes. Overall, 82 per cent sized firms; indeed, medium-sized, export-oriented of the firms drew on at least one external source for these firms are less likely to use these sources in India than in purposes. Analysis shows that the extent to which firms the other countries. This may reflect skepticism by firms use these sources depends on their internal capabilities, of "insulated" TIs having the capability of supporting corporate strategy and structural characteristics. The export-quality technology. It may also reflect Indian TIs' main relationships are summarized below. inadequate attention to marketing their service to other than the largest companies. Firms that have used an in-house department to develop products or processes are far more likely to In Korea, the larger, more successful firms tend have also used external technological sources than those to internalize their technological capabilities. In particu- that did not use an in-house department. Presumably lar, those with high world market shares make less use this reflects these firms' greater absorptive capacity. A of external help in developing new products and pro- further "push" to outside sources comes from the per- cesses than their size and other characteristics might in- ception that their internal facilities are inadequate both dicate. The Korean firms most likely to make some use in terms of infrastructure and of practical problem-solv- of external sources are significant exporters and view ing ability. themselves as technology leaders domestically but not internationally. For these firms, which are frequently 4 Larger firms use external sources more frequently mid-sized and which tend to rate themselves as facing holding all other characteristics constant. intense competition, external sources would seem to pro- vide a means of bolstering a vulnerable market position. 4 Despite the implication of strong complementarity between internal capabilities and the use of exter- In Japan, on the other hand, the use of exter- nal sources, large firms with very high ratios of train- nal sources tends to be more evenly spread across 22 firms, including those lacking in-house development Firms with an in-house development depart- departments. In contrast to the Korean model, firms ment are most likely to have received help from long- with high shares of world markets tend to make term customers in product or process innovation. De- greater use of external technology sources. At the spite this complementarity of in-house resources and ver- same time, external sources are also important to firms tical support, the largest firms (those in the top decile of with mostly domestic sales--for example firms in the firms ranked by firm size) make less use of customer metal castings and machine tool industries, which sell relationships. Firms that rank their technological capa- to larger firms within the same keiretsu. Many of the bilities more highly compared to domestic firms than firms in these industries lack an in-house development when connpared to foreign rivals (a difference referred department; external sources seem to play a greater to here as the firm's technology gap) make greater use role as replacements for such departments in Japan of these relationships. Customer relations, in other than they do elsewhere. This may be because the range words, seem important to firms that are "catching up" of Japanese TIs--national, regional, and local--better intermationally, especially in Japan, India, Taiwan, and satisfies the needs of firms of different sizes, sectors, among medium-sized, export-oriented firms in Korea. and capabilities. Similar decentralization and special- ization by size and type of TI may be valuable else- Analysis of the services that customers and sup- where. pliers provide to innovating firms also highlights the differing roles of vertical links. Overall, medium-sized firms do not make especially heavy use of vertical rela- Patterns of External Use: Customers and tions, but they do rely on customers and suppliers for Suppliers information about technological and market develop- ments. In contrast, the smallest firms use customers and Among external sources of technical assistance, suppliers for problem-solving, commercial advice and long-term customers and suppliers are the most com- building contacts. monly used. Firm responses highlight some of the problems Foreign lLicensers that reliance on vertical links for help in product and process development entails, notably the lack of Whilerelianceoncustomersandsuppliersiswide- confidentiality, the difficulties of securing "ownership" spread, the use of foreign licensers for technical support is of innovations, and the intermittent nature of assistance. concentrated. Some 17 percent of firms in the sample have But long-term customers, in particular, are seen as readily drawn on foreign licenses or partners. Foreign licensers available ("fast and easy") providers of assistance, ca- are regarded as the best source of ideas about new prod- pable of solving problems, useful in developing prod- ucts and processes, and as having the strongest capacity to ucts and improving processes, and an important means solve problems. But they are costly to deal with because of securing contacts. (Benefits and problems of various of high fees, distance and sheer technical complexity. external technological sources are summarized in Box 2.) Few firms have the resources to overcome these obstacles. The primary users of foreign licenses are larger The perception of vertical links is widely held, firms (thouagh the size effect is reversed at the top end of but the actual use of these links varies between coun- the size-class distribution) that operate in oligopolistic tries and across the industrial structure. Given the im- home markets, have in-house technical capabilities, ex- portance of long-term customers, the analysis focuses port, do so as "technology followers," and are tied into on the factors that shape their role. vertical links (including with foreign firms). Institutions and Policiesfor Industrial Technology Development 23 The extent of reliance on foreign sources also of technical support: the smallest firms are about half as depends on the sector characteristics. In the techno- likely to use consultants as the firm population overall. logically complex polymer industry, where products and processes are demanding and amenable to codifi- cation, firms are twice as likely to acquire foreign li- Technology Institutions censes. Firms in the relatively mature foundry sector, where "tacit" know-how tends to be more important, In contrast to the private sources, public TIs-- are less than half as likely than other firms to have for- national, regional and local technical laboratories, stan- eign licenses. dards institutions and universities and technical colleges- -are widely used in all the countries except Canada, Finally, firms in Canada-which has high pro- where private sources predominate. Public sector domi- ductivity by world standards and is open to foreign in- nance is, naturally, pronounced in China, but even in vestment (which may be a substitute for contractual li- Japan, Korea and India firms draw primarily on public censes)--are least likely to rely on foreign licensers. Finns sector institutional5 support. in India, whose Government restricted foreign licenses for many years as part of its inward orientation, also There are substantial differences in how and by make less use of foreign licenses, despite low produc- what sorts of firms public TIs are used. Even correcting tivity levels. for the effects of in-house capability, the largest firms are the primnary users of national technical laboratories, universities and technical colleges, and research and aca- Consultants and Private Contract Facilities demic associations. The gap in use by larger firms and others is much smaller for local and regional technol- Consultants and private contract R&D also play ogy centers and industry associations. The contrast a limited role among firms, although the extent differs between the use of national institutions on the one hand, among countries--Canadian firms rely more heavily on and local and regional institutes on the other, is seen in these sources, Japanese firms rely less. Firms recognize several patterns: that they are a good source of entirely new ideas and are generally technically competent and reasonably timely, 4 National TIs are most used in Korea and least used but they are expensive. hn addition to high fees, they in Canada. Taking country differences into account, seem to involve sizable transactions costs, because of the typical clients of national TIs are dominant firms, the difficulties in specifying the problem, selecting some of which are government-owned or subsi- among consultants, and assessing their value. As much dized. as the out-of-pocket costs, it may be these transactional difficulties which lead many firms to dismiss private 4 Mid-sized firns, most notably those without an in- consultants. house development department, make less use of national Tls, which are frequently viewed as too slow Reliance on private contractors is therefore lim- and bureaucratic. National TIs in fact were the most ited by the firm's capacity to manage the relationship frequently cited for "red tape". and to learn from it as well as the cost. Survey results highlight these constraints. Large firms, and firms with 4 The gap between large and small firms is consider- an in-house technical department, are about three times ably reduced for frequency of use of local and re- more likely than others to use private contractual sources gional TIs. They are viewed as having less techno- 5 'Institutional' in contrast to the more informal support provided by customers, suppliers and foreign licensers. 24 logical capability than national institutes, but are Overall, this diversity of patterns of use ensures considered easier to use. They are also least fre- that contacts with TIs are relatively pervasive through- quently cited as incurring problems of confidential- out the industrial structure - although it is only in Ja- ity. pan that even the smallest firms seem able to draw on public sources of support for developing new products 4 Medium-sized firms are substantial users of local and processes. There is an important contrast here with and regional bodies. Use is especialLy high among the distributive consequences of other instruments of firms without an in-house technical department, technology policy. whose functions the TI would seem at least partly to replace. Government Policies: Services versus Financial 4 Strikingly, the smalLer firms in Japan are nearly five Assistance times more likely than the average firm to use local and regional technical institutions. This may reflect The largest firms were nearly three times more their extensive availability in Japan and their respon- likely than average to have received government grants, siveness to industrial needs (see Chapter 4 below). tax concessions or special technology loans (government- owned firms in turn were twice as likely as private firms These differences partly reflect the different to benefit from these instruments). The disparities in purposes of respective TIs: access to government procurement and protection from competition were somewhat smalLer. Only those forms 4 The development of entirely new products can re- of assistance most frequently provided through TIs and quire complex skills, which users perceive many of aimed at promoting technology diffusion - help in de- the smalLer bodies (such as regional and local TIs veloping standards and subsidies for education and and industry associations) to lack. National labora- training -- were used by small as welL as by larger firms. tories and universities and technical coLLeges have the skills but are difficult to work with, especially for firms that lack in-house capabilities. It is conse- Policy Implications quently the larger firms which tend to draw on the latter bodies, frequently for the development of new External sources of technical support play an products or as a source of substantially new ("stimu- important role in all the countries and industries. They lating") ideas. tend to complement rather than replace internal firm capabilities, though they can help smaIL or medium-sized Regional and local TIs are seen as unable to develop firms compensate for gaps in skills and facilities. new products or generate ideas. But they are the least Whether the full potential of external sources is realized distant source, and are viewed as good at solving prod- depends on supply- and demand-side factors - that is, uct or process improvement problems - both by pro- on whether firms are encouraged by their environment viding expertise and by alLowing third-party use of fa- to develop internal technical resources and to improve cilities. Medium-sized firms especially use intensively products and processes, and whether support institu- these TIs for applied problem-solving, as welL as for tions for innovation are welL designed and managed. technical information, help with standards and educa- tion and training services. Industry associations are Economic and social policies that encourage firms seen as poor at problem-solving but important for se- to enter world markets provide in themselves incentives curing contacts, building networks and providing ex- for technoLogical development. Even exposed to these in- posure to new ideas. centives, though, firms are unLikely to have all the skills Institutions and Policiesfor Industrial Technology Development 25 they need. The private, largely contractual, TIs are really to use external suppliers of technological information. accessible only to the largest, most sophisticated firms; They need reliable information brokers as much as in- smaller- and medium-sized firms face substantial obstacles. formation providers. Small, less technologically ad- Government support such as loans, grants and tax conces- vanced firms neither are able to judge whether a pro- sions, as well as procurement and protection from imports, vider can meet their needs, nor can they use the exper- are also of greatest use to the largest firms. The survey tise offered, lacking the skills to assimilate this informa- results show that public Tls can reach at least some of the tion. The effectiveness of government programs can be other firms, but often they are ineffective. Designing TIs greatly enhanced by a recognition of this. that can meet these needs is a central challenge. Distinguishing betvween large and other firms is The biggest handicap of smaller firms is their not enough. Firms' diverse needs are broad. This is il- lack of technological expertise, which limits their ability lustrated by the survey on R&D. R&D is most likely to Box 2 Benefits and Problems of Technological Sources Firms were asked to cite the benefits and problems of each external technology source used. Interestingly,the patterns apply across countries and are useful in drawing the conclusions of the study. The raw data indicated that (as expected) benefits and problems were most frequently associated with the most used sources: customers and suppliers. However, weighted by usage, foreign sources, research associations (RAs) and in-house labs were cited as frequently as customers and suppliers for benefits, and foreign sources, private labs and consultants, national Tls and RAs had the highest percentage citation of problems. Speed and ease of use and the ability to solve problems were the most frequently cited benefits (each 20 percent) from external technological sources. Only private contract labs offered faster and easier service than average. While universities and technical schools, private consultants, and local and regional TIs were all good at problem solving. Other benefits cited were product development, building contacts, idea stimulation, and quality improvement (each with about 10 percent). As expected industry associations were associated with building contacts, as were academic associations. The latter excelled at idea stimulation, as did universities, technical colleges, and consulting firms; local and regional TIs and research associations stood out for their shared facilities, and firm laboratories had low cost. (However, shared facilities and low cost were the least cited benefits.) Long-term customers and suppliers, foreign sources, and national TIs were not strongly noted for any particular benefit. Long-term suppliers were seen as somewhat cheaper than average (their costs are hidden); foreign sources were better at product development; and national TIs were noted for shared facilities. The most commonly cited problems were high fees and the inability to solve problems (18 percent each). Lack of timeliness (16 percent) was next. Private contract labs and consultants stood out for excessive fees, also a complaint about foreign sources. Technical disability and lack of timeliness were relatively evenly spread among sources, though academic and industry-related sources tended to suffer most. Red tape characterized national TIs and to a lesser extent local and regional TIs, while private contract labs were cited least in that regard. Long-term customers and suppliers, and especially in-house facilities, tended to have problems with personnel continuity. Customers and suppliers failed to protect confidentiality, for which in-house labs and local and regional TIs received relatively fewest complaints. Inadequate rules for research ownership was a serious problem of research associations, and to a lesser extent of long-term customers. Sources: Global survey results. 26 be carried out by larger firms with strong in-house tech- in national TIs, toward firms that are relatively unsup- nical capabilities in sectors with rapid changes in tech- ported, are likely to be high. Careful analysis of the costs nology, such as polymers and pharmaceuticals. Such and benefits of programs is needed, particularly of firms use national TIs and foreign sources, but see local projects and policies that could benefit firms that are cur- and regional TIs as poor at generating ideas or develop- rently ignored. This may be especially so in industries ing products. Most small and medium-sized firms, on such as foundries and textiles and for technology diffu- the other hand, are rarely interested in cutting-edge R&D, sion, in which scientists have little interest, and for small partly because they simply cannot assimilate it, or even and medium scale industries with limited political clout. formulate the questions that might bear profitable an- swers. As a result, they make less use of national TIs Smaller firms have difficulties in other areas too, and characterize them as slow and cumbersome. But making disproportionately little use of foreign sources, small firms are relatively interested in applied problem private commercial consultants, and TIs in general. This solving, and medium-sized firms desire information suggests that the common focus of governments, uni- about technological and market developments. These versities,, industry and research associations, and non- services are accessible from local and regional TIs. profits, on provision of technological information along with more sophisticated services is inadequate for many No doubt the benefits of government programs potential clients. For firms that have little in-house ex- may be increased with more careful targeting of firms. pertise, the real need is for reliable advice in assessing The survey suggests that small and medium-sized firms and managing technological information, as much as for (SMEs) fail to exploit government programs adequately. that information. SMEs underutilize the high profile and well-funded na- tional TIs. But that is even more true of broad industry ]3roadening the focus of firms to encompass the policies, especially grants, tax breaks, and technology world stage is likely to encourage innovation. Innova- loans. In many instances, the greatest potential for de- tion may as much follow as allow entry into world mar- velopment lies with smaller firms that lack the resources kets. This suggests that an important role for TIs is to to build internal technological expertise or to lay claim help domestic firms breach international markets in ad- to funds offered through government programs. The dition to helping them innovate. Government policies benefits from shifting a few dollars of expenditure away aimed at integration with the world economy can help from generously supported firms' activities such as R&D strengthen domestic technical expertise. C]H[A]PTIER 4 MAKIFNG TECHNO]LOGY INSTI[U11T]ONS WORK Introduction What Constitutes Good Performance? This chapter highlights the lessons of experience An effective TI reaches out to serve its clients and for effective TI performance. Its conclusions rely princi- foresees their future needs so that it can serve those as pally on the survey results, but also on case studies and well. It can adapt to changes in industry and competition. the historical analysis of institutions and technology advance in the countries and sectors examined. The particulars will vary according to the ser- vices the Tl provides and the sectors and circumstances The suggestions for TIs depend on the culture in which it operates. For example, a national applied of the country, the role of the TI, and the sector involved. research laboratory for machinery automation (includ- Having appropriate, quality staff is necessary, but staff- ing machine tools) will need to master a range of ma- ing needs varies by type of institution and sector: the chines, become familiar with the industries its clients skills needed for improving a foundry training and test- serve, stay abreast of relevant and rapidly changing as- ing center in Mexico cannot be the same as for a phar- pects of informafion technology (IT) and inform clients maceutical R&D center in China, even if the principles of the technological directions of their competition. Much of management are. of this goes beyond the daily work of responding to cli- ents' particular needs and problems, such as attaching Broadly the findings are similar to those in nu- automation equipment to jacquards and lathes, testing merous management studies of business enterprises. Tis the reliability and the precision of the machines and train- must have a service orientation, know their clients, be ing operators and maintenance personnel. focused, know their capabilities, including strengtis and weaknesses, and orient their organization and activities In contrast, a well functioning productivity cen- accordingly. They also need to identify the innovations ter for SSI might offer two services: training and prob- influencing their client industries future and design pro- lem-solving on technical and organizational issues. To grams, accordingly, to build the capabilities required. the extent that most SSI are labor intensive, much of the TI's work will relate to industrial engineering and busi- The government's most important role is to en- ness and financial management. As industry develops sure that the business environment is competitive, so that and wages grow (as has occurred in Taiwan and Korea), industry is constantly seeking to maintain its competi- more of the center's activities will be concerned with tiveness and improve its technology. It must ensure in- machinery, automation, quality control, and standards. stitutional support for technological development and see to it that TIs and mechanisms exist to respond to Standards institutions in most countries have failures of the market and impediments to competitive- evolved from setting and enforcing safety standards (of- ness. It should also create incentives for firms to use TIs ten at a level below industrialized countries). A good and other sources of technology effectively. The gov- institute helps companies reach export standards erinent is particularly important for encouraging the through outreach, training and other mechanisms and development of specialized support and institutions for works to make local standards compatible with export small scale industry (SSI). And it should be catalytic market norms. It also helps industry reach quality man- rather than managerial in all these endeavors. agement standards, like the ISO-9000 series. 28 Most TIs offer multiple services. Measures of SSIs in particular require special TIs; otherwise success include clients' use of and satisfaction with ser- they obtain little or no support. But the same point ap- vices. Revenue from clients is an indication of the ser- vices' value, but it should be tempered by a TI's need to Box 3 work on future issues and develop capabilities for which clients are not likely to pay directly. Another indication Problems with the Best and Brightest of its effectiveness would be a combination of how it has changed and demonstrated its "dynamism" in the In the survey, firms were asked to name TIs and past, and how it has trained staff and changed cultures associate particular benefits and problems with them. to adapt to new industry requirements. In Korea and Taiwan two themes emerged that were not unique to those economies. First, TIs cited as providing excellent service were also roundly criticized (by fewer but still significant numbers of firms) for high How TI Managers Can Improve TI Performance fees, slowness and complex procedures, and impractical technical know-how. Second, large, well regarded R&D The art of running a TI is not altogether differ- centers such as the Korean Institute for Science and ent from that of running any other business. Both must Technology (KIST) and the Industrial Technology know and reach their markets, and have products the Resea:rch Institute (ITRI), with their recognized market wants today and will need tomorrow. In a scientists and engineers, were generally not considered skills-based service institution or business, manage- effective in working with small firms. ment must be able to recruit, develop and keep the right These results are not unexpected. KIST and people. For TIs that means high quality technical ITRI scientists are often at the forefront of their people. It also needs a "user friendly" approach to its disciplines and well regarded overseas. They prefer to work and to bring in outside ideas and people to work on state-of-the-art problems of interest only to the broaden the awareness of the TI and prevent its most technologically advanced, generally larger firms stagnation. or on projects of their own creation. Management must be clear who their clients At the same time these institutions receive are, so that they can match their resources to client preferential funding and some of their researchers are preceived to be spoiled and arrogant. Their usefulness needs. This is particularly crucial for publicly sup- to the economy is consequently hotly debated. ported TIs. Too often the definition of the clients var- Diffusion of technology and more basic problem- ies according to the audience. Consider a typical na- solving do not excite their scientists and engineers, but tional TI whose vision statement includes an entire in- simple applications are what most firms require. dustry or range of industries as "clients." For politi- cians, the clients are the small firms and the region These centers can and do serve as depositories where that same TI is located. For the TI's scientists of technological capabilities that produce technological and engineers, the most dynamic firms are perceived synergies, economies of scope and scale, and a more to be the clients, as they are most likely to be able to dynamLic work environment. They should not be absorb technology or communicate with them. Ulti- expected to carry out functions such as supporting a labor-intensive, small-enterprise dominated sector for mately fevr clients are served well. Dynamic, aggres- lbritesv, saletrrs o-ntdsco o which they are ill-suited. They, must, however, be more sive firms get some of what they want, but most SSIs accountable to the society, earn more of their resources are inadequately attended. Rarely can hi-tech, large from industry, and be more responsive to clients. firms and labor-intensive smaller firms be properly served by the same TI. Box 3 illustrates the manifesta- Source: Global survey results and intervews by Kee Young Kim and tion of this problem even in well regarded TIs. San Gee. Institutions and Policiesfor Industrial Technology Development 29 plies widely. If the TI's clientele and focus are clear, then Only when the client orientation was seriously mani- the technical staff can use their skills appropriately and fested did performance improve. prepare to address future problems. For example, a tex- tile technology center that attempted to cover all aspects A TI manager should attend seriously to reju- of textile production and all sizes and shapes of produc- venating the institution's capabilities. Ways in which this ers lost sight of the effect of information technology on can be accomplished are through training and exposing textile mills and was unable to help its client mills pre- staff to relevant new technological developments and pare for competition in overseas markets. industrial programs abroad, sabbaticals, sending staff for periods to work in industry and bringing in visiting in- TI management should encourage a range of dustrialists, lecturers and researchers. The dynanic qual- approaches to ensure that their technical staff know their ity of TIs surveyed was enhanced dramatically by expo- clients and the clients' needs. Effective TIs provide a sure to the outside. Again, TIs are similar to firms. TI range of services that incorporate many means for cli- staff need to be exposed to the outside world, to visit ent interaction, including extension programs, open trade fairs and conferences and receive their equivalent houses, surveys and seminars. When these activities of technical training. were all extensively used, the survey found that TI per- formance improved significantly. Having multiple pro- Survey results suggest that incentives matter, grams of industrial interaction is an integral part of the and they will vary from country to country. Essentially work in effective TIs, where good scientists get involved they should encourage technical staff to be client-ori- in these activities, learn about their clients and problems, ented, stay on top of their field, and work in teams. Staff and sometimes redirect the longer term projects in light advancement based partly on practical achievement and of the market. contribution to income seemed to have the impact in- tended. Such incentives schemes are important tools for Another way of keeping in touch with clients is encouraging client-orientation and income generation, doing work for them. Generating income from clients provided society's norms permit this sort of differentia- makes TIs more practical. The practical activity of India's tion in staff remunerations. National Chemical Laboratory has grown along with its contract revenues-which grew from less than 20 per- Technical staffing must match the vision and cent to about nearly 40 percent of total revenue in five role of the TI. It makes no sense to seek theoretical sci- years and have exceeded one half of operating expendi- entists to staff a standards, testing and problem-solving ture over the past two years (see Box 9 in Chapter 5). laboratory for machinery industries. But good mechani- cal engineers, materials scientists, and technicians are Certain formal aspects of client orientation and needed. Firms' most frequent criticism of TIs is their association did not seem to affect performance. In par- failure to solve technical problems. ticular, formal organizational links with industry by themselves, such as membership on boards of advisers Two illustrations show good TI practices for dif- or directors, seemed to have no impact on the quantity ferent markets, roles and sectors. One type is a national of TI services provided to industry. Similarly, whether laboratory in a science-based sector. The leadership of a the CEO of the TI came from industry or research seemed chemicals R&D laboratory, to serve the budding chemi- to make little difference on performance (except that an cals, plastics, pharmaceutical, fertilizer, and polymer industrial background of a CEO did on average add 20 industries, must have a staff that can foresee develop- percent to industrial contract earnings). Perhaps these ments. It needs chemists and chemical engineers that findings illustrate that client orientation must be inter- can help design polymers and other new chemicals, ad- nalized. Superficial links by themselves will not suffice. vise on process, and carry out pilot-scale experiments. 30 Some will need experience in technologically advanced Variationls by Sector and Technology6 multinational companies, and a few should have car- ried out internationally recognized research. Confiden- Sectors vary by the nature of their key technolo- tial work for some clients will include testing and infor- gies, whether they are mature or emerging, rapidly or mation searches, helping adapt imported processes to slowly changing, sophisticated or simple, science or en- local equipment and/or materials, and designing poly- gineering-based, and the extent they use generic tech- mers. The laboratory will also advance technology by, nologies such as information technology, industrial en- for example, developing know-how and products in ad- gineering, and materials. They also vary according to vanced engineering polymers, or engaging in research their economies of scale and their industrial structure-- with a university in an advanced country. Perhaps one- the number and size distribution of firms. third of its budget will be devoted to these types of projects, funded by research councils and government The findings suggest that effective institutions and industry organizations. The other two-thirds would as well as policies must vary by sector. The nature of emanate from projects and services-testing, training, con- support and mix of services offered will also need to tract R&D-for mostly domestic industrial clients. change as firms grow and become more technologically sophisticated. A second well-run TI supports a cluster of small scale firms in a mature sector in a region. Serving the Engineering industries, such as auto parts and local foundry and machinery industries, the TI will be machine tools, involve many technological areas and staffed by engineers who understand industrial process, complex production processes that can be mapped and production organization, and materials. Staff will fre- disaggregated. Most innovations (product or process) are quently visit clients to identify and interpret their prob- small and incremental and nornally only affect part of lems. Testing, information provision, and trouble shoot- the production line. They can emanate in many ways ing are its main activities. The staff might spend 15 from many sources. Most are not proprietary, and the percent to 20 percent of their time at a national metallur- few exceptions are frequently better kept as trade secrets gical center or a university mechanical engineering de- than pate,nts. Hence patent regulations are not major partment to leam about developments in other parts of factors in these sectors. Firms or countries that are far the country, and every year at least one staff person will behind the frontier have many opportunities to obtain spend time abroad learning about processes, materials, technologies, skills and advice cheaply. But most inno- and research in advanced countries. Perhaps a small vations re(quire the firm to have technological capability amount of R&D is supported by a research council and to build upon. Not all innovation, however, need be car- the local industry association, who also might support ried out within the firm. Wide-ranging scale economies regular visits and collaboration by outside lecturers and can be obtained by collaboration and joint learning (see researchers at the TI. Clients pay a large part of the cost Boxes 4, 5 and 6 for example). of services, but the local and national governments sub- sidize some activities, including training, technical ex- Engineering industries are built on a wide tech- change with other institutions-perhaps a quarter of the nological base, and therefore need a range of TIs. Some budget. specialize in a sector, others in industrial or production 6 This section relies on the contributions of Greg Felker, a doctoral candidate at Princeton University. Institutions and Policiesfor Industrial Technology Development 31 engineering and work with firms in many sectors. Large problems with the aim to publish eventually. government-run national laboratories are not normally relevant. In auto parts, for example, advanced research In sectors where the technology is mature, labs are more likely organized by or with an assembler. changing slowly and largely not proprietary-some en- gineering sub-sectors, foundries and textiles--there are In science-based sectors much of the technology many opportunities for using TIs to diffuse known tech- is codified, and if it is to be understood or improved, nologies widely and thus offer scale-economies for tech- scientists need advanced training and equipment. Phar- nology transfer. Instead of companies importing know- maceuticals and polymers are two such sectors. Patent how individually, the TI imports and transfers it to many protection and economies of scale are important consid- companies. Sector studies of Japanese foundries, Chi- erations in pharmaceuticals because of the high cost of nese textiles and Korean auto parts illustrate this. Tex- developing, testing and marketing drugs. In pharma- tile research associations (see Box 4), Japanese-type busi- ceuticals, India and Hungary7 illustrate that high-level ness associations (see Box 6) or cluster-type technology government support can help build underlying capabili- service organizations are models for different circum- ties. Both countries had relaxed intellectual property re- stances. gimes, and built public-sector corporations and research institutes. While few new drugs have been developed, These cluster-type technology transfer organi- many were reverse-engineered, and processes were de- zations are likely to be most relevant for early veloped for existing pharmaceuticals. In India's rapidly industrializers since industrialization normally starts in growing sector, some entrepreneurs and managers were sectors with mature technologies that use unskilled, low- trained as engineers in public-sector companies before cost labor and tends to develop in selected regions. a "modern" pharmaceutical sector existed. Whether companies in either country can make it into the big leagues is questionable, but multinationals are already Cooperatives tapping the local talent. Cooperative research associations (RAs) are Polymers have some similarities to pharmaceu- most commonly found in the textile sector. Box 4 pre- ticals, but developing a polymer is much quicker and sents an example from India, but relatively successful requires no extensive trials.8 The returns to scale are in examples can be also found in other surveyed econo- the production process, and changes usually come in an mies, including Taiwan, as well as all over Europe. Co- entirely new plant or product. Efforts in Japan to build operatives are generally difficult to organize. Compa- research associations to carry out joint R&D by and large nies need to perceive that they stand to benefit more than failed. In India by contrast, national R&D centers have they risk by joining together with their competitors. The helped firms develop polymers and solve material prob- promoter should be a leader trusted by competing firms. lems. The interests of companies and scientists seem to In textiles, comparable companies may join together, as converge. Companies are interested in making prod- the perceived competition may be largely from foreign ucts and profits from the intellectual property, and sci- companies, whether in domestic or foreign markets. And entists want to work on complex chemical and material the perception of foreign competition may be needed to 7 Insights concerning pharmaceuticals are based on the work of Professor Shekhar Chaudhuri of the Indian Institute of Management in Ahmendabad, India and Ms. Katalin Gyorgy of the Budapest University of Economic Sciences in Hungary. 8 This section owes insight to Professor Shinichi Watanabe of the International University of Japan and to Fumio Nishikawa, an industrialist and an advisor to the study. 32 spur the formation of cooperatives. Many textile manu- ment). Of those, eleven were in textiles9, and the rest facturers compete domestically on design and quality, were spread among software, pharmaceuticals, synthetic based on trade secrets and employee expertise. In those rubber, auto parts and textile machinery. Interestingly, circumstances, firms see a joint need to stay current on although there were no cooperative RAs in foundries processes and design, to share testing facilities, and to Japanese business associations (akin to RAs) were an im- help firms solve their problems. Of the 150 TIs inter- portant component of Japan's catching-up phase in viewed in China, Japan, Korea and India, twenty-one foundries (see Box 6). Cooperatives take on many forms; were cooperatives (with or without government involve- an exam.ple of a club-like keiretsu cooperative in auto parts is presented in Box 5. Box 4 Cooperatives: A Relative Success Reflects Potential and Pitfalls Cooperatives have the advantage of having captive industrial dients and interest but must be service oriented to survive- members will stop paying dues and purchasing services if they get no value. Cooperative Tls more frequently suffer from inadequate funding from members after the initial stage of cooperative spirit and the passing of the motivation for their founding. Industry will fund projects that have an identifiable early benefit. Beyond that, even successful cooperatives have had trouble supporting R&D in new areas. The Ahmedabad Textile Industry Research Association (ATIRA) in India was modeled on the British Research Association, in which industry and government joined forces for technological cooperation. A couple of far-sighted large mill owners mobilized the backing of industrialists, government, and technical staff. AkTIRA offers the full gamut of services-training, testing, consulting, R&D-and is considered well-run. In its fifty years, ATIRA has changed with the textile sector. Its role has varied over time on a narrow band of providing services to industry and attempting to lead industry into the technological future. Like many TIs, it has had difficulty integrating its various roles, and its high-level staff have been more at home doing basic research that "reinvents the wheel" than anticipating the implications for its xnembers of technological advances and changes in trade rules and patterns. Until recently, its sources of revenues varied only modestly. Roughly a fourth each has come from government grants, membership dues, services to industry and sponsored research, largely from government and international agencies (like IJnited Nations Development Program). Untied government grants are now being phased out, and services are being emphasized. ATIRA is also likely to rely more on consortia of key members to sponsor R&D and build new capability, as government and international aid funds become scarce. That would also encourage industry to be more involved in strategic, industrial and technological thinking. The danger, of course, is that when the going for industry gets tough, industry will cut support, particularly for the long term activities. ATIRN's well-regarded services, organization and staff have not ensured sufficient foresight. Like many institutions, sectors and countries, better mechanisms of funding and governance for TIs are required to ensure they remain responsive and provide strategic advice to their sectors and companies. Source: Interviews and report of Dr. Shekhar Chaudhuri of the Indian Institute of Managernent-Ahmedabad. 9 There is no evidence from the survey, but the persistence of the textile coops may have been enhanced by their ability to serve as lobbyists for the members in allocating quotas. Most of the coops preceded the international textile quota arrangements. Institutions and Policiesfor Industrial Technology Development 33 Box 5 Keiretsu Cooperative The Japanese auto parts industry provides an example of a cooperative TI and the evolution of Tls. Established in 1960 at the behest of the parent auto assembly company (which had its own large R&D lab) and key first-tier parts suppliers, this automotive TI has grown to 919 staff. Seventy-five percent are technical specialists, performing a range of activities, with advanced and contract R&D being most important. The TI is closed to outsiders, working with and sharing information only among its forty members (the auto maker, nine first-tier suppliers and thirty secondary suppliers). Work comes by individual or joint contract, through technical-advisory groups in a range of fields covered, and from the ideas of R&D staff. Most of the budget is financed from membership fees; less than ten percent is from service fees. The nature of TI activities evolved from information exchange to joint projects on fuel efficiency and emission controls after the 1970s oil shocks, to greater contract and advanced R&D work. In contrast to public sector TIs, the automotive TI became far more important to parts suppliers, once they caught up to the most advanced in the world. Today, joint R&D is common, since members view technology sharing and alliance as helpful even as they are concerned to achieve the right balance between openness and confidentiality. The lack of openness to non-members does act as a restraint to top class researchers who prefer to share their research findings. The TI has made major contributions in emissions control R&D and in robotics. It offers expanded approaches to problem-solving and R&D, increases the efficiency of technological development by building on synergy, and raises technological capacity among members. Source: Based on interviews and reports of Tatsuya Ohmori of Seigakuin University and Yoshitaka Okada of Sophia University. CIH[A]P]ER 5,- (GOVERNMENT IROLE I[N RIlE FORMI[ NG T]EC]H[NO]LOGY ][N]FRASTRIU[NCTU[R]E Government Role in Technology Development Japan, where an extensive technical support infrastruc- ture is dedicated to SSIs, do SSIs use TIs (which in Japan Government's role in TIs is, first and foremost, are heavily subsidized) extensively. It is generally a mis- to create an environment conducive to technology ac- take to tack on support for SSIs to general TIs other func- quisition and development, and to offer the right incen- tions. Most SSIs have different needs and lack the tech- tives. Regardless of ownership, firms need to operate in nical staff to communicate with sophisticated TIs (small a competitive environment to generate concern for im- hi-tech firms are exceptions). TIs that focus on SSIs need proving product quality, designing new products, and to expose companies to the benefits of change and gen- increasing production efficiency TI services will then erate demand for technology improvement, as well as have a potential demand. resolve problems brought to them. The appropriate environment also includes an T'wo approaches for supporting SSIs seem to be education system that produces quality technical man- effective. Japanese TIs are decentralized and directed at power, and a technology infrastructure. Government's geographic clusters of industries. They are supported role is to promote both but, at least with regard to infra- by prefectural (county) governments, and their purpose structure, not to manage them. Government might ini- is to suppFort local SSIs. TI staff work with their clients in tiate action itself, stimulate industry to form jointly TIs, many waiys, and most work relates to diffusion. The and/or stimulate an educational institution or private clustering of related industries permits TIs to specialize. group to start TIs. Government can, as a catalyst, offer the right incentives and financial assistance that will Taiwan's productivity center offers a different stimulate growth and service to industry. For its finan- approach and also is successful. This center is an SSI cial contribution, government should demand results in model of a TI that develops technological expertise in a achieving the objectives set, but avoid the pitfalls of at- generic area with wide applicability to a range of SSI tempting to control how they are achieved. across in,dustries. The center's expertise is industrial engineering and automation. Its services are well-re- Where the infrastructure exists, but fails to sup- garded, but they are used by only a low percentage of port industry, government is frequently part of the prob- the enormous number of SSIs, a rate of TI usage much lem. The way it finances TIs--for example, budgetary lower than in Japan. A similar model is used in Korea, transfer for line item expenses, rather than programs-- and is being developed in parts of China. frequently acts as a disincentive to industry support, as does the control it exercises over TI activities. Success- ful instances of institutional transformation have had in The Importance of a Diffusion Orientation common a "hard" budget constraint and a clear vision of the institution's role and procedures, including incen- Dliffusion of known technology to firms should tives and programs for industry and effective leadership. be central to the work of TIs. Industry most often de- mands information, skills training, and solutions to prob- An important role for government relates to lems. Their needs can be met by those with appropriate small-scale industry (SSI), which is inadequately sup- technical background. Even where the need is to learn ported by Tls in almost all countries studied. Only in an entire technological package (as when it is imported), 34 Institutions and Policiesfor Industrial Technology Development 35 the country or company must consider how to train its lack of associations may also reflect the complexity and staff, companies or subcontractors to become thoroughly the science basis of polymers. Associations would not familiar with the know-how involved, as well as to en- attract the talent normally found only in basic research sure adaptation. laboratories or in the R&D labs of advanced chemical companies. Narrow research associations have also Japanese busi- failed in Japan's poly- ness associations illus- Box 6 mer sector. They lacked trate how imported the wide chemistry and technology can be dif- TIs For Catching Up chemical engineering fused to catch up with Japanese Business Associations talent of broader basic foreign competitors. research labs. Business associations To transfer, assimilate and diffuse foreign technologies, imported and diffused groups of Japanese companies, with and without govermnent help, organized various institutional arrangements. The Diffusionfor Small (foundries) tetnologies Japan Business Association (BA) was one type of TI. BAs Firms were widely used in foundries. in the 1950s and 1960s (Box 6). An association The Japan Shell Molding Association was set up by twenty- Diffusion encom- may license a foreign six firms in 1956 to import as sole licensee shell molding passes the use of ma- process, master it and technology and disseminate it to its members. By 1961 it chinery, the application then sublicense it to had about 250 members and had acquired and disseminated of quality enhancing member firms. Or, once many other technologies. systems (even manage- the patents or licenses ment practices and in- lapse, an association On the other hand, Toshiba licensed the technology for formation technology), may sublicense a tech- spheroidal graphite cast iron from a U.S. company and the solving of familiar nology from the original developed the process. Once the patent expired, the Japan licenologye then ornapt aDuctile Cast Iron Association was set up to disseminate cast problems, technical licensee, then adapt and iron technology and carry out related R&D. An Association training, the testing of diffuse it to subcontrac- member used centrifugal casting to apply spheroidal graphite products and materials, tors of the original lic- cast iron to pipes. and the provisions of ensee. By becoming the technology, standards loci for mastering tech- Direct transfer within keiretsu by an importing firm, as well and related informa- nologies, the associa- as by prefectural institutions to small-scale industry tion, in books, by tions can improve them independent of keiretsu, are other mechanisms for diffusion. people or on line. or carry out R&D in re- Source: "Catching Up to Leadership," by Ms. Sakura Kojirna, Tokai University, lated areas. and Professor Yoshitaka Okada, Sophia University, Tokyo, Japan. Even orga- nized, large companies, TIs then can with significant techni- play a major role in facilitating import and absorption cal know-how, have difficulty keeping abreast of their of foreign technologies, but to do so, they must quickly fields. For small and medium-sized firms, it is difficult learn the art and legalities involved in intellectual prop- to know the questions to ask. It is not enough for a TI to erty and confidentiality. Business associations, used ex- support small firms. They need organizing principles tensively in castings, could not be used successfully in and expertise-specialized skills such as industrial engi- every sector in Japan. They were not used for polymers; neering for a range of industries, or experience in one or chemical companies in the U.S. and Europe guarded their more sectors in a region. Successful examples of the clus- technologies and did not permit easy sublicensing. The ter approach are Japan's prefectural R&D centers, while 36 Taiwan's productivity center applies its specialized ca- this reason, governments are increasingly interested in pabilities in industrial engineering and automation reform of Tls. Results have frequently been disappoint- across small-scale industries. ing. Partial solutions--setting up a marketing arm for TI innovations, changing the charter and composition of Which approach is suited for an economy de- the TIs' board, or adjusting project selection criteria--do pends on many factors: the size of the sector, the struc- not alter the incentives that shape TI activities. ture, and the institutional culture, including the feasi- bility of industrial firms working in cooperatives. And Successful restructuring is feasible in most situ- the two types of TI orientation for SSI are not mutually ations. There is no simple formula for TI reform, and exclusive. Taiwan has sector-specific TIs that compliment the approach will depend on many internal as well as its productivity center. As the sector grows, technologi- external factors. The key is to understand the TI's cli- cal needs become more diverse, and the institutional re- ents, how to reach them, and how to support them. quirements become greater. Sometimes ownership, but more often internal incen- tives, legal structure, management, and the direction of Even diffusion-focused institutions need to be programs, need change. forward-looking. Garment, electronics, and castings firms must be able to compete as wages rise and buyers A common weakness is that a TI is funded en- demand higher quality. TIs specializing in small firms tirely or largely by government, without any mechanism must stay abreast of developments in their areas, or they to ensure that it is serving well-defined clienteles. Staff will stagnate, and demand for their services will decline. rewards are based principally on criteria like publica- tions and research recognition, that are unrelated to cli- ent services. Priorities are then set by the scientists, who Reforming or Restructuring Tls generally have never worked in industry. This may be appropriate for basic research, but not for assisting in- In most countries TIs are established by the gov- dustrial clients. Another weakness stemming from the erinent to build the nation's knowledge base and de- TI's "insularity" is the frequent lack of understanding of velop technical manpower, as well as assist industries the impact of a more open, competitive market on the that are not ready to pay for know-how and technologi- needs and competitiveness, including technology and cal information. After investing in the TI infrastructure, product quality of local industry. however, many governments find that its contributions to development remain minimal and its technological Three cases demonstrate the routes to reform. expertise is of limited practical re}evance to industry. TI The Industrial Technology Research Institute (ITRI) in managers need to bridge the communication gap be- Taiwan (Box 7) illustrates how effective restructuring is tween owners and managers in industry, who may be possible by changing leadership, funding and project sophisticated technology users but use obsolete technolo- approval process, without altering the structure. gies, and their own scientists and engineers who are interested in cutting-edge problems and research. China chose another route to reforming more than 5,000 TIs belonging to various government minis- When TIs become divorced from the needs of tries, municipal and local authorities, and state enter- industry, firms are tempted to curtail support for them. prises. Government squeezed their budgets and decerw- On the other hand, governments are hesitant to close tralized decisions to authorities and to the TIs. Some TIs the door on a national resource, especially one that could have down-sized; others have set up spin-off plants, and help improve the country's competitive position. For some have become demand-driven by serving well-de- Institutions and Policiesfor Industrial Technology Development 37 Box 7 Changing Directions in Taiwan (China) The Industrial Technology Research Institute The Industrial Technology Research Institute (ITRI) was formed in 1973 from three TIs to serve as the premier TI of Taiwan. ITRI was to lead technology development and carry out advanced R&D beyond industry's capability. Funded entirely by the government, ITRI grew isolated from industry and, except for a few instances of absorbing and transferring imported technology (for example, it helped set up Taiwan's first integrated circuit producer and designer in 1979 and 1982, respectively), it ignored industrial demand. ITRI worked in metals, chemicals, mechanical engineering, mining and energy. In 1982 the government initiated reforms. To lead the organization it recruited an experienced researcher with industrial experience abroad, pushed for financing from clients, and insisted that the bulk of government funding would be on a contractual basis--researchers had to submit concise strategic proposals outlining why a project should be undertaken, identifying potential clients for the results, and proposing budgets and timetables. Proposals are assessed by a technical group from industry, academia, and government. Researchers are expected to live with their commitments. Today ITRI is considered one of the most successful TIs in supporting industrial development. By the early 1990s, only 5 percent of the budget comes in untied government disbursements. Half comes from industry contracts, and the remainder from government contracts with long-term links to industrial development. Source: Based on interviews cerried out by San Gee of the National Central University in Taiwan. fined clienteles. There have been costs. Some talented The National Chemical Laboratory (NCL), and its poly- staff have quit, and a number of institutions have closed. mer division in particular, did transform themselves. China also fostered the creation of new institutions to That, however, was a consequence of exceptional lead- fill gaps. To develop basic and strategic research, the ership, technical capacity, and a technical agenda that government designated a number of national key R&D was not radically different from NCL's previous one, programs. The designation is competitively assigned and along with limited outside support. Other Indian insti- is meant to ensure that a portion of resources is reserved tutions, particularly less technically able ones, have not for developing long-term capability in critical technolo- adapted. But as salaries remain guaranteed, there is no gies. The signals to TIs were unequivocal: they had to imperative to change. Reforms since 1991 have gener- serve clients and earn a sizable portion of their budgets. ated increased competition and fostered increased de- The budget squeeze on most TIs was substantial. Given mand for improved technology. This changed environ- the enormity of the task, the Chinese Darwinian squeeze ment now calls, perhaps, for further institutional reform may have been the least bad option. measures. The squeeze in India was softer, and the results have been more modest. By Indian standards, reforms Lessons for Government Policy and the TI System of the key group of industry-related TIs in 1988 were substantial, including encouragement of consulting by While the reform process is complex, success is research staff, reducing (marginally) untied government certainly achievable if incentives can be reoriented. TIs' funding, and mandating increases in contract income. performance and funding, along with the advancement 38 of staff, must be linked to the quality of service to indus- lessons are similar. A service orientation is the founda- trial clients. Achieving this alignment depends on the tion. In early stages, most industries are likely to de- industrial context. Governments with established TI sys- mand basic, diffusion-related services. Building institu- tems must change budgetary and internal staffing in- tions for diffusion-standards, testing, information dis- centives, as well as become responsive to a changed en- semination, training--will lay the groundwork for re- vironment. search capabilities. The economies of scale in technol- ogy diffuision are considerable. Governments and TIs For countries with nascent industrial sectors should constantly look for opportunities to foster diffu- who are just establishing a technology infrastructure, the sion, whether through direct services or by stimulating Box 8 Restructuring, Chinese-style China's efforts to restructure its technological infrastructure have been diramatic. TIs were previously simply given a mission by their parent authority and a budget. Whatever the results, the system was slack. Beginning in 1989 the budget for more than 5,000 TIs was slashed in half over three years. The pain was not distributed uniformly; more prestigious and better-connected TIs were less affected. Coping with cuts was left to the leadership of the TIs. About one-third of the cut was funneled back through banks as low interest loans, and other funds were channeled to national programs for S&T planning or key studies. Further reforms and subsidy reductions required additional belt tightening in the 1990s. The response by TIs has been varied. The Research Institute of Automation for Machinery Industry (RIAMB) in Beijing has weathered the cuts. Before 1989 RIAMB's work was determined by government and was largely scientist- and laboratory-driven. Even in 1986, its revenues were Yuan 12 million from government and Yuan 2 million from clients. By 1993 its budget had risen to Yuan 60 million, of which only Yuan 3 million came from government, in the form of contracts. It has obtained thirty patents, won thirty national and intemational prizes and eighty rninisterial prizes. It has diffused many technologies all over China and solved many problems. It lost some of its most talented staff to multinationals and joint ventures and the new private sector before its salary structure was liberalized. The Beijing Research Institute for Plastics (BRIP), established by the Beijing Municipal Government in 1964 to support local plastics manufacturing firms, did not fare as well. It had to shed thirty percent of its 500 staff and reduced the pay of an additional 100 staff. It even had to rent out some of its building. It has not been able to keep up with some of Beijing's more technologically dynamic polymers and plastics manufacturers, and it has had no resources to invest in equipment and know-how. Its financial situation has improved slightly with extreme belt-tightening, and with the introduction of incentives to staff who bring in contracts. But its viability remains unclear in the long run. It could become a combination of cooperative production unit and advisor or joint venture with rural firms (township and village enterprises) that need intermediate technical know-how. Some TIs have set up factories and dropped their institutional roles; others have become production and technology development organizations; some have shed the weakest staff or the least well-connected. Many have adopted merit salary and incentive schemes, sometimes only after losing qualified people. What is clear is that all have had to move toward market demands, whether in technological know-how or real estate. The adjustment seems to be a product of the initial endowment and the vision and capability of the leadership. Source: Based on interviews carried out by and under the guidance of Professor Xu Zhao-xia.ng of the National Research Center for Science and Technology for Development, Beijing, People's Republic of China. Institutions and Policiesfor Industrial Technology Development 39 Box 9 Institutional Reform in India National Chemical Laboratory The National Chemical Laboratory has had excellent technical leadership from its inception in 1953. It was a science institute where personnel spurned practical work and considered those in academia abroad, particularly in the U.S. and U.K., their peers. Like many TIs in India, it was huge and its work was of mixed quality, although some of its research was world-class and was published in intemational journals. It is located in an industrial city, with a pleasant climate and good schools--an attractive place for scientists to work and live. Major areas of concentration are catalysis, biochemistry, organic chemistry and polymers. Through the 1970s and 1980s the chemical industry was maturing, but in the late 1980s, NCL was doing only a small portion of work for industry, eaming slightly over 20 percent of its operating budget from contractual services, much of it with govermnent. Contracting with industry was not valued like publications, and there were no monetary or other incentives for doing anything beyond publishable work. Extemally, however, the role of TIs like NCL was being questioned, along with their automatic funding at a time of budgetary stringency. Govemment reforns (based on the controversial but highly regarded Abid Hussein Commnission report) beginning in 1986 included restraining budget support, setting guidelines for contract revenue performance, and authorizing monetary incentives to staff for contractual work. Following those policy changes and the appointment in 1988 of a new director, NCL rethought its vision and business plan and secured a (World Bank financed) loan to help develop its practical capability. It hired consultants to help set its priorities and instituted a reward system that values industrial service, team effort and income generation, while still supporting outstanding science. It is seeking a balance of foreign and Indian industrial contracts, so that it remains at the forefront of its field at home and abroad. Revenues have climbed: contract revenues have exceeded 50 percent of recurrent expenditures over the past two years. The success of the institute is so widely recognized that in 1995 the director was appointed director-general of India's network of scientific and industrial laboratories whose forty Tls include NCL. Source: Based on interviews carried out by Professors S. Ramnarayan and Shekhar Chaudhuri, Indian Institute of Management, Ahmedabad, India. the private sector to exchange information through busi- and they can promote efficient use of scarce technical ness associations, sub-contracting networks, or other resources. However, they must be structured, staffed forms of private sector organization. and funded to support industry's needs including those of smaller firms. The study has illustrated various, but The evidence is overwhelming that a country's all too rare, successful institutions helping smaller firms industrial development benefits from a well-designed develop technologically. However, the institution and technology infrastructure and policy that aim to meet program models need to be appropriately refined, the needs of the range of industries and firms. TIs can adapted and experimented with, before applying them complement, substitute and facilitate industry efforts, in the vast majority of inadequately served countries. RJE]FRERIEN(C]ES Amsden, Alice H. 1989. Asia's Next Giant: South Korea Ergas, Henry. 1986a. The Importance of Technology Policy. and Late Industrialization. New York: Oxford University Paris: OECD. Press. . 1986b. Does Technology Policy Matter? Centre for Arnold, E., J. Bessant, M. Hobday and H. Rush. 1993. European Policy Studies, Paper No. 29. Background Benchmark Studyfor the Venezuelan Institute of Engineering. Brighton. Ernst, Dieter, and David O'Connor. 1989. Technology and Global Competition. Paris: OECD. Bell, Martin R. 1984. "Learning and the Accumulation of Industrial Technological Capacity in Developing Fransman, Martin. 1990. The Market and Beyond: Coop- Countries." In Martin Fransman and Kenneth King, eds., eration and Competition in Information Technology Develop- Technological Capability in the Third World. New York: St. ment in the Japanese System. Cambridge, England: Cam- Martin's Press. bridge University Press. Bell, Martin R. 1993. Integrating R&D with Industrial Pro- Fransman, Martin. 1986. Technology and Economic Devel- duction and Technical Changes: Strengthening Linkages and opment. Boulder, Colo.: Westview Press. Changing Structures. Science Policy Research Unit, Uni- versity of Sussex, Brighton. Fransman, Martin, and Kenneth King, eds. 1984. Techno- logical Capability in the Third World. New York: St. Branscomb, Lewis, et al. 1991. Empowering Technology: Martin's Press. ImplementingA U.S. Strategy. Cambridge, Mass.: MIT Press. Freeman, Christopher. 1988. "Japan: A New National Braunling, Gerhard. 1986. Publicly Funded Organizations System of Innovation?" In Giovanni Dosi Christopher Assisting Innovation and Technology Transfer in Small and Freeman, Richard Nelson, Gerald Silverberg, and Luc Medium Sized Companies. Luxembourg: TII. Soete, eds. Technical Change and Economic Theory. Lon- don: Pinter. Dahlman, Carl J., Bruce Ross-Larson, and Larry E. Westphal. 1985. Managing Technological Development: Les- Goldman, Melvin. 1994. "Technology Institutions: sons From the Newly Industrializing Countries. Washing- When Are They Useful? Lessons from Europe." The ton, D.C.: The World Bank. World Bank, Asia Technical Department. Washington, D.C. Dahlman, Carl J., and Ousa Sananikone. 1990. Technol- ogy Strategy in the Economy of Taiwan: Exploiting Foreign Goto, Akira, and Ryuhei Wakasugi. 1988. "Technology Linkages and Inventory in Local Capacity. World Bank Tech- Policy" In Ryutaro Komiya Masahiro Okuno, and Kotaro nical Paper. Washington, D.C. Suzumura, eds., Industrial Policy of Japan. Tokyo: Aca- demic Press, Inc. Dertouzos, Michael L., Richard K. Lester, and Robert M. Solow. 1989. Made in America: Regaining the Productive Hobday, Michael. 1995. Innovation in East Asia: The Chal- Edge. Cambridge, Mass.: MIT Press. lenge to Japan. Aldershot, UK: Edward Elgar. 41 42 Kakazu, Hiroshi. 1990. "Industrial Technology Capabili- atory Study." Center for Japan-U.S. Relations Working ties and Policies." Asian Development Review 8, No. 2 Paper No. 5. Intemational University of Japan. Kech, Otto. 1993. "The National System for Innovation Organization for Economic Cooperation and Develop- in Germany." In R. Nelson ed., National Innovation Sys- ment. 1992. "Technology and the Economy: The Key Re- tems: A Comparative Analysis. New York: Oxford Uni- lationships." Paris. versity Press. Pavitt, Keith. 1984. "Sectoral Patterns of Technical Kodama, Fumio. 1991. Analyzing Japanese High Technolo- Change: Towards a Taxonomy and a Theory." Research gies: The Techno-Paradigm Shift. London: Pinter. Policy, No. 13, pp. 343-373. Lall, Sanjay. 1987. Learning to Industrialize: The Acquisi- Priore, Michael J., and Charles F. Sabel. 1984. The Second tion of Technological Capability by India. London: Macmillan Industrial Divide. New York: Basic Books, Inc: Press. Ramesh, Jairam, and Charles Weiss, Jr., eds. 1979. Mobi- Levy, Jonah, and Richard Samuels. 1992. "Institutions lizing Technology for World Development. New York: and Innovation: Research Collaboration as Technology Praeger Publishers. Strategy in Japan." Science, Technology, Management. MIT Japan Program. Cambridge, Mass.: MITJSTD. Rosenbe:rg, Nathan. 1990. "Science and Technology Policy for the Asian NICs: Lessons from Economic His- Meyer-Krahmer, Frieder. 1990. Science and Technology in tory." In Robert E. Evenson and Gustav Ranis, eds., the Federal Republic of Germany. Harlem: Longmans. Science and Technology: Lessons for Development Policy. Boulder, Colo.: Westview Press. Meyer-Krahmer, Frieder, and Philippe Montigny. 1989. Evaluations of Innovation Programs in Selected European Ranis, Gustav. 1990. "Science and Technology Policy: Countries in Research Policy. Karlsruha, Germany: ISI. Lessons From Japan and the East Asian NICs." In Rob- ert E. Evenson and Gustav Ranis, eds. , Science and Tech- Mody, Ashoka. 1989. "Institutions and Dynamic Com- nology: Lessons for Development Policy. Boulder, Colo.: parative Advantage: Electronics Industry in South Ko- Westview Press. rea and Taiwan." Washington, D.C.: World Bank/IENIN. Schimank, Uwe. 1990. "Technology Policy and Technol- Najmabadi, Farrokh, and Sanjaya Lall. 1995. "Develop- ogy Transfer from State Financed-Research Institutions ing Industrial Technology: Lessons for Policy and Prac- to the Economy: Some German Experiences." Science and tice." Washington, D.C.: World Bank/OED. Public Policy, vol. 17, no. 4, pp. 219-228. Nelson, Richard, and Sidney Winter. 1982. An Evolution- Schmitz, Hubert. 1993. Small Shoemakers and Fordist Gi- ary Theory of Economic Change. Cambridge: Harvard Uni- ants: Tale of a Supercluster. Discussion Paper 331. versity Press. Brighton, England: University of Sussex, Institute of De- velopment Studies. Nelson, Richard, ed. 1993. National Innovation Systems: A ComparativeAnalysis. New York: OxfordUniversityPress. Simon, Denis Fred. 1992. "Taiwan's Emerging Technologi- cal Trajectory: Creating New Formns of Comparative Ad- Okada, Yoshitaka. 1990. "Technological Development and vantage." In Denis Fred Simon and Michael Y.M. Kau, eds., Growth of Japanese Integrated Circuit Firms: An Explor- Taiwan: Beyond the Economic Miracle. Taipei: M.E. Sharpe. Institutions and Policiesfor Industrial Technology Development 43 Tassey, Gregory. 1992. Technology Infrastructure and Com- Technology Development." Beijing: Center for Science petitive. London: Kluwer Academic Publishing. and Technology Development. Beijing: Center for Sci- ence and Technology Development. Wade, Robert. 1990. Governing the Market: Economic Theory and the Role of Government in East Asian Industrial- Wang Huijiong and Lu Wei. 1995. "China's Textile In- ization. Princeton, NJ: Princeton University Press. dustry: A Profile of Development and the Impact of its Technological Institutions." Development Research Cen- Westphal, Larry E., Linsu Kim, and Carl J. Dahlman. 1995. ter, State Council. Beijing, People's Republic of China. "Reflections on The Republic of Korea's Acquisition of Technological Capability." In Nathan Rosenberg and Hungary: Claudio Frischtak, eds., International Technology Transfer: Concepts, Methods, and Comparisons. New York: Prager. Gyorgy, Katalin. 1997. "R&D and Cooperation in the Hungarian Pharmaceutical Industry." Innovacios Kutato Womack, James P., Daniel T. Jones, and Daniel Roos. 1990. Kozpont (Innovation Research Centre), Budapest. The Machine that Changed the World. New York: Rawson Associates. Havas, Attila. 1997. "The Re-Emergence of the Car Parts Industry in Hungary." Innovacios Kutato Kozpont (In- Westphal, Larry E., Yung W. Rhee, and Gary Pursell. novation Research Centre), Budapest. 1984. "Sources of Technological Capability in South Ko- rea." In Martin Fransman and Kenneth King, eds., Tech- Inzelt,Annamaria. 1994. "Institutional Support for Tech- nological Capability in the Third World. New York: St. nological Improvement: Back to Cooperation from Martin's Press. Rigid Separation-Hungarian Institutional Adventures in the 20th Century." Innovacios Kutato Kozpont (In- Westphal, Larry E., Yung W. Rhee, and Gary Pursell. novation Research Centre), Budapest. 1981. "Korean Industrial Competence: Where It Came From." World Bank Staff Working Paper No.469. Wash- Mosoni-Fried, Judith. 1994. "Industrial Research Insti- ington, D.C. tutes in the Transition Period in Hungary." Innovacios Kutato Kospont (Innovation Research Centre), Budapest. Reports from the Institution and Policies for Industrial India: Technology Development Study. Chaudhuri, Shekhar. 1995. "The Evolution of the Indian Canada: Pharmaceutical Industry: Government Policy, Indus- try Structure and Technological Development." Landry, Rejean. 1994. "Technological Services Provided Ahmedabad, India: Indian Institute of Management. by Institutions to Canadian Manufacturing Firms." Re- search Group on Government Interventions, Political Sci- . 1994. "Technology Development in the Indian ence Department. Quebec, Canada. Textile Industry: Interaction Between Government Policy, Firms and Cooperative Research Associations." China: Ahmedabad: Indian Institute of Management. Xu, Zhao-xiang, Wang Huijiong, Lu Wei, Tang Shi-guo, Chaudhuri, Shekar, and C.B. Dasgupta. 1994. "Indian Chen Xiao-Long, Ma Chi, Yang Tien-zheng, Zhao Jie. Software Industry." Ahmedabad: Indian Institute of 1994. "Policy and Institutional Priorities for Industrial Management. 44 Chaudhuri, Shekhar, M. R. Dixit, S. Ramnarayan, C. B. Korean Textile Industry." Yonsei University. Dasgupta, et al. 1995. "International Study on Policy and Institutional Priorities for Industrial Technology Devel- Yoon, Se Joon, and Kim, Sang Pyo. 1996. "Institutional- opment India," Vol. I. Ahmedabad, India. Indian Insti- ization o{E Technology Development in the Korean Au- tute of Management. tomotive Parts Industry." Yonsei University. Dixit, M.R. 1996. "Indian Auto Components Industry Mexico: Learning to Export and Grow." Ahmedabad, India: In- dian Institute of Management, 1994 (rev. 8/96). Unger, Kurt, Javier Jasso, Victor Paredes, and Arturo Torres. 1995. "International Study on Policy and Institu- Ramnarayan, S. 1994a. "Technology Development in the tional Priorities for Industrial Technology Development: Indian Foundry Industry: A Case of Choked Potential." Mexico." Mexico: CIDE. Ahmedabad, India: Indian Institute of Management. Taiwan (C) .1 994b. "Tough Challenges Facing the Indian Poly- mer Industry: Unshackling the Follower Mindset." San Gee, Chen, Ying-chen et al. 1995. "Technology Sup- Ahmedabad, India: Indian Institute of Management. port Institutions and Policy Priorities for Industrial De- velopment in Taiwan, R.O.C." Chung-Hua Institution Japan: for Economic Research (CIER). Maeda, Yasuo. 1996. "The Case of the Japanese Textile . 1997. "Taiwan Auto Parts Industry." Taipei. Industry." Osaka University, Nigata. Forthcoming Publications Ohmori, Tatsuya. 1996. "Japanese Auto Parts." Seigakuin University, Tokyo. Kojima, Sakura, and Yoshitaka Okada. 1997. "Catching Up to Leadership: The Role of Technology-Support In- Okada, Yoshitaka. 1997. "Cooperative Learning and stitutions in Japan's Casting Sector." World Bank Tech- Technology Development in Japan," Tokyo. nical Paper No. 384. Washington, D.C. Wada, Masatake. 1994. "Contribution of Technology Sup- Felker, Greg, Shekhar Chaudhuri, Katalin Gyorgy, with porting Institutions to Technological Innovation in Japa- Melvin Goldman. "The Phamaceutical Industry: Policies, nese Software Industry." Tokyo. Institutions and Technological Development in India and Hungary." World Bank Technical Paper No. 392. Wash- Watanabe, Shinichi. 1995. "Evolution of Technologies and ington, D.C. Institutions: A Case Study of the Polymer Industry." In- temational University of Japan, Nigata. Felker, Greg, Se Joon Yoon, Sang Py Kim, Tatsuya Ohmori, M. R. Dixit, Attila Havas, and San Gee. 1997. Korea: "Technological Development in the Automobile Parts Industry--Five Paths." World Bank, Washington, D.C. Kim Kee Young, Sqo Oh, Se JoonYoon, and Young Phil Park 199£ "Policies and istitutions for Industrial and Technological Okada, Yoshitaka, Sakura Kojima, Tatsuya Ohmori, DevelopmentAKoreaStudy."Seoul,KoiraYonseiUniversity. Shinichi Watanabe, and Yasuo Maeda. 1988. "Institu- tions and Cooperative Learning for Technology Devel- Kim, Kee Young, Moon Il Lee, and Young Min Jang. 1996. opment in Japanese Industry." World Bank, Washing- "Institutionalization of Technology Development in the ton, D.C. Distributors of COLOMBIA GERMANY ISRAEL NEPAL PORTUGAL SWEDEN Distributors of Infoenlace LUda. UNO-Vedag Yozmot Literature Ltd. Everest Media Intemational Services (P.) Ltd. Livrara Ponugal Wennergren-Williams AB W orld Bank Carrera 6 No, 51-21 Poppelsdorter Allee 55 PO. Box 56055 GPO Box 5443 Apartado 2681, Rua Do Carmo 70-74 P0. Box 1305 Apartado Aereo 34270 53115 Bonn 3 Yohanan Hasandlar Street Kathmandu 1200 Usbon S-171 25 Solna Publications Santald de Bogotd, D.C. Tel: (49 228) 949020 Tel Aviv 61560 Tel: (977 1)472 152 Tel: (1) 3474982 Tel: (468)705-97-50 Prices and credit termvsvanifrom Tel: (57 1)285-2798 Fax: (490228) 217492 Tel: (9723) 5285-397 Fax: (977 1)224431 Fax: (1) 347-0264 Fax: (468) 27-00-71 country to cntrtry Consutt your Fax: (57 1) 285-2798 URL: hrtp)Avww.uno-verIag.de Fax: (972 3) 5285-397 E-mail: mail@wwi.se local distriutor befiir placing anE-mnail: unovedag@aol.com NETHERLANDSRMAA localdlstibutr beore lacig anCOTE D'IVOIRE E-mail: unoverlag@aol.com R.O.Y International De LindeboomnlnOr-Publikaties Compani De Librarii Bucuresti S.A. SWITZERLAND order. Center d'Edition et de Diffusion Atricaines GREECE PO Box 13056 PO. Box 202, 7480 AE Haaksbergen Str. Lipscani no. 26, sector 3 Librairie Payot Service Institunonnel (CEDA) Papasotiriou S.A. Tel Aviv 61130 Tel: (31 53) 574-0004 Bucharest Cotes-de-Montbenon 30 ARGENTINA 04B.P541 35, Stoumara Str. Tel: (9723)5461423 Fax: (31 53) 572-9296 Tel: (40 1)613 9645 1002 Lausanne Oficinadel Libro Intemacional Abidan 04 106 82Athens Fax: (972 3) 5461442 E-mail: lindeboo@worldonline.nl Fax: (40 1) 312 4000 Tel: (41 21) 341-3229 Av1 Counoba 1A77 Tel: (225) 24 6510;24 6511 Tel: (301)364-1826 E-mail: royil@netvision.net.il URL:thttp:llwww.woridonline.nv-lindeboo Fax: (41 21) 341-3235 1120 Buenos) 1res Fax: (225)25 0567 Fax: (301) 364-8254 RUSSIAN FEDERATION Tel: (54 1) 815-8354 Palestinian Authority/Mikdle East NEW ZEALAND Isdatelstvo ADECO Van Diemmen EditlonsTechnkItues Fax: (54 1) 815-8156 CYPRUS HAITI Index Information Services EBSCO NZ Ltd. 9a, Lolpachniy Pernulok Ch. de Lacuez 41 AUSTRALIA, FIJI, PAPUA NEW GUINEA, Center tor Applied Research Cus0ur Diffusion P0.B. 19502 Jerusalem Private Mail Bag 99914 Moscow 101831 CH1807 Blonay SOLOMON ISLANDS, VANUATU, AND Cyprus College 5, Rue Onpols Tel: (972 2) 6271219 New Market Tel: (7 095( 917 87 49 Tel: (41 21) 943 2673 WESTERNSAMOA,FIJI,PAPUANEWGUIN 6, Diogenes Street, Engomi C.P 257 Fax: (9722)6271634 Auckland Fax: V7095)9179259 Fax: (41 21) 943 3605 WESTERN SAMOA P~~O. Boa 2006 Pad-an-Prince Tel: (649) 524-8119 DA. Information Services Nicosia Tel: (509)23 9260 ITALY Fan: (64 9) 524-8107 SINGAPORE, TAIWAN, THAILAND 648 Whnehorse Road Tel: (3572)44-1730 Fax: (509)234858 Licosa Commissnbnaria Sansoni SPA MYANMAR, BRUNEI Central Books Distribution MVcham 3132 Fax: (357 2) 48-2051 Via Duca Di Calabria, /1 NIGERIA Asahgate Publishing Asia Pacitc PIe. Ltd. 306 Silom Road Vlctorel HONG KONG, MACAO Caselba Postale 552 University Press Limited 41 Kallang Pudding Road #04-03 Bangkok 10500 Tel: (61)392107777 CZECHREPUBLIC Asia 2000 Ltd. 50125 Firenze Three Crowns BuildingJericho Golden Wheel Building Tel: (66 2)235-5400 Fax: (61) 3ec210i77 c8 National Infornation Center Sales & Circulation Department Tel: (55) 645-415 Private Mail Bag 5095 Singapore 349316 Fax: (662) 237-8321 E-mall:servc0dadireccomau prodejna,Konvikdska 5 Seabird House, umnt 1101-02 Fan: (55) 641-257 Ibadan Tel: (65)741-5166 URL: lhttp:lwww .dadirect.com.au CS -113 57 Prague 1 22-28 Wyndham Street, Central E-mail: licosa@fthbcc.it Tel: (234 22) 41-1306 Fax: (65) 742-9356 TRINIDAD & TOBAGO, AUSTRIA Tel: (422) 2422-9433 Hong Kong URL: httpJww.ftbccitAlicosa Fax: (234 22) 41-2056 E-mail: ashgate@asianconnect.com AND THE CARRIBBEAN GeSTRAl and Co.Fax: (42 2) 2422-1484 Tel: (852) 2530-1409 Systematics Studies Unit Weihburgganse 26 URL: http://wwW.nis.cz1 Fax: (852)2526-1107 JAMAICA NORWAY SLOVENIA 9Wa8sStreet A-lollrgrassen26 E-mail: saleo@asia2000comEhk Ian Randle Publishers Ltd NIC Info A/S Gospodarski Vesnik Publishing Group Curepe A-1011 27ien DENMARK URL: http/1www.asia2000.com.hk 206 Old Hope Road, Kingston 6 Book Depanment, Postboks 6512 Eterstad Dunajskacesa 5 Trinfhad, West lndies Tel: (43 1) 512-47-3140 SamfundsLideratur Tel: 876-927-2685 N-0606 Oslo 1000 tLubljana Tel: (809)662-5654 Fax: (43 1)p512c47-31-2s RosenoemsAlle 11 HUNGARY Fax: 876-977-0243 Tel: (4722)97-4500 Tel: (38661) 1338347 1321230 Fax: (809)662-5654 URL: http1Avww.gero1d.co1at.cm1ine DK-1970 Fredeniksberg C Euro Info Service E-mail: irpl@colis.com Fax: (47 22) 97-4545 Fax: (386 61) 133 80 30 E-mail: tobe@trinklad.net BANGLADESH Tel: (45 31) 351942 Margitszgeti Europa Haz E-mail: repansekj gvestnh.si UGADA Micm IndusLries Development Fax: (45 31) 357822 H-1138 Budapest JAPAN PAKISTAN Assrstunce Socinly (MIDAS) URL: hllp:llwwwS.bSl.cd.k Tel: (36 1( 111 6061 Eastem Book Service Mirza Book Agency SOUTH AFRICA, BOTSWANA Gustro Ld. Assistance Society (MIDAS) Fax: (36 1) 302 5035 3-13 Hongo 3-chome, Bunkyo-ku 65, Shahrah-e-Quaide-kAzam Forsingle tites: PO Box 9997, Madhvani Building Hounmondi RoArea ECUADOR E-mail: eunoinfoumail.matav.hu Tokyo 113 Lahore 54000 Oxford Universiy Press Soalhem Aftia Plt 16/4 Jhja Rd. Dhanmondi R/Area Libri Mundi Tel: (81 3) 3818-0861 Tel: (92 42) 735 3601 Vasco Boulevard, Goodwood Kampabl Dhaxa 12u9 Librerialntemachonal INDIA Fax: (813)3818-0864 Fax: (9242) 576 3714 PO. Box 12119, NI Ciy 7463 Tel: (256 41) 251 467 Tel: (880 2) (36427 PO. Box 17-01-3029 Allied Publishers Ud. E-mail: onrders@svt-ebs.co.jp Cape Town Fax: (256 41) 251 468 Fax: (8602)811188 Juan Leon Mera 851 751 Mount Road URL: htp-Jtwww,bekkoame.or.p/-svt-ebs Oxford University Press Tel: (27 21) 595 4400 E-mail: gus@switugandai.oom BELGIUM Osuo Madras - 600 002 5 Bangalore Town Fax: (27 21) 595 4430 Jean De Lannoy Tel: (593 2) 521-606, (593 2) 544-185 Tel: (91 44) 852-3938 KENYA Sharae Faisal E-mail: oxnord@oup.co.za UNITED KINGDOM An. du Roi 202 Fax: (593 2) 504-209 Fax: (91 44) 852-0649 Africa BDok Service (E.A.) Ltd. PO Box 13033 Micronto Ltd. 1060 Bduneoi2 E-mail: librimul @librimundi.com.ec Quaran House, Mtangano Street Karachi-75350 For subscptlion onrfers: PO. Box 3, Afton, Hampshire GU34 2PG Tel: (322(53-5169 E-mail: librimu2@1ibrimundi.com.ec INDONESIA P.. Box 45245 Tel: (92 21) 446307 Intemational Subscription Service England Tel: (32 2) 538-5169 ~~~~~~~Pt. Indira Limited Nairmbi Fan: (92 21( 4547640 P.O. Boo 41665 Tel: (44 1420) 86848 Fax: (32 2) 538-0841 EGYPT, ARAB REPUBLIC OF Jalan Borobudur 20 Tel: (254 2) 223 641 E-mail: ouppak@TheOffice.net Cmaighall Fax: (441420) 89889 BRAZIL Al Ahram Distdbution Agency P.O. Box 181 Fax: (254 2) 330 272 Johannesburg 2024 E-mad: wbank0ukanio.demon.co.uk BRcAcZeL TcinlomcoacLda Al Galaa Street Jakada 10320 Pak Bask Corporation Tel: (27 11) 880-1448 URL: hffp-Jtwww .microinfo.co,uk Pubkacs es Tecnicas 2ntemacionais Llda. Cairo Tel: (62 21) 390-4290 KOREA, REPUBLIC OF Aziz Chambers 21, Queen's Road Fax: (2711) 880-6248 01409uSaPeo Gomide, 209 Tel: (20 2) 578-6083 Fax: (62 21) 390-4289 Dae)on Trading Co. Ltd. Lahore E-mail: ius@is.coza VENEZUELA 01409 San Paulo, SF Fax: (20 2) 578-6833 PO. Box 34, Youida, 706 Seoun Bldg Tel: (92 42) 636 3222; 636 0885 Tecni-Cencria Libros, S.A. Tel: (55 11) 259-6644 IRAN 44-6 Youido-Dong, Yeongchengpo-Ku Fax: (92 42) 636 2328 SPAIN Cenlo Cuidad Comescial Tamanco Fax: (55 11) 258-6990 The Middle East Observer Ketab Sara Co. Publishers Seoul E-mail: pbc@hrain.net.pk Mundi-Prensa Lrbros, S.A NMel C2, Caracas E-mall: postmaster@pti.uol.br 41, Sheri Street Khaled Eslamboli Ave., 6th Street Tel: (822)785-1631/4 Castell37 Tel (582)9595547;5035:0016 URL: htp://WwwuoI.brCairo Delafmooz Alley No. 8 Fmx: (82 2(784-0315 PERU 28001 Madrid Fax: (58 2) 959 5636 CANADA Tel: (20 2) 393-9732 PO. Box 15745-733 Ediorial Desarmollo SA Tel: (341) 431-3399 Rensuf Publrshing Co. Ltd. Fax: (20 2(393-9732 Tehran 15117 MALAYSIA Apurtado 3824, Uma 1 Fux: (34 1) 575-3998 ZAMBIA Renoui Publishing Co.RLtd. Tel: (98 21) 8717819,8716104 University on Malaya Cooperative Tel: (51 14) 285380 E-mail: lir3ath @mundiprenua.es Unrversity Bookshop, Universry of Zambia 53t9 Canotek Road FINLAND Fax: (98 21) 8712479 Bookshop, Limied Fax: (51 14) 286628 URL: http-A/www .mundipmensa.es/ Great East Road Campus el:awa, Ontario Kl J 93 Akateeminen Kirjakauppa E-mail: ketab-sara@neda.net.ir PO. Box 1127 P.O. Box 32379 Tel: (613) 745-2665 P.O. Box 128 Jalan Pantai Baru PHILIPPINES Mundi-Prensa Barcelona Lusaka Fax: (613) 745-7660 FIN-00101 Helsinki Kowkab Publishers 59700 Kuasl Lumpur Intemational Booksource Center Inc. Consell ide Cent, 391 Tel: (260 1) 252 576 URL h9p:ll urwerd renoafbooko.com Tel: (3580)121 4418 P.O. Box 19575-511 Tel: (603)756-5000 1127-A Antipolo St, Barangay, Venezuela 06009 Barcelna Fax:(2601)253952 URL: hftp:/f www ranoutbooks.c m Fax: (3580)121-4435 Tehran Fax: (603)755-4424 MakatiCity Tel:(343)438-3492 CHINA E-mail: akatilaus@stockmannJi Tel: (98 21) 258-3723 Tek(632)8966501;6505;657 Fax: 343)487-7659 imbabwe China Financial & Economic URL: hipK/www.akateeminen.com/ Fax: (98 21)258-3723 MEXICO Fax: (632)8961741 E-mail: barcelona@mundiprens.e ToureRoad Ardbennie Publishing House FRANCE IRELAND Av. San Femando No. 37 POLAND SRI LANKA, THE MALDIVES P.O. Box ST1 25 8, Da Fo Si Dong Jie Wodd Bank Publications Govemment Supplies Agency Col. Tonetlo Guerra Intematinal Publishing Service Lake House Bookshop Southerton Beijing 66, avenue diena Odtig an tSoiahair 14050 Mexico, D.E Ul. Piekna 31/37 130, Sir Chitampalam Gardiner Mawatha Harare Tel: (86 10) 6333-8257 75116 Pads 4-5 Harcout Road Tel: (525) 624-2800 00-677Warzawa Cobmbo2 Tel: (263 4) 6216617 Fax: (86610) 6401-7365 Tel: (33 1)40-69-30-56!57 Dublin 2 Fax:(525)624-2822 Tel: (482) 628-6089 Tel: (94 1)32105 Fax: (2634)621670 Fax: (331)40-69-30-68 Tel: (3531)661-3111 E-mail: infotec@r9.nef.mx Fair (482)621-7255 Fax: (94 1)432104 Fax: (353 1)475-2670 URL: http/rtn.net.mx E-mail: books%ips@ikp.atm.como.p E-mail: LHL@nrilanka.net 08AN97 URL: htpYhvww .ipscg.wawpIpts'exporto RECENT WORLD BANK TECHNICAL PAPERS (continued) No. 345 Industry and Mining Division, Industry and Energy Department, A Mining Strategyfor Latin America and the Caribbean No. 346 Psacharopoulos and Nguyen, The Role of Government and the Private Sector in Fighting Poverty No. 347 Stock and de Veen, Expanding Labor-based Methodsfor Road Works in Africa No. 348 Goldstein, Preker, Adeyi, and Chellaraj, Trends in Health Status, Services, and Finance: The Transition in Central and Eastern Europe, Volume II, Statistical Annex No. 349 Cummings, Dinar, and Olson, New Evaluation Proceduresfor a New Generation of Water-Related Projects No. 350 Buscaglia and Dakolias, Judicial Reform in Latin American Courts: The Experience in Argentina and Ecuador No. 351 Psacharopoulos, Morley, Fiszbein, Lee, and Wood, Poverty and Income Distribution in Latin America: The Story of the 1980s No. 352 Allison and Ringold, Labor Markets in Transition in Central and Eastern Europe, 1989-1995 No. 353 Ingco, Mitchell, and McCalla, Global Food Supply Prospects, A Background Paper Prepared for the World Food Summit, Rome, November 1996 No. 354 Subramanian, Jagannathan, and Meinzen-Dick, User Organizationsfor Sustainable Water Services No. 355 Lambert, Srivastava, and Vietmeyer, Medicinal Plants: Rescuing a Global Heritage No. 356 Aryeetey, Hettige, Nissanke, and Steel, Financial Market Fragmentation and Reforms in Sub-Saharan Africa No. 357 Adamolekun, de Lusignan, and Atomate, editors, Civil Service Reform in Francophone Africa: Proceedings of a Workshop Abidjan, January 23-26, 1996 No. 358 Ayres, Busia, Dinar, Hirji, Lintner, McCalla, and Robelus, Integrated Lake and Reservoir Management: World Bank Approach and Experience No. 360 Salman, The Legal Frameworkfor Water Users' Associations: A Comparative Study No. 361 Laporte and Ringold. Trends in Education Access and Financing during the Transition in Central and Eastern Europe. No. 362 Foley, Floor, Madon, Lawali, Montagne, and Tounao, The Niger Household Energy Project: Promoting Rural Fuelwood Markets and Village Management of Natural Woodlands No. 364 Josling, Agricultural Trade Policies in the Andean Group: Issues and Options No. 365 Pratt, Le Gall, and de Haan, Investing in Pastoralism: Sustainable Natural Resource Use in Arid Africa and the Middle East No. 366 Carvalho and White, Combining the Quantitative and Qualitative Approaches to Poverty Measurement and Analysis: The Practice and the Potential No. 367 Colletta and Reinhold, Review of Early Childhood Policy and Programs in Sub-Saharan Africa No. 368 Pohl, Anderson, Claessens, and Djankov, Privatization and Restructuring in Central and Eastern Europe: Evidence and Policy Options No. 369 Costa-Pierce, From Farmers to Fishers: Developing Reservoir Aquaculturefor People Displaced by Dams No. 370 Dejene, Shishira, Yanda, and Johnsen, Land Degradation in Tanzania: Perceptionfrom the Village No. 371 Essama-Nssah, Analyse d'une repartition du niveau de vie No. 373 Onursal and Gautam, Vehicular Air Pollution: Experiencesfrom Seven Latin American Urban Centers No. 374 Jones, Sector Investment Programs in Africa: Issues and Experiences No. 375 Francis, Milimo, Njobvo, and Tembo, Listening to Farmers: Participatory Assessment of Policy Reform in Zambia's Agriculture Sector No. 376 Tsunokawa and Hoban, Roads and the Environment: A Handbook No. 377 Walsh and Shah, Clean Fuelsfor Asia: Technical Options for Moving toward Unleaded Gasoline and Low-Sulfur Diesel No. 382 Barker, Tenenbaum, and Woolf, Governance and Regulation of Power Pools and System Operators: An International Comparison No. 385 Rowat, Lubrano, and Porrata, Competition Policy and MERCOSUR No. 386 Dinar and Subramanian, Water Pricing Experiences: An International Perspective No. 387 Oskarsson, Berglund, Seling, Snellman, Stenback, and Fritz, A Planner's Guide for Selecting Clean-Coal Technologies for Power Plants No. 388 Sanjayan, Shen, and Jansen, Experiences with Integrated-Conservation Development Projects in Asia No. 389 International Commission on Irrigation and Drainage (ICID), Planning the Management, Operation, and Maintenance of Irrigation and Drainage Systems: A Guidefor the Preparation of Strategies and Manuals THE WORLD BANK I SI I I StircCi. .\\ \\islhmgon, I ).(. 20)433 .. S \ ICLph1ml.. 21)2-477-1234 IF,icsimilc: 202-477-63(91 TCie\: \1(.1 (4145 \\( )R I)BIANk1 \1(I 24S423 \\ OR 11)1I \VNk \\oi-IdV\\IdC \\cl: Iittp: xx xx.11i ;,, 1nk.,r I :I-i W .1k I _ _ (I 'II I i ~ LI III) ',IIA (II aju , . cbui of terr"a \l:c 'I ,"M I(CII ISBN 0-8213-4052-2