.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4!~11 44 *tT *I L,' 1. O'i; EAi . ("'t M # k A C l., ,,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~WI,-r it,;t4 '' I!WSbi7S'XtWs6Ni SHS'WlIF .8e,.l.e iXS4 NE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4 F ~~~~~~~~~~~~~~~~~~~~~~~~~~~1 S ~ ~~~~~~~~~.'' s~~~~~~~~~~~~F f) Cll rOS ( :.; ] I X ,0 g r X~~~I Lul=lpph. ; - .r. _' d 0 ;. 0 X\ 0 fib~~~~~~~~~~~~~~~~~~~~~- Jakarta z * *NWLI - . I) 6 :6 S - . MF C)~~~C 0 The East Asian Miracl e Economic Growth and Public Policy Published for the World Bank CxCORD UNIVERSrrYPRS Oxford Univeirm Pressw OXFORD NEV YORK TORONTO DELI BOMBaY CALC;UJrA MADRAS KARACHi kUAA LUMPUR SINGAPORE HONG KONG TOKYO NAIROBI DAR ES SALAAM CAPE TOWN MELBOURNE AUCKLAND and asidad compania in BERLIN IBADAN 0 2993 7lmriarmnd Ba&nkfor Reomtruawn wdDevaop mewl OPJE U'2LD HAAh 1818 H Sreet N.W. Mathinog D.C 20433. USA Pub hed by Oxfrd UniAnityPrs, fINC. 200 MAdiwrAvemd4 New York N. Y 10016 Oxford is a rqwra bradeark of Oxford UnirsiyPreaz A(1rgth raerd Noparjrh5pub&icatum 6be reprw l .woredin a rrtrudqsfem or Drwmm&mc4 in any farm or by any meas, deamn, merhank4 pumwpy4 rewonmg& ore.z se,i w etvepdorpermrnuonfxf0uirdUriviy Prm Mamftiazdn she Unisd s /Ameria Fon paning&Spemner 1993 Ciogun ph s-At the wp. a ricefud inJava, Indoniaig; c resy ofMatricrAiser. At r beraomI , the por ofPuan, Rephfir ofKar=cowty ofJindre &pjDA The map tnappars o n the izside rova, wbib shws t6c eight eom wigs diiwed in Mis bok wsprqmred soyfortke onvenie ofthe reader the dres-arions andpresramt of maeial in it d wet imply the owrr ofrmy opbun warsoener on the pavt of th World Bank its affiiner or isr Bad ormember conries onwMtming the egl as1 0f ay ewonM, mq rroj cirt or or /t&e anthrifia xref, or rcwemning rthedeimiwdon qfirs boitames or in natiolaie rfCwFs Caragtoing-Publadon Dar The EasAswaw mirade w :omwgrvww5 =d; bicpAl p. cn.-(MordBankpdiy r=* reponw) Incdaw r 3uietpieal ner ,IrN 0-19-5209934I L. AsEa;rnee-Ecwmic pli 2- tE a-Econidc cendo-, 3. Asi Sawi&s-Ecenamicpole 4. Ask CitiomS L. WorldBank f. Se,ie HC4625.A275 1.993 338.95-c20 9S33046& ismv 1G20-0851 Torprepaper anom the Ameria Natnax SxdrdfrPawzw*w wf P1pfr1Jiwd MreLw ialr Z39.48-1984 Foreword EBATES ON THE APPROPRIATE ROLE OF PUBLIC POUCY IN economic development have occupied policymakers and scholars since the study of developing economnies bcgan in earnest at the dosc of World War IL. The success of many of the economies in EastAsia in achieving rapid and equitable growth, often in the context ofacivist public policies, raises complex questions about the relationship between government, the private sector, and the markern Seningly, the rapidly growing economies in East Asia used many of the same poficy instruments as othcr developing economies, but with greater success. Understanding which policies contribuced co their rapid growth, and how, is a major question for research on development pol- icy. Forthese rcasons I announced at the time of the 1991 Annual Meet- ings of the Board of Govemors of the World Bank in Bangkok, Thailand, that our Development Economics Vice Presidencywould un- dertake a comparative study of economic growth and public policy in East Asia. This volume is the summary ch. appears as the first in a series of Policy Rcsearch Reports, which arc intended to bring to a broad audience the results of research on development policy issues carried out by staff of the World Bank.As reports on policy issues, we intend that they should help us to take- stock of what we know and cIealy identifir what we do not know; they should contribure to the de- bat in both the academic and poli-y communities on appropriate pub- lic policy objectives and instruments for developing economies; and they should be accessible tX nonspecialiscs. Because they sunmarize re- search, we also anticipate that Policy Research Reports will provoke far- ther debate, both within the Bank and outside, conceming the methods used and the conclusions drawn. What does this report tell us about the East Asian miracle? The re- seawh shows that most of EastAsias extraordinary growth is due co su- perior accumulation of physical and human capita. But these economies Were also better able than most to allocate physical and buman resources to highly productive investrents and to acquire and master txchnology. In this sense there is nothing 'miraculous" about the East Asian economics' success; each has perfirmed these essential func- tions of growth better than mosr other economies. The eight economics studied used very different combinations of policies, firom hands-ofTto highly interventionist. Thus, them is no sin- gle 'East Asian moder of development. This divcrsity of experience re- inforces the view that economic policies and policy advice must be country-specific, if they are to be effectivec But there arc also some com- mon threads among the high-performing East Asian economies. The authors conclude that rapid growth in each cconomy was primarily due to the application of a set of common, market-friendly economic poli- cies, leading to both higher accumulation and berter allocation of re- sources. While this condusion is not strikingly new, it reinforces other rcsearch that has sressed the essential need for dcveloping economics to ger the policy fundamentals igIL The research also firther supports the desirability of a two-track approach to development policy emphasizing macroeconomic stability on one hand and investments in people on the ocher. The importance of good macroeconomnic management and broadly based educational systems for Easr Asia's rapid growth is abun- dandly demonstrated. The reporr also breaks some new ground. It condudes that in some economics, nainly those in Northeast Asia, some selective interventions contibuted o grwth. and it advances our undersmading of the condi- tions required for interventions to succeed. The authors argue thar where selective intervcntions succceded they did so because of th o es- sential prcrequisites. First, they addressed problems in the functioning of markets. Second, they took place within the contex of good, finda- mental policies. Third, their success depended on the ability of govern- merts to establish and monicor appropriate economic-performance criteria related to the interYcntions-in the authors' terms, to create eco- nomic contests. These prerequisites suggesr that the institional con- text withLin which policies are implemented is as important to their success or failure as the policies themselvcs, and the report dcvotes sub- stantial artention to the institutional bases for East Asi-a rapid growth. While these factors help to explain why apparently sinilar policies did not succeed in mrany other economies, the report also leaves unan- swered many important questions. The market-oricnted asj ctxs of East Asia!s polides can be recommended with few reservations, but the more vi institutionally demanding aspects, stich as contest-based interventions, have nor been successfully used in other settings. Noneconomic factors, induding culture, politics, and history, are also important to the East Asian success story. Thus, there is still much to be learned about the in- teractions between policy choices and inscitutional capability and be- tween economic and noneconomic factors in development. Work in these areas will continue beyond this report. The support of the Government ofjapan for the research program on the high-performing Asian economics is gratfilly acknowledged. The report is a product of the staff of the World Bank, and the judgments made herein do not necessarily reflect the view of its Board of Directors or the governments they rcprescnt. Lewis T. Preston President The World Bank Augur 1993 vii Contents 7Te Rear Team xi Ack3wwkdgmnae xiii Defmiufons xv Overview-r The Making of a Miracle The Essence of the Mirale: Rapid Growth with Equity 8 Policies for Rapid Growth in a Changing World Economy 23 Note 26 1 Growth, Eqrity, and Economic Change 27 Rapid and Sustained Economic Growth 28 Declining Income Inequality and Reduced Poverty 29 Dynamic Agricultural Seaors 32 Rapid Growth of Exports 37 Rapid Demographic Transitions 38 High Investment and Savings Rares- 40 Creatng Human Capital 43 Rapid Productivity Growth 46 Appendix 1.1: Accounting for Growth 60 Appendix 1.2: What Do Tess of Cogniive Skills Show?- 70 Notes 76 2 Public Policy and Growth 79 PolicyExplanations 81 The Functional Growth Framework 87 Notes 103 3 Macroeconomic Stabiity and Export Growth 105 Pragmatic Orthodoxy in Macroeconomic Management 106 Creatng an Export PuLsh 123 Appendix 3.1: Economric and Political Timedines 148 Notes 156 ix 4 An Insitutional Basis for Shared Growth 157 Achicving Legitimacy through Sharcd Growth 158 Insulating the Economic Technocracy 167 Wooing Big Business 181 Notes 189 5 Strategies for Rapid Accumulation 191 Explaining EastAsia's High Human Capital Forrnation 192 Explaining East Asia's High Savings Rates 203 Explaining East Asia's High Investment Rates 221 Appendix 5.1: Granger Causaliy Tests forSavings Rates and Growth Rates 242 Appendix 5.2: Technical Note on the Relationship beween Interest Rates and Growth 245 Notes 255 6 Using Resources Efficienty: Relying on Madrets and Exports 259 Explaining EastAsiais Efficient Resource Use 259 Using the Marken Labor Markets in East Asia 261 Assisting the Markct Financial Markets and Allocation 273 Using the Internarional Market Trmde and Industrial Policy 292 How Manufactured Exports Increased Productiviy 316 Appendix 6.1: Testing dhe Impact of Indusaial Policy on Productivicy Change 326 Appendix 6.2: Tests of the Relationship berween TFP Change and Trade Policies 337 Notes 342 7 Polices and Pragmatism in a Changing World 347 Foundations of Rapid Growdh-Getting dhe Fundamentals Right 347 Creating Institutions to Promote Growdi 352 Intervening in Markets 353 Note 368 Bibligraphic Note 369 The Research Team T HIS POLICYRESEARCH REPORTWAS PREPARED BYATEAM LED by John Pagc and comprising Nancy Birdsall, Ed Campos, W Max Corden, Chang-Shik Kim, Howard Pack, Richard Sabot, Joseph E. Stiglitz, and Marilou Uy. Robert Cassen, William Easrerly, Robert Z. Lawrence, Petcr Petri, and L-nt Pritchett made major contri- butions. Lawrence MacDonald was the principal editor. Case studies of the seven devcloping high-performing Asian economies werc under- taken under the direction of Danny Leipziger. The tcam was assisted by Maria Ltiisa Cicogniani, Varuni Dayaratna, Leora Friedberg, Jay Gon- zalez, Jennifer Keller, May Khamis, Sonia Plaza, Myriam Quispe, Carol Strunk, and Ayako Yasuda. The work was initiated by Lawrence H. Summers and was carried our under the general direction of Nancy Birisall. The editorial-production team for the report was led by Alfred Imlhof The support staff was headed by Sushrma Rjan and included Milagros Divino, Jan-Marie Hopkins, Anna Marie Maranon, and Linda Oehler Polly Means gave graphical assistance. Bruce Ross-Larson and Meta dc Coquercaumont provided additional editorial support The Map was designed by jeffrey N. Lecksel. Mikla Iwasaki coordinatcd work in Japan. XI Acknowledgments REPARATION OF THIS REPORT DREW ON SLVERAL OTHER RE- laced research projects. Studies of the Japanese main bank sysrcn and civil service werc carried out under the dirnction of Hyung-Ki Kim. Yoon-Je Cho and Dimirri Vittas directed the project on the effec- tiveness of crcdit policies in EastAsia. Brian Levy directed the project on support systems forsmall and medium-size enterprises in Asia, andJohn Page coordinated projects on lapanese perspectives on public policy dur- ing the rapid growth period, and tax policy and tax administration in East Asia. Many individuals inside and outside the World Bank provided valu- able contributions and comments. (Specific acknowledgmcnt of their efifrts is made in the Bibliographic Note at dce end of the book.) Par- ticular thanks are due ro the following senior officials of the economies studied who reviewed and comment"d on an earlier drafc Seung-Chul Ahn, Isao Kubota, Shijuro Ogata, J. B. Sumarlin, Teh Kok Peng, Chilao Tsukuda, Tan Sri Dato Lin See Yan, and Cesar Virata. Vinod Thomas comnmented on several drafts and coordinated the comments of his colleagues in the East Asia and Pacific Region of the Bank. Com- ments by Cad Dahlmaan, Kemnal Dervis, Mark Gersovitz, Ralph W Har- bison, Magdi kskander, Hyung-Ki Kim, Danny Leipziger. Johannes Linn, Gobind Nankani, Mieko Nishimizu, Guy P. Pfeffermann, D. C. Rao, ant Michael Walton contributed to the improvement of the book. Preparation of the report was aided by seminars organized by the In- donesian Economists Association, the Economic Society of Singapore, the Monetary Authority of Singapore, Stanford University, and the World Institute for Development Economic Researeh. The financial as- sistance of the Govemrnment ofJapan is gratefully acknowledged. ... Definitions Economy Groups OR OPERATIONAL AND ANALYTICAL PURPOSES. THE WORLD Bank's main criterion for dassifying economies is gross national F product (GNP) per capita. Every economy is dassified as low-in- come, middle-incomc (subdivided into lower-middle and upper-mid- dle), or high-income. Other analytical groups, based on -:gions, exports, and levels of extemal debt, are also used. Because of changes in GNP per capita, the economy composition of each income group may change from one Bank publication to the next Once the dassification is fixed for any publication, all the hisrorical data prescnted are basod on the same economy grouping. The income-based economy groupings used in this sEudy are defined as folows * Low-income economies are thosc with a GNP pcr capit of $635 or less in 1991. * Middle-inme ecoomzies are those with a GNP per capita of more rhan $635 but less than $7,911 in 1991. A fijrther division, atGNP per capita of $2,555 in 1991, is made between Iouer-midlk-in- cow and upper-midl-income conomies. * High-income economies are those with a GNP per capit of $7,911 or more in 1991. * Word comprises all economies, including economies with sparse data and those with fewer than 1 million people. Low-income and middle-income economies axe sormetimes referred to as developing economies. The use of the term is convenient; ir is not intended to imply that all economies in the group are experiencing sim- ilar development or that other economies have reached a preferred or final stage of development. Classification by income does not necessar- ily reflect development status. 'V -Anabtical Groups F OR ANALYIICAL PJRPOSES, THIS STUDY GROUPS ECONOMIES into several regions, defined as fIllows *High-perfirmingAsian economies (HPAEs), led byjapan, arc iden- tified by several common characteristics, such as very rapid ex- port growth. The HPAES are subclassified roughly according to the duration of their succcssfiul record of economic growth: . TheF : vur igers usually identified as Hong Kong, thc Republic of Korea, Singapore, and Taiwan, China, have been growing rapidly for decades and have joined or approached the ranks of high-income economies. * The newly industrialidZWg economies (NIEs) are Indonesia, Malaysia, and Thailand. They have joined the group of HPAEs more recently, within the last two decades. * EastAia comprises all the low- and middle-income economies of East and Southeast Asia and the Pacific, east of and including China and Thailand. * South Asizr comprises Bangladesh, Bhutan, India, Myanmar, Nepal, Pakistan, and Sri Lanka. * S&zb-&harn Af5ca comprises all econoniies south of the Sahara including South Africa but ecduding Mauritius, Reunion, and Seychelles, which are in the OtherAsia and islands group. * Europe Middle F."t and North Africa comprises the middle-in- come Europan economies of Bulgaria, the former Czechoslova- kia, Greec, Hungary, Poland, Portugal, Romania, Turkey, and the forrner Yugoslavia, and all the economies of North Africa and the Middle East, and Afghanisan. m Latin America and the GCaibcan cDmprises all American and Caribbean economies south of the United States. The regional grouping of economies in occasional parts of the text or tables may differ from that used in the main tt of this study, described above. Such variations are noced where dthy occur. Data Notes * Bilion is 1,000 million. * Trillion is 1,000 billion. - Dolars are current U.S. dollars unless otherwise specifiecd. * Growth ratesare based on constant price dart Historical data in this study may differ from those in other Bank pubLications because of continuous updating as better data become available, because of a change to a new base year for constant price data, or because of changes in economy composition of income and analyti- ml groups. Acronyms and Initials CPI Consumer price index DFI Direct foreign investment GDP Gross domestic product GNP Gross national product rlPAEs High-performing Asian economnies 1MF International Monetary Fund ISIC Intemational Standard Industrial Classification mI Ministry of Trade and Industry Japan NIs Newly industrializing economies OECD Organization for Economic Cooperation and Develop- ment (Australia, Austria, Belgium, Canada, Denmark, Fin- land, France, Germany, Greece, Iceland, Ireland, Italy, Japan, Luxembourg, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, United Kingdom, and United States) PPP Purchasing power parity R&D Research and devclopment TFP Total factor producdivity UNDP United Nations Development Programme UNESCo United Nations Educational, Scientific, and Cultural Organization -n. Overview: The Making of a Miracle AST ASIA HAS A REMARKABLE RECORD OF HIGH AND sustained economic growdh. From 1965 to 1990 the twenty-three economliies of East Asia grew faster than all other regions of the world (figure 1). Most of this achievement is attributable to seemingly miraculous growth in just eight economies: Japan; the 'Four Tigeser-Hong Kong, the Republic of Korea, Singapore, and Taiwan, China; and the three newly industrializing economies (NiEs) of South- east Asia, Indonesia, Malaysia, and Thailand. These eight high- performing Asian economies (HPAEs) are the subject of this study.* Selectinganysecofeconomiesand attemptingto undesand theorigins of their successful growth are necessarily arbitnary processes.' Botswana, Egypt, Gabon, and Lesotho in Sub-Saharan Africa have also been among the world's top growth performers in the past two decades, as have such diverse economies as Brazil, Cyprus, Greece, and Portugal (see figure 2). Why focus on eight economies in East Asia? In part the choice reflects popular interest; it has become common to see references to the "Asian Economic Miracle., In part it reflects recent artention by the aca- demic and development policy communities to the relationship be- tween public policies-which some authors have argued have a number of common threads in the eight economies, especially Japan, Korea, *Recendy China, particularly southern China, has recorded remarkably high growth - ates using policies that in some ways resermblc those of the HPAEs. This vtay significant development is beyond the scope of oursudy, mainly because China's ownership seruc- . ure, methods of corporate and civil govcrnance, and reliance on markets are so differ- ent from the those of the HPAEs, and in such mapid flux, that cross-economy comparison is problematic. We touch on China's recent development in chapters 1 and 3. Te eco nomic transiion in China is thesubject of current reseach by the Policy Research De- parment ofthe World Bank (see Bibliographic Note). -- : ~~~~~~~~~~~~~~~~~~~~~~~~~~~I 5 1jRRL CLE Figure 1 Average Growth of CNP per Capita, 1965-90 East Asia | _ ; _____ HPAEs East Asia without HPAES South Asbi Midae East and Meditermean.. Sub-Saharan Africa OECD ec__ , ; Latin Amedcm and Carlbbean 0 1 2 3 4 5 6 GNP per capit growth rate (percent) Skwre World Bank (1992d). Singapore, and Taiwan, China-and rapid growth. And in part it re- fleas the belief of those involved with this study that the eight economies do share some economic characteristics that set them apart from most other developing economies. Since 1960, the HPAEs have grown more rhan twice as fist as the rest ofEast Asia, roughly three times as fast as Latin America and South Asia, and five times faster than Sub-Sahann Africa. They also significandy outperformed the industrial economies and the oil-rich Middle East-North Africa region. Between 1960 and 1985, real income per capita increased more than four times in Japan and the Four Tigers and more than doubled in the Southeast Asian NEs (see figure 2). If growth were randomly distmbuted, there is roughly one chance in ten thousand that success would have been so regionally concentrated. The HPAEs have also been unusually successfiul at sharing the fiuits of growth. Figure 3 shows the relationship between the growth of gross do- mestic product (GPD) per capita between 1965 and 1990 and changes in the Gini coefficient, a statistical measure ofthe inequality ofincome dis- tribution. The HPAEs enjoyed much higher per capica income growth at the same time that income distribution improved by as much or more than in other developing economic% with the exceptions of Korea and Taiwan, Chna, which began with highly equal income distributions. The HPAEs are the only economics that have high growth anddedining 2. Figure 2 Change in GDP per Capita, 1960-85 Taiwan, China Indonesia Hong Kong . - Singapore Rep. of Korea Japan Matta Lescotho Egypt CYPMUS . o Gabon Greece Brazil Syria Portga :wA Malaysia Yugoislavia China __ _ - - - ~ - Thailand __ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Norway Itay Spain ~-.-.-.--- Israel Jordan ___ Germnany.. . Mexico Pakistan Colombia .Middle 20: United Statbes United Kingdom -----________- _ Philippines -----,.-----.------------ Venezuiela .-. .. Keniya . -. Peau Argentina Uganda___ India __ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ Benin__ _ _ _ _ _ _ _ _ Senegal Haiti Uberla Sudan Somalia I___ Zaire_ __ _ NIgerfia _ Afghardstan ~-* Midi '~~~~~~~~Bottom20 Central African Republic Ghana . .. Guyana ___ - Madagascar .- Chad__ _ _ _ _ _ _ _ _ Zambia ------- ---- H A s-- Angola_ Mozamibique I - Kuwalt __ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -2. 0 1. 2 3 45 Percentage change In GDP Soarae:. Summers and Heston (1988). -RACLLE F.gum 3 Change in Inequity and the GDP per Capita Growth Rate Change In avenage Glni coeffcient (1980s minus ISGOs) 0.2-------- ---- I - . Ranking by G Ini coefficient 1980 * Most equal third ofsample 0Q15 - ------------ - U Middle thibd A Least equal thlrd 0.05 - - - - - - - - . - - - - - - - - - . . - - - - - - - - - - ~ -- - - - A Chilet Brazil A 0- ; .Rep- of Korea Argentina ! Mexico v Colombla .Taiwan. China ______________ ______A_Colom bia _ _ _ _ __ _ _ _ _ _ _ _ 40.05 - _ _ Thailand * Venela . Malaysia Srigapore * Philippines -J1- Indonesa Hong Kong APeru -0.2- 0 0.03. 0.02 0.03 0.04 0.05 0.06 0.07 GDP per capita growth rate (average, 165-90) NYet:Figurc3 plots the relationship bcetwen averagc per capit incomc groth and changes in the decade avege ofthc Gini cocfficient fim rde 1960s to the 190s; a negative number indicates thar income becume lss concentraredL The decade avcge is used because data are available for different years in diffacmr conornies the decade average fbr the 1960s begins with dat from 1965. Sowrer World Bank dam. inequality. Moreover, the fastest growing East Asian economies, Japan and the Four Tilgers, axe the most equal. As a result of rapid, shared grovth, human welfare has improved dra- matically. Life expecancy in the developing HPAES increased fiom 56 years in 1960 to 71 years in 1990. (In other low- and middle-income economies, life expectancy also rose considerably, fiom 36 and 49 to 62 and 66 years, respectively.) In the HPAEs, the proportion of people living in absolute poverty, lacking such basic necessities as dean water, food, and shelter, dropped-for example, from 58 percent in 1960 to 17 percent in 1990 in Indonesia, and from 37 percent to less than 5 percent 4 in Malaysia during the same period. Absolute poverty also declined in other developing economies, but much less steeply, fiom 54 to 43 per- cent in India and from 50 to 21 percent in Brazil fronm 1960 to 1990. A host of other social and economic indicators, from education to appli- ance ownership, have also improved rapidly in the HPAEs and now are at levels that sometimes surpass those in industrial economies. 'What caused EastAsia's success? In large measure the HPAEs achieved high growth by getting the basics right. Private domesric investment and rapidly growing human capital were the principal engines of growth. High levels of domeLic financial savings sustained the HPAEs' high investment levels. Agriculture, while declining in relative impor- tance, experienced rapid growth and productivity improvement. Popu- lation growth rates declined more rapidly in the HPAEs than in other parns of the developing world. And some of these economies also got a head strt because they had a better-educated labor force and a more ef- fecrive system of public administration. In this sense there is littde that is Kxmiraculousr about the HPAEs' superior record of growth; it is largly due to superior accumulation of physical and human capital. Fundamentally sound development policy was a major ingredient in achieving rapid growth. Macroeconomic management was unusually good and macroeconomic perfornance unusually stable, providing The essential firnework for private investment. Policies to increase the integrity of the banking system, and to make it more accessible to non- traditional savers, raised the levels of financial savings. Education poli- cies that focused on primary and secondary schools generated rapid increases in labor force skills. Agriculural policies stressed productivity and did not tax the rural economy excessively. All the HPAEs kept price distortions within reasonable bounds and were open to foreign ideas and technology. But these fimdamental policies do not tell the entire story. In most of these econoniies, in one form or another, the government intervened- systematically and through multiple channels-to foster development, and in some cases the development of specific industries. Policy inter- ventions took many forms: targeting and subsidizing credit to selected industries, keeping deposit rates low and maintaining ceilings on bor- .rowing rates to increase profits and retained earnings, protecting domestic import substitutes, subsidizing declining industries, establish- big and financially supportng government banks, making public investrments in applied research, establishing firm- and industry-specific 5 r1A!N. -M IRACLE export margets, developing export marketing institutions, and sharing in- formation widely between public and private sectors. Some industries were promoted, while others were not. At least some of thesc interventions violate the dicturn of establishing for the private scaor a level playing field, a neutral incentives regime. Yet these strategies of selective promotion were closely associated with high rates of private investmenc and, in the fastest-growing economies, high rates of productivity growth. Were some selective interventions, in fict, good for growvh? In addressing this question, we face a central methodological prob- lem. Since we chose the HPAAEs for their unusually rapid growth, we know already that their interventions did not significantly inhibit growth. But it is very difficult to establish statistical links between growth and a specific inr.w..tion and even more difficult to establish causalityr Because we cannot know wrhat would have happened in the absence of a specific polii); it is difficult to rest whether interventions increased growth rates. Other economies attempted similar interven- tions without success, and on averge they used them more pervasively than in the HPAEs. Because the HPAEs differed from less successfiil economies both in their doser adherence to policy findamentals and in the manner in which they implemented interventions, it is virtually im- possible to measure the relative impact of findarnentals and interven- tions on HPAE growth. Thus, in attempting to distinguish interventions that contributed to growth from those that were either growth-neutral or harmfil to growth, we cannot offer a rigorous counterfactual sce- nano. Instead, we have had to be content with whar Keynes called an &essay in persuasion," based on analytical and empirical judgments Our judgment is that in a few economies, m ly in Northeast Asia, in some instances, govenment interventions resulted in higher and more equal growth than otherwise would have occurred. However, the prerequisites for success were so rigorotus that policynakers seeking to follow similar paths in other developing economies have often met with failure. What were these prerequisites? First, governments in Northeast Asia developed institutional mechanisms which allowed them to esab lish dear performance criteria for selective interventions and to monitor performance- Intervention has taken place in an unusually disciplined and performance-based manner (Ainsden 1989). Second, the costs of interventions, both explicit and implicit, did nor become excessive. When fiscal costs threatened the macroeconomic stability of Korea and 6 Malaysia during their heavy and chemical indusrries drives, govern- ments pulled back In Japan the Ministry of Finance acted as a check on the ability ofthe Minisrry of Intcrnational Trade and Indusry to carry out subsidy policies, and in Indonesia and Thailand balanced budget laws and legislative procedures constrained the scope for subsidies. In- deed, when selective interventions havc threatened macroeconomic sta- bilty, HPAE governments have consistently come down on the side of prudent macroeconomic management. Price distortions arising from se- lective interventions were also less extreme than in many developing economies. In the newly industrializing economies of Southeast Asia, govem- ment interventions played a much less prominent and frequently less constructive role in economic success, while adherence to policy finda- mentals remained importamt These economies' capacity to administer and implement specific interventions may have been less than in North- east Asia. Their rapid growth, moreover, has occurred in a very different international economic environment from the one that Japan, Korea, and Taiwan, China, encountered during their most rapid growth. Thus the problem is not only to try to understand which specific policies may have contributed to growth, but also to understand the institutional and economic circumstances that made them viable. Indeed, the experience of the Southeast Asian economies, whose initial conditions parallel those of many developing economies today, may prove to have more rel- evance outside the region than thar of Northeast Asia. The book is organized as follows: chapter 1 describes the distinguish- ing characteristics of the East Asian economic mirade, rapid growth with equity and uscs economic models to attempt to account for this growth. Chapter 2 reviews policy explanations for East Asia's economic success and introduces the framework that we will use throughout to ex- plore dte relatonship berween public policy and economic growth. Chapter 3 discusses pragmatism and flexibility in the formuation of policies that led to two impormnt characteristics of the HPAEs' economic performance: macroeconomic stability and rapid growth of manufac- tured exports. Chapter 4 disLcusses the role of institutions. Chaprer 5 looks at the role of public poliqc in the HPAEs' unusually rapid accurnu- lation of physical and human capital, while chapter 6 analyzes the means used to achieve efficient allocation of resources and productivity growth. Chapter 7, in condusion, assesses the success of East Asian polices and their applicabiliy in a changing world economy. The remainder of this 7 -' - .--V IICACLE overview parallels the organization of the book, highlighting the central arguments and conclusions. The Essence of the Miracle: Rapid Growth with Equity T HE EIGHT HPAEs ARE HIGHLY DIVERSE IN NATURAL RE- sources, population, culture, and economic poliqc What shared characteristics cause them to be grouped together and set apart from other developing economies? First1 as we noted above, they had rapid, sustained growth between 1960 and 1990. This in itself is un- usual among developing economies; orhers have grown quickly for periods but not for decades at such high rates. The HPAEs are unique in that they combine this rapid, sustained growth with highly equal in- come distributions. They also all have been characterized by rapid de- mographic transitions, strong and dynamic agricultural sectors, and unusually rapid export growth (see chapter 1). The HPAEs also differ fiom other developing economies in three fac- tors that economists have -traditionally associated with economic growth. High rates of investment, exceeding 20 percent of GDP on aver- age between 1960 and 1990, indcuding in particular unusually high rates ofprivate investment, combined with high and rising endowments of human capital due to universal primary and secondary education, tell a large part of the growth story. These factors account for roughly two- thirds of the growth in dte HPAEs. The remainder is attributable to im- proved productivity. Such high levels of productivity growth are quite unusuaL In fact, productivity growth in the FHAEs exceeds that of most other developing and industrial economies. This superior productivity performance comes from the combinadon of unusual success at allocat- ing capimt to high-yielding investments and at catching up technologi- cally to the industrial economies. Public Policies and Growth What was the role of public policy in helping the HPAEs to rapidly accumulate human and physical capital and to allocate those resources to high-yielding investments? Did policies assist in promoting rapid produc- 8 tivity growth? There are several explanations for East Asias success. Geog- raphy and culture were dearly important; however, dthy do not entirely account for the high-performing economies' success, as the presence of unsuccessful economies in the same region attests. Among the variety of policy explanations, two broad views have emerged (see dcapter 2). Adherents of the neoclassical view stress the HPAEs' success in getting the basics right. They argue that the successful Asian economies have been better than others at providing a stable macroeconomic environment and a reliable legal framework to promote domestic and international compe- tition. They also stress that the orientation of the HPAEs toward inter- national trade and the absence of price controls and other distortionary policies have led to low relative price distortions. Investnents in people, education, and health are legitimate roles for govenrnment in the neoclas- sical framework, and its adherents stress rhe importance of human capital in the HPAEs' success. Adherents of the renisionist view have successfilly shown that Easr Asia does not wholly conform.to the neodassical model. Industrial pol- icy and interventions in financial markets are not easily reconciled within the neoclassical fiamework. Some policies in some economies arc much more in accord with models of state-led development. Moreover, while the neoclassical model would explain growth with a standard set of relatively constant policies, the policy mixes used by East Asian econ- omies were diverse and flexible. Revisionists argue that East Asian gov- ernments "led the m.narke"e in critical ways. In contrast to the neoclassical viev, which acknowledges relatively few cases of market &il- ure, revisionists contend that markers consistendy fail to guide invest- ment to industries chat would generate the highest growth for the overall economy. In East Asia, the revisionists argue, governments remedied this by delibemtely "getting the prices wrong"-altering the incentive struccure-to boost industries that would not otherwise have thrived (Amsden 1989). The revisionist school has provided valuable insights into the history, role, and extent of East Asian interventions, demonstrating convinc- ingly the scope of government actions to promote industrial develop- ment in Japan, Korea, Singapore, and Taiwan, China. But, in general its proponents have not claimed to establish that interventions per se accel- crated growth. Moreover, as we shall show, some important govemment interventions in East Asia, such as Korea's promotion of chemicals and heavy industries, have had little apparent impact on industrial strucmrt. 9 -4iE$AS,kAN MIRACLE In other instances, such as Singapore's efflort to squeeze out labor-inten- sive industries by boosting wages, policies have dearly backfired. Thus neither view fiully accounts for East Asia's phenomenal growth. The Market-Frendly View. In describing the policies associated with rapid growth, World Development Report 1991 (World Bank 1991 b) ex- pands on the neodassical view, darifyring systematically how rapid growth in developing countries has been associated with effcctive but carefully limited govcrnment activism. In the 'market-friendly" strategy it articulates, the appropriate role of government is to ensure adequate investments in people, provide a competitivc climate for private enter- prise, keep the economy open to intemational trade, and maintain a sta- ble macroeconomy. Beyond these roles, the report argues, governments are likely to do more harm than good, unless interventions arc marker fiendly. On the basis of an cxhaustive review of the experience of devel- oping economies during the last thirty years, it condudes that attempts to guide resource allocation with nonnarket mechanisms have generally failed to improve economic performance. The market-friendly approach captures important aspects of East Asia's success. These economies are stable macroeconomically, have high shares of international trade in GDP, invest heavily in people, and have strong competition among firms. But these haracteristics are the out- come of many different policy instnuments. And the instruments cho- sen, particularly in the nordteastern HPAES, Japan, Korea, and Taiwan, China, sometimes induded extensive government intervention in mar- kets to guide private-sector resource allocation. The success of these northeastern economies, moreover, stands up well to the less interven- tionist paths taken by Hong Kong, Malaysia, and more recendy In- donesia and Thailand. A Functional Appmach to Undemtaing Grwth. To explore these varying paths to economic success, we have developed a framework that seeks to link rapid growth to the attainment of three functions. In this view, each of the HPAEs rnaintained macroeconomic stability and accomplished three fiunctions of growth: accumulation, efficient allocation, and rapid technological catch-up. They did this with combinations of policies, ranging from market oriented to state led, that varied both across economies and over time. We classifjr policies into two broad groups: fundamentals and selec- tive interventions. Among the most important fimdamental policies are those that encourage macroeconomic stability, high investments in I0 human capital, stable and secure financial systems, limited price disror- dons, and openness to foreign technology Selective interventions in- dude mild financial repression (keeping interest races positive but low), directed credit, selective industrial promotion, and trade policies that push nontraditional exports. We try to understand how government policies, borh fundamental and interventionist. may have contributed to faster accumulation, more efficient allocation, and higher prnducriv- ity growth. We maintain as a guiding principle char for interventions that at- tempt to guide resource allocation to succeed, they must address filures in the working of markets. Otherwise, markets would perform the allocation function more efficiently. We idenrify a dass of economic problems, coordination failures, which can lead markers to fail, espe- cially in early stages of development. We then interpret some of the in- terventionist policies in East Asia as responses to these coordination problems-responses that emphasized cooperative behavior among pn- vate firms and clear performance-based standards of success. Competitive discipline is crucial to efficient investment. Most economies employ only market-based competition. We argue that some HPABs have gone a step further by creating contests that combine com- petition with the benefits of cooperation among firms and berween gov- ernment and the private sector. Such contests range from very simple nonmarket allocation rules, such as access to rationed credit for ex- porters, to very complex coordination of private investment in the government-business deliberation councils of Japan and Korea. The key feature of each contest, however, is that the government distributes rewards-often access to caedit or foreign exchange-on the basis of performance, which the government and competing firms monitor. To succeed, selective interventions must be disciplined by competition via either markets or contests. Economic contests, like all others, require competent and impartial referees-that is, strong institutions. Thus, a high-quality civil service that has the capacity to monitor performance and is insulated from po- litical interference is essential to contest-based competition. Of course, a high-quality civil service also augments a government's ability to design and implement non-contest-based policies. Our framework is an effort to order and interpret inlformation. We are not suggestng that HPAE governments set out to achieve the func- tions of growth. Rather, they used multiple, shifting policy instruments II R AC LAN MIRACLE in pursuit of more straiglhtforward economic objectives such as macro- economic stability, rapid export growth, and high savings. Pragmatic flexibility in the pursuir of such objectives-the capacity and willingness to change policies-is as much a hallmark of the HPAEs as any single pol- icy instrument. This is well illustrated by the great variety of ways in which the HPAEs achieved tvJ important objectives: macroeconomic stability and rapid export growth (sec chapter 3). Achieving Macroeconomic Stabilty and Export Growth More than most developing economies, the HPAES were characterized by responsible macroeconomic management. In parricular, they gener- ally limited fiscal defcits to levels that could be prudently financed without increasing inflationary pressures and responded quickly when fiscal pressures were perceived to building up. During the past thirty years, annual inflation averaged approximately 9 percent in dtese economies, compared with 18 percent in other low- and middle-income economies. Because inflation was both moderate and prediaable, real interest rates were far more stable than in other low- and middle-income economies. Macroeconomic stability encouraged long-term planning and private investment and, through its impact on real interest rates ai, 4 the real value of financial assets, helped to increase financial savings. The HPAEs also adjusted their macroeconomic policies to terms of trade shocks more quickly and effectively than other low- and middle-income economies. As a result, they have enjoyed more robust recoveries of pri- vate investment. Many of the policies that fostered macroeconomic stability also con- tributed to rapid export growth. Fiscal discipline and high public sav- ings allowed Japan and Taiwan, China, to undertake extended periods of exchange rate protection. Adjustments to exchange rates in other HPAEs-validated bv policies that reduced expenditures-kept them competitive, despite differential inflation with trading partners. In addition to macroeconomic policies, the HPAEs used a variety of approaches to promoting exports. All (except Hong Kong) began with a period of import substitution, and a strong bias against exporEs. But each moved to establish a pro-export regime more quicldy than other developing economies. First Japan, in te 1950s and early 1960s, and then the Four ligers, in the late 1960s, shified trade policies to encour- age manufactured exports. In Japan, Korea, and Taiwan, China, govern- 12 ments established a pro-export incentive structure that coexisted with moderate but highly variable protection of the domestic marker. A wide variety of instruments was used, including export credit, duty-free im- ports for exporters and their suppliers, export targets, and tax incentives. In the Sioutheast Asian NIEs the export push came later, in the early -1980s, and the instruments were different. Reductions in import pro- tection were more generalized and were accompanied by export credit and supportinginstitutions. In these economies exportdevelopment has relied less on highly selective interventions and more on broadly based market incentives and direct foreign investment. Building the Instutional Basis for Growth Some economists and political scientists have argued chat the East Asian miracle is due to the hiigh quality and authoritarian nature of the regions institutions. They describe East Asian political regimes as "devel- opmental statee' in which powerful technocratic bureaucracies, shielded from political pressure, devise and implement well-honed interventions. We believe developmental stare models overlook the central role of government-private sector cooperation. Whle leaders of the HPAEs have tended to be either authoritarian or paternalisdti they have also been willing to grant a voice and genuine authority to a technocratic elite and key leaders of the private sector. Unlike authoriruian leaders in many other economies, leaders in the HPAEs realized that econornic develop- ment was impossible without cooperation (see chapter 4). The Principle of Shard Growth. To establish their legitimacy and win the support of the sociey at large, East Asian leaders established the princi- ple of shared growth, promising that as the economy expanded all groups would benefit. But sharing growth raised complex coordination problems. First, leaders had to convince economic elites to support pro- growth policies. Then they had to persuade the elites to share the bene- fits of growth with the middle class and the poor. Finally, to win the cooperation of the middle class and the poor, the leaders had to show them that they would indeed benefit from future growth. Explicit mechanisms were used to demonstrate the intent that all would have a share of future wealth. Korea and Taiwan, China, carried out comprehensive land reform programs; Indonesia used rice and fer- filizer price policies to raise rural incomes; Malaysia introduced explict wealth-sharing programs to improve the lot of edtic Malays relative to '3 - IRfLMIRACLE the better-off ethnic Chinesc; Hong Kong and Singapore undertook massive public housing programs; and in several economies, govern- ments assisted workers' cooperatives and established programs to en- courage small and medium-size enterprises. Whatever the form, these programs demonstrated that the goverment intended for all to sharc the benefits of growth. Cualmg a Business-Fiendly Environment To cacide coordination prob- lems, leaders needed institutions and mechanisms to reassure compet- ing groups that each would benefit from growrth. The first step was to recruit a competent and relativcly honest technocratic cadre and insu- late it from day-ro-day political interference. The power of these tech- nocracies has varied greatly. In Japan, Korea, Singapore, and Taiwan, China, strong, well-organized bureaucracies wield substantial power. Other HPAES have had small, general-purpose planning agencies. But in each economy, economic technocrats helped leaders devise a credible economic strategy. Leaders in the HPAs also built a business-friendly environment. A major element of that enviromnenr was a legal and regulatory strucmre that was generaly hospitable to private investment. Beyond this the HPAiEs have with varying degrees of success enhanced communication between business and government. Japan, Korea, Malaysia, and Singa- pore have established forums, which we call deliberation councils, in which private sector groups are invited to help shape and implemnent the government policies relevant to their interests. In contrast to lobbying, --where rules are murky and groups seek secret advantage over one an- odter, the deliberation councils are intended to make allocation rules dear to all participants. Using Deliberation Council. In Japan and Korea technocrats used delib- eration councils to establish contests among firms. Because the prvate sector participated in drafting the rules, and because the process was transparent to all paricipants, private sector groups became more will- ing participants in the leadership?s development efforts. One by-product of these contests was a tendency to reduce the private resources devoted to wasteful rent-seeking activities, thus making more available for pro- ductive endeavors. Deliberation councils also &cilitated information ex- changes betwren the private sector and govecnment, among firms, and between management and labor. The councils thus supplemented the market's inforrnation transmission function, enabling the IPAEs to re- spond more quickly than other economies to changing markets. '14 Institutions of business-government communication have not bem sta- tic in the HPAJ. The role of the deliberation ouncil is changing in Japan and Korea to a more indicatve and consensus-building role, along finc- tional as opposed to industry-speific lines. In Malaysia the councils appear to be increasing in importance and scope. In Thailand the formal mecha- nisms of communication have generally becn used to present businesses' positions to goverment and to reduce the private sector's suspicion ofgov- ermnent. In institutional development, as in economic poliymaking East Asian governments have changed with changing circumstances. Accumulaing Human and Physical Capital Drawing on the strength of their institutions, East Asian economies used a combination of fundamental and interventionist policies to achieve rapid accumulation of human and physical resources. Funda- mentals included such traditional government obligations as providing adequate infrastructure, education, and secure financial institutions. In- terventions induded mild repression of interest rates, state capitalism, mandatory savings mechanisms, and socialization of risk (see chapter 5). Bilding Human Capiti. The East Asian economies had a head start in terms of human capital and have since widened their lead over other de- veloping economies. In the 1960s, levels of human capital were already higher in the HPAEs than in other low- and middle-income economies. Governments built on this base by focusing education spending on the lower grades; first by providing universal primary education, later by in- creasing the availability of secondary educations Rapid demographic transitions facilirated these efforts by slowing the growth in the number of school-age children and in some cases causing an absolute decline. Dedining ferdlity and rapid economic growth meant that, even when education investment as a share of GDP remained constant, more re- sources were avilable per child. Limited public fimding of post- secondary education focused on technical skills, and some HPAEs imported educational services on a large scale, particularly in vocation- ally and technologically sophisticated discplines. The result of these policies has been a broad, technically inclined human capital b±ve well1- - stited to rapid economic development. .IHPAE education policies also contributed to more equitable income distributions. To be sure, initial conditions helped to set up a virtuous circle: initial low inequality in income and education led to educational I5 .:~~~~~~~~~~~~~~;n 11 :t4iSjA-N MIRACLE expansion, which reinforced low inequality. In addition, by focusing spending on primary and secondary education, and leaving demand for tertiary education to be largely met by' a self-financed private system, governments served large segments of the population that otherwise would have lacked access to education. bcreasing Savniug and ImnestneuL The HPAEs increased savings and in- vestment with a combination of findamental and interventionist poli- cies. Two fundamenral policy areas provided a foundation for high and rising savings rates. First, by avoiding inflation, the HPAEs avoided volatility of real interest rates on deposits and ensured that rares were largely positive. As a result, the HPAEs have generally offired higher real interest rates on deposits in the financial system than other developing economies. Second, they ensured the security of banks and made them more cotnvenient to small and rural savers. The major instruments used to build a secure, bank-based financial system wvere strong prudential regulation and supervision, limits on competition, and institutional re- forms In Japan, Korea, Malaysia, Singapore, and Taiwan, China, postal savings systems lowered transaction costs and increased the safey of sav- ing while making substantial resources available to government. These initiatves promoted rapid growth of deposits in financial instiutions Figu 4 Savings Rates of HPAEs (see figure 4) anda Selected Eiconomies, 19704 (se fiur 4) Some governments also used a variety of more interventionist mecha- HPAEs nisms to increase savings. Singapore and Taiwan, China, maintained un- I - - - usually high public savings rates. Malaysia and Singapore compeled high private savings rates dtrough mandatory provident fund contributions. Othor _ ' Japan, Korea and Taiwan, China, all imposed stringent contrls and high interest rates on loans for consumer items, and levied stiff taxes on so- 0 2.0 20 30 40 called luxury consumption. Whether these more interventionist measures 'Percenta,ge ar GDP to increase savings improved welfare is open to debat On one hand, Nenar. Finlind,uds Austria Bthee um. making consumers save when they would not have otherwise imposes a Dcrnmark, Finland, Fmncc, the E:ederol Republic of Gcmany bebore reunificaion, welfare c -r. On the other, the benefits are apparent in the rapid growth of Greece, Iceland, Iredand. Imly, Luxem- these ecovomies. Savings, forced or not, generaed high payoffs based on bour, dte Netherlands, Norny, Portugal. consistendy high rates of return to investments. In conwnt to other Spain, Sweden, Switxrland, and the United Kingdom. 'Orie indudes these economies that have used compulsory savings, such as the fonrer Soviet &-veoping cconomies: Argenona Brazil,T developin eColombia. Carenina SEl, Union, welfare costs were dearly offset by substantive benefits. Chile, Cdlombia, COre divoire, Egyp4, Ghana, India, Mexico, Morocco, NiScria, The HPAEs encouraged investment by several mans. First, they did a Pakistan, Pem, Sri Lnka, Turkey, better job than most developing econom>es at creating -infastructure Urugua. Ycnzuels the fone upEra Urgan y Venezuela. the YugosIiaI that was complementary to private investment. Seco.d, they created an and Zaire SawreuSummer and Heson (199l). investment-friendly environment through a combination of tax policies I6 ivoring investment and of policies that kept the reative prices of capi- mal goods low, largely by avoiding high tariffs on imported capital goods. These fiundamental policies had an important impact on pnvate invest- ment. Third and more controversial, most HPAE governrents held de- posit and lending rates below market dearing levels-a practice termed financial repression. Japan, Korea, Malaysia, Thailand, and Taiwan, China, had extended periods of mild financial repression. To be sure, increasing interest rates from negative to zero or mildly positive real rates and avoiding fluctua- tions (by avoiding unstable inflation) encourages financial savings. But because savings are nor very responsive to marginal changes in positive real interest rates, HPAE governments were able to mildly repress interest rates on deposits with a minimal impact on savings and to pass the lower rates to final borrowers Because savers were mostly households and bor- rowers were mostly firms, this resulted in a transfer of income from households to firns and in a change in the form in which savings were held, from debt to corporate equity. Holding down interest rares on loans increases excess demand for credit, which in turn leads to rationing of credit by the government it- self or by private sector banks working with goverrunent guidance. This heightens the risk that capital will be misallocated. Thus there is a trade- off between the possible increase in investment and the risk that the in- creased capital will be badly invested. There is some evidence that in Japan, Korea, and Taiwan, China, govemments allocated credit to activ- ities with high social returns, especially to exporcs. If this was the case, there may have been benefiEs from mild financial repression and government-ided allocation; mirroeconomic evidence from Japan supports the view thar access to government credit increased investment (see chapter 6). Generally, financial repression is associated with low economic growth, especially when real interest rates are strongly negative. But test of the relationship between interest rates and growth in Japan, Korea, and Taiwan, China, do not show the negative relationship between in- terest rate repression and growEh found in cross-economy comparativ studies (see chapter 5). While we cannot establish condusively that mild repression of interest rates at positive real levels enhanced growth in northeast Asia, it appaently did not inhibit it. Finally, some governments, especially in the northeastem Asian tier, have encouraged investment by spreading private investmcent risks to the '7 s;~ A? -o m v y LRC ~~~~~~~~V- . ZiN [.IA RC L E public In some economies the govern-ment owned or controlled the in- titutions providing investment funds, in others it offered explict credit guarantees, and in sdll others it implictly guaranteed the financal via- bility of promoted projects. Relationship banking by a variety of public and private banking institutions in Hong Kong, Japan, Korea, Malaysia, Singapore, Thailand, and Taiwan, China, involved the banking sector in the management of troubled enterprises, increasing the likelihood of creditor workouts. DirectEd-credit programs in Japan, Korea, and Tai- wan, China, signalcd directions of government policy and provided imn- plicit insurance to private banks. Effcient Allocation and Pmductity Change Some policies thar favored accumulation in the HPA,s, induding fi- nancial repression and the socialization and bounding of risk, could haVe adversely afFected the allocation of resources. Similarly, industrial targeting could have resulted in extensive rent-seeking and great ineffi- dency. Apparently rhey did not The allocadional rules folowed by H4PAE govemments-particularly the devices used to shift mareke incentives-- are therefore among the most controversial aspects of the Eat Asian suc- cess story (see chaprer 6). Like policies related ro accumulation, policies affecting allocation and productiviy change fall into findamental and interventionist cat- egories. Labor marker policies tended to rely on fundamentals, using the market and reinforcing irs fiexibility. In capital markets, govern- ments intervened systematically, both to control interest rates and to direct credit, but acted within a framework of careful monitoring and generally low subsidies to borrowers. Trade policies have indcuded sub- standal protection of local manufacurs, but less than in most other developing countries; in addition, HPAE governments offiet some dis- advantages of protection by actively supporting exports. Finally, while interventions to support specific industries have generally not been suc- cessfil, the export-push straregyE-the mix of fundamental and inter- ventionist policies used to encourage rapid manufacured cxport growth-has resulted in numerous benefirs, including more efficient allocation, increased acquisition of foreign rechnology, and more rapid productivity growth. Feible Labor Markets. Govenment roles in labor narkets in the suc- cessul Asian economies contrast sharply with the situation in most I8 other developing economies. HPAE governments have generally been less vulnerable and less responsive than other developing-economy govern- ments to organized labor's demands to legislate a minimum wage. Rather, they have focused their efforts on job generations effectively boosting the demand for workers. As a result, employment levels have rnsen first, followed by market- and productivity-driven increases in wage levels. Because wages or at least wage rate increases have been downwardly flexible in response to changes in die demand for labor, ad- justmenr to macrocconomic shocks has generally been quicker and less painfiul in East Asia than in other developing regions. Rapid adjust- ments helped to sustain growth, which in turn made more rapid real wage growth possible. IHigh productiviry and income growth in agriculture helped to keep EastAsian urban wages close to the supply price of labor. In contrast to many other developing- economies, where the gap between urban and rural incomes has been large and growing, in the HPAEs the incomes of urban and rural workers with similar skill levels have risen roughly at the same pace; moreover, the overall gap between urban and rural incomes is smaller in the HPAEs than in other developing economies. In Sub-Saharan Africa, Latin America, and South Asia, where wages in the urban formal sector are often pushed up by legislated minimum wages and other nonmarket forces, urban wage earners often have in- comes twice their counterparm' in informal sectors. In contrast, the gap between the formal and informal sectors in East Asia is only about 20 percent. Smaller income gaps contribute to overall social stability, thus enhancing the environment for growth. Caprtal Markets and fllocatio. Most HPAEs influenced credit allocation in three ways: (i) by enforcing regulations to improve private banks' pro- ject selection; (ii) by crearing financial institutions, especially long-term credit (development) banks, and (Li) by directing credit to specific sec- tors and firms through public and private banks. All three approaches can be justified in theory, and each has worked in some IIPAES. Yet each involves progressively more government intervention in credit markets and so carries a higher risk. Goverrunent relationships with banks in the HPAEs have varied widely. In Hong Kong banks are private and regulated primarily to en- sure their solvency. In Indonesia, Malaysia, Singapore, and- Thailand, banks are prvately owned and exercise independent authority over lend- big. While governments have broadly guided credit allocations through I9 isACLWMIRACLE regulations and moral suasion, project selection is generaUly left to bankers. In other HPAEs, banks have been subject to direct state control or stringent credit allocation guidelines. For example, Indonesia, Korea, and Taiwan, China, tightly controlled the allocation of credit by public commerial banks. Each of the HPAEs made some attempts to direct credit to priority ac- tivities. All East Asian economies except Hong Kong give automatic ac- cess to credit for exporter. Housing was a priority in Singapore and Hong Kong, while agriculurc and small and medium-size enterprises were trgeted sectors in Indonesia, Malaysia, and Thailand. Taiwan, China, has recently targeted technological development. Japan and Korea have used credit as a tool of industrial polic34 organizing contests through deliberative councils to promote at various times the shipbuild- ing, chemical, and automobile induscries. The implicit subsidy of direcred-credit programs in the HPAEs was generally small, especially in comparison to other developing econ- omies, but access to credit and the signal of goverunment support to if- vored sectors or enterprises were importanL In Korea, the subsidy from preferential credit was large during the 1970s, resulting in a large gap be- tween bank and curb market interest rates. This gap has dedined sharply in recent yea, as Korea has shifted away fiom heavy credit sub- sidies to selected sectors. In Japan imnplicit subsidies were small, and the direction of credit may have been more important as a signaling and in- surance mechanism than as an incentve. Although East Asias directed-credit programs were digned to achieve policy objectives, they nevertheless induded strict performance criteria. In Japan, public bank managers chose projects on basic economic crite- ria, employing rigorous credit evaluations to select among applicants that fell within government sectoral trgets. In Korea, the govemment individually monitored the large conglomerates using market-oriented criteria such as exports and profirabiliy. In some cases, major enterprises that filed to meet these tests were driven into bankruptcy. Recenr as- sessments of the directed-credit programs in Japan and Korea provide microeconomic evidence that directed-credit programs in these econo- mies increased invesnnent, promoted new actvities and borrowers, and were directed at firms with high potential for technological spillovers. Thus these performance-based directed-credit mechanisms appear to have improved credic allocation, especially during the early stages of rapid growth (see chapter 6). z0 Directed-credit progras in other HPAES have usually lacked strong, per- fbrnance-based alloation and monitoring and therefore have been largely unsuccessful. Even in the norther-tier economies, the increasing level of financial sector development and their increasing openness to international capiml flows have meant that directed-credit pograms have declined in irm- portance, as the economnies have liberalized their financial sectors. Openness to Foreig Technology. The HPAEs have actively sought foreign technology through a variety of mchaisms. All welcomed technology transfers in the forn of licenses, capital goods imports, and foreign training. Openness to direct foreign investment (DRi) has speeded tech- nology acquisition in Hong Kong, Malaysia, Singapore, and, more re- cently, Indonesia and Thailand. Japan, Korea and, to a lesser extent, Taiwan, China, restricted DFI but offier this disadvantge by aggressively acquiring foreign knowledge through licenses and other means. In contrast, odter low- and middle-income economies such as India and Argentina have adopted policies that hindered the acquisition offor- eign knowledge. Often they have been preoccupied with supposedly ex- cessive prices for licenss They have refised to provide foreign exchange for trips to acquire knowledge, been restictive of DFI, and have at- tempted prematurely to build up their machine-producing sectors, thus forgoing the advanced technology embodied in imported equipment. Prmofing Specific lidusies. Most East Asian governments have pur- sued sector-specific industrial policies to some degrce. The best-known instances includeJapan's heavy industry promotion policies of the 1950s and the subsequent imitation of these policies in Korea. These policies induded import prorection as well as subsidies for capital and other im- ported input& Malaysia, Singapore, Taiwan, China, and even Hong Kong have also established programs-typically with more moderate incentives-to accelerate development of advanced industries. Despite these acions we find very litde evidence that industial polices have af- fected either the secroral structure of industry or rates of productivity change. Indeed, industrial structures in Japan, Korea, and Taiwan, China, have evolved during the past thirty years as we would expect ven factor-based comparative advantage and changing factor endowments. It is not altogether surprising that industial policy in Japan, Korea, and Taiwan, China, produced mainly market-conforming results. Wile these govemments slectively promoted capital- and knowledge- intensive industries, they also took steps to ensure that they were foster- ing profitable, intermationally competitive firms. Moreover, their 21 I rRAA CVLIACLE industrial policies incorporated a large amount of marker information and used performance, usually export performanc, as a yardstick. Ef- forts elsewhere to promote specific industries without better informa- dion exchange and the discipline of intemational markets have not succeeded. This has been the case with the ambitious industrial policy programs in Brazil and India, and with the more limited but also disap- pointing efforts to build an aerospace industry in Indonesia and to pro- mote heavy induswtries in Malaysia. Export Push: A Winning Mix of Fundamentals and Interentions One combi- nation of fundamental and interventionist policies practiced in the HPAEs has been a significant source of rapid productiviy growth: the ac- tive promotion of manufactured exports. Although all HPAEs cxcept Hong Kong passed through an import-substitution phase, with high and variable protecrion of domestic import substitutes, these periods ended earlier than in other economies, typically because of a compelling need for foreign exchange. In contrast to many other economies, which tried to preserve foreign exchange with sticter import controls, the HPAEs set out to earn additional foreign exchange by increasing exports. Hong Kong and Singapore adopted trade regimes that were dose to free trade; Japan, Korea, and Taiwan, China, adopted mixed regimes that were largey free for export industries. In the 1980s, Indonesia, Malaysia, and Thailand have adopted a wide variety of export incentives while gradually reducing protection. Exchange rate policies were liberal- ized, and currencies frequently devalued, to support export growth. Overall, these policies exposed much of the industrial sector to interna- tional competition and resulted in domestic relative prices chat were doser to international prices than in most other developing economies. The northem-tier economies-Japan, Korea, and Taiwan, China- halted the process of import liberalization, often for extended periods, and heavily promoted exports. Thus while incentives were largely equal between exports and imports, this was the result of countervailing sub- siclies rather than trade neutrality the promotion of exports coexisted with protecion of the domestic market. In the Southeast Asian HPAEs, conversely, govermments used gradual but continuous liberaization of the trade regime, supplemented by institutional support for exporters, to achieve the export push. In both cases governments were credibly committed to the export-push strategy; producers, even those in the proteced domestic market, knew that sooner or later their time to cx- port would come. 22 East Asia's sectoral policies were usually geared toward export perfor- mance, in contrast to the inward-oriented policies of less successful de- veloping economies. Japan, Korea, Singapore, and Taiwan, China, all relied on economic performrance criteria, usually exports, to judge success. For example, in Taiwan, China, the government suspended domestic- content requirements that interfered with the exports of foreign in- vestors. In addition, sectoral policies were closely monitored and frequently adjusted. Thus, many of East Asia's "industrial upgrading' programs of the lare 1970s and early 1980s were substantially modified or abandoned when they failed to produce satisfactory results. Using the export rule meant that even programs of selective industrial promotion were indirectly export promoting. Manufacmred export growth also provided a powerfil mechanism for technological upgrading in imperfect world technology markets. Be- cause firms that export have greater access to best-practce technology, there are both benefits to the enterprise and spillovers to the rest of the economy that are not reflected in market transactions. These information- reated extemalities are an imporrant source of rapid productiviy growth. Both cross-economy evidence and more detailed studies at the industry level in Japan, Korea, and Taiwan, China, confirm the signifi- cance of exports to rapid productivity growth. These experiences suggest that economies that are making the transi- tion from highly protectionist import-substitution regimes to more bal- anced incentives would benefit from combining import liberalization with a strong commitment to exports and active export promotion, es- pecially in those cases in which the pace of liberalization is moderate. Policies for Rapid Growth in a Changing World Economy 5 IT>71AT ARE THE BROAD LESSONS OF SUCCESS IN THE HPAEs? 'sKI/ Their rapid growth had two complementry elements. Firz, v v getting the fundamentals right was essential. Without Jhgh levels of domestic savings, broadly based human capital, good macro- economic management, and limited price distortions, there would have been no basis for growth and no means by which the gains of rapid pro- ductvty change could have been realized. Policies to assist the financial 2-3 - XIAIEMI R AC LE sector capture nonfinancial savings and to increase household and cor- porate savings were cental. Acquisition of technology through openness to direct fbreign investment and licensing were crucial to rapid produc- tivity growth. Public investment complemented private investment and increased its orientation to exporcs. Education policies stressed universal primary schooling and improvements in quality at primary and sec- ondary levels. Second, very rapid growth of the type experienced by Japan, the Four Tigers, and more recently the East Asian NIEs has at times benefited from careful policy interventions. All interventions carry costs, either in the direct fiscal costs of subsidies or forgone revenues, or the implicit taxation of households and firms, for example, through tariffs or inter- est rate controls. Unlike many other governments that attempted such interventions, HPAE governments generally held costs within well- defined limits. Thus, price distortions were mild, interest rare controls used international interest rates as a benchmark, and cxplicit subsidies were kept within fiscally manageable bounds. Given the overriding im- portance ascribed to macroeconomic stability, interventions that be- camne too cosdy or otherwise threatened stability were quickly modified or abandoned. Whether these interventions contributed to the rapid growth made pos- sible by good fundamnentals or detracted from it is the mosc difficult ques- tion we have attempted to answer. It is much easier to show that tie HPAEs limited the costs and durtion of inappropriately chosen interventions- itself an impressive achievement-than to demonstrate condusively that those interventions maintained for a long time accelerated growth. Our assessment of three major uses of intervention is that promotion of specific industries generally did not work and therefore holds litde promise for other developing economies. Mild financial repression com- bined with directed credit has worked in certain situations but carries high risk. Export-push strategies have been by far the most successfil combination of fundamentals and policy interventions and hold the most promise for other developing economies (chapter 7). But are these approaches feasible in the early 1990s? While limited repression of interest rates may have contributed to ovemll higher rates of investment in Japan, Korea, and Taiwan, China, these three north- eastern economies undertook their initial growth spurts-and their most sustained and forceful repression of interest rates-during a period when it was possible for a developing economy to dose its financial mar- 24 Icets to the outside world. Furthermore, strong bureaucracies and a gen- eml climate of government-private sector cooperation meant that their restrictions on capital outflows were more effective than similar restric- tions in many other economies. In today's increasingly global economic environment, few governments have the ability or desire to close their fi- nancial markets. Indeed, many East Asian governments are in the -proes of liberalizing restrictions on capital flows. In such circum- stances, the scope for repressing interest rates without provoling capital flight is sharply narrowed. However, in some exceptional instances, very mild financial repression of short duration to increase corporate equity remains a viable optiOII. This has been the case in Malaysia, which has wide open financial markets but nonetheless succeeded with very mild financial repression for more than a year. The export-push strategy appears to hold great promise for other developing economies. But the conditions of market access under the General Agreement on Tariffs and Trade (GArr), and odlier trading arrangements, will hamper developing economies' use of policies viewed as unFair in major industrial-economy markets. Subsidies to exports and directed-credit programs linked to exports are not generally consistent with the GATr and may therefore invite retaliation from trading part- ners. Furthermore, like financial repression, these highly directed inter- ventions require a high level of insttutional capacity now lacking in most developing economies. Fortunately, many powerful instruments of export promotion are nor only within the institutional capacity of many developing economies but remain viable in today's economic environ- ment. Creating a free trade environment for exporters, providing finance and support services for small and medium-size exporters, improving trade-related aspects of the civil service, aggressively courting export- oriented direct foreign investmnent, and focusing infrastructure on areas that encourage exports are all attainable goals that are unlikely to provoke opposition from trading partners. Indeed, some or all of these have been part of the export push in Indonesia, Malaysia, and Thailand. These three economies, the most recent participants in the "economic mirade,' may show the way for the next generation of developing economies to follow export-push strategies. The phenomenal success of the HPAEs is already inspiring attempts at imitation. We have shown that the HPAEs used an immense variety of 25 ' _Ni1A CCLE policies to achieve three critical functions of growth: accumulation, al- location, and productivity ,rowrh. The sheer divenrsiy of these policies precludes drawing any simple lcssons or making any simple recommen- dations, exccpt perhaps that pragmatic adherence to the fundamentals is central to success. These market-oriented aspects of East Asia's experi- ence can be recommended w;ith few reservations. More institutionally demanding strategies have often filed in other settings and they clearly are not compatible with economic environments wlhere the fuindamen- tals are not securely in place. The use of contests in Japan and Korea re- quired competenr and insulared civil servants. In parts of Sub-Saharan Africa and Latin .Amnerica, and elsewhere in Asia where such institu- tional conditions are lacking, activist government involvement in the economy has usually gone awry. So the fact that interventions were an element of some East Asian economies' success does not mean that they should be attempted everywhere, nor should it be taken as an excuse to postpone needed market-oriented refonn. The success of the HPAEs broadens our understanding of the range of policies that are consistent with rapid development. It also teaches us that willingness to experiment and to adapt policies to changing cir- cumstances is a key element in economic success. In the following chap- ters we explore more fully the contribution of fundamental and interventionist policies to East Asiis remark-able growth, and the crucial role that institutions have played in their evolution and application. As we shall see, maaking a miracle is no simple matter Note 1. Japan, which has been firmly in the ranks of indus- have been used subsequently by devdoping economies. trial economics arguably for all ofthis century, may at first Thus, notwithsranting Japan's longer history of modem seem to be an inappropriate subject for study. However economic growrh, it may provide some useful insights many of the policy instruments used by the Japanese gov- into the relationship between public policy and growth. emmnent during the period of rapid growth, 1950-73, 26 Growth, Equity, and Economic Change ml .HE EIGHT ECONOMIES OF OUR STUDY ARE HIGHLY diverse in natural resources, culmre, and political in- stitutions. Japan, unlike the others, was already a rela- tively mature industrial economy at che beginning of the postvar period. Moreover, the eight differ in the degree of government intervention in the economy and the manner in which their leaders have shaped and implemented poticies. Korean policymakers, for example, have intervened heavily in industrial, labor, and credit markets, while policymalers in Hong Kong have been consistently hands-off. Despite many differences, however, the eight economies have much in common. In a number of ways, their postwar experience distin- guishes them as a group. Their most obvious common characteristic is their high average rate of economic growth. During the same period, income inequality has declined, sometimes dramatically. These two outcomes-rapid growth and reduced inequality-are the defining characteristics of what has come to be known as rhe Easr Asian eco- nomic miracle. The eight economies share six other characteristics chat set them apart. Compared with most other developing economies, all have had: * Morerapidoutputandproductivity growthinagriculture - Higher rates of growth of manufactured exports * Earlier and steeper dedines in ferility * Higher growth rates of physical capital, supported by higher rates of domestic savings a* IHigher initial levels and growth rates of human capital * Generally higher rates of productvity growth. 2-7 * A C LkkYI&CE These charaaeristics are all related to their rapid, more equitable growth. Some are sources of growth, some are outcomes of growtl, and some are unique features of HPAE growth, but most fall in more than one of thcse categories. Rapid and Sustained Economic Gmwth T HE EIGHT HPAEs GREW MORE RAPIDLY AND MORE CONSIS- tendy than any otler group of economies in the world from 1960 to 1990. They averaged 5.5 percent annual per capita real income growth, outperforming every economy in Latin American and Sub-Saharan Africa (except diamond-rich Borswana). Another East Asian economy, China, has grown 5.8 percent a year since 1965 and could stake a claim to join the ranks of the HPAEs.1 Figure 1.1 shows the relationship between income level relative to the United States in 1960 and per capita income growth for 119 econormies during the period 1960 to 1985. Developing economies were noc catch- ing up with the advanced economies; more than 70 percenr of the de- veloping economics grew more slowly than the high-income-economy average.2 More disturbingly, in thirteen developing economies, per capita income actually fell. Growth among the eight HPAEs is quite dif- ferent. Their growth rates are significantly above the high-income-econ- omy average. Unlike most of the rest of the developing world, the HPAEs were catching up to the industrial economies. Hong Kong, Japan, the Republic of Korea, Singaporc, and Taiwan, China, were particularly notable. Other developing economies have grown ist fr several years, partic- ularly befbre the 1980s, but few others have sustained high growth rates for three decades.3 Figure 1.2 shows the growth rates in per capita in- come for 119 economies in two periods, 1960-70 and 197045. The 11 that achieved rapid growth during both periods are in the northeast corner. Of these, five are East Asiau success stories: Hong Kong, Japan, Korea, Singapore, and Taiwan, China Japan's shift from exatremcly rapid growth in the 1960s to rates more typical of high-income economies in die 1970s is apparent. The other three IIPAEs-Indonesia, Malaysia, and Thailand-all show accelerating growth, with higher growth rates in the second period than in the first Indonesia is one of only three economies '-28 GROWTH, EQUITY. AND ECONOOM Fhgure 1.1 GDP Growth Rate, 196085, and GDP per Capita Level, 1960 CDP grwth rate (percent, average, 1.960.85) Taiwan, China i ~*Rep.f 'aHongKong Kor 6 - --*Rep.-f Ko ma~~~~~aa lballand 111i hOOh8SIB* j i - 2-----O---- A -*--_ --* 0 **. S S -. - ...... ... . .. *S U HIgh4I.come econondes -2 -PEs ----- - -* HPAEs S Other developing econonies *~~~~ I . I III 0 0.2 0.4 0.6 038 1.0 1.2 elative per capita GDP, 1960 (percentag of U.& GDP per capita, 1960) Neo-This figuroplots this egession equation: GDPG = 0.013 + 0.OG2RGDP60 - 0.061RGDP602. N = 119: R7 = 0.036. (0.004) (0.027) (0.033) SorseSumnmers and Hestn (1991); Barnf (1989);Word Bank danL to mnove from the bottom to the top of the distribution of growth rates between the two periods. Declining Income Inequaliy and Reduced Poverty HPAES HAVE ALSO ACHIEED UNUSUALLY LOW AND DE- lininig levels of inequality, contrary to historical experience and contemporary evidence in other regions (Kuznets 1955). The positive association between growth and low inequality in the HPAEs, and the contrast with other economies, is illustrated in figure 1.3. Forty economies are ranked by the ratio of the income share of the richest fifth of the population to the income share of the poorest fifth and per capita real GDP growdt during 1965-89. The northwest corner of the figure 29 ATSIYA!SLAN. M I RAC L E Fruie 1.2 Growth Rate Perisbence GDP growth rate (percent), 1970-8S - Sin.s .oret Taiwan, China 6$|- _.!._ ._ _>_______ : * @ ~~~Singapore** 6%- d *Mai*a.siaRep. of Korea 4% - w*,*f . t *Thland Japa 0%~~~~~ 0 I . | * U High-Income economies 4%- * *HPAEs a Otherdeveloping economies -6%-- ! .4% -2% 0% 2% 4% 6% 8% TEM GDP growth rae (percent), 1960670 Nne: Boxes are seventy-fifth percenrilc of growdt ratcs in each period. S&natvSununcrs and Heston (1991); Barn (1989); World Bamk datL identifies economies with high growth (GDP per capit greater than 4.0 percent) and low relative inequality (ratio of the income share of the top quintile to that of the bottom quintile less than 10). There are 7 high- growth, low-inequality economies. All of them are in East Asia; only Malaysia, which has an indexc of inequality above 15, is exduded. When the East Asian economies are divided by speed of growth, the distribution of income is substantially more equal in the fast growers (Birdsall and Sabot l993b). For the eighr HPAEs, rapid growth and de- clining inequality have been shared virtues, as comparisons over time of equality and growth using Gini coefficients illustrate (see figure 3 in the Overview). The developing HPAEs dearly outperform other middle- income economies in that they have both lower levels of inequality and higher levels ofgrow&h Moreover, as figures Al .7-A1.9 at the end of the chapter show, improvements in income distribution gmerally coincided with periods of rapid growth. 30 GROWTH. EQUITY. AND ECONOX-rO Figure 1.3 Income inequalit and Growth of GCP, 1965!89 GOP growtb per capita (percent) Rep. of Korea * Botswana 7 . . Taiwan. China Singapore*. Hoagon ' Gabon 5 ----------apan---------- - -- ] e Indonesie . M . , , i Th. lland :- Malaysia _ _ _ _ _ __ . ;_ _*I Brazil I Italy * Austria - ;. .Spain o France * Belgium United Kingdom AuS ilia i * Colombia e Sri Lanka Switzerland' * iine Mec, Ia Pakistan ' 0Kenya . India * Venezuela 1- ._ lMalaw_ _ _ .A.t-- Ne * ChIIe Argenttna'j Bangladesh *PBolivia * *eruge!tn - Cote d'lvoErr * Maurita nia *. Ghana I a Sudan * -1 * -- - - - .- . __--... Za~~mbia $_ _ _ _ _ _ I _ _ _ _ _ _ I _ _ _ _ _ 0 5 10 15 20 25 30 35 40 45 Income Inequality N- et Income inequality is measured by the rado of thc income shares of the richesc 20 pexnr and the poorest 20 percenr of the popularion. Sm:n. World Bank dacaL 31 ECLE Two qualifications should be noted herc. First, some studies of Korea have shown increasing inequality in recent years; however, most of this is due to dte rising value of assets, particularly land, rather than in- creased variation in incomes. Second, reductions in inequality in Thai- land have been relatively minor compared with chose in the other HPAEs, although Thailand's performance is still better than that of most devel- oping economies. Given rapid growth and dedining inequality, these economies have of course been unusuallysuccessful in reducing poverty Table 1.1 compares the declines in poverry, defined as the inability to attain a minimal scan- dard of living (World Bank 1990b), in some HPAs and other selected economies (the period varies depending on data available). ncreases in life exptcy have also been larger than in any other region (see table 1.2). Dynamic Agricultural Sectors T iYPICALLY, AS AN ECONOMY DEVELOPS, AGRICULTURE'S SHARE tof he economy dedines. The six HPAEs with substantial agri- cultural sectors-Indonesia, Japan, Korea, alaysia, Thailand, and Taiwan, China-have been mak-ing this transirion more rapidly than other developing economies.4 But the decline in the relative im- portance of agricUlture in the HPAEs is not because agriculture has lacked dynamism. Across developing regions, agriculture's share of output and employment has declined most and fistest where agicultral output and productvity have grown the rnost (see table 1.3). From 1965 to 1988, growth in both agricultnual output and agriculmtal productvity was higher in East Asia than in other regions. Many factors contributed to the success of agriculture in these economies. Land reform (notably in Korea and Taiwan, China), agriculural extension services, reasonably good fastructure (especially in the former Japanese colonies), and heavy investments in rural areas (notably in Indonesia) all helped. East Asian governments have actively supported agriculatral research and extension services to speed diffilsion of Green Revolution technolo- gies. Their substantial investments in irrigation and other rural infia- struct hastened adoption of high-yielding varieties, new crops, and the use of manufactured inputs, such as fertilizer and equipment, to cul- tivate dthne In Taiwan, China, during the 1950s, 45 percent of rhe 32 GROWTH, EQUITY.- AND ECCOiL Table 1.1 Changes in Selected Indicators of Poverty PTerntage ofpoprdauion below the poverty line N mber ofpoor (millions) F.rst Last Fis - Last Percent A&vnenq Year year yar aAnge 'ar year chang HPAEs" Indonesia 197242 58 17 -41 67.9 30.0 -56 Malaysia 1973487 37 14 --23 4.1 22 -46 Singapore- 1972-82 31 10 -21 0.7 0. -71 -Thailand: ` 1962-86 59 26 -30 16.7 . 13.6 -18 Ot,ers : :- Bra4YAb - 0 1960-80 50 21 -29 :36.1 25A 29.6. Wolombia -1971-88 - 41 25 -16 8.9 7.5 -15.7 : . CostaRica 197146. 45 24 -19 0.8 . 0.6 -25 C6reds'Iire 1985-86-1 30 31 1 -3.1 3.3 .6.4 I-ndia 1972-83 54 43 -9 311.4 315.0 1 Morocco- 1970-84. -43 34 -9 . 6.6 7.4 :12.. -R; .-. > Pikismn0; . 1962-84 . ; 54 23 -31 26.5 21.3. -19 Sri-Laik 1963-82 37 27 -10 3.9 . 4.1 5 :NoeThiis able uses economy-specificpovertylines Official orcommonlyused povertyrlines have been used whenav2iblec In other C3SC5 die poveyie has been set as 30 percent ofean income or expeaditrc. The enge of povertylincs, epressed in tems ofewendi- . per householdmemberand in tersofpursingpowrpari (P) dollars, isapproximnatdy$3004S700aycarin 1985ececptfor- CosaRic (S960),-Malaysia (SI 420), ind Singapore (S860). Unless othenise indicatd, the cble is based on expexliure per household.s - ember. - - - - - -- -- -- -:- - - :-.. - a. Measures for thse enties use inmcmc cAher than expenditure. b. Mcasurs fir dtse entries are by hhold aier than by household memnber. Sm wz World Bk (1990b, tabics 3.2 and 3.3). - growth of agriculure was due to rising productivity, much of which re- sulted fiom government programs (see Ranis 1993). Information on the allocation of public investment betveen rural and urban regions is limited, and it is difficult to make good compar- isons among economies, but available data suggest that the HPAEs have allocated a lager share of their public investment to rural areas than did other low- and middle-income economies. Of critical importance in this respect has been the build-up of infrastructure-roads, bridges, trans- portation, electicity, water, and sanitation. Table 1.4 shows that there has been a more even balance between rural and urban public invest- ment in sanitation and water &cilities in Indonesia, Korea, and Thailand than in other developing economies. The data on rural electrification also suggest that the HPAEs with rural sectors have, on average, more ef- fectively provided electricity to rural areas. Since the early 1980s, elec- 33 I :1R.ACLE Table 1.2 Life Expectancy at Birth, 1960 and 1990 Lif expeeane, birthnh 6a) :- &onoal rfreon -1960 1.990 -;HPAEs- :,-- Hong Kong 64 78 Indonesia 46 59 -K(orea, Rep. of. 53 772 Malaysia 58 71 Singapor . 65 74 Th- Nailand: - -: . ;- . -: - - 52 .68 : OtherAsW: .0 .62 -Chin. 43 -69 In&dia .47 58. -Sub-SaharnAlea . 43 52 Lad Ame7ia and Cribbean . 54 . 70. :Low-Inome econom .s : 36 - 62- -Middle-lacnc ieconomies . 49 66 - -Industnaleamenimie 70 77:. N - TAhe rCgionalpavsarweghreda.; - - -t Exdudes Chinaand ndi:L SountWorld Bankdra.t -- tricity has been universally available in the rural areas of Korea and Tai- wan, China. Malaysia and Thailand have made great strides in rural electrification. Endonesia has not done as well, but even there die relatve disparity between the urban and rural sectors is smaller than the dispar- ity in econorries with approximately the same per capita income (Bo- livia and Liberia) or the same population (Brazil [see table 1.4]). Table 13 Growth Rates of Agcultural Income, Labor Force, and Share in Output, 1965488 (percent) * - Decline,Average anased' growth rv&eof 1965-88l. A t,i - d Atcrds :,,-;,: 196 198 -irc ,--: .;-:-;:e kiborfirce rru8,,,A A'd'-:.-.:'..' " t SouhAsa ,3 2. ' <2., 6 1.l9U .:: :'4asrArn 41 22 : :-t _ - 2.2 :~xree..LLL2jt.... ~~c:~;'nA2, 34 LO) 21~~3 34,., S :10" ~.. 37 2 17 34 GROWTH. EQUITY. AND ECC Equally important, however, were the typically low levels of direct and indirect taxation on agriculture in East Asia. During the past three decades, dozens of governments in other regions, eager to promote m- dustrial growth, have funneled surpluses from agriculture to industry through taxes, food price controls, and pro-industry allocations of pub- lic investment. Less overtly, govemments have fivored manufacturers, and hurt agriculture, by overvaluing currencies and protectng domestic industries thar manufacture agricultural inputs and the goods purchased by rural households. The exchange rte chat results firom restrictions on Table IA Comparison of Rural and Urban Public investnent ARnral-urban dipadrics in access to gervies, 1987-90 (100= rural-urbanpaitj) -c y - : . - . Watrr S- - ti don S PA Es Korea, Rep. of ;54 101 Thailand 1- -26 102 Indoncsia -168 113 OtcrAia - - : - 64.6 38.6 Latin America 58 4665 SubS - ibw bn Africa 43 30 .- Prcatage of rrurl adnd uran pop ukonser&d by ecaicity co~namy Ur Rura ' PAEr Indonesia,-1984 39 10 :M.a:ilays.;1;983 - : -:0: g 5- 85 55 Thailand, 1984 . 78 :40 -OrhlerAsik: - Bangradash, 1981 :20 2 :India, 1981 <25 15 -S -Lakna,j982 -..35 8 - LeninAme-ic *~ArcndinaJ 1982- ->95 : 5 Bolivia, 1981: 72 . 9 .. Bra;Ail, 1981 N :-.i .......... :0 0- - : >95 19. ;C6tc -dvoir, 1981 93. : 20 i iberia,1982 . ^ ~-- -~ 86 ; ; 4 ~0~ - Senegk1982 . * ; 83 . . 12 Soercertop panel: UNDP ('r fariousyis); Bottom pan. Muasii (1987) - .; . . 3S _RACLE manufacured imports reduces thc domestic currency proceeds of agri- cultual exports. Industrial protection acts as a hidden tax on agrculture, raising the price of agricultural inputs to subsidize industry. Figure 1.4 il- lustrates the lower taxation ofagriculture by contrasting three HPAEs with large rural economies with three South Asian economies. Direct inter- ventions include export taxes and price controls, while indirect interven- tions also take account of industrial protection policies and real exchange rate overvaluation. Both Korea and Malaysia have substantially lover tax- ation of the agricultural sector than the cornparators, and in Korea the Figue 1A Intervembon and Gmwt in the Agricultural Sector .7=,,,_T; r-- ,. Rep. of Kore, Malaysia -------w------ _, , Pakis-tn' w __. ......... _. _---- --- -'i-~~R 19709~~ PSilippines K. -. _ . . 1 j _ _ +_ _ _ _-1_ _j___________________________ _ __I___._________ Sri lanka '::H-~ ! : I I ; : I ! | I :~~~I * -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 Level of piotection (percent) Rep. of Korea alys ... . i -............I i I Thaland . -- PsidalsiU =96 i ___ ____ I ___ ~~~~~N1970s I - 3---- 197ks PhEippines . . I k .~ ~ ~ ~~~ ~ ~ ~ ~ _ l_ __ _ _ _ _ _ _ 1 ~ ~~~~~~~~~~ I Sri lanka IE~ -2 0 2 4 6 8 10 Agrutural growth mte (percent) Sour= Schiff(1 993). 36 GROWTH. EQUITY. AND EC_OO agricultural sector receives positive protection. Thailand's taxation of the agricultural sector was similar to South Asian levels in the 1960s and 1970s but fell in the 1980s while taxation in South Asia was rising. We are nor suggesting that the HPAEs avoided transferring any re- sources from agriculture to industry. Rarher, in contrast to ocher devel- oping economics, they transferred fewer resources. Moreover, transrs were often voluntary-financial savings, for example-and were nearly always limited to levds that did not choke off agricultural gwth. Rapid Growth of Exports A NOTHER STRIKING FEATURE OF THE HPAEs HAS BEEN THEIR superior export performane This is reflected in their steadily Ilk rising share of woldd exports (see table 1.5). As a group, the HPAEs increased their share in world exports from 8 percent in 1965 to 13 percent in 1980 and 18 percent in 1990. Manufactured exports have provided most of this growth. From 1965 to 1990, Japan emerged as the world's biggest exporter of manufactured goods, increasing its share of the world market fiom nearly 8 to almost 12 percent. In the 1970s and 1980s, the locus of growch shifted to the Four Tigers, whose share of manufactured exporrs grew nearly fbur tmes faster than Japan's (see table 1.5). Beginningaround 1980, the three SoutheastAsian HPAEs (In- donesia, Malaysia, and Thailand), which had been historically depen- dent on commodity exports, recorded a sirnilar but so far smaller surge in manufactured exports. Some analysts have, with hindsight, attributed these achievements to unique cultural and geographical circumstances. But there was litde evi- dence at the outset that East Asian economies would achieve such spec- tacular results. In the 1950s even trade optimists were export pessimists and did not anticipate that Korea's exports would grow four times as fist as world trade during the next thirty years (see, for example, little 1982). One obvious efFec of rapid export growth has been a marked in- crease in the openness of thcse economies, the share of exports plus im- ports in GDP (see table 1.6). Malaysia has been heavily tade oriented throughout its post-colonial history while Korea, Thailand, and Taiwan, China, had average trade propensities untdl 1965 but have since moved wel above intemational norms. The trade orientation of Indonesia, 37 I1RA c IrE Table 1.5 Export Penetration, Selected East Asian Econonies, i965-90 Share in wuorld eports S-are iL developing-economy xporm Eeonq - :1965 1980 1990 1965 -1.980 1990 -Totalmg,parn Japian 5.0 -7. 9.0 - 7 Fourflgcr- 1.5 3.8 6.7 6.0 13.3 33.9 SoutheastAsian NEs - 1.5 - 2.2 2.4 6.2 7.8 12A HPAE subtoal- 7.9 -13.1 18.2 12.2 21.1 56.3 All developing economies .242 28.7 198 100.0 100.0 100.0 World 100.0 100.0 100.0 NA NA NA : --i - a-u.-:; -- Japan> -. =.- ;- 7.8 11. 11.8 - Fourr rsa - 1.5 5.3 7.9 13.2 44.9 61.5 SoutieasAsian NIEsb - 0.1- O 1.5 1.1 3.8 12.0 HPArEsubtotal 9.4A 17.3 21.3 14.- 48.6 73.5 * AIIdevc1opingeconomies 11.1 11.8 12.9 100.0 100.0 100.0 *Wodd . 100.0 1;00.0 -100.0 NA NA NA - -Nor available. NA-.Nor applicable. -ai Republic of Korea. Hong Kong, Singapore. and Taiwan. China. b. Indonesia. Mala a. anlThallu:xL : ort U.N. Trade Systems darn. w-ith its vast economy and only relatively recent export drive, is only sightly above the world average but growing. Rapid Demographic Transitions THE DEMOGRAPHIC TRANSMON FROM HIGH TO LOW BIRTH and dearh rameS began in Europe and North America with the .1. Industrial Revolution, and it took nearly two centuries to com- plete. That same transition is occurring much faster in the developing world. In virtually all economies, deati rates have fillen substantially in the postwar period; in ali eccept Sub-Saharan Africa, birth rates have also fllen significantly (see table 1.7). However, compared with other developing regions, the transirion to low fertility began sooner in East Asia (in the 1960s in the north and in the early I 970s in the south-and of course even earlier in Japan) and has gone firther. 38 GROWTH, EQUITY, A-ND E Table 1.6 Ratio of Total Trade to GDP Ecnousn" ; EO1970 1980 1985- 1.988 HPA& 'Hong Kong. 1.50 1.52 1.78 2.82 Indonesia- . - 0.25 0.46 : -0.38 -0.42 -lorea, Rep. of: - 0.32 0.63 :0.66 0.66 Japan 0.19 '0.25 0.23 0.11 Malysia 0.89 .1.00 0.85 -1.09 Singapore 2.12 3.70 -2.77 3.47 -Taiwan, China -.0.53 0.95 ... 0.82 0.90 Th1ailand 0.28 049 0.44. 0.35 SubR -SSI,arnAfr-k - 0.24 : 0.30 0.27 0.45 5out6Asia --; ; -: ; -0.11 :; .-0.17 - 0.16 0.19 A-men -Amc'sa ad Caribbema 0.20 025 0.22 .0.23 N..-eTotlrade .-value ofxports and valueofiimpomr divided bygrass domesdc - -.- produce :- -. - -t' -.SowrwWodd'Bandata. Taiwan, China, vanious issues. Nadionial Accounts SMtaddcs: Analysis ofMainAggegates, 1988-89 (Unietd Nations). - During the period 1965-80, all the developing regions of the world experienced a marked decline in crude death rates (see table 1.7). The de- cline in most economies was 30-40 percent and did not vary much among regions. There was, however, substantial regional variation in the extent to which declines in birth rates held in check the potentially ex- plosive growth in population from the rapid mortality decreases. In Sub- Saharan Afriica during this period, the declines in birth rates were around 2-10 percent; in South Asia, 10-30 percent; and in Latin America, 3040 percent. The sharpest declines were in East Asia: 40-50 percent. As a result, the rare of population growth declined in all the East Asian economies, in some cases quite sharply. For example, in Korea it fell from 2.6 percent a year in 1960-70 to 1. I percent in 1980-90; in Hong Kong, from 2.5 to 1.4 percent; and in Thailand, from 3.1 to 1.8 percent. In Latin America, fertility dcdines were also sufficient to reduce population growth rates, though generally not to the levels observed in East Asia.- In South Asia the picture is mixed, with fertility declines suf- ficient to reduce the rate of population growth in Bangladesh but insuf- ficient in Nepal or Pakistan. In Sub-Saharan Africa, the combination of sharp declines in death rates and modest declines in birth rates resulted in an acceleration of population growth-for example, in Ghana from 2.3 percent a year in 1960-70 to 3.4 percent in. 1980-90, and in Kenya from 3.2 to 3.8 percent. 39 Table 1.7 The Demographic Transition (percent) Change in -ange in Aventgeannul cr-de crude grovtt of birth . eadl popadation rante, rate, Eonomy - 19650 1965-80 160-70 1980-90 EaswAsia * Hong Kong --52 -54 2.5 1.4 Indoncsia -40 -55 2.1 1.8 . Korea, Rep. of - -54. - 2.6 1.1 Malaysia -25 -58 2.8 2.6 Singapore 45 -16 2.3 2.2 Thailand -46 -30 3.1 1-.8 Latin Amen'ca Bazil - -31 -36 -2.8- 2.2 Mexico -40 -55 :3.3 2.0 Peru -33 -50 2.9 2.3 Venezueda -31 -38 3.8 2.7- Sub-Saaran Afri-. Ghana -6 :-2 23 -3A- Kenya -13: -50 -3.2 3.8 SierraLeone -2 --29- -1.7 2.4 Tanzarta -2 -22 2.7 5.1 SouthAj~a Banglad -27 --33 2.5 2.3 India -33 , -45 2.3 - 2.1 Nepal -13 -: 42 - 1 . 2.6 .-Pakisran. I -13 -43 2.8 31 -Nor avaibbk.- So5Wrrw World Ba dara -- - - :- High Investment and Savings Rates HYSICAL INVESTMENT INCLUDES ALL OF THE ECONOMY'S OUT- P Jput that is not either direcdy consumed or used up in the pro- ducuion of other goods. Machines, buildings, and infrastructure are physical capital, bur elements ofworking capital, such as inventories, are also important. Economists traditionally have viewed investnent as one of the driving forces of economic growth. In a dosed economy, sav- ings is the only source of invesrmenr, and the two, by definition, must be equaL But in an open economy, investment can be financed by bor- 40 GROWTH, EQUITY. AND ECOQi rowing from abroad as well, that is, with foreigners' savings. Even so, in- adequate domestic savings will eventually pull down investment rates, either direcdy or through constraints on the continued build-up of for- eign liabilities, which must eventually be repaid firom domestic savings. Between 1960 and 1990, both savings and investment increased markedly in the HPAEs, outstripping -',e performance of other develop- ing regions (see figure 1.5). Savings rates in the developing HPAEs were lower than in Latin America in 1965, but by 1990 they exceeded Latin America's savings rates by almost 20 percentage points. Investmenc lev- els were about equal in Latin America and East Asia in 1965; by 1990 East Asia's investment rates were nearly double the average for Latin America and substantially exceeded the rates for Scuth Asia and Sub- Saharan Africa. The HPAEs are the only group of developing economies in which savings exceeds investment, making them exporters of capital. When we compare the HPAEs individually to all 118 economies for which investment data exist, the picrure is more complex (see figure 1.6). During the period 196045, the HPAEs' investment levels were in Figure £5 Savings and Investnent as a Percentage of GDP Gross domestic investrnt Developing HPAEs Latin America and Cauibbecn South Asla Sub-Sahaan Africa Gross domestic sawings Developing HPAEs _-_ _ _| I --_____________________________ - i .1-. ..- &----- Latin America nd Caribbean South Asia Sub-Sahoran Africa : - . g. e10 20 -30 40 N-eo: The regional avages arc unweiglhted. Sepurc World Bank darn. 4' RAC L E Fgut 1.6 Average Investment Rate as a Pementage of GDP, 1960-1985, and GDP per Capita Level, 1960 Average Investment rate (percentage of GOP, 190-8S) 40 : 30----- * S bnga*Japan 0 *Malaysia - e Rep. of Korea * * Talwan. Chilna 207 '.- -6-- - '… E S SXhalland - U 111gb-Income economnies * HPAEs i 1' ! . * Other developing economies 0-- 0 0.2 0.4 0.6 0.8 .0 2.2 Relative per capita GDP (percentage of U.S. GDP per capita, 1960) Note. Regression equation: 16085 = 10.125 + 59.12ORCDP60 -51.88IRGDP602. N = 119: R2 = 0.295. *(1383) (10-344) (12.593) Swrre: Summers and Heston (1 991); Bam (I989): World Bluk dat. the top quartile of all low- and middle-income economies, but since in- vestment generally rises with incomes, this is not remarkable.5 What is remarkable is their high share of private invesrment. Figure 1.7 com- pares average private and public investment shares in GDP for 1970-89 in the developing HPAEs and other middle-income economies. Private investment is about 7 percentage points higher in the HPAES than in other middle-income economies. It rose from about 15 percent of GDP in 1970 to nearly 22 percent in 1974, then declined and held at about 18 percent between 1975 and 1984. Private investment contracaed sharply between 1984 to 1986, reflecting the global recession, then re- covered by 1988.6 In contrast, private investment in other low- and middle-income economies has remained relatively stable at about 11 percent of GDP. .427 GROWTH. EQUITY, AND The story is different with respect to public investment. In the 1 970s overall levels of public investment did not differ markedly between the HPAEs and other developing economies; during the decade public investment rates in all economies rose from about 7 to 10 percent (see figure 1.7). But during the 1980s the HPAEs and other developing econ- omies diverged. In other economies, the fiscal contraction of macro- economic adjustment was reflected in lower public investment races. In the HPAEs, conversely, public investment shares actually rose between 1979 and 1982 and then remained at a level nearly 4 percentage points above their 1970s average. Only after 1986 did they begin to dedine toward historical levels. In short, in striking contrast to elsewhere, pub- lic investment in 1980-37 in these East Asian economies was counter- cyclical to the reduction in private investment. Creating Human Capital I-N NEARLY ALL THE RAPIDLY GROWING EAST ASIAN ECONOMIES, the growth and rransformarion of systems of education and training during the past three decades has been dramatic The quantity of ed- ucation children received increased at the same time that the quality of schooling, and of training in the home, markedly improved. Today, the cognitive skill levels of secondary school graduates in some East Asian econormies are comparable ro, or higher than, those of graduates in high- income economies (see appendix 1.2). Figures 1.8 and 1.9 present a stylized summaxy of the results of re- gressing primary and secondary enrollment rates on per capita national income for more than 90 developing economies for the years 1965 and 1987.7 Enrollment rates are higher ar higher levels of per capita income. But the HPAE's enrollment rates have tended to be higher than predicted for their level of income. At the prmary level, this was most obvious in 1965, when Hong Kong, Korea, and Singapore had already achieved universal primary education, well ahead of other developing economies, and even Indonesia with its vast population had a primary enrollment rare above 70 percent8 By 1987, EasrAsia's supenror educarion systems were evident at the secondary level. Indonesia had a secondary enroll- nment race of 46 percent, well above other economies with roughly the same level of income, and Korea had moved from 35 to 88 percent, 43 jA C L EitACLE Fgure 17 Public and Private Investnent Private Investment/GOP (percent) 25 .5 -D- HPAEs (excluding Japan) Other LMlEs 0 1969 1974 1979 1984 1989 Public Investmfent/GDP (percent) 20 - ------ 5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I 10 -U111- FPAEs (excluding Japan) -4- Other LMIEs I ' I } t i ~~~~i ; ! i . ; 1 : i I I i i , I 0- 1969 1974 1979 1984 19 Note: LMIEs = low- and middle-income cconomics. So=eue World Bank dami. maintaining its large lead in relative performance- Only in Thailand was the 28 percent secondary enrollment rate well below the income- predicted 36 percent and the 54 percent mean for middle-income economies.9 In recent years Thailands weak educational performance has been felt, as serious shorages of educated workers have begun to threaren 44 GROWTH. EQUITY. AND ECC Firm 1.8 Crss-Ecnony Regressin for Prmaiy Enrolment Rates, 1965 and 1987 Pdmaty enrollment rates 140' - - :120 .. ......... . 120- - n -- . lindonesia (1118V Brazil (108) Singapore (1053 [37.0] Hong Kong (106) Brazil (103) U [21.9] * ~~~~~~~~~~~~~[-7.5] ioo Rep. of Korea-(101)u - [29.2,1 Hong Kong (103) - - _- i7aysb (102) Rep. of Korea (101) . Malaysia (90) [15.61 Tbai95) [4.3] [-7.8 Thailand (78) [17-9_ [7.01 8 [10.8] Indopes15 72)1 -; I60 ~ pesI5ji2y -- '[8 d _ _ _ _ _ _ _Banigtadesh (59) *Bangladeh(4.9) ,Pakistan (52) 40-its . 1965 - -.[O79]1 9 7 4-Palkistan (40) 6$2 ........... . -------197--- 1-24.5] 20-- Lw . 0 250 500 750 1,750 2,000 2,250 0 500 1,000 1*500 3,500 Per capita Income (198 U.S. dollar) Nater Figures in parentheses arc enrollment rates brackretd numbers show rcsiduaLs Sourre: Bchrnan and Schneider (1992). continued very rapid growth. In part as a function of their success in in- creasing enrollment, the East Asian economies have also been faster to dose the gap between male and female enrollments (see box 1.1). A common, though imperfect measure of educational quality is ex- penditures per pupil. Between 1970 and 1989, real expenditures per pupil at the primary level rose by 355 percent in Korea- In Mexico and Kenya, expenditures rose by 64 and 38 percent, respectively, during the same period, and in Pakistn expenditures rose by only 13 percent be- tween 1970 and 1985 (Birdsall and Sabot 1993b). These dramatic dif- ferences reflect mosdy differential changes during the period in income growth and in the number of children entering schools, both of which favored the East Asian economies. A somewhat better measure of school quality is the performance of children on tests of cognitive skills, stan- dardized across economies. In the relatively few international compar- isons available from such tests, East Asian children tend to perform better than children from other developing regions-and even, recendy, be'tter than children from high-income economies.'0 45 & 4TIRACLE Fimre 1.9 Cross-Economy Regresin for Secondary Enrollment Rates, 1965 and 1987 Secondary enrollment rates 100 - . -. ,U , ,~- - - , _- ; , . , . ¢ | I , ', . . Rep. of Korea (88) so - -- -- i 60 - -.. Malaysia (59) Singapore (45) Indonesia 1 _- 40 BraA 9*-[18. .119.5] >I 40~~~~~~~~ - - Rp 're\35 --lt- ril3s *! [19.1] T k .-~slhailand (28) Bangladesh (13) *Malaysla (28) g 9) I - 77--{ ~ir3jidnf .BalI (16)f Bangadesh (18) Ft8 0- 2 1 akstan (12) 119 0 250 500 750 1,750 2,000 2,' 0 500 1000 1500 3,500 Per cpl Income (In 198 US. dollrs) Nomt Figures in parentheses arc enrollment raes; bracketed numbes showv residuals. Seurce Bhrman and Schneider (1992). How much a child leams is also influenced by the nature of the leam- ing environment in the childcs home- Again, children in the East Asian economics had advantges in the 1970s and 1980s. Using a simple index taking into account the mother's education and the number of children at home, we estimate that the learing environment in Korea during twenty years was enhanced 114 percent more than in Brazil and 147 percent more than in Pakistan (Birdsall and Sabot 1993b). Rapid Prdu Growth HE FINTAL OUTSITANDING CHARACTERISTIC OF THE HPAFs IS their rapid productivity growth. That labor productivity in the HPAEs increased rapidly is self-evident per capita GDP grwth depends primarily on rising outpur per worker Some of this increase in output per workler results from increases in physical capital per worker (for exarrplck more machines) and some reults from increases in human 46 GROWTH. EQUITY, AND:,F-COX0 MT VT!"12 rT- 77 Ze. ,i Mz w,Was-,FAA IF.AJRL'' 11AAHI'.. MRICALLY-k6---,----!: eci c.- ene tv o cu!,adng:V 7 .9 es Lrsu to - oyt i'an biin subiiiiia 4. d. lop.1 o y a=,;-.,us J--- ygra uaUj;'..'ib;gdsj gender' Th d il iffii:ntii f6cyiii-,., in ca-d6i. 1.o-- n";' evc -rnenr iv aryan sc6ndzrj'k:& -th -1--'-'6d---' c, ower ataun.-..,,.-,,,to-aiaineas-incomesmaeast --- --bf thi Aia tir).Ji-I 9 65i at c pd-n-.---- yztdj,.,1mpr6ve, I the' 'c[ p rnaryp_ c ccorioriuq,-,gnder V V '1987'thi,&ur ,,;,.t-',"Jt)r,outst e,--.tlii --hi" The- 'j!aI[70Wid-,- e:gcn-':-'.,-',;;,' e. omm`- e' --HPAB_ WM Swl"cr-,,E=ticxpected.,,-By- pmg jan!eqonoa,Lies-f6i.W.W&6ca. ot er eV 0 aM I r; ;cconomjqs,' splteZwr'u''iat, norms-,,ttl2t. put, greater--, achieved universai p6mary cation C , :r p-at' a-'t-1e'%'rib7SLLbi6iL Wqq01L- ucation-ofs6diaahi, tisci; -Ov- iiiiirii'n'aungthe''gendir th' C gender-gaps, continik46" dvt: I Mm ii ai -d6cationl., dau M' um mtes 10 , i'. E. -'. r -;- " -, - -.  - , -- -.4.-, -"s. I. .......... Sl g M e; ra er. C CXCCPtI0r4.-- in-liow-tncom"C' t Ca-1r0m-1 SUCCCSS 0r-t1MVCrsJid[7-, '. -,; Z:. - , - P-M 7-,C ... PUS Is 'y ucauon Sphe' thi! m c3q%=on- o oni ive4rgu -64c4,,th qarrip e,, f G61 64i S31S =&T.1fij,'-1991 :F6v 7 lecon6nues, GNP pq.qLpibi '$500'-% 'I. A -in aTsmoolv 0'ff -o m Miienr, etwccwiboys .AD"Mak-Yea 6f 'Edudfialfl ges-2 6fita- i pio,in, fs-,--=,a, as, 'no-r4n p In!; 1965, 1-ir '(6c, 6 Thi -.4 cq 4 gcn..d&,gaps-,-avtqe-k' di nAmi Iiid 'd'-" -16iill: cc aryschool-pp increastrig.second e;M as;mu .2- C Ij 'an C5 7 0, v-166'.1-970i cl -,q80s!2,S1t-.1S rhl - - ey opmg. -wo & -Ccon6mli SUCCCSS Ilh laL%ngj cntsto PV .98 V4.1 gn !cr:ga,p -wJ- `6ii6, er 1) Ja ii I 4-. -j k capim-41 -6 A-siadeconornies.5nour,stu Z.1 fi capit2l. (for cample, better schoolecl workers). In addition, however, somc may resulr fiom greater efficicncy, fiom changes in producdon practices that result in greater output from the same stock of physical and human capital. We caU the increase in productiviry thar cannot be accounted for by measured increases in inputs total factor procluctivity 47 r1;CTACLE (TFP) growth." Among its sources are becter technology, betrer organi- zation, gains fiom specialization, and innovatons on the shop floor. As we demonstrate below, about two-thirds of East Asia's extraordi- nary growth is attributable to rapid accumulation; that is, to unusually rapid growth in physical and human capital. The remaining third of ti-s growth cannot be explained by accumulation and is tierefore actribut- able to increased efficiency or TrP. This is large relative to other econo- mics, both absolutcly and as a share of output growth, and therefore partly explains why these economies have been catching up writh the in- dustrial economies, while most other developing economies have not. In box 1.2 we review two broad interpretations of the relation between ac- cumulation and output growth and the link between productvity change and technological catching up. Evidence from Cross-Economy Regressions of per Capita Income Growth In this section we use two statistical techniques to examine the rela- tionship between accumulation and output growth Our first model is a cross-economy regression in which we estimate the relationship between the rate of rcal per capita income growth and the share of investment in GDP, and two measures of educational attainment, controlling for dte rate of growth of the economically active population, and the relatve gap between per capita income and U.S. per capita income in 1960 at 1980 U.S. dollar prices.'2 Our measures of educational attainment are primary and secondary school enrollment rates in 1960, the area of human capiral accumulation in which the HPAEs showed a substantial difference from other low- and middle-income economies. The results of applying our statistical analysis to 113 economies are summarized in table 1.8. The estmates indicate that fcror accumulation matters. Investment in both physical capital and schooling contribute sig- nificantly to economic growth.'3 An increase of 10 percentage points of GDP in the rate of investment-about the difference in private investment rates between HPAEs and other low- and middle-income economies- would raise the growdt rate of GDP per capita by 0.5 percent An increase of 10 percenmge points in the primary or secondary school enrollment rare would raise per capita income growth by 0.3 percent Beyond accumulation of physical and human capitl, initial income also has a significant relationship with per capita income growth. 48 GROWTH, EQUITY, ANDE'CO.'i M77 i 7, x ,Zrl%evjc,;i-.Ii_-.Ip"-."- _2__ A A )NOMISTS, ROAD, mmmenr uman an 1-4 accumuw- er: onj'am Murpur VC.'. .,Jf. -w-V h ton;t,&I rexten --o 3K - Flj e 3Vlcwlls"_b"l V4, 1. _T'. y a IC um2n ppi ajt it r=en ini p nonluesi- irk*, -o'wth' an A ysi &pfil- &72ACt 1.06 ".w -,,-,,.amo4ccon6n3jcs_ cr cromki&m rz-k iE66-Er; esame, 6i&bffinva= ccononiic,_-or -4 gan,=qon;,-jnnovau nonuc-.,gr -0 o ogyarc o arm y.14666i assem- 'lik" r tLon MR P 31i iAiii '446 ccum non-, cKami) t M _-llFAE:f a ii'ibilbimer vict-, ruo have "d zin 4 "W o .. ........ CoDrifth lporpcmicnVor economc gm. M, Om ccum 07Ww"Ah -,-tc con-._t-,,, oil ficior- Z.- Q _9M Y ---- 2ssum rec i=cs 0 ,p LW2'CX (SOIW49- 5.7) N -nu e-mcoifi 0 f -T, P, P'. J:i _7 coiiiii.econoun v ll,z, Om r onspi c Oj_ Cj;V;. aupajmong de ,-,'2ni t V 19P -J ;_2 t crcmeat&jj06, 'gro er, 0 gt 6ibasji-,tVoutpu poon 0 .4;_ f ItaL A econo n V4 OD 'p t Sd' 49 L i-MIRACLE S j ': . $ ':~~~~~~- --.r~ l'--'rr' ~i. r:*rt ',, ,.. :.;,..:,, ?%e7 aWe to acqwre technologyhniore~ =Žie4iy; daii nical efnaen4 chan ge- Bo6h tho y -and empi ic prgres&more rapidlj than induiial --economics. evidence: leadus to, nclude chat tol&cr pruc- K>Thisiis pnid4crrybased catcbiiig: up.. vy growthl thdeindusirialecmomiioicilrer U- radbe- i" a ; < . mov dc j-- r'BuLiiiwrnadbsn4a1 be fpraice is uiseif amoving, t. technologicacg the-. m target Tow fiictor. producvity cbangeii; el.igh t -nonal e practic4 While inflow-and middle-icome i--- OM '-Co .-di Z = -sc toie to, oc g5wrkXCh enry 1986;B.n an 192 m iic j(Pacad P 1 1993X Forpductrv- -liI;pm~~~~~~n a &uncV~A9)--4 l .nocr; i, pendix in resents acformiantan4is" oEfi rea~onow-bs > slp bcwce ni#eicrsKn ~do iia b hs' mov4iidj occr, 'lo-Cand ; - pracice, whitWChe temi "ltec-noloejcl r banmcemanon d ccInomits, cam-b gaicing on, 6te ~~~~~~~~~ O& 4' - -moIn Ass* zzG* 9; >.~movemenrs)owardbestphiice,whicWweolraech'- -.mu s t .: .-Ceposk,iv---: Economies that were relatively poor in 1960 grew significandy fister than relatively rich ones, controlling for levels of education and invest- ment So although, as shown above, poor economies did not do better on average than the high-income economies-income levels were not converging-this was partially because poor economies invested less. But an economy at 50 percent of the level of U.S. per capita income in 1960, and at average levels of education and investment, would have grown about 2.1 percentage points more rapidly thian the United Stares. We call this conditional convergence, because economies with low rates of investment and school enrollment would not catch vp despite the "advantages" apparently offered by being relativly poor. This conditional convergence can be interpreted as a measure of the gains realized as a consequence of moving from lower to higher techno- logical levels, or "catching up" (see box 1.2). But rhe relative income variable may capture more dtan the productivity-based catch-up effict that some initially poor economies benefited fromr. One of the impor- tant empirical regularities found in the early literature on structural transformation (Kuznets 1959; Chenery 1960) is the discrepancy in the average product of labor between traditional (agriculture) and modem (industry) sectors at low income levels. A dominant share of producriv- ity growth in low-income economies can be attributed to intersectoral realocation of labor from agdriclture to industry (Pack 1993c). Thus the initial income variable may also be capturing the reallocation effect of this structural change on per capita GDP growth. 50 GROWTH, EQUITY, AND ECO:4I- Table 1.8 Basic Cross-Economy Regression Results (dependent variable: average rate of real per capita income growth, 1960-85) 113 113 ~~~~~~~113 * Variable O&Osedns a obsenmign obsen'aions Inte'rcept 4).0070 -0.0034. 0.0042 (0.0079) (0.0075) .(0.0081) Relative GDi' to U.S., 1960 -0.04O -0.0293 ~*-0.0320 (0.01 18) (.15)(0.0110) Priuary enrollment, 1960 0.0264M 0.0233 0.0272 (0.0065) ~ (0.0062) (006) Secondary enrolmenr, 1960' 0.0262 0.0160 0.0069 * (0.0139) (0.0132)- (0.0131). *Growth of pop ulation, 1960-85 0.1015, 0.0201 0.0998 (0.2255) ~~~(0.2095) (.03 Averageinvestmendr/CD. 1960-85 0.0578 0.450.0285 (0.0224) (0.021 1) (0.02.07) I{HPAFS'. 0.0230 0.0171 -(0.0056) (0.0056) *LatiAmrca *-0.0131 t (0.0039) Sub-aharan Africae 0.0o9.oo Adusted R. 0300424 0.48211 ~Stadrisda1y.signfl=wacrtiC 0.0f lVeL. nnsdca1y.sniflc~nt2tt die0.051V& - '6. Nr:coefficlent is: top number in cach ceIll Standard crror is in p crnhses t.Also indiude Tun".isiand South.AMica. S6an Woild Bank snff csdnnrcs. Armned with thes e statistical results, we can examine two facrs about the role of accumulation in growth in the HPAES. First, our estimates of the parameters of the cross-economy growth equation enable us to do a simple "accounting for growth" in the HPAEs. Table 1.9 shows the pro- dicred contribution to growth of investment, human capital, population g'rowrh, and relative income for eac-h H-PAE. It also shows the share of ac- tald growth predicted by these variablest Excep t for Hong Kong (44 per- cernt), 60 percent or mare of the actual growth rate in the HPAES is predicted by the accumul-ation of physical and human capital, initial in- come levels, and population growth, ranging as high as 87 perce-nt for _ E FML~"xN I R-A-C L E Table 1.9 Coutnbution of Accumulation to the Gmwth of the HPAEs, 196045 ,~~~~~~~~~~mm Iha . w . , ; fl,.~A, ' 11- os r_ P0- f +: -aaf hiv --fa-d - aeihsh ral -..&vh If-d rmd Pear Mg. .6 PedSI MlUe6 prial 5W. pen&tu ugaqh pia *VWzh& mlal (vi uik(si gush (rf-) guhw. s Ian~~~x ~40676 4270 (.26) 476 (.1) 4710 4.151 473 (-191 - - . - .X7e .e70 {.Nz wu78 402 -141 .1.1 -W48 17-191 car US.2560 4.030 -ISe (7) 4 - 32 (.242 -2.16 (1 +X4 (- -f-mapa,Imau 15%6 O-.Jf 20o (a5) 1.77 -9) .72 (sw 243 10 S 5 -cmXular md . 19S0 i e.o2 - 0.3 (231 0.16 (7) 1.4 1421 e, a19) - -: Gcpq&u.n2104 0.1025 0X3 - s1) -.2 (16) - 0.20 (2l 0.79 - GM* AwpiPdUwunt 160.85 ex57 l (4) 2.10 (.1)la M .4 3 :~ ~ ~ ~ ~~~~~J 1._D.108 -L0N12 -1q --It {1918 D 1.40 Lin -- - AaW Ewd.A 69 5.7 =.6P 5 - U6 P- M& - 2A 223 - 4.69 30 F- gn pdiicS (44)( * .vAwM cmibdn wpz.n vacsylwdwwAaeesI .iwnyim XI nb dal a enIa dwbEnul(nr dc bienps. flqnJul rend Grcceyqlsdwwm JaUsihIa- yr. -:dbc .pi .m.bl-dn. Malaysia. On average, about two-thirds of the observed growth in the HPAES is predicted by our model. Primary education is by far the largest single contributor to the HPAEs' predicted growth rates. Between 58 percent (Japan) and 87 percent (Thailand) of predicted growth is due to primary school enrollment. Physical investment comes second (berween 35 and 49 percent), fol- lowed by secondary school enrollment. Japan's high secondary school enrollment in 1960 makes a particularly strong contribution to its growth (41 percent), more than investment, while the laggards in sec- ondary enrolhment rates, Indonesia, Malaysia, and Thailand, have the smallest proportion of their predicted growth attributable to secondary enrollments (less rhin 15 percent). Investment, conversely, is most im- portant in Hong Kong, Indonesia, Malaysia, and-Singapore, where it ac- counts for more than 40 percent of explained growth. Second, we are also able to look for regional patterns in growth rates that are unexplained by physical and human capitl investment. Con- trolling for their performance in education and investment, and initial income, the HIPAEs have a significantly higher rate of growth than all other economies. Latin America and Sub-Saharan Africa, in contrast, have significantly lower underlying growth rates (about 1 percent), con- trolling for the same variables. Thus, the expected growrh rate differen- tial between the high-performing East Asian economies and Sub-Saharan Africa or Latin America, even if they had the same accu- mulation and initial income, is nearly 3 percent. To assess the contribution of the HPAEs' superior accumulation rel- ative to other groups, we can predict growth rates in four groups of 52 GROWTH. EQUITY. ANDND .=: '- i. ::, . :. 4.~~~'~ .I ., - : .fw C.gi7 -~i X .Jimut.- ; ,- -;,., ,.- .r., c>*d*.s_ ,^- - p- G- *rr -. * ...P -. 70. (-4q .0.7 j19J - 7 (.)3 die a(' 4NB -"I' '' a.77 -2- i) .L4 442. t 21) 0,0284: - '-;) QOb - 25 .S (69' 231; .%t Om72 :. - - . f X 0 ,- 1141 - W;t2 (tv 04 Om oil (231 .. . lwt u J 0624 - fi)ls. X MIS2s{ -.57 '.;4 1473 13z (36) IA9 (431 au Ax 1; - '(3m 638c 6.03 3.45;SC e 3.70 3310 151 -; Z -R - S ' ' , . .(6) IG - economics based on level of investment, primary and secondary school enrollment, relative income in 1960, anrd population growth. The differences in actual growth rates between the HPAEs and other groups can be divided into the predicted differences (due to group dif- ferences in accumulation or initial income) and to an unexplained residual (see table 1.10). Accumulation explains only some of the dif- ference in per capita income growth rates between the HPAEs and ocher groups. For example, between 1960 and 1985 average primary and secondary school enroUment rates and investment rates in the HPAEs were below those in the oECD, leading us to predict that growdt would be lower by 0.68 percent. Offsetting this is a conditional convergence gain of 1.2 percent and the contribution of the HPAEs' higher rate of population growth, -0.1 percent We woulld predict that the HPAEs would grow about .67 percent faster than the OECD, primarily due to their initially lower income level; in fact, they grew about 2.3 percent faster. The effect of the HPAS' higher rates of accumuladon is more telling in the comparisons benween East Asia and Latin America or Sub- Saharan Africa. Between the HPAEs and Latin America, 34 percent of the predicted difFerence in growth rates is due to higher investment levels and 38 percent to higher carollment rates. Far and away the major dif- ference in predicted growth rates btween HPABs and Sub-Saharan Africa derives from variations in primary school enrollment rates. Investment accounts for only about 20 percent of the diflerence and is offset by the oonditional convergence advantage of Sub-Saharan Africa and its more rapid population growth. Educaion is the main theme of the story of 53 ACLE Table 1.10 Accounting for Differences in Growth Due to Differences in Accumultion: The HPAEs, Latn America, Africa, and the OECD Economies : . .: : f . ; - f : f - u X : -. : : ;~~~~praw./..P --AI -- .- -. ..Pa . : - K r -e - aSt ;UA Afie.- O ; V Ab~~~- twst .lar 41. OEC :AAim-0D G4CPIZdISIW I CtLS..9l60 -013- 4tE574 0a7m7 : y IrInay.enmIm-.-. 1960 -:D612 . X4774 - - 774 - r,tuy - a9 uIbnnv.1910 -0- - . O. 0*3 .£. - - - - 2427 .-2 66 . . Aq.whFdpla. 156045 M.U095 .0*015 4003 -(. l *-- PAEn Dl. 39.--AS - Q25 *73 . .I3: :4o : iHr:E. am. . I.3 - b- mt ME *fa - : w -: et - . . 40 Irk dLmImtMIbimriu tjwdkm3cW.rhto ;-- T hkm6dundanJSwhA-. :. c ,: -DiPkmr ad posh . : M,mtilhedSehthcrCn.lnfoawsehnwairlEa Irk & m.uqe wdwin nrlm EbnkrX .wsqewmbldlucnlueln DNbuuhudaSdnb.wbmebn& - : the diferences in growth between Sub-Saharan Africa and the East Asian high performers. WhIiat is most striking, however, is how lirde we are able to account for differences in growth rates between the HPAEs and other economies on the basis of conventional economic variables.14 We are able, in the end, to prcdka only about 17 percent of the actual difference in growth rates between the HPAEs and Latin America. We do somewhat better be- tween Sub-Saharan Africa and the HPAEs, predicting about 36 percent of the difference. Controlling for their superior rates of accumulation, the HPAEs stil outperform while Sub-Saran Africa or Latin America - derperform the statistical relationship between accumulation and growth, leaving much of tie regional difference in per capita income growth unexplained (even though a large fiacton of HPAE success is ex- piained).15 They have been apparendy more successful in allocating the resources that they have accumulated to high-productivity activities and in adopting and mastering catch up rechnologies. Evidence from Estimates of TFP Gnxvb iTP change captures both of these important aspects of productivity growti. Te is estmated in a neoclassical fiamework by subtracting from output growNt the portion of growth due to capital accumulation, to human capital accumulation, and labor force growth. In keeping with this framework we estimate TFP change for eighty-seven economies 54 GROWTH, EQUITY. AND ECONONIIMC 9C1H . *~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ j .,. iup,2 pwe pfnwi UFAA. .ad = wu,f t- _ li2 . . ,s''0' Rel-" i - -"' '';' A.Mm ~~~~A4i.ORD fm. 0.17 (;0) 130 H rs ; - - - ( -1- 136) 0.2 , ,,;:.-. () 4b , - ., ,h M- . P11 (4). -013 CM! (11i2 .- L9 .< . 15i200 ; - 4 (I 3,0'.. ~~~~4.IM 2.32 using a produc-ion function esdimated with cross-economy data. (The techniques are discussed in appendix Li..) Figure 1.10 plots the result- ingr Tr growth rstimates for 196049 against relative income levels in 1960. There is great va-riance in the rates of n.' growth among low-in- come economies. A number of develo'ping economics show higher rates of TEP growth tha industrial economies, consistent with the possibility- of large catching-up gains. Many others, however, have low or even neg- ativi rates of productivity change. The East Asian economies stand out sharply, witha high absolure levels of TF. Five HPAEs-Hong Kong, Japan. Korea, ThAand, and Taiwan, China-are in the top decile. The other three HimaEs-Indonesia, Malaysia, and Singapore-aze closer to TF' growth rates in high-income economies (about 1.5 percet), but are still in the top third of alldeveloping economrues. Presumably, in the high-income economies most ofthe estimated'r growth is due to advances in best practice, which explains dthir relatively compact distribudton ofTFP hrates around 1.5 percent a year, and the tendency for IT? growth to dedine with rising income among the high-income coonomies (see figure 1.10). In low- and middle-income econozPrwst, howevr, changes in *rTFP must rcflect more than technical progress, othierwise we would never find negative TF growh ratev. We have already argued that TFP growtd for low- and middle-income -economlies conains an element of catching up to (or fing behinde) bT-pracice technologies. (The relationship between technical change and catching up is discussed in appendix 1.o1.) Bu TFP growth rates in an 55 Figure 1.10 Total Fadeor Pmodndivty Growth, 1960489, and GDP per Capita Relative to 11.S. GDP, 1960 1FP growth (percent par year) Taiwan. China * ~~*Hong Kong *Japan *Rep. of Korea Thai!rd 3 - 2 -- Indonsia *Malaysia U * *5 -2~~~ -2 0 0.2 0.4 0.6 0.8 1.0 1.2 Ratio of GDP to U.S. GDP, 1960 Source.World Bank dama. one-sector, cross-economy estimate will also contain an element of al- locative efficiency; economiecs char allocate physical and human capitcal to low-yielding investments will have low or negative estimat-ed TFP growth .rates. The dclarest -demonstrations of this ane the thirtee-n econ- omies with negative output growth. and positive accutmulation. Thus our estimates of TEP growth answer the following question: on the basis of the average efficiency with which physical and human capital are used in the world economy, does accumulation over- or underpredict income growth? The answer is that for most low- andt mididle-income economies it overpredicts, while for the HPAEs it underpredicts.' Under very restrictive assumnptions (see appendix 1. 1), we are able to offer some speculations concerning which of the H-PAEs we.re catching up 56 GROWTH. EQUITY. AND ECONOMICŽC to best practice between 1960 and 1989. To do this we assume that the elasticities of output, which should be used to calculate TIP change, are those that apply to high-income economies only. These more alloca- tively efficient economies have elasticities of output with respect to both physical and human capital that are higher than those for the whole cross-economy sample. We subtmct the average rate of TFP change for the high-incomne economies, which we have associated with movements in international best practice, from TFP change estimated using those higher elasticities to get an estimate of technical efficiency change. Using this metlhod, Hong Kong (2.0 percent); Japan (1.0 percent); Taiwan, China (0.8 percent); and Thailand (0.1 percent) are the only HPAEs catching up to international best practice. Korea (-0.2 percent) was es- sentially just keeping pace with technological progress in the high- income economies, while the investment-driven economies of Indonesia, Malaysia, and Thailand were falling behind international best practice ac rates of 1.2 percent (Indonesia) to 3.5 percent (Singa- pore) a year (see table A1.3 in appendix 1 .1). When the HPAEs are contasted with other developing regions, how- ever, their ability to keep pace with intemational best practice seznms somewhat more remarkable. Using the same method, we have estmated rhe average rate of technical efficiency change for Latin America (-1.4 percent) and Sub-Saharan Africa (-3.5 percent). Against these bench- marks, all the HPAES except Singapore stand up well in their ability to keep pace with the world's shifting technological frontier.7 Although the absolute magnitude of TFP in HPAEs is higher, so is growth. Is the proportion of growth due to TFP high? On the basis of an analysis ofsources of growth forasample of economies, Chenery (1986) devc-cped a typology of the contribution of TFP growih to total output growth by income. A typical low- and middle-income economy had a relatively small contribution of TFP growth to total output growth, be- tween 10 and 20 percent. High-income economies, conversely, derived about 30-50 percent of dteir total output growdt from TFP growdt. Figure 1.11 largely confirms this pattern for low-, midcle-, and high-income economies. Only seven of the fifty-nine non-HPAE low- and middle-income economies have contributions of TFP growth ex- ceeding 33 percent, while for the high-income economies it was much higher. The HPAEs fill into two distinct groups: investment-driven and productivity-driven economies. The investment-driven economies- Indonesia, Malaysia, and Singapore-conform to the developing-econ- 57 &EAAS4 ksA N M I RAC L E Figure 1.11 Total Factor Producti Growth and Part of Growth Due to Gmwth of Factor Inputs, 196049 TFP Growth 1960-89 percent par yer) v , \ s s~~~~~~ 10% 4 3[ . 4. ' 8% 4. 4. 4. 4. '. 2966 ". . ': .4 .4 '4 4.npan *Ta"n.Ch1na - 0% .* Hong Ki 2 . 4. . . 64 4.an D4 .4 - 4. .4 . - 0 4 - .4. '' 4 *. j Highincomeecon . om. * HPAEs % | Othiordevelopingecornomles ^ * \ '2%' .4 . .4 1E 4 -4~~~~~~4 r _ ~ N .4 . 4.__ ' .4 -2 4 2 4 8 TOtal bector hipLrt growth (percent per year) Note: Dashcd line:s tcprcscnc: toul 2vcngF GDP growch raus, 19650 89. Soli rays reprewnsi Ehe contribution Eo toclJ growth by TFP growth. S#tirees.Nchru and Dh2rcshw:ar 11993): World B3ankd2m. omy pattemnwith alowTFP contriburion. Conversely,the productvity- driven economies-Japan, Korea, Hong Kong, Tliailand, and Taiwan, China-are more unusual and look- more like industrial economies, widi a highi cont-ribution of TFP (above 33 percent). China's recent growth (see box 1.3) shows some of these characterisEics. The EIPAEs are unusual arrong developing cconlomies because of the relatively important rolc of TFP growth. Even so, we find, consistent =: ~~~~~~~~~with our previous findings with cross-economy regressions, that be- tween 60 and 90 percent of rtheir output growth ctcrivres from accumu- lation of physical and humana capital. Producdviry change has been higher than in other cleveloping economics and is importat to the East Asian success stor. But it is nor the dom-inant factor. ''- s8 GROWTH, EQUITY, AND O ,~~~~~~7 -: - . -..- -1 EN.FECNMI(.;OWFI.'EAST..o; ln im.est:- an-astoundiiien3-... G..P.. d a ;:.tc~ ~~~~CtW Ati:wud-btcmle dout.'i cosdeaontf,: 1 men smc 17.ase ii. rkia improvement: in 'S;! I;anoa45!"26 ;--tic t '@-' C6ns1ump'non.ldf-I on..of.- m ,111a i -. ;. r -. -, - - i o 7 ] ' =;^ i-- - ;!. ...- ,jp, - oa~v>s Y: S . - -. - -- --4. B-soxn Tnhe Chnese em e.- M ac :- A su'' i y -?-ea;.e; .. -aea. .n-ca...r y .& o"' ow- ..u b NO--. ASSESSMENT OFapn=en tECns-Ofi GNOMiC Odm sIlas T nvest aatu n 30erearonspnit op. g~~~~~~~~~~~~~~~1gtCnM-;.. -, ,P;A,E_ ,/~ _..;ie-.t ze. ^ ,5m ;9 1978:t.v :. v_of e19i8_8 t- Z' t.o. - - 4 - 7 1 - 1 4-49 f4..r.; . !- .9.s .tame~~nt fsh pronne: fyn te.- ;...s wld .a>mos', o - tuc,:2E, -rpDi ,: gr.d.m7 t -~~~~~ _ . , = -r { to t; 'v --- '- .. - a >; -> T - x - i .. - l - _ f 1, Asia.e wofd: b&&mpledy,tcu.wir-thourconsideadolsm.he oft met sinledo 17hasoa see lnlfunaln a akdimrvmnt in mwlgg -wcma.a-. gpal e -s gntao %AAuciMAdannbaawaL.I5> CWfiyP w; .i-heiAEwdbledYfroi 198 't? Ona - : ; , -j .4<. .4.;¢;>t;!-<,.+<-=.<-2. .&zt ,'4- . !, ' "-f ;!n ;i'9 ^W3- >-.- , : . , - 4' -s Jdp be-at termeo d an' eo p w .:erenr'n~lso0,o ie :onowicjmeŽifChina nuiFxceibk4f%lorX televi ksit ?r 9.leibusexios nS.-' i½Scr&Did1t0 ' -r.carmEa-relm- -'n .t .; eoetsn: : 1d16&arnvtlper. alddyinnually ill aulEzdiD S t- eds- B-, wmgatn,oterenmdyrai&y.avo&hlltaed .~R-<- ftrnEm l ba T in.ms. Uk8 J x:o5 in 1990.-uds,-, ass, -t zessment ph-:h^S ' &rf or iah&isi Asih'soCre H'Fi .ii growdi-n; ;'ba and evemen government-fbszeehxci& slXRoiinsgwing v-~~~~~~~~~~~~~~~frerLNqs g , ,-wzA;§ rou, {tsIea'te cottE St*n Ex tro ovr th e Ian.1,! . Bt'.*.urdiW4oremteonents d. ud tliey ti lled,c-;l -~ ton' 'health, and Cinacrnoeuhhiiiicsr.'bili'ty, .,China --- ~~~gecrber wdt 2 jrcenr ia ihcreincrTpjancs fropcrct}en t,;saviing ta armcsinvsted inXrw nsip-nd Af&0MF1982wjt 198~Fo bier-retreii.chmin il aeti~se.Czi tharS.- comc.cauamlgyihsEhEr Droblcmanc. rcM'4&44eii& 8iehi - mg 3)1125 h~iidM c suFpn4dealbi.tiowih ndshi anL p n n iY-.U:-.-e.,;nomy;Tas ,e5Lrv po*k-^nm~uvcs Tt- n --as ixbiJqjig rapid 4-breald&&ckfTt2r|n onpmentn,,T-i.rW s -: p"-'u '. tt -~ X-.-&WZ- i -ikA'r.-.--oer oftheecopnomies in-i nour stud_ - ..., - DeoDkallewedEom Dvmosmce-EF &n2 :S, jDiledyLbr.i of -Ci s 4- t_.h.._ .:y~~~~ ~~ ,sfxpd nu"C-S eee r s, A e&oiiozfies i aeynwichiuecnvinwg - high rates dof terms .nd~ reductions ininequali wererrm a cmpa- r ip our-p si un captala ind a W I -- r- od wheno %ro noive ofsms Ch-inniaksj flSnr -C$Qeiiid oAtes ";- iebhn i comesstI chea tbknaii& ghptr,wth ha mb e t44 efiuiiveN~rpsgy grwhanx-du&v aIst..aa'~~~~~~~~~~~~~~~~~~~~~5 .w r 4.- ~~~~~~~~~~~~~~p~~~~~~6~~~~--Ie ; t,s. r.. - . -. W -- ' We have described the East Asian economxies' success in achieving hligh rates of growth and redluctions in inequality- These were accompa- mied by-.agricultural transbrmnaions, rapid fertility dedinci, andmau factured. export: growth; and accounted for by rapid ac cumulation of physical'and human capital and by productivity growth. What policies lie behind this success? In the next chapter, we examine alternative cx-. 59 CLE planations of the relationship between public policies and economic success and particularly the ways that public policy can lead to improved economic outcomes. In subsequent chapters, we explore the relevant policies and their apparent links to the inputs and outcomes we have de- scribed in this chapter Appendxc 1.1: Accounting for Growth IN THIS APPENDIX WE BRIEFLY DESCIBE THE MET HODS USED TO estimate the relationship between accumulation and growth using both cross-economy and producton function methods. We also compare our results with other resulrs available. Cross-Econonm Regressions The cross-economy regressions employ the basic specification of other cross-economy growth smudies (Barro 1991; Dc Long and Sum- mers 1991; Dollar 1992): (1.1) GDPG = f(L ED, LFG, RGDP6O) where GDPG is the average rate of real per capita income growth using Heston-Summers' measures from 1960 to 1985; INV is the average share of investmenr in GDP for the period 1960-85; ED is a measure of educational ataiunment LEG is the rate of growth of the economically active population, and RGDP60 is the relative gap between per capita in- come in 1960 (at 1980 U.S. dollar prices) and U.S. per capita income in 1960.'; Mankiw, Romer, and Weil (1992) have recently demonstrated that this specification corresponds to the transitional dynamics of an en- hanced neodassical growth model with human capital. Table Al. below reports our basic results on the relationship be- tween accumulation and growth. The estmated equations compare fia- vorably with other studies using similar specifications. The overall fir of the regressions is good, and the coefficients of the variables are ofthe ex- pected sign and are significant at conventional levels (0.05 level).'9 We have also estimated the basic relationship fir two subperiods, 1960-70 and 1970-85. The fit of the regression is markedly better in the earlier 6o GROWTH. EQUITY, AND, Table A.1 Deterinants of Growth (dependent variable.-per capita CDP growtl) Smp-pe rio 196 -965 .19160-70 1.970-85 1960-5 -1960-85 .196-55 Nwnberofobsennrtianr 1413 .11.3 11.3 61 "98 5 Intercept-, -0.0070. 0.0064. -.16 -.04 004 0.0243* *(0.0079): (0.0092) :(0.0109), (0.0113) 0.8): (0.0094) GDP relative to U.S 1960 --0.0430" :-0.0444* -0.0422* 0.0408** -0.0292' 0.25 (0.018) 0(0.0103146 (0.0133) 016) * Drw.xynrolircnt, 1960 l 0.0264*" 0.0169' 0.034" .0.0247"* (0.0065) (0.076) - 0090 0.0082) ~Secondary enrollment, 1960 0.062 0.0192 0.09 .07 (0.0139) (0.01G2) (0.0.192) -.(0.0180) .Educational attimemnt, 1960 0.03 -0.0002 0000) 0.0013) Grot of population, 1960-85 0.1015, 013 0.2738. -0.0450 -0-3135' .-0.4290 - . - .: K ~~~~~(0.223) (0.2592) 7(0.3083) 0281 (.251 (O.64 Average InvestmendGD,. 1960-85' .00578* 0.1 153"' 0.0201 0.086" 1'.- . - - , ~~~~(0.0224) .(0.0260) (0.0309) (0.0245)- Avrge'qupmentinvestment * GD?, .1960-85 -0.3050" 0.3100" (0.0721) ~ (0.0743) Adjusted R2 * ~~~0.3480 0.3424 0. 1921 0.36 0, 1893 0.2614 Safl1ilM sgtfintt the 0.01 II~ Staisdeally iignicattth 0.5 lve * Ne Coffiien istop number mn each ceIL.Smiinkdanirrris in parahee SOUIrcZWorld Bank staffestiniares. two periods, but in most respects t-hc' parameter estimates have similar values and significance. The major exception is the coefficient estimate on investment, which becomes insignificant in the period 1970-85. Table AM. also reporcs on two e%frcs to examine the impact of changing the basic specification. DeLong and Sununers (1991, 1993) have argued that equipment investment, rather than 'total investment, is a superior explanatory variable for per capita income growth- When the investment share in GDP is replaced with the share of equipment invest- ment, the explanatory power of the regression is improved, and the co-. efficient on the share of equipmrent investment is highly significant. Our sample size is much reduced, however, and we condude that the reduc- tion ini residual variance is not sufficient to justifyr a change in our basic specification'. Because school enrollment may not be a good indicator of Gi ,* 1IRA C LE human capital accumulation, Barro and Lees (1993) measure of educa- tion stock, based on population censuses, can be used in an alternative specification. Deviations fiom the regression are estimares of labor productivity clange that cannot be attributed to accumulation-investment in phys- ical or hunan capiral or to the component of TFP change associated with relative income levels. The patterns of productivity change resulting from the basic specification of the cross-economy growth regressions can be seen in figure Al.1. The figure plots the partial scatter of the relationship between growth of output per capita and investment, controlling for human capital and the component ofTFP change related to relative back- wardness. Per capita income is an increasing function of the share of in- vestment. Observations are plotted relative to tle estimated regression line, and the HPAEs are identified The estimated productivity perfor- mance of the HPAEs is remarkably siillar to the pattern derived fiom the Fgm Al.l GDP Growth Rate and Average Investment Ouithagonal component of GDP, 1960-85 Y.06 ----- Hong Kong * - *Taiwan; China Rep. of Korea*: * *Japan 0.02 - - indon ia - ngpo . F i @ e ; , s*! __ e O002- --* - - -- I * , . .halwidi r ~~ 0 ''*' -0.04 _ . - .. H* HPAEs * * OUuerd.velopIugeconomies .0.06. -15 -10 -5 0 5 i0 i5 20 Orthaganal component of I/GDP, 1960-85 Note:The Xand Ycomponents rc orriagouai on RGDPBO, PRIM60, and GPOPGO85, wherc RGDP = eal GDP pcr c2pim relativec o US. GDc per capita, 1960; PRIM6o =primy school enl]ncart am, 1960; G0PP6085 = grwh mre ofpopulatian, 1960-85. SourccWodd Bank dat. 6z GROWTH, EQUITY, AND EGo TFP estimates. All seven of the economies are positive outliers. A dummy variable taking on the value of I for HPAEs is positive and significant. How We Estimate TFP Change Suppose that every economy has access to an international cross- economy production function of the form: (1.2) Q =AF(K,E,L) where A is total fictor productvity, Kis a measure of capital services, Eis a measure of human capita endowments, and L is a measure of labor services in natural units Then growth of output per head can be represented as: (1.3) q--a + s k-l) + s (e-l) where lowercase letters indicate rates of change and sA and sE are the dlas- ticities of output with respect to physical and human capital. The sham- weighted growth rates of plhysical and human capiml per head give the contribution ofaccumulation to growth in output per worker. TEP change can be then found as the residual of gowth of output per worker after de- ducting the contributions of human and physical capitl accumulaton: (1.4) a = (q-1) - s, (k-l) - SE (e-0. Under assumptions of competitive factor markets and constant returns to scale, s,K and SE are equal to the income shares of factors. Thus most empirical applications of equation 1.4 estimate the output elasticiy co- efficients with income shares.20 Since income share data are not available for most economies in our sample, we instead estimate s5 and sE direcdy using a simple, cross- economy production fnmction. The data for this analysis indude a new constant price capital stock data set (Nehru and Dhareshwar 1993). Measures of human capital are incorporated in the specification using Barro and Lee's (1993) measure of educational atainment. We regress annual log output growth on log capital growth, log human capital growth, and log labor growth during the 1960 to 1990 period, constraining their coefficients to sum to unity (that is, specilig the production function to be Cobb-Douglas). We also indude cconomy- 63 - RLIRACLE specific dummy variables to estimate individual rates of TFP change for each of the sample's 87 economies. Table A1.2 reports the production function parameters and the esti- mated TFP growth rates. One striking finding of our analysis is rhe low elasticity of output with respect to capital from the wholc cross- economy sample. This is not altogether surprising. As we pointed our in chapter 1, there is a subset of developing economies that has positive net investment and human capital growth rates but negative output growth rates. In effect, the marginal product of both physical and human capi- tal was negative in these economies, reducing the elasticity of output of both physical and human capital in the production function based on the whole sample. We have reestimated the production function on the basis of only high-income-economy input-output relationships. These re- sults and the estimated TFP growth rares are displayed in table A1.2. The elasticity of output with respect to capital in the high-income-economy production function rises to more conventional levels. TFP estimates, particularly for those economies with rapidly growing capital stocks, are Table A1.2 Elasticity of Output with Respect to Capital ISKI, Labor (SL), and Human Capital (SH): Full Sample and High-Income Economies Obsations SK rtaO SL (t-)m . SH (f-stat Fullsample 2,093 0.178 10.895 0.669 6.411 0.154 1.49 High-income economics 460: 0.399 10.237 0.332 1.679 0269 1.476 Ruling toasaifacuerprodsadvi4ygtou*h &ximafi r t HPAEs TFPrwsA TFPwtb1 ; - - - - 0 ; ; - 0t;t(filsa5Jmpk, - . (ldg-im v:j -Eonomy/region. parpameer ates) .te Hong Kong 3.6470 2.4113 Indoncsia 1.2543 -0.7953 Japan 3.4776 1.4274 Kore,.Rep.of 3.1021 0.2355 M2;ysia - 1.0755 -1.3369 sing2pore1 1.1911 -3.0112 Taiwa,China - 3.7604 1.2829 - -ThaiLnd 2.4960- 0.5466 ~ + - ~ 0 LatinAmerica ~ ; 0 0.1274; -0.9819 - -SSub-Sahv fric' -0.9978 - --3.0140 a. Alsoincludes Tunisia and South AfNic:- Sor- World Binkda:L--- 64 GROWTH. EQUITY. AND ECONOlT correspondingly reduced. The pattern of productivity change, however, with the five productivity-driven H'AEs fialling as outliers in the distrib- ution of developing economics is maintained. (See figure A1.2.) It is possible to compare the consistency of the TFP estimates pre- sented above with other independent estimates of TFP growth in a large sample of economies to test their robustness. Using a data base derived from World Bank real product data and alternative estimates of con- stant price capital stock, Elias (1990) calculates economywide TFP growth rates for a subser of 73 economies using standard growth ac- counting assumptions. The relarionship benveen his results and those presented above are summarized in figure A1.3, which plots the scatter of Elias's estimate of average TIP growth during 1950-87 against our estimates. Fischer (1993) has also presented a set ofTFP growth rates for a large sample of economies. A similar plot of his results against ours is shown in figure A1.4. Figure A112 Per Capita Output Growth and Capital Stock Growth, 196089 Per capita output growth (percent) Taiwan, China *Rep. of Korea__---i; *S.ong Kong *Singapore i @ Thailand *Japan _ - 4 _ , -------- - *MalaWsia-----' * Indonesia * 2 - '-__ _ 2 -- ~ ~ - 4 i t ' ; - - - - -- _ _ ----Capita-growth pL - _ _ _ 9--- ~ --Fl sa.k 0D et .06@(2* rwh is 0- MS 6 I U N4igh-Income economies Se S *NPA~~~~~~~~~~~~~ tes eelpn economies 0 510 1520 capita growth (percent) --High-Income economies GDP grOwth = 0.010 + (0.395*K growth) - - - Full sanuple; GOP growth =0.0076.+ (0.23*K growth) Soure.rWod d Bank data 65 MA.-IvIR AC LE Figure 41.3 Compadson of Total Factor ProductiNt Cmowth Estmates TFP growth camparator (percent) (Ella) 4. 3 0* 2- ~~~~~~~~~~~~~~Singapore -Taiwan. Ohina* 0-- a Hwighicomneeconomies. -2 -----A * Other developing econiomies -3 -2 4. 0 ± 2 3 4 TFP growth, 1960-99 (percent) Snuru:v Elias (1 990a); World Bank data Both samples show sim-ilarly great variance in rates of TrP growth among low-income economies, while the high-income economies are rather closely distributed around a mean ThFP growth rare of about 1.5 percent a year. There is also great consistency between the relative rank- ing of TEL? growth rates for the HPAES. The same productivity~-driven economies-Japan, Korea, Thailand, and Taiwan, China (Hong Kong is a bsent from Fischees TF' estimrnaes)-are in the upper quintile of both distributions, with the except:ion of Thaiand, which under the Fischer estimate is still in the upper half of the distribution but is not in the upper half under the Elias esrimara- Technical Prmgress and Technical Efficiency Change As we pointed out in the conduding section of chapter 1, it is difficult to disentangle the relative contributions of allocative mistakes, technical progres, arnd technological catch-up to rm' growth, especially in a cross- economy, one-sector setidng. One way of separating the concepts of tech- 66 GROWT1, EQUITY. AND E£CCN Figure AlA Comparison of Total Factor Productivity Growth Estimaites lVP growth comparator (percent) (Fischer) 3 --__ ---- - O- ~ ~ ~ ~ ~ ~~ ~ ~ ~ ~ ~~~~~~~~~~~ Hlg-,h4ncldoesau econoiesn -3-- -2 ---O * I~~~~~~~~~~~~~~~~~~~~~~~~~~~awn Ch_ ,:in /~~~~~~~~~~~~~~~~~~~~~ W het devlfoping economies t.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Rp of Korea 4 -3 -2 -1. 0 1 2 3 4 TiFP growth. 1960-89 (percent) Sources: Fischer ( 199,3); World Bank data. nlological prgress and changing efficiency in the use of technology istO model formally the relationship between observed output and best- prIactice technology; We address the relationship between accumnulation, productivity change, and growth using a simple neoclassical model. Following Nishimizu and Page (1982), we define an internationlally accessible best-pralctice production function of the form (1.5) Qf(W-=F7(t#]t where Qf(t) is porential OUtpUt at best practice, and Z(t) is a vector of inl- puts in natural units at time t. We assume that the frmncrionx F(.) satisfies the usulal neodlassical properties and that an appropriate aggregate index of OUtpUt exists. The best-practice function defines the "stare of the art" in the sense that fulrther increases in output at given levels of inputs can.not be achieved without the introduction of new techniques. Firms can move along pract fnctin, incrain the use esult of accu- -67 MTRA CL E mulation of inputs. The introduction and dissemination of new tech- niques move the best-practice frontier and is tchnological progress as defined by Solow (1956). Observed performance in a sample of economies or firms reveals that few are at best practice.21 Rather, most -lie below the production frontier due to use of dominated techniques or to inefficient use of best-practice tcchniques. Observed output Qfit) For a vector of inputs Z(:) can be expressed as (1.6) Q4) = Q %ft)eu (t) - ON, t) e u where U) is the level of technical efficiency [0 c eIt() = Q(t)lQf(t) < 1] corresponding to observed output Q(t). The derivative in logarithms of equation 1.6 with respect to time yields: (1.7) Q (tOIQ( = (t)IZ(t) + Ft + u- ()(t) where F and Eare the output elasticities of F(Z(t);t) with respect to in- puts Z4-) and timnc e and dotted variables indicate time derivatives. Output changes in cquation 1.7 are decomposed into three main elements. The first one gives output changes due to input changes, weighted by the elasticity of output with respect to each input. This is the component of growdt due to accumulation. The second ele- ment is the rate of technological progress of the best-practice frontier, and the last element, u-et), is technical efficiency change during period t. While i; is always non-negative, u *t) can be either positive or negative. We define the rate of total flacor productivity change as the variation Figure A1.5 The Growth in output nor explained by input changes. Thus for any observation, i: of Inpuis and Outputs Output-labor (1.8) TFPJ(t)= F+ u -(t). A- is the sum of technological progress, measured at the frontier, and the change in efficiency observed at the individual level. fh These concepts are represented in figure Al.' 22 We assume con- stant returns to scale in capital and labor. The international best- :: f- - qo fflz _ practice production fitnction, fJ, relates output per worker to capital -0 S - r ; - k (induding human capital) input per worker. Economies that are tech- capitalout-u nically inefficient operate along functional relationships such as f in 68 GROWTH, EQUITY, AND. EC figure A1.5.23 Catch-up can be achieved by moving from a point such as A to D, combining accumulation with a movement toward best practice. T!-,s reinterpretation of TFP diange is useful in understanding the sources of rapid catch-up in technologically backward economies. In- dustrial economies, which employ international best practice, are lim- ited to rates of TFP change determined by the rate of technological progress, zvi(t) = 0. Economies that do not employ best practioe can have TFP growth rates exceeding the rate of technological progress if technical -efficiency change is positive, icit) > 0. It is also possible for TFP change to be negative, if technical efficiency change is negative and greater in absolute value than technological progress. A rapid shift from average practice to best practice-positive technical efficiency change-can pro- vide a powerful engine of growth that is recorded as high rates of (TFP) change. There are two possible approaches to the estmation of the model outlinecd above. One wvould be to estimate the best-practice production function aiLd derive estimates of both the rare of technological change and the rate oftechnical efficiency changc34 This approach is notsuited to the cross-economy data available, and we do not employ it. An alter- native is to measure TFP change direcdy by growth accounting. Since measured TFP change consists of both technological progress and tech- nical efficiency change, we impose the strong assumption that techno- Table A1.3 Technical Efficienq logical change, the movement of best praaice, is constant and does not Change Estimates for the HPAEs vary across economies.25 Under this assumption all of the variance in I enbikid rates of TFP change derives from variance in the rate of technical effi- effdciency ciency change. w --60-89 lkEcoJomylrqov/g t-.1960L9 ' We have argued, however, that allocative mistakes were partly re- HongKong - : 9714 sponsible for the low output elasticities in the cross-economy produc- Indoneia -1.2352' - ton function estimates. Hence the high estimated TrrP growth rates of Japan -.9876. those economies with rnid factor accumulation may represent alloca- Ko Rep. of -0.2044 CM a~Mlaysia 1L7767- Eive efficiencv more than either technical progress and technical effi- - -3;4510 ciency change. To address this we use the parameter estimates of the K Tanvan, China 0.8431- high-income economy production function on the grounds that these Thaiaind 0-OI67-i-t economies are the most allocatively efficient. We then sub tract from the Latin Ameca- --i :17 estimated TPP growth rate the average TFP growth for the high-income SubSdiaAfi 34539, -j economies, our estmate of technical change.: The residual estimate of 1 a.AI inudesi,Ttnisiada tedical efficiency change is presented for the iihPP and other regional .. . - grouping in table Al .3. - - -: 0 ::~ ~ ~ ~ ~ ~~~~~~~~~~~~~6 MTRA CL E Appendix 1.2: 'What Do Tests of Cognitive Skills Show? Tests provide a convenient quantitative measure of diflFerences in per- formnance. By providing a consistent instrument, it is possible to com- pare the performance of students in different learning environments. The key is objective versus subjective assessmenr. It is important, how- ever, to recognize the weaknesses of standardized tests. First, rests evalu- ate only a limited range of skills, and there is no reason to be confident that the mastery of skills tested is highly correlated with the mastery of other skills that may also have an importnt influence on subsequent productivity in the labor market. Second, tcsting skills, not just cogni- tive skills, may influence performance; if there is significant "reaching to the test," both the predictive ability of test scores and levels of skill in areas nor tested may fail. Differences in the curriculum may also account for some of the observed differences in performance (see Hanushek and Sabor 1991). Tlhese criticisms suggest thart whenever possible, evidence firom a va- riety of rests should be reviewed before reaching a general conlusion. While internationally comparable test data are not abundant, they are, however, consistent with the results already noted. Table Al .4 shows the performance of thirteen-year-old students on a stndardized test of Table AI.4 Percentage of Children Perfoming at or above Each Level of the Mathemacs Scale, Age 13 Years two-step' Udrtns nria' - ::and simple ; ab/s : -o.c-ptP - . - Ar -*U ,,,,m/nvnc _ m&mac!;,t, |^s - ,'problems ! p0kn -Biish b . 100 ;a - - 69 - :- 24- - 6 2 ,Irelirdant, - -. - e -.98 8t . 55 14 . I ea;-.Rp- 100- --. . 95- 78 - . :4 5 Newf wkC lh 100 - .5 ; -65 18 1 N-w Bf` Ilik. (Frenc) ;100 95 58 12 -. J Onrado JE~igIish~ . 92 5 .16- 1 .99 tquebe(Fwnch) - . ~100. 7 3222 :::-87 55 .tg :-78 70. GROWTH. EQUITY, AND'E- mathematics administered by the Educational Testing Service in a num- ber of different economies, including Korea.26 Students from Korea ranked first in cach of the five categories of problems. The performance gap between Korean students and the others, all of whom were from high-income economies, was greater for higher-order skills. The International Association for the Evaluarion of Educational Achievement also administered mathematics tests to thirteen-year-olds in 1980-82 and included children from Hong Kong, Japan, and Thai- land among the children from twenty economics in the sample. Figure A1.6 summarizes the results for algebra, which are very similar to those for arithmetic, geometry, and measuremenL Japanese students rankled first, students fiom Hong Kong were in the top half of the distribution, and students fiom Thailand -were near the bottom, at about the same level of performance as students fiom Nigeria. Figure A1.6 Test Scores e:: Algebra Tests: Selected Economies Swazland Luxenboug _ Nigei _ ____ . - _______ ______ _ -. L Sweden _ Thailand _ Now Zealand - - England and Wales Canada (Ontado) _ _; United States Scotland Hong Kong I . j Finland Israel _. Canada (Br Columbia) . Belgium (French) i Hungary ____ Netherlands - i _ __ ____ * _*- _ _ __ _ Belghirn _ _ __ France Japan . _ 0 20 40 60 80 100 Mom score Souwe Hauvsheh and Sabor (1991). 7' - M 1 RACLE These results reinforce the conclusion that in some East Asian economies cognitive achievement levels exceed those in other develop- ing economies and match, or exceed, those in high-income economies. The results also suggest that just as Thailand has lagged the rest of East Asia witl respect to rates of enrollment in basic education, Thai chil- dren also lag with respect to levels of cognitive achievement. Figure AL7 #-Lii Coefficient and CW per Capita Growth Rate, 1965-70 G1nl coefficint 0.7 ;- * F ' * Gabon -16 . --- - ~~~~ i Peru L-eio 'Brazil *Colonmbia j Venezuei au Chile Malaysia Singapore 0.5 ,- -~- U~~~~~~~ -~~~s~Phiil;Dlr - ~ EZamba EHong Kang U Philoianes Sudan m N~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ThailandI t Sudanu E Argentina * 0.4A- -4--- *lnldia - *SrlLankau Olcoe- I * Pakistan ' X3---X--i TaIan, Chinau *Rek'rwa CaJapanU 0.2-1 - ---- -----'----- -… * . O , I I : ' ! ' :~~~~~~~~~* -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 GDP per capita growth rate SaurceWorld Bank darm. 72 GROWTH, EQUITY. AND _|O Fire Al.8 Gini Coefficient and GOP per Capita Gmwth Ralte, 197180 rlnl coefficlent 0.6 ----- - . -enya NCOI IPeru iBrazil EZambla Nepal CWIie MeVdcof enezuela i I *Malaysia I i r j EPhillippines * Singapore I E~~~~~~~~~~~Horg Kong OA t -- i- . . .: , ,_,,_hdiaE *ArgenU;tinda Indong 0.4j - ------Idnea * Rep. of Korea * Banfjadesh 1111halland *~ L.~ Taiwan, China . 02 -1 - -- -- -- |1apa - i 0.2-h . I 0--, , .0.04 -0.03 -0.02 -0. 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 GDP per capita growth rate Source World Bank dataL 73 MPI RACLE Figre A1D9 Gini Coefficient and GDP per Capita Gowth Rate, 1981-90 Gilnl coeficent XMexiao EChile Venezuela ! Argentina * alyi 0.4 . - Peru- - ;> _ _-ngapo re O l-~----------EPeru-- * Philippines - - Ehilane - M Hong Kong EThailand Rep. of Korea U 0.3 --r - _ - -- - I n-- ,donesiaE **Talwan.-China - I~~ ~ ~~~ ~ ~~~~~~~~~~~~ . .- . U.1~~~~~~~ I ~ ~~~~~~ . - '- 0 -0.02 -.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 CDP per capita growth rate Soun:Wo rid Bank data 74 GROWTH. EQUITY. AND EC Fgure Al.1G Cross-Eonomy Regression for Gender Gap in Prknary Enrollment Rates, 1965 and 1987 * Pakistan I 1987 120.61 I I 1965 987 di *Pahistan 6 ~~~~~~~~~~-~~~~~~I ~~~~~[19.21 E Indonesia >- I * [-5, I .- E maliand i 1nd %ralaysna I i [-10.t6] [5.1] N '- jRlof Korea A-ndnei t as21 Singaporeu; \ [8 a I - -~~~~ 14-31:3 . I ^ Hong kng * \ , X 7 [-2.2] 0 2,50 2,0070 2,750 0 o ° 500 1,000 1,S0 3,500 P^er capita Incomer (1958 U5S doflars) Note Figue in brackets show resiEdua Sowre: Behrnman and Schneder (1992). Figure M41.1 Cross-Economy Regression for Gender Gap in Seconldaiy Enroilment Rates, 1965 and 1987 . , , } 1 ~~~~~~ ~ ~ ~ ~ ~ ~~~~1965 I, 1 - 987 6, [7-7]- 4 * - Rep. of KoreaK, .a, i a *__Indonesia E37o.'j]§. . . di - .L[.1, *-61 J- - - *P'akistani ,, , -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~[22 0 250 Soo 750 :1,750 2,000 2,M5 0 SOD 1,000 1,500 30500 P.r cap6a Mcame 1h US. do laysa) Not- F4urcs in brucker show residual S 5ware. Bdhmnan and Schncidkr (1992). Figre At1 CosEnoIy RersinfD edrGpin eodr noletRts 95ad .~~~~~~~~~~~~~~~~~~~16 .98 * 70 77 E a IndonesaKo3g. 913 ~ ~ ~ ~ ~ ~ ~ ~ 29 PoakKontan E ~~~~'Thailand -J~~~~~~~~~~~~~ 18.2]I [-.2 Sn apore__ _ _ _ _ _ _ _ * Av~~~~~~~~~~~~26 N Maayi 0 250 500 750 1,750 2,000 2,250 0 500~ ~~~~~~ ~~~~~~~ 1,00 1,003,0 Per capita Inconie (In 1.998 U.S. dollars)Kore Nate Figures in brackets show residuals. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~3.4 Scarce Behaan and Schnidern(1992) 1[2.91 rig N~~~~~~~~7 S I R A~~~~~XRC L E Notes 1. As we pointed out in the Ovcrview, a strong argu- 6. This basic pattern is observed in fbur individual ment could bc made for induding China among thc "mir- economies-Korea, Malaysia, Singapore, and Thailand. ade econornies of East Asia. We lhave nor done so for The pattern fir Indonesia difers; real private investment three reasons: (a) China's recent very rapid growth has dedined continuously during the 1980s from a peakof2O been highly regionally concentrated; (b) the experience of percent ofCDP to a low of 13 percent in 1989. very rapid growth is more reccnt than in Japan and the Four Tigers and than thar of Malaysia and Thailand 7. Behrman and Schneider (1992). The regressions among the southeast Asian NIEs; and (c)-and most control for a polynnomial in average per capira income in important-the nature of the economic policy framevwork the relevanr year. The authors used per capita CNP at offi- in China is sufficiendy different from that of thc other cial cxchange rates as the measurc of income. eight economies, and is in such rapid flux, that it makes ar- tempts to understand the links bctwen public policy and 8. The residuals for Sri Lanka and Egypt are also above growch on a cross-economy basis even more problematic. the line in 1965. We look atsome aspects of China's rapid growth in the r- mair.der of the book but have not tEratcd it as systenat- 9. Despite increasing its secondary enrollment rate ally as the other economies. from 29 to 74 pcrcent, Hong Kong also fell wedl below the predicted level in 1987. This is becaLse, at $15,000, 2. High-income economies are defined according to the per capita income of Hong Kong was so high that the rclative incomie to the United States in 1960. They in- OECD economies were now its comparators. dude Australia, Austria, Belgium, Canada, Denmarkc, Finland, France, Germany, Iceland, Isrel, Italy, Luxem- 10. Birdsall and Sabot (1993b) cite Stevenson and bourg, Netherlands, Norway, New Zealand, Sweden, Stigler (1992), among others, who report results oftests in Switzerland, Trinidad and Tobago, United Kingdom, cities ofJapan, the United Stares, and Taiwan, China. Ap- and United States. Simple catching up by low-income pendix 1.2 reviews the results ofother tcsts comparing the economies, with growth rates exceeding those in indus- cognitive skills levels of East Asian chidren to other trial economies, is defined as 'unconditional conver- students. gence' in the new literature on growth dteory (see box 1.2) and would be expected to occur only under very re- 11. See appendix 1.1 for a formal definition of TPP stnricve assumprions, for example, equal savings rates and growth. technological levels. Below we exarmine the concept of "conditional convergence," catch-up controlling for dif- 12. This specification is the simplest of a group of ferences in variables contributing to growth. cross-economy regressions that have been used in recent years. These studies have also introduced other variables 3 This issue, termed "persistence," is explored by Easterly to address different questions-equipment invesrment, and others (1993), who find even more stiking evidence of De Long and Summers (1991); trade orientation, Dollar the uniqueness of persistent growdt among the HPAEs. (1992); endogenous investrment, Barro (1990). We have - not included these variables because they are not central to 4. Hong Kong P-:d Singapore, which almost entirely our theme. This approach, while generally associated with lack agricultural seaors, are exduded from this discussion. rests of endogenous growth theory is also consistent with neoclassical assumptions (Mankiw, Romer, and Weil 5. Staristical analysis indicates no significant (in the 1992).Arecenr conference, HowDo National PoliciesAf- statistical sense) difference of these eight economies firom fect Long-Run Growdt, summarizes the state of the art in other cconomies in their share of total investment, con- this field (Eastely and Schmidc-Hebbel, forthcoming). troling for relative income levels. 76 GROWTH, EQUITY. ANND-EC 13. We have also estimated the basic relationship for 18. The studies cited arc listed in note 13 above. We two subperiods, 1960-70 and 1970-85. The fit of the re- hlave not included these variables because they arc nor grsion is mnrkedly better in the earlier period, but in central to our theme. most respects the parameter estimates have similar values and significance. The major exception is the coefficient 19. The magnitude of the coefficient on investment is estimate on investment/GDP, which becomes insignificant lower than that reported in Dollar (1992) for his sanmpie of in the period 1970-85. 114 economies (0.113). The magnitude of the coefficient for education is abour that given in Dollar's results, and the 14. Easterly (1993) finds similar results fr the Four coefficient for RGDP60 is essentially similar to the results Tigers. obtained by Dollar and by De Long and Sumrmers (1991). 15. Use of a variable measuring equipment investment 20. This literature is briefly surveyed in Nishimizu and (DeLong and Summers 1991, 1993) enhances somewhat Page (1987). the explanatory power of the regression, though in a smaller samnple. An alternative education variable (Barro 21Farl(95)wsoefthfittoneted- san Lee.19) amasreo education stock, b . a ron vergence between observed behavior and best practice, and Lee 1993), mcasurc of education StOCK, based on whcheadtcnilicfcecyTrcisbnoa population censuses, fails to perforn better than the p- which he cllled technical inefficiency. There is by now a maty enrollment rare. large literature on technical inefficiency. Pack (1988) summarizes much of rhis literature as it applies to devel- 16. Figures A1.8 and AI.9 (in the appendixes to this oping countries. chapcr)r compare our <:simatcs of TFr vvidir twihTO. chapter compare our e srmaLes of . r growh I 22. Mankiw, Romer, and Weil (1992) also employ a other independently derived estimats for a large sample Solo of countries (Elias 1990; Fischer 1993). The pattem Of wtype neoclassical mode. Proponents of endoge pr.oductivity growth rates in figur 1.10 is remarkably ro- nous growth theorywoul l not accept the depiction of the - i , ,^ ^ , . ~~~~~~~production function vAr. Aminishing returns to cpita. busr to the specification of the growth accounting equa- tion and to the capital stock series used. IMF (1993), 23. A comprehensive review of the large literature on contrasting Asia with other devcloping economies, te measurement of best-practice or frontier producton reaches similar condusions, both with respect to the esci- fiunctions and their relationship to traditioral estimates of mated magnitudes of TFP change and the relative contri- the production funcdion is contained in Aigner and bution of TFP change to output growth. Thomas and Schmidt (1980). Wang (1993) and S. Edwards (1992) also reach broadly similar results concerning the pattern of productivity 24. This is the approach adopted by Nishinmizu and change in the HPAEs compared with other economnies. Page (1982) and recently applied to high-income economies by Fecher and Perelman (1992). 17. Young (1992) finds similarly disappointing results with respect to Singapore's Tr perfiormancc. Some cau- 25. This assumption, while strong, may not be fir fiom tion is needed in interpreting these results, however. Much the truth. Industrial sector estimates of TFP change in in- of Singpore's investmrent between 1960 and 1990 was in dustrialeconomiicsgenerallyyieldacompactdistributionof housing and social infrastructure, outputs of which are no- ras with a mean value near 1.5 percenr a year, both within ntoiously difficult to meas;:e. It is possible that we have: and across economnies. This may therefore be a good first therefore undervalued the rate of growth of output and approximation ofthe rate of technological change. hence the rate of rvp change. Similar derailed criticsms could be made for other economies, both HPAE and non- 26. The rests were administered to a random sample of HPAE, aLw our TFP results are rherefore best regarded (as in- 2,000 students in each county. The test questions were deed are the :ross-economy regression results) as indicatve drawn from the National Assessment of Educational of broad internatonal trends. (On problens of estimates Progress, a test designed for U.S. students. See Hanushek - of cross-economy regressions, see Sriniasan 1992). and Sabot (1991). 77 R.TE 2 Public Policy and Growth HAT ACCOUNTS FOR THE REMARKABLE record of rapid income growth and declin- ing inequality described in chapter 1? Some observers attribute success to geography and culture rather than economic or other policies. But geography and culture cannot explain it all; East Asia indudes Myanmar and the Philippines as well as the Republic of Korea and Malaysia. Nor is there a simple policy story. Of the successful East Asian economies, Hong Kong and Thailand have had relatively little activist intervention in economuic sectors by their governments, while Korea and Malaysia have had a lot. in this chaprer, after examining the significance of geography and culture, we suggest a simple framework for understanding growth in the eight HPAEs. This framework explicidtly allows for diffcrenc, bur po- tentially successful, mixes of policies across time and across economies. To set the stage for latef chapters, we then show how very different policies can be used, if circumstances are propitious, to address the three critical functions of economic management mentioned in chap- ter 1-accumulation of resources, efficient allocation of those re- sources, and productivity growth, that is, increasingly grearer output for given resources. How much of East Asia's success is due to geography, common cul- tural characteristics, and historical accidenit Certainly some-but def- initely not all. Ready access to common sea lanes and relative geographical proximity are the most obvious shared characreristics of the successful Asian economies. East Asian economies havc dearly ben- efited from the kind of informal economic linkages geographic prox- imity encourages, induding trade and investment flows. For example, 79 S_ MIRWACLE throughout Southeast Asia, ethnic Chinese drawing on a common cul- tural heritage have been active in trade and investment. Intraregional cvnomic relationships date back many centuries to China's relations with the kingdoms that became Cambodia, Japan, Korea, Laos, Myan- mar, and Viet Nam. In South and Soutleast Asia, Muslim traders sailed from India to Java, landing to trade at points in between, for several hundred years befre the arrival of European ships. Thus tribute missions and tradi- tional trade networks, reinforced in the nineteenth and twentieth cen- turies by surges of emigration, have fostered elements of a common trading culturc, including two lingua francas, Malay and Hokein Chi- nese, thar remain important in the region today. In our own cenrury, key Asian ports were integrated into the emerg- ing worid economic system as the result of European military and trade expansion. Cheap ocean transport and shared historical experiences fir- ther knit together a far-flung, culturally disparate region. U.S. assistance in rebuilding Japan after World War 1i, followed by massive U.S. eco- nomic assistance and military spending in the region rbroughout the cold war, also helped to set the stage for rapid gowth. Positdve impacts included enhanced security, which enabled governments not directly in- volved in the combat to focus on economic development Perhaps as im- portant, the U.S. military's exensive purchases in Asia provided a ready market for emerging export industries. Japanese industry received a sub- stantial boost from the provisioning of U.S. troops in Korea; just as some of Koreis biggest conglomerates got their start selling goods and services to the U.S. nilitarv during the war in Vietnam. Regional linkages fadilitated the adoption of imitative strategies, in both public and private sector activity. Policy imitation-specifically of Japans industrial strategy-was an explicit objective in Korea and Malaysia. Korea borrowed Japanese rechniques for building large trad- ing companies and directing the structure of industry; Malaysia fo- cused first on developing heavy industry and more recendy on building business-govcrnment relationships. Furthermore, the general model of Japanese success undoubtedly impressed policnfakers throughout East Asia, engendering a sense of confidence as well as providing models of potential instruments of gowth. Fmally, geographical proximity has facilitated capital flows, particu- larly in the past decade, as Northeast Asian manufacturers of labor- intensive exports moved their fiictories south to take advantage of lower 80 PUBLIC POLI wages. A surge of inves-tment (since the real appreciation of the Japan- ese yen following the Plaza Accord)-flowing first from Japan and later from Hong Kong, Korea, Singapore, and Taiwan, China-has con- tributed significantly to the dynamism of Indonesia, Malaysia, and Thailand. From the policy viewpoint, these linkages have been encout- aged by the generally liberal treatment of foreign investment But even where foreign investment policies have been restrictive, informal credit and information networks have helped investors to move capital rela- tively freely. In addition tO direct linkages and imitation, East Asian economies may have benefited from positive regional externalities. Through their earlier trade with Japan, Western importers had become farmiliar with Asian business, established networks for sourcing East Asian products, and gained respect for East Asian quality. Later, U.S. trade policies im- posed quantitative restrictions on Japanese products and created rare opportunities for other East Asian producers to enter international markets. Producers of garments, shoes, television sets, automobiles, and other products in Korea and Taiwan, China, took advantage of these episodes to establish lucrative market positions (Petri 1988). Of course, regional cl'2racteristics alone cannot account for East Asia's remarkable success. If geography, history, and culture were an ad- equate explanation, other economies would have little to learn from Asia's success stories. Fortuntately evidence suggests that this is not the case. Indeed, cconomies that are part of the same matrix of geography, culture, and history as the HPAEs but followed different economic policies-the Democratic People's Republic of Korea and the Philip- pines are two widely divergent examples-have yet to share in the East Asian miracle. We turn, then, to the heart of this book, an examination of the policies that have shaped East Asia's success. Policy Explanations TEMPTS TO IDENTIFY THE POLICIES THAT HAVE CREATED the East Asian economic miracle fall inco several broad cate- A gories. Below we present a summary of these, focusing on ihe differences in interpretation that have arisen among observers con- fronted with the same set of economic facs. -8i SvA C LSYXN MIRACLE The Neodassical View In the neoclassical interprectaion of East Asia's success, the market takes center stage in economic life and governments play a minor role. For ea- ample, Wolf (1988, p. 27) found it 'a striking fact that the few relatively successful developing [economies]-Hong Kong, Malaysia, Singaporc, the Republic of Korea and Taiwan, China-have greatly benefited from decisions and policies that limit government's role in economic decision making, and inscead allow markecs--notwithstanding their imperfec- tions and shortcomings-to exercise a decisive role in determining re- source allocation." Similarly, Chen (1979, pp. 183-84) argucd that in Japan and the Four Tigers (Hong Kong, Korea, Singapore, and Taiwan, China) state intervention was largely absent. "Wha. the state has pro- vided is simply a suitable environment fior the entrrc -ne.iurs to perform tieir functions." The neodassical interpretation of the axperiencc of die HPAEs, and e; pecially of Korea and Taiwan, China, presents a coherent and powerul view of one path to successful industrialization.' In this view, govern- ments in all of the HPABs provided a relatively scablic macrveconornic en- vironrment, characterized by limited inflatinn (except at times in Korea). Real effective exchange rates rarely appreciated, and such episodes were quickly corrected (S. Edwards 1988). Interludes of intensive import- substituting industrialization in Korea and Taiwan, China, were brief Manufacturers were thus able to concentrate on improving productivity perfbrmance rather than coping with mpidly changing relative prices of inputs and outputs. Although neoclassical interprettions of the success of Hong Kong, Japan, Korea, Singapore, and Taiwan, China, are ofren nuanced-and advocates offer them with varying levels of consumer warnings-they stress the benefits to the HPAEs of limited policy distortions in the for- eign trade regime and domestic factor markets. In the neodassical view, East Asia's export success is due to similar rewards for selling in the do- mestic and foreign markets. iBecause variation in incentives across sec- tors, as measured by the effc-tive rate of protection of value added, has been limited, inputs havr- flowed to sectors roughly on the basis of static comparative advantage, anti inrernational competition has provided the impetns for cost disciplne and technological upgrading. Traded inputs have been made available to exporters at international prices, and ae- porters have iced an n.ternational price regime in making their deci- -St PUBLtC P OTC P sions. Finally, factor markets have been roughly competitive, so positive real rates of interest have prevailed, and thcre has been an absence of du- ality in the wage structure by size of firm or sector of production. The Revisionist Vww In the past six years, the neoclassirl interpretation of the sources of rapid growth has been criticized for its lack of factual validity, at least as it applies to Japan, Kor=a, and Taiwan, China (Pack and Westphal 1986; Amsden 1989; Wade £990). Advocates of this new view, sometimes dubbed revisionist, have systematically documented that governments in these three economies cxtensively and selecivelypromored individual sectors. They have convincingly shown that levels of protection and the variation of protection across sectors has been greater than recognized in &> i rthe neoclassical interpretations.2 Indeed, governments in each of these three economics at times inter- vened forcefiilly in marker Korea, for example, strongly encouraged heavy and chemical industries by setting targets and offering a variety of financial incentives (Pack and Westphal 1986; Westphal, Rhee, and Pursdl 1988). Japan promoted the development of several weak indus- tries in the first fifteen years after World War II, offering protective mr- ifis and financial incentives to encourage the introduction of advanced technology and establishing rationalization cartels to &cilitate the exit of inefficient firms?3 Taiwan, China, used public investment in large-scale manufacturing enterprises to ensure inputs for predominantly small and medium-scale exportng industries. Moreover, capital markets were not free in these three economies. Proponents of the neodassical view have focused on the reduced inter- - vention, particularly in the 1960s, that allowed real interest rates to shift from negative to positive levels (McKinnon 1973). Buit while interven- nion declined, it nonetheless contnuedL In fact, Japan, Korea, and to a lesser extent Taiwan, China, did not rely solely on markets to allocate savings. Rather, they repressed interesr rates and directed credit in order to-guide investments. - The revisionist view, proposed by Amsden (1989), Wade (1989, 1990), and others, sees market failures as pervasive and a justification for - governments to lead the narket in crtidcal ways In this view, the experi-- ences of Japan, Korea,. and Taiwan, China, provide evidence that gov- ernments can foster growth by "goveriing miarker and "getting prices 83 ; * - : RA C ~~~~~L E wrong" and by systematically distorting incentives in order to accelerate catching up-that is, to fadlitate the establishment and growth of in- dustrial sectors chat would not have thrived under the workings of com- parative advantage. Amsden(1989, p. 14), for ex-ample, asserts that all "economic expansion depends on state intervention to create pricc dis- tortions that direct economic activity toward greater investment State intervention is necessay even in the most plausible cases of comparative -1 - -advantage, because the chief asset of backwardness-low wages-is counterbalanced by heavy liabilities." The Market-Fdendly View World Devlopment Report 1991 (World Bank 1991 b), in a compre- hensive attempt to describe the policies needed for rapid growth, falls in the middle ground between the neoclassical and revisionist views. It condudes that rapid growth is associated with effectdve but carcfully de- limited government aciivism. In the sMarrkea-tendly"suategy it artica- lates (see box 2.1), not only do governments "need to do less in those areas where markets work," namely the production sector, they also "need to do more in those areas where markets cannot be relied upon" (p. 9). The appropriate role of government in a market-friendly strategy is to ensure adequate investments in people, provision of a competive cdimate for enterprise, openness to international trade, and stable macroeconomic management. But beyond these roles, govemments are likely to do more harm hn good. On the basis of an exhaustive review of the experience of most de- -eloping economies, World Development Report 1991 condudes that in -gneral governments have been unsuccessful in improving economic performance through attempts to guide resource allocations by other than market mechanisms. Attempts to guide resource allocation in international trade, financial markets and labor markets have reduced competitive discipline, guided resources into low-productivity and in- ternationally uncompettive sector, and resulted in widespread rent- seeking. In short, though market failure is an importrant impediment to rapid growth, so is government failure-and governent failure can have high costs. A cental contribution of WorI Development Refort 1991 is the ar- -gument that sustined growth :esults fiom the positive interaction of four critcal aspects of econoomic policjr macroeconomic stability, 4.4 PUBLIC POLICY AND GROWTH Box -2.1 MakingtheMostofMarkets; IN-THE;AST;-TWEN YYEAS A CONSENSUS.HA -;trenologi c nged 'produ y. TeRepor- 'aenergkmdaiongeconomists on the best approac' to .stressd at omesticpolCy. ou d epeennepre- - economic dedoprnetn -. This zonsensus' was dp -.-...cissed' az. ngth:in WoMDev p'nt Rport 1991 -provide die infasuucure- and institutions'-at will The- TSChalnge oDv ne (WdrId Baiik.1? ' r 1199 Ib). ,-h' 'help he respond.i_ The Repor+t highlighted te iportance of a helty-' ' Oprness to ona tr invesren nd' privarc:sector, which. -rss m investments in -ides has been cra in encouraging omnstic po ,people,ta muchreduced role or.,goveiment, ope- du-r' toucutrcosts byw inrdi holopos-;:r -,nessroXd,'o eom ndvs w t rds ) Ce,oft and deveoping newand tcerpruct. By' esame: K' world;,and; macroeononmc-stability TIh~s' idcas -token,3protonSr erc ind usa ol -'''av c#lled into;whAt is now cae te-he mark' d6elom decades .He -gon np ' rv.m rriL- ' okvole. Injapan an.d Korea, foruinstace,gov~- 9' Y -;appro&; 0'- T eReporrcnotedtrhtare economicufronim. ement :agen and-indut asaton co' bo- t investng mi peoppl can be high, bur many govern-. rate to gahier and. disseminate i:normation on:.-J,. 'no..t- on] a r rd-'- tt and deop qu aity control 'Ibrp& t.. ' -ntarget thcir snbapkpmropacrod ecoanro,-in':',-Ale'rocr micc foundtionoi. on of the Brazil 2nd Pakistan,'f alone5 did' m,sost i ' OOpublic oosihagormenrs ca - derto; 'w hie sot l indicators;i Ce.and .rovide,- as o. Wt%an .unprovt. ~f th neotpule.ot jamnaica,howevr c to CarS inprov eVe. i mnt ocr th o times, ofKislogw.wth. .Governments must ais - . Mr overspens, terdui senoonowinaj '- st monetayryl opnin n fnniscordfiute,' .-prov- th 4ini fhes To, hled by inflation, choni i '''o% o pvisios, for -recurrent: pen ' are ina de- of ,thel y alssof o 'p ? i, quam te rEsuldjiawasrefilu undemrutli2ation> Thd economies with a:bs-orof macn&o'c u"r> Motsessfuuonlei nudhJai;Si-" iaiitprruesleionthaiidd ArgeCntn.'olivia < =ic o' m s~~~~~~~~~~~~~~~n -sf- .y>ir l - - ~~~iposr-,adoc~hv salihdter~o on tabedili ie d hnOh&nu~rj ItcP%%Y :avarae.a through rhen riors,- of~' ke,'..' doeia Koe,Maas; nTalidhiemn K L .markeslcqi .k%versely; retuctionson competition' gecL tod "keep. thefir rmacroononic:. polid&s on.' I (fo e5pe;Indsiadcensn o nappopriate. course adteir-.bioaUednoic jerfbr~ceha ihes ?P opeoi) .biddk'.b ' betarizlyo S > , ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~., human capital CtiaaCion, openness tO nternational trade, and an en'n- romnmenr that encourages private invrestment and comjpetition. Effective policy in one diension (such as human capital f hrmation) imprves. the results from effectve policies in others (such as openness or mac o economic stability)t In this view, the success of many economies in East Asia has been due tO teinforcing policy feedbacks. No single policy has ensured success; strong and effictive policies in all four critincal aEcas, and over a sustained period, have been hkey. - f -p . 'csumThe marke-ffiencdly. approach captures important aspects of Eastcr Asia's succe:m. These economies are macroeconomnicallystable, have vEry Asia-s suc=ss These .oniies are macroe-onomiay stabl, he v85, ... -, ,,. . X - .- ' ' ' 0' - :'~ ~ ~ ~ ~ ~~~8 high leves of human capital, are thorughly integrated into the world economy, and have very high levels of competition among firms. More- oe,East Asian success sometmsocurdin spite of rather than be- *,~~~~~~~~~~~vr tmc ocure cause of market interventions. Koreas heavy and chemical industries (Hcr) drive and Japan's computer chip push did not live up to expecta- tions. Even so, other interventions combined with export targets appar- ently were consistent wick rapid growth: quota-based protcaion of domestic industries in Japan and Korea; targeted industrial policies in-. dluding directed credit in Japan, Korea, Singapore, and Taiwan, China; heavy reliance on large state enterprises in Japan, Korea, and Taiwan, China; and so on. Furthermore, the successes of these three northeast- em economies compares favorably with the successes of Hong Kong, Malaysia, and more recently Indonesia and Thailand, where policy choices have been less interventionist. -NMultip Paths to Gmwth As these three views show, while it is easy to see how elements of East Asian experience contributed to rapid growth, it is nor easv to identifyr a single recipe followed across the region. An alternative approach is to recognize this diversity ofpolicies and to assess whether and how various mixes of policies contributed to successfil implementation of three cen- tral functions of economic management accumulation, allocation, and ---productivitygroth.4Thi4 s approach recognizes that policies, like tac- tics, can and should vary depending on the situation, while the central functions, which are crucial to development, musr always be addressed. One of the hallmarks ofeconomic policymaking in the HPAEs was the pragmatic -flexibility with which govenments tried -policy instruments in pursuit of economic objectives. Instruments that worked were re- tained. Instruments that failed or impeded other policy objectives were abandoned. Thus, the Korean and Malaysian governments reduced in- tensive promotion of heavy and chemical indusries when the fiscal costs and strains on the financial system thrdatened macroeconomic stability. Indonesia abandoned capital marker controls ancd attempts to control interest rates when they werc found to be ineffeccive. Japan shifted from higly selective interventions to promote exports in the 1950s and 1960s to industrial policies designed tO reduce trade friction in the. 1980s (see box 2A later in the chapter). Taiwan, China, gradualli priva- tizel key public sector investments-made to support the rapid growth of 86- PUBLIC POLI exports by small and mediutm-size firms. Singapore shifted from a policy of repressing wages to one of promoting mpid wage growth, finally con- cluding by coordinating the increase in wages with rising productivity. And Malaysia shifted from policies that resultcd in business-government conflict to highly structured mechanisms to encourage business- government cooperation. Chapter 3 discusses the pragmatic flexibility of HPAE governments in pursuing two goals, macroeconomic stabiliry and growth of manufactured exports. Obviously, our three functions of growth-oriented economic man- agemnent were not objectives in themselves for policymakers in the rapidly growing EastAsian economies. Instead, they are simplya frame- work for analyzing how the varying mix of policies and instruments chosen by HPAE governments contributed to or detracted fiom rapid growdt. At the same time, we do not want to imply that growth was due to mere luck in the HPAEs. Though policies varied over time and place, the formulation and implementation of policies were purposeful, with an overriding objective of 'shared growth." (This tcrm is described and discussed in chapter 4.) In contrast to the diverse, changing, but purposeful mix ofpolicies in the HPAEs, ocher economies have either rigidly adhered to policies that were demonstrably unsuccessfil-fbr example, extended periods of im- port substitution in Latin America and Sub-Saharn Africa, policies to develop domestic technology in India, and state-led industrialization in Easteem Europe-or have lacked a purposeful direction, as the history of adjustment pograoms in some economies in other regions attests. In- deed, as we illustrate in the next chapter, the flexibility and pragmatism ofpolicyrnaking in most of the HPAEs reflected governments' willingness to try new approaches to achieve particular purposes, discard unsuccess- fuil ones, and keep those perceived as successful. Th Functional Growth Frameworlk F _ IGURE 2.1 PORTRAYS THE FUNCTIONAL APPROACH TO UNDER- standing growth in the HPAEs The figure shows, in four columnns, the interaction among two sets of policy choices (fiudamentals and selective interventions); two methods of competiive discpline (market and contest based); the three central fimctions of econonmic 87> Figure 2±1 A Functional Apprach to G :owh Comnpetitive Policy choices discipline Growth functions Outcomes Fundanmentabe Accnnmulatn Rapid and sustained growth . Stable macroeconomy * Increasing human capital * Rapid growth of * High human capital *Hg,aigsexports -~~~~~~~~~~~~~~~~~~~ Hihsa.g : Effective and secure MdHgll Rapid demographic financial systems Metetransti Um.t-ig price distortions isin Export RFoapid agricultural Openness to foreign Rad competiin transformation * Hlgh human capital ~~~~~~~~~~~~~capital I ai ho S Openness to foreign It *technolrtY p Domesc . _ A Ficultuial depelopment I Contit ron . industalizadon - § ~~~pollicis v Alboatin : Directed credit credit-Reduceyfective use of } . ~~~~~~human capital in. : Selective Intprventions - labor rnart I rI\ * High reurts on I Export push _ Ilmestmrent < >Equal Ineeme I vz Financial repression Contest4edtu clstributbon * Directed cedit > *Export credit . Reduced poverty I * Seiectivem prormotion I * Lfwestmnent . Improving sorlal coordination Productty change inditrs * Information * PmduftJvpbased W InsUtUdons S exchange catching up o Technocratic Insulation R tec_ & 0- . _ * H gh-ualfty civil service +_______________________.______________ ____ - Monitoring I managemenrt and tie outcomes of growth and equity discussed in chap- ter 1. Institutions are also shown as critical to the sucossful definiton and implementation of policies and to supportng high levels of com- :pettive discipline. The solid lines in the figure show how policy choices contributed to outcomes via attainment of the three functions. Many policies, such as high human capital and openness, contributed simultaneouslyto two or three functions. For example, stable macroeconomic management con- tributed to vigorous accumulation through higher rates of investment and to improved allocation by reducing instability in relative prices. High and growing investmenEs in human capital contributed both to accumulation, since human capital is an essential input to economic grwth, and to productivity-based catchiing up by permiting better nmastery of technology. Effective and secure financial systems helped to increase the level of financial savings (accumulation) and channel it to high-productivity investments (allocation). Limited price distortions, by keeping domestic relative prices relatively dose to intrnational -pices, were good for alloaton andmay have provided incenives for 88- :~~~~~~~~~~~~~~~~~~~~~~~ : LI C7-: : firms to adopt technologies and innovate. Openness to foreign rechnol- ogy was a major vehide for productivity-based catching up, and agricul- tural sector policies that promoted rural development were cental to both growth and improved distribution ofincome. The arrows indicate that the system has numerous self-reinforcing feedbacks. For example, rapid growth and relatively equal income distri- butions contributed to the HPAEs' superior accumulation by increasing savings -rates and generating largr and more efective investments in human capital; they also contributed to superior allocation of human capita by fostering labor mobility and marker determinaton of wages. Similarly, they helped in the creation and protecton of instimtuions, par- ticularly the dcvi service, by reducing incentives for corruption. The six policies listed as fumdamentals are so defined in the sense that they affect the attainment of growti functions primarily through market- based mechanisms of competitive discipline. Three-macroeconomic smbifiry, effective financial. systems, and limited price distortions-assist markets. The two others-high investment in human capital and open- ness to foreign tecLnology-require efficient markets to operate. But the HPAEs also went beyond the fundamentals, intervening with varying degrees of intensity to alter market incentives. The broadest in- rervenrions were generous incentives for manuhactured cxports. (These are discussed in some detail in chapter 3.) As part of this export push, all the HPAES permitted exporters automatic access to imported intermedi- - are inputs ar intemational prices. Most also offered subsidized credit, often performance-based, for exporters. In Hong Kong and Singapore, Where international prices guided the domestic economy, massive pub- lic housing programs helped to keep down labor costs. In Japan, Korea, and Thailand, implicit or explicit export targets were used as the basis for awarding access to foreign exchange, investment licenses, or credit In Taiwan, China, large-scale public investments, induding in public enterprises, along with extensive support senices to small and medium- size exporters, were used to support the export drive. Most HPAEs also controlled interest rames on deposits and bank lend- -ing. These financially repressive policies enabled governments to ration credit; favoring some would-be bonowers over others. In addition to the importance of these directed-credit schemes to export promotion, some T governments also directed credit to other activities, induding agricul- ture, small and medinm-size enterprises, and, in Japan and Korea, large- scale imvestments. By intervening in credit makets, governments - f - - - - - - . ~~~~~~~~~~~~89 - selectively promoted industries, ownership groaps, and, in some in- stances, individual firms. (These issues are analyzed in detail in chapters 5 and 6.) Some HP'AE chose very narrow promotion iargets. selectively assisting knowledge- and capital-intensive industries and firms in hopes of upgrading the economjs overl level of industrial technology: In Japan andt Korea the governments at timnes concentrated on promoting private domestic investment in) heavy and c-hemical industries (Hcas). In Ta-iwan, China (and to some extent in Indonesia), the public sector in- vested directly in technologically sophisticated industries. Singapore and Thaiand have attempted to attract foreign investment in capita- and knowledge-intensive sectors. Malaysia initially undiertook public invest- menus in HCI firms but has recendly shifted to a strategy of promoting private investmnent more akin to that of japan and Korea. Some selective interventions went beyond helping markets perform better. Rather, theyr guided and in some cases even bypassed markets. In the following section of this chapter, we explain why, in light of market failures, some interventions make sense even within the neoclassical framework. We then describe how some of the interventions under- taken by HPAE governments addressed these marker fihilures, effectively (if not necessarily intentrionally) reducing the gap between social and privat-e gains. Market Failures: The Coordinatiom Prohiem A primary function of markets is coordination. The price systemn is a mechanism by w-hich the production decisions of the myriad firms that make up the economy are coordinated; for instance, they signal to inter- mediate goods producers what outputs are required by final goods pro- ducer. Whe marets arc incomplete or missing. they cannot efn this signaling function. Since the prices of outputs and inputs (including wages) depend critically on what- other firms are doing, there ar-e great potential benefits ofsharing information. Even in well-developed market economnies, information is conveyedi in many ways other than prices. Trade journals, trade association meehings, and information newsletters are al inportant minsitutions for the trsnsmission of informiation. Adam SmiteHs invisible hand paradigr argueso hat ach indcividual, in pussuing his or her self-inters also maidmizes the common welare; cooperation is therefore unne lssary- -In rafiuty modern economies are characterized by extensive cooperation. Among firms im economis that '90 PUBLIC POLI'CY 44$Dt½N` ':belong to the Organization for Economic Cooperation and Develop-- ment (OECD),for example, formal and informal sharing of information is common. Cooperation among firms is maintained whenever the gains from cooperating outweigh the gains fromi cheating. Coope-rative rela- tions are ensured because of benefits from maintaining a good reputa- tion. If a firm lidls to cooperate, its reputation will be damaged, and future business chat depends on its reputation will be lost.' Institutional. arrangements for cooper-ation and information ex- change in developing economics are weak-er than in industrial econo- minis, yet the needs for these forms of coordination are undoubtedly gre'ater. When economic change is.siow, making predictions about fu- ture prices is relatively easy: a simple projection of the recent past will do. But when economic change is rapid, predicting the future is much more diffi'cult. There is often need for coordination. beyond what mar- kers can provide; hence there may be a greater role for govermecnt to create institutions and facilitate coordiination. Missing Infonnatio and Credit Markets. Among the markets most af- fectedi by information problemns are capita markets. Even in industrial economies, equity markers generally do not finance much new invest- ment, largely because of information asymmetries (Stiglitz 1993a). In economies where'equity markers are weak- or absent, credit mecha- nisms become the primary vehicle for raising capital and. diversifyring and spreading risk- But credit markects-even those flee of interest rate controls-are often characterized by credit rationing. Credit is seldom allocated to the highest bidden. This is for an obvious reason: bidders * are bidding promnises; those who bid the most may not be able to fulfill teir promiss Lenders are concerned nor with the promised return but the actual return. As a result, capital is allocated by a screening and evaluation process that is quite different from the anonymnity that char- acterizes resource allocation in an idealized marker. In some oftdie IHPAEs, governiments intervened aggressively to address this problemn in the credit marker, going beyond the normal regulatory functions and prudential supervision that help ensure consumer confi- dence. For example, in Japan, Korea, and Taiwan, China, government helped establish andt aggresively supported development banks and other financial institutions,-and only later encouraged equity and bond markets. In Japan a-nd Korea, government set up and managed directed- credit programs,.and through explicit and implicit guarantees or other forms of interventions, reduced the risks borne by investors. A' Scale Econmnies, Externalities, and Ciooperation. Coordination may yield substantial benefits when large indivisibiliries in investment lead to economnies of scale. One such case arises when investments are inter- dependenc. For example, if a steel plant and a steel-uising industry are needed concurrently, it does nor pay to develop the plant unless there is a steel-using industry, and it does nor pay to develop the steel-using industry unless there is a plant. If each awaits the other, nothing hap- pens. Market failures due to incomplete markets, such as the absence of capital and risk markets, exacerbate thi~s situation. With such large- scale investments, no single entrepreneur could amass the capital required for both investments, and capital marker imperfections mean that investors hiave dlifficuilty obtaining the funds required. Moreover, there are likely to be large risks, and the market provides no mnecha- nisms by which these risks can be divested. Economists have traditionaly found exteralities a persuative argu- nent forsintervention in marker allocation. di My of these xtemalities are related to learninge spiovers associated with developing markets, spillovers associated wit discovering what goods can be produced, and spillovers assocated wevith the incomplete appropriability of tecnological knowledge. In each case, fims undertaking a new actvity face long-term, noncollateralizable ivestments that are difficult to finance. Coordination failures loom especialy large where "diFuse externalities exist and there are mutiple linkags between firms (Stiglitz s993a). It is in situ- ations where there are many participants tat markets are particularly use- -fi and where the absence of markets is particularly cosdty Examplies of hese diffuse e mrnalities indlue, for insance, dith development of hgh- technology services and nontraded intermediare goods. In the earlier developent liteature, the coordination problem aris- ing from interdcpendenc investments, economies Of scales and exteral- ities was giene a grea dal paromipnec tThis was the argument for planning in the 1950s and 1960s. Such planning ulled in part because it attempted to concentrate all rclevant infonadvon nm the goverumenit planning bureau. This was sinply inmpossible; infrroation is too dif fuse, too complex. Plnmning ministries were not adept at generating the infornation needed to support investmnt dceisions; as a rexsultemny planned investments filled. Moreover, because many of the planned in- vestments were in monopolistic public sector companies, severe prob- lenms of economIiic performance arose related to- corporate govemance and market powec: 92. PUBLI 0 L14 Some HPAEs tackled problems of ;cale and externalities by setting up mechanisms through which business and government could exchange information and coordinate investment decisions. Japan and Korea, and later Singapore and Malaysia, developed institutions and market rela- tionships that facilitated this process. In doing so, however, these gor- ernments did not attempt to supplant cxisting information networks. Rather, thcy built on the superior information that firms had as dcci- sionmakers Their approach to these problems, emphasizing informa- tion and cooperation through formal and informal interactions1 is discussed in chapter 4.5 Creating Contests Cooperation raises several problems, however. First, cooperative be- havior may become collusion, if firms act together to raise prices. Scc- ornd, cooperation may inhibit competition, leading to managerial slack or a more general loss of efficiency. Third, business-government cooper- ation may encourage firms to seek &vors from government. How did the East Asian economies that encouraged cooperation avoid these problems? They combined cooperative behavior-indcluding sharing of information among firms and between the private and public sectors, coordination of investment plans, and promotion of interde- pendent. investments-with competition by firms to meet well-defined Table2.1 Concentration RtoinManufacturing -conomic performance criteria They developed institutional structures Ratiosm in- - in which firms competed for valued economic prizes, such as access to - - credit, in some dimensions while actively cooperating in others; in 7%em0cewhflia-.1 short, they created contests (see box 2.2). Market-based compeiition J S i8 56Z'. K&R9:oi962A * and contest-based competition both included prohibition of monopo lies, although the number of competing firms was sometmes small .' Even though Japan and Korea have tended to have high levels of con- nI9 4; . 0; f centration in their manufacturing sector (see table 2.1), domestic com- KBraz4 1980 ; AtiZN petition has usually been vigorous. The Japanesc government has proceeded- on the assumption that competition among fewer, more '-- evenly matched firms is preferable to having one large firm competing s4 with many smaller rivals, a principle that is well-recognized in athletic . - -competitions (NalebuffadSii193.v Contest-based compeution included dear well-enforced rules, and & ~~~~~~ prizes for winners. Table 2.2 briefly describes some of the contests orga- nized by HPmE governments.6 The simplest and most widey used were 93 b . A R RA C L E -RC- j:~~~~~~~~~~~~~~~~~~~~~~~~~~~'t.- ,. ..W 'X47 '- . '. ' Y'Oi'3m2 Running a Contest, Rewais;,Ruies, n ?Rdees . JT-H-lE,ECONOTC CONASTS.IT GOVERNMENTS . Contests in NortheastAsian.economics have suc- :i.japan,.Korea, and Taiwan,.Chi'Ta,-ser up-amoi g ceeded pardy becauserthe prerquisits were right. *flins eo'combime die b&eherts orfcohpetidod and co- 1.:Preferential- access~ to .credit .and- Ibrdgn excchange: month , J - wri; sQi ian -d * . coh: l ies-. montns and7somchmes arrl&hi' base, however, these :-centere 'caconomicpertormance,primarilawull - ::Cj_ -!j;-- , - .>. '''.; .;--: . ... .............. . - .Y ..!.;,. e 5..,| 1_...- rcontestis,anddiir-role-mthe economy resmble chl7-" understoodd:mpcrauve to ,expor; Referees, he gov-- ren,sprc gaes - .''ernmenc officials-wh&hbave design'ed ~andsprie ti.- .O:ganiihg conrc s ts is a more t5m pccaned way to . ,ihetonresrs -have:been gFnerally comperciu and fir., run! ;a e:,oomv than,el:'i.i r:i Esez-fi*i; , j a r t . - W .' ."' . been - .L ,, nA rC economy nng.on ssz-talre, just as same time, ;parcipation has nortbeen ,; -b'f , °p s s ' - r r:,Complcate, anatoq. c at prce cd rrtosit out'a, contest g. ! Thul IC ffi.l1iezi dobeeni phdinr pAovA KaA~ieana-p ayingfid etunga-hcxiuo nave not punished igo.; In some in- tnAy-fied, L both 'vk. - .-srancessuchas Hondasshifr toHau- - -; l easc., i-fa ie -fi; . s;-l--vm *i*o -*e-toau ut7laeyp"' and-u&ec"oplay game,:e-. 6ioniobile' maniffctrang .at a1 time whenzfJapines :^- .: . Bu--a-wdll-run.contesr, luce a well-mrri pgarta ugb .p ; X i can:erite6er nibec&r ul&by prpVidhig i .-..i~ offiaiaIs consideredthe eautnomotiv fieldsatruraced th i X for cwii.c peduiioh~nL aidnducingpartipantsto cooper- !. . ffhtolerance ofz5nconfdrmss has d offandsomely 0 qX~~cop ta es .pnt - -icac or iLn :llE pal ; P.;- CYb - - l, m C ' -J . I .Cnet. won't w cvcryw er,.4 , c: - - mnzatr as.. onttr. steprpoi 6I lioihjmstances there arc .threc prerequisites for.. deraince ot:uuois m o:dvec6mga. success:~ rewards,'rules,p andr&rees.Rewairds mutts`tic be econiorreine'tat: .-p.of~ la' tck.a.e 0~ ~~~S . ; f -g ec _6 e reoo -st uak-clear orov r,,,veJ-.;l- -r- :- o-u-h/t i-. ; .eliit .broa' .partiipationan 'r.ues. and, retre m . M e in eA&rt&c .oCnbeutidon :Rulesr ebe dearcut so -; economies with-strong government instirutions,an- k .iW ,vih:fehvllb e , > ; 6 otestu 'e- ing contst be -.cd.es harde a, fs growand ther, : -wardedL-ahcL. whih puiserh'Fnuly cur"acy +;:z iy&.wardyAandL.who ch -om1 b~c- punssned*- ...L_ a powurncreascin icladon to. tat othe bureau y * . ay .w hsrneclespary gamnes-- .: As Japanese officials who ate nowjmoving4ay.fom., : .- ,knosi;cet,aieraare cmcal. 'conrescard a morefinuonal ow, :..cd & h :pa1ntnz r niz thed_`hi rel- -: ;pevs ghavdicoverd.runmng pakygaineTor". -- -peten'to!-W/nLZhaeF i dZ...-.&voin.dow d,ild. Uiildren- Si -9 ,lgdi.byrfessil'- ,C'r ;. -w _ . . . - scr rexport contests fior access to credit and sometimes foreign exchange. rum -t aWhile eaprt credt schemes varied widely, all made credit available to -: : . - ~~~~~irms with confirmed export orders. For many small and medium-scale p- ts n entrprises, export credits were their only access tO the fbral financialS : - .- sector (Levy 1993). Monitoringperfomiance was simple; firms were re-i- ;quired to show evidence of export orders to receive credi In Korea and Taiwan, China, export contests were more elaborate. Nor only did export projects er easier access to fids, but firms char ex-s ported successfilly had an easier time getting fiure loans. Because such credit was fungible and credit generally was tight, firms sometimes dil verted export financing to higher-yielding nonexport investments. its '.4- f . - - -; ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~P U B Li l-C. '2 exports became a performance criterion for projects that were not di- rectly aimed at serving the export market. In these cases monitoring problems were greater, since both export perfomnance atid the related projects needed to be monitored. A second type of contest used the power of government to grant li- censes. Table 2.2 shows two of these contests, based on Japanese experi- ence. In the first, the Bank of Japan created a contest in which commercial banks competed to provide financial services in line with government policy-for exampic, by opening rural branches-in hopes of winning highly sought-after urban branch licenses. The contest also gave banks a strong incentive to comply with technically nonbinding Table 2.2 Examples of Contests -It. w * d.t . . . . ... ... . . ..- _,p Voooehaaniernaef £ndn to; Fctdmi eX.einI ses Cana bob a,nno - ea ditM in. Noon ep. .r1 t- niaja lmeroroulbe.C). 1. Nomadidnbvn UCaMmnd MaIqt3 bonge po.lpmeoe u ofkrlnnL, : 7.1..,;:-:-. -Ctnin. lndrn±npons Cnrsqjdknm spnennd mqr an .tita, -vo' .4 unto &smk LC. C. *11.CUU~~Lordto iI1 mOtmuanmati deump kl RM11112NEW411W ~~~~~(siwtubmncnwCa)ocscaonI t ~~~,;1* - . <' lftvo.epena&f pn- - r r.h 7'_,>: 'JI, . ; ~ zb a 0 ~ 1 '' 9 * n-iai es i1 ' iR : - : ...::............ : - ', io an. .ed- . n, -; nt,htMepmpm- ( . o u .w *tep u *a-;k. * 6 F-v to.m , ; _. n 'mut : dexpnplocmfpeypm a- , ee.r ,e at ,inInn ,n hav ,pntoe pre edm hymn An ;seadeomnu (Ja gonuesu . -&hudue eem ankmpnda-I ,-ef , .C -: : cadSwt-e nepnder-idsbanwud.......................r-hsnnnccenahe ..... . ... p is etttitre'-'-\ '>biw .,boglCaeg .- ,s'. I, , '. , 7'-7'' " ' ' . '4 t ' 'i _ kfim Muhinewrder Pamnodom tow (194141 B onSk- DOz3 < b.;-.. : . . *, . ' Oneith obetves.deFl Tuyo;e.ayMemrnao e , -rr * .~~~~~~~~~~~~~i~f pei, l .;-t dkmi danh"okma cmin UPbnie n '~~~~~~~ ~~~~ ' -0pcmr'ti1567 :--- - : - . - . - Z . . . : ..6 .'' ' J - - , - 'M .7 ' - ' :,7: ~~~~~~~~~~TegIdaPund.,,'-: '..' - :C - , : .9 : C, .. .r . . -. . .g . - .j--g. : . -" t., ,- .. . . ..... yoteJvbgrnnueppn. . esnndafdSepne&u .¼ ,$ ,.-.-7. .c . . '.-''' AYf 4Z~- , CL{~, 7hr.r I' S a ~ Ct~. 9 ,e . - . -. -. : . ' .. .~ -,: . .. . , . . :- - '.: , :,: -;5 RA C LE 'Table 2.2 (contfinued) IN- Jt rP o u a u h d S i d h m :~ T iw h nk k u fj p uimppn wkteu m at h d m h u u m u lnabSInInxhlnhm.um udncmhhAnadit . -,two?" 44-. ~~~~~~~~~~~~ . t~~~~~iflncrwimm"AS1 ain ls kMhlngaahmla. Am* rudpmiadl bmulmyrnL * h~iwIUcDlvsnn andvbmmrad. Mise a-. h. zdlIc.M lie. *. -~~~~~~~~~~~~~~ko (1d ~ wte d,nknm o" A.).. -...~kiW*rnnd , (dr iime rd rJrmmck rW61 ~~~~Th ~~~~~~~~~~ I . .. I~ ~ ~ ~ ~ ~~~~d yKrjymn& initu cm mm, dwhmmsd&dtuW ~~ bmm.Kq~ .1 Damr mguhcflsp-dykdLn p"c apwtnruaiiitli uilidwaI ade .1 *lugemlmerpand ammafc SlpI.nsio.la ~ uhidgl*p*lluuca ~ Nlazm afirms SsAflCs. .,. dto mul. ~ ddr£rtmu aaa aS . lca ;'L, Iwam ais mas, huh alps . -'h Z- PC f'Wi Wdurn .. tmunymhiepiRaIMacMftWOI. 1 tapumumnnsuuchatmaecdlmsrniiza~~~rnaum.- . '"Mr .'Z J uilobmalbr mapnhdma.* 'V *1~~~~~~~~~~i (C- i Mm*mm:armshV,.a lafaryolavcn - 2 ItS nheppamatn4talmUd. tedon Amaik C ', '. - Nandi ladmim'strativ. guidoMancemf thcauirau S ulending banksotbar failneydIm oco pl it. cm dehpmim Cumitla mmihqClr3lesc-usuanmkn.y hev-n hniclidstis r h 90 n 19 -s admsinistrpatsiveguda of cpithei ledig:bak thatwas feqired to complye wirsthaguidaceiabou the spreatsdubetwee depositiand lni-rngrtes were- nrepermitteo t rexpandt thesepoir brnhnewrs (Yohin o19) Duringtherapidgrothperiod the acMtinsr ofd Ineratonlt Trad andonduisrysal (Mrr lieamnsedIt capacit Thpasio in sevral keytibindu-t deries-usueall heavyc andichemicald induestiesgI makt'hea195s ands 1960s, Dmflpimted entrydoandi mrkestrce compewtinga irmpworts. ofim in-the pro- tected ndustrywerereuireds tos cordnaeivsmnstopeeta pcsite expansion.1offupacit ibutby tyheeas equiedtn to ompete 796~ ~ ~ ~~~~Frs htrcie etaprft let euaio"(us-et)wr PUBLIC POLI 0: economies of scale or leaming permitted industries to cxpand export market sharcs. Our final example is that of the coordination of investments in Korea under the Ha drive. This is the most complex example we have of a con- test, one that combined licensing of capacity, access to credit, and pro- tection against competing imports with both physical and economic targets fbr performance. Its economic success is still not clearly estab- lished, and many observers have judged it a failure, but it illustrates very well the use of contests. At the heart of the HCI drive was the goal of changing industrial struc- ture toward more capital- and knowledge-intensive industries through coordinating public and private investments. Large private companies, the "chaebol," bid on individual, large-scale investments-fior example, electronics, shipbuilding, and machinery-for which thev '2ceived ac- dusive licenses combined with generous access to credit. The rewards consisted of quasi-rents due to import restrictions and restrictions on entry, access to credit in a highly credit-constrained system, and govern- ment support during business cycle downturns. These incentives have been the source of disastrous economic perfor- mance in other economuies. The check on poor perfornance in Korea was the result of rwo performance-based rules. Fi-st, government estab- Iished a timetable for the attainment of international competitiveness in each industrial sector Firms that failed to maintain the pace of cost re- cductions in line with the international norms faced both political and economic sanctions (Kim and Leipziger 1993). Second, because access to credit, even at early nonexporting stages of the development of HCI investments, was linked to the export performance of other products produced by the chaebol, efforts to maintain export Competitiveness across a wide range of products were encouraged. Mhe Rules of the Came: Exports as a Yanistici Contests are only as bene- ficial as their rules. If perfbrnance criteria channel effort into unpro- ductive activities, there will bt little benefit to the economy.8 If rewards are unrelated to effort, contests will not increase compe'tiion. There is one common thrrad among all of the contests outlined above. Exports, and especially manufactured, nontraditional exports, were the yard- stick ainst which the success of other allocation decisions-for exam- ple, credit allocation, domestic content requirements, and industrial licensinwere judged. Accordingly, there was a high degree of com- petiion among finns, in spite of the &fct that in certain domains and '97 rttC LVRcE at certain times they acted cooperatively. The export performance rule was broadly shared among HPAE governments and is almost unique to the East Asian economies. Chapter 3 illustrates how each of the HPAEs used shifting combinations of fundamental policies and selective inter- ventions to maintain the pro-export orientation. Box 2.3 describes how some HPAES used apparent constraints on their ability to export to create export contests to new markets. Using exports as a performance yardstick generated substantial eco- nomic benefits. A firm's success in the export market is a good indicator of economic efficiency-a much better indicator, in faict, than success in a domestic market. Export markets are likely to be much more compet- hiive than domestic markets. Even if the firm's success is based on find- ing a niche in a foreign market, its contribution to the domestic economy is still as large as if it had succeeded by developing new pro- duction processes or othenvise boosting efficiency9 ''-~~~~~~~* '~ '~t W;'X 5 l'i . - r ~~~~~M ~R zmo.g C}:*ahezspmrheJsccPIAaEasJencdtixar&icellzngSowoFuha*cm irk'~~~~~~~~~~~~k u -~~~~~~~~~~~~s 'n.amamiarllot YE;uats~amcRA orthers lP.asy¶V : ~~~~~~ ~ ~ ~ S " __tij.;!str* p8 p_no R X FrsRER>^*z.zbV.F ^gg;2ii5X;t> ~~~~~~~Isoii f , - - . Eiwa a~~~goyrmmen< tioo D or amodo Zsa&UO%%OOi5UDnaV3flCPeSw reihogn. .r emuncess,U .'-~~~~~~~ .'~4r~ -0 ,g p8~~~~~~~~~~~~~~~~~~~~~~~M - V §9 ,*>w~~~~M.14oo c uc nt areusofdlereamorun , . . gs,s BB,..N.X>.^*,>t S *.co - : - - - lR a*t; M S~~~ . . . t>;I:6Sy 3 ncollnpSaEpovtcotowDhsnRneurApmanceMDo6luonsqo a~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 9s- oXBg,.sp ttXs +;X:j 8 PUB LIC PO LI C.YAbrD4rO z There are, of course, other advantages associated with exports. In the process of entering international markets, firms learn a great deal and nor only about the particular markets they are entering. There are spill- overs related both to mnarketing know-how and to producrion know- how, topics to which we shall return in chapter 6. For instance, success in the production of intermediate goods requires producing to standards that are typically higher than those tat prevail within developing economies. The contacts made in the process of exports may also be of value when the firm decides to enter related markets. It will know to whom to turn to acquire advanced technology. While from a social perspective, success in exporting may be a better indicator of whether a firm merits additional fuinds than success in selling domestically, banks have (in the absence of govemrnment prodding) typi- cally preferred domestic lending to foreign lending and for a simple rea- son (Stiglitz 1993b). Banls are likely to be less informed concerning external markets than they are concerning internal markets, and from the banks perspective, there is greater risk associated with lending for export- oriented projects than with lending for products for the domestic market. HPAE governments were also less heavy-handed than some others that have attempted selective interventions. Though they made mistakes of judgment, they generally did not force decisions on others who were will- ing to risk their own capital (The creation of the Mitsubishi Automobile Company in) 965, rejectng m's gtidance ro refrain from entering the production of automobiles, is a case in point.) This is one ofthe strengths of decentralized decisionmalcing it provides insurance against mistaken views being given too much dominance. Running the Gamen The Role of Referees. Referees are also fundamental to contests. Someone must enforce the rules, administer the rewards, and discourage cheating. If one group of participants in a contest captures it and turns it to their advantage, they will reap rewards without the necessary effort. If some participants feel that others are excessively favored, they will refuse to abide by the rules. In the contests described abovei the referees varied. For exporc credit, the contest was largely self-enforcing. Banks had the incentive to moni- -tor export performance, since execution of the export order was essential to repayment Banking supervisors and customs authorities could also .-,use their coercive power to enforce the rules. In the case of banldng in Japan, the supervisory authorities of the Bank ofJapan were responsible for running the con Tlhe officials of mm were the arbiters of indus- .99 RAC LBE trial policy in Japan, and in Korea during the HCI drive both the political leadership and the economic administration monitored performance. The. Liits and Coustaints of Cootests What determines whether contests can be effiecively used to promore better allocation and fitster growth? There appear to be two important elements the relative benefits of coordinated behavior and the institu- tional costs of implementing nonmarket, contest-bared compeiitive discipline. Where the benefits of coordinated behavior are not great-for exam- ple, in small, highly open economics with good entrepreneurial skills and small nontraded goods sectors-the benefits of coordination of in- vestment decisions or sharing of information are likely to be small. International prices convey sufficient infiormation for markets to be efficient, and social and private returns will coincide- This may explain why HPAEs such as Hong Kong and Singapore have not found it worth- while to run contests and establish the types of highly structured information-sharing mechanisms (deliberative councils) found in Japan and Korea, despite the high quality of their bureaucracies. Even in cases where coordinadion benefits are potentially large-in nontraded goods, externalities, or interdependent investments- govermnents may not be able to enforce performance-based coordination rules. Taiwan, China, is an example. Despite benefits fiom coordinated behavior that were presumably potentially as large as in Korea, and the presence of a high-quality-civil service, contests of the types employed in Japan and Korea were not used. This may have been due to Taiwan, China's, different industrial structure. With a large number ofsmall and medium-scale firms, in contrast to Japans and Korea's relatively small number of major industrial groups, the institutional costs of attempting to coordinate all but the simplest activities were prohibitive. Instead, Taiwan, China, addressed the coordinaton problem through public in- vestment in large, indivisible industries. The absence of effective conrests-except for the export contest-r the Southeast Asian economies, Indonesia, Malaysia, and Thailand, can also be interpreted as a case of high institutional costs. Civil services in these economies are nor well insulated fiom political or economic inter- - ference It is not surprising, then, that effirts to coordinare economic ac- t-vity tirough selective interventions have been less sucssfL The *100- contests cithcr bave had poorly defined rules or have bcen bijacked by -their pardicipants. - - These nwo factors may also explain why the reliance on contest-based - :competition may change over time in the samne economy. Box 2.4 doc- urnents the changing nature of industrial potigr in Japan, from selective tO functional incentives and fiom extensive reliance on contests during the rapid growth period tO market-based competiuon in the 1990s. In -large measure this is due to perceptions among poliyrnakers that the large benefits from coordination characteristic of the rapid grwth pe.- 4waiJ.wISYt~OFuOsDEt etdf6i-r.iu --et"N'Vith. e eonos(ni ; . - ;v,==auSmat luc>S as. <,<-3 V ; -rt-4r4---R,-F,3 98,' ;;~- w-t C q i£M t- sovv - )sD~~~~~~~ ~~~~ inreqmnemure,nnn;otdaterneationalreemisvaiSSnieEt R ;uncustue,Xandh8D~silldadrmcdl m&m i;>snonaerDqns tua~an-dc Oro!E, B t n>- FSl.gr.-otXZ ^se-i>t>tzr~¢<<;Qv lRitFl-r|J-t -s LtR g2J;i-*; stX-lr -9 5fS=§tX 9 - f~~~~~~~~~~~1~l g>l~-leimR^ow fZpno;MTp ZEnSLEool>qcao^adeeomnhpnhat 3;is iR 9¢+j=,-t%rin- iC g>z. sF^ ,>:* .>+; i1L e t 5s- ,>:~ "rn|y;E>.fi#.ana z+mmal DoMan;.was~~~~~ ~~~~~~~ sQaxoamolaI4- mos. > $t05Ar '>4XAS-rW3^DRbRX 2*^t ttti >,== 84tSitt 8 -u^'§.^&a> t J-8 *iX45r£W .+5=i 9s.+N :-~ ~ ~~~~~~~~~~~~Pq . Drexnod7sw;:nEnUte,uedM ms;n. iotlkhEdasaloyEt - 'gS 95,>uW,-^Q* ............................... Rr".wrl,>,,oE.isi,4e,a.2*,sw044> i' -z-'tt.Z'^: [^sl^ttt>'}Y'-'^F4Ea Z,oAaodnra:$rDvdcpcyTa-ytt '-ru blcne le.d>ndu ngn l 5W.^grLtg'@eS=< @X<' ncg- 4cot.2-Xti.>C{vlsgt/4't. =.E$*-.w'xjl2>(g,.an >i : i Q. ut9. W a | = 4 t , A |R ,e ; 4 g ' %M l*jE;^;s > f ->4n4:,.,- Yu* Br n ea Ostbrecnsructo,lSycmDeesacm - coinasoofiie or d- ots : ~~~~~~~~~~~~~~~~~~~~~~io _ A MI~RACLE riod had diminished as- the economy grew and became more complex, and that the insttutional costs of administering contes had increased as power shifted from government admininstmraors to enterprises. Simnilar concerns appear to be driving recent changes in Korean industrial poli- ces (Km and Leipziger 1993). .;e -t i,_ xe_^,r,>,U.le.x-U.. The searc hfrpolicyexplanations forEast AsiA succss hasnotbeen completey successfulL Each of the broad views of the relationship be- tween public policy and rapid growth-neoassical, revisionist, and markcet-friendlyadds important elements to our undcrstandingl but none fuiily captures the compleity of public poiicy and rapid growth in the Hfl'AE. We have proposed using a fuinctional approach to under- standing the relationship between policies and gowth. In this vieew policies contribute to t oh atteauunt v of t hree centml Sanctions: aci- nulation, allocation, and produc growth-neocl. Markcr-orienited poi- cies, the fundamentas, were widely used by al the HPAEs and formed the basis for their rapidh growth. More interventonist policies have the potenti to contribute to growth in cases wohrec the address economic coordination problems, but to succeed they must combine the benefits of cooperation with competitive discipline by creating contests. Effective implementation of contests depended on two sets of&fictors - almost unique to the northern HPAEs. The first is the competence, insu- lation, and relative lack of corruptibility of the public administrations in Japan and Korea. The historical origins of these economies' high-quality economic administrations and the mcchanisms that they used to main- cain quality and limit political influence and corruption are the subject of chapter 4. The second is the pragmatism and flexibility of govern- ments in the high-perfonning East Asian economies where contests were tried and abandoned or continued dependig on their results. We .examine this flexibility in policy formulation in two importnt areas, macroeconomic management and promotion of export, in the next chapteL 102 PUBLIC. PO'LICyA i Notes 1. Forgoodexamples,seeBalam(1982) Bhgwvati (1978), 5. Popular discussions of the success ofJapan and sev- Krueger (1978), and Litle, Scitovsky, and ScoEr (1970). erl other economies of East Asia have stressed the coop- eranve relations between government and business, 2. Moreover, some of the measures presenord, such as between workers and cmployers, and among businesses. those of Itoh and Kiyono (1988) for Japan, are probably This phenomenon is sometimes rerred to as 'Japan, minimum estmates insofar as their calculations of efieclve Inc." or "Malaysia, Inc," conveying an impression of a rates of protection (Em's) are based on triffdata, not on es- single-minded direcrion to economic activity, promoting timates of nominal levels of protection derived from com- the collective inuerescs of the economy. Clearly, dtis pop- parisons ofdomestc and international prias. On Korea, see ular imge caggeractes. Like all governments, govern- PadcandWcstphal (1986) andtherefrencrscitedthercin. ments in EastAsia lack the control t cnforcea common On Taiwan, China, see Wadc (1990). On Japan, see Itoh set of goals. and Klyono (1988). Wade does not present or dce estimates of effecive rates of protection in Taiwan, China, for the pe- 6. The list is by no means exhaustive, nIor does it in- riod afer 1969, while the other two present evidence on dude examples of failed contests, of which there are ERPS forjapan and Korea Conversely, Wade does offer care- many. frfly documented qualittivc evidence ofthe extent ofselec- tive intervention. Despite the frequent use of Taiwan, 7. See, for example, Kim and Leipziger (1993) for a China, as a modd of limited intervention, therc are no esti- dicussion of the instiutional features of the Ha drive mates of which we are aware of effectrive rates of protection and an evaluation of the success of the program. or dome resource cost for the period after 1970. 8. Amsden (1989) in a seminal workl oudines the na- 3. See Yamawaki (1988) on the steel industry, Mutoh ture of perfiormance critera used in Japan and Korea to (1988) on cars, Yamnazawa (1988) on textiles, Yonezawa discipline interventionist policies. (1988) on shipbuilding, and Tanaka (1988) on alu- minum refining. 9. Again, this is perhaps not quite accurate; there may be spillovcrs from Iarning how to produce some com- 4. A formal exposition of this concept is contained in modity more cheaply. R%g and Peuri (1993). 103 M4acroeconomnic StabIlt an ort ACROECONOMIC STABILiTY AND RAPID Ex- port growth were two key elements in scarning the virmous cirdes of high rates of accumula- tion, ficent allocation, and strong p roductiv- ity growth thiat formed the basis for East Asia's suiccess. Governments achieved macroeco- nomic stability by adhering to orthodox policy prescriptions-in partic- ular, by holding budget deficits to levels that could -be prudently fin2nced. They achieved rapid export growth through an export push- a combination of orchodox, market-oriented mechanisms, miany linked to Macconomic stabilitye and omplementary pro-export incentives.' Policy initatives in these two areas shared two features that furtier illustrate the framework outlined in chapter 2: respect for certain findamentals-for example, fiscal pridence and avoidance of exchange rate overvaluation; and quick and flexible responses to changing eco- nomic circumstances. Policies to achieve macroeconomic stability and rapid export growth differed over time and across economlies, partly be- cause governments were responding to a changing economic environ- ment and paryld because they wcre seeing efficient policy packgens thdrugh trial and error. We define macroeconomic stability to mean that inflation was kept under control, inti eal and external debt remained manageable, and macroeconomic crists tat emerged were resolved quiddy, usuaiy within a year or two. Short recessions and policy adjustments to macro- economuc stress somtimces sq the private scector. But these weare transidonalperiods to new episodes of rapid growth, quite unlike the yeam of recession andi uncertainty that have plagued.many other devel- Opig coonomes, particularly in Sub-Saha ran Africa and Latin America. 1105 Export push is a more complex phenomenon than macoeconomic st- bility and the variety of policies used is greater. Mechanisms range from broad, export-friendly measures such as avoiding an appreciated exchange rate, which are esmployed in all the HPAEs, to government-mn export con- tests, which have been used primarilyin Japan, the Republic of Korea, and Taiwan, China. Universal export incentives, such as tax breaks and credir guarantees for all exporters, have been the main instruments in Hong Kong Indonesia, Malaysia. Singapore, and Thailand. Despite this diver- sity of approach, as a group the HPAES are unique among developing coun- tries in the attention they havc devoted to export promotion and the success they have achieved. The second half of this chapter reviews the evolution of export-push policies in each of the HPAEs. Pragmatic Orthodoxy in Macroeconomic Management ry"IHE SHPA WERE MORE SUCCESSFUL THAN MOST OTHER DE- veloping economies in keeping public deficits within the limits -. the economy could absorb; as a result, ,hey were better able to restrain inflation and manage both internal and external debL Low in- flation and manageable debt in turn facilitated realistic exchange rates and the avoidance of the appreciation that elsewhere undermined export performance. When the macroeconomy did go awry, usually due to ea- ternal shocks, govermments quiddy implernented orthodox solutions, reducing the fiscal deficit and, when necessary, devaluaring the currency- In contrast, many other developing economies have been less successful in keeping deficits within bounds and have therefore had more trouble managing inflation, debt, and exchange rates. As a result, policymakers in these econonmies have often had less room to maneuver when con- fionted with a macroeconomic shock; perhaps pardy because of this, their response has often been hesitant and ineffecrive The HPAEs macroeconomic management superior is reflected in less severe imbal- ances and generally lower variance in key indicators, including teal ex- change rates, ral interest rates, and inflation. To be sure, macroeconomic conditions varywidely among the HPAEs. Malaysia and Singapore have long-run inflation rates comparable to Switzerland7s, at below 4 percent, while inflation in Indonesia and Korea xo6 MACROEC'ON exceeds the South Asian average of around 8 percent. Singapore has con- sistendy avoided fiscal deficits, while Malaysia's fiscal deficit peaked at 18 percent of GDP in 1982. Hong Kong, Singapore, and Taiwan, China, -do not borrow abroad, while Korea was the worlds fourth biggest -debtor in 1980 and Indonesia's foreign debt tripled during the 1980s. Exchange rate regimes have varied from rigidly fixed to managed floats. Arnid the diversiy, however, are some common themes. While some governments have run substantial deficits, none has financed a deficit in a manner that destabilized the economy. The level of the deficit that is affordable, and hence nor destabilizing, is specific to each economy. It is generally larger the faster the rate of growth and the larger the pool of private savings (both at home and abroad) relative to private investment. In both these dimensions, the HPAEs have perforned better than many of their developing-economy counterparts during the past thirty years. Because of this, while some economies have had higher inflation than others, none has had to endure the very highi, debilitating inflation that has troubled other developing economies. The HPABs chose a variery of macoeconomic policy paths because of different economic conditions and preferences. All lay within the bounds of prudent stabiliy, and whenever the macroeconomy appeared to be in danger of moving out of control, swift action was taken to re- store stability. This was true even when the source of macroeconomic instability was policies intended to promote growth in the real economy. For example, the heavy and chemical industries drive in Korea, to which we shall return below, was modified when its adverse impact on the key indicators of macoeconomic stability, inflation, and the real exchange rate became excessive. In the following section we discuss how low deficits enabled the successful Asian economies to keep key macroeconomic indicators rela- invely stable, and we contrast their performance with the macro- economic instabiliy that plagues many other developing economies. We then rum briefly to four HPAE case smudies that illustte the variey of rapid and effective responses to macroeconomic shocks. Adhering to Macroonomic Fundameuias In contast with many other developing economies, where boom- and-bust cycles have cauised wild swings in macroeconomic indicators, the HPAES have been remarkably successul in creatng and sustaining 107 macroeconomic stabiliy. This has been a potent encouragement for pri- vate savings, investment, exports, and growth, since the private sector could count on relatively constant prices and interest rates. Here we consider the HPAE's successfiul management of four macroeconomic fin- darnentals budger deficits. inflation, external debt, and exchange rates. Keepi.g Budget Deficits Manageable. The HPAES' budget deficits are not dramatically better as a group than other developing economies'. But there are two distinctive things about the HPAEs. First, dtey almost al- ways kept the deficit within the limits that could be financed without macroeconomic destabilization. Second, these limits were higher than in other developing economics because of the beneficial feedback fiom other good policies. International experience suggests that the macoeconomic conse- quences of pubLic scaor deficits depend on how they are financed.2 Ex- cessive monetary financing of deficits leads to inflation; heavy govermnent domestic borrowing drives up interest rates and crowds out private borrwing; large external financing of the deficit leads to debt aises The -PAFEs kept each type of financing within bounds, avoiding the corresponding macroeconomic disease. The means for restraining deficit financing have varied widely. Some governments established institutional watchdogs, such as the currency board in Singapore. Others took rule-based appmaches. such as In- donesias balanced budget law and Thailand's exchange rate manage- ment framework, which undl the early 1980s resembled a gold standard. Sill others reed on the discretion of economic policymakers, as evidenced by the macroeconomic adjustment process in Korea and Malaysia. Table 3.1 shows consolidated public sector deficits for the 1980s fbr three HPAEs for which there are good data compared with a sample of OECD and developing econornies. As a percentage of GDP, Korea's bud- get deficits were below even the OECD average. This helps explain why Korea was able to keep inflation, extrnal borrowing and interest rates within bounds. Malaysia and Thailand are more complicated. Thai- lands budget deficits were about average for devdopi economies in the 1980s, while Malaysiais were substantally bigger than average. Both ran bigger budget deficits than such troubled economies as Argentina, Brazil, Mexico, and the Philippines. Unlike these and other economies that encountered difficulties, however, Malaysia and Thailand success- fully financed their deficits. This was possible for the foilowing reasons: io08 MACROE.ON-O, Table 3.1 Consolidated Public Sector Deficits, Selected East Asian and Other Economies -Averag pliRaikmo ng 4O . . def20tcii pcet 0>;0vj6wP 'age: dkevelpngncotrnries * Eir -.:- -gi ^. ofDP~ '980-88- ( ; .-D,a9r (I ehi 4stefx) .,~ ~ ~~~~O ., -. Korea,Rep.of 1^89 ; ;u0 034- r ;Mn.Iaysia .* 10980' 6 .$Thai -nd e ' 5,8023 Auemge~~~ 6.39, -via g. OECDeronomnies - : :. 2.82 -- Oderieconomi& ~~ Az~~cndna ~~9.62- .BIUI :4.02- Yhilippines ~~~~~z.4.30 £. Sewtcasticr1y Rdgc -and S i - Hcbbdl (lirrhcomi-g). :* Frst, diere was feedbackfrom higbgrowh. Since growth was higher in Malaysia and Thailand than in less successfil economies, a higher budget deficit could be financed. Because high growth in- creases the demand for financial assets, Malaysia and Thailand were able to absorb higher levels of monetary financing without a rapid rise in inflation. Moreover, their rapid GDP growth raised the level of sustainable domestic and external borrowin (an economy can borrow more for a given debt to GDP ratio when GDP is rising rapidly). In contrast, many Latin American economies fell into a vicious cycle of low growth and unsustainable deficis a Second, there was fredback from bigb financial savings. Savings rates were high in Malaysia and Thailand, and much of this sav- ings went into the domestic financial system (as opposed to real as- sets or capital flight as in Latin America) due to the pro-savings financial policies described in chapter 5. This firther increased the demand fir money and other domestic financial assets, making increased domestic financing of the deficit possible without re- sortng to inflationary financing. In Malaysia, the government's Provident Fund mobilized domesuc savings for the goverrment's use in noninflationary financing of the deficit 109 ~~~~~~~LE - - _~~~~~rIR C L E- * Third, there were low init debt ratios. In Thailand, the initial levet of external debt to GDP was very low, which meant that exter- nal financing was available when needed. Because of this, the HPAEs have avoided the inflation-inducing burscs of money aceation that afflict other developing economies. Figure 3.1 shows money creation as a ratio to GDP in Korea, Malaysia, and Thailand and in three unstable comparators-Argentina, Mexico, and Zairc. The contrast is striking while money creation has been relatively constant among the HPAEs, each of the comparators experienced two episodes of rapid money creation when fiscal balances deteriorated or external fi- nancing dried up. The impact on inflation has been equally dramatic. maintaining Moderate to Low Inflation. Unlike many Latin American economies, the HPAEs kept inflation from spinning out of control. Table 3.2 shows the low inflation in Malaysia, Singapore, Thailand, and Tai- wan, China. Indonesia and Korea have higher inflation but sdll far below Latin America. Jnternational experience suggests inflation below 20 percent, a level not breached by any of the FPAEs during their rapid growth periods, can be maintained for long periods without generating Table 32 Inflation Rates macroeconomic instability (Domrnbusch and Fischer 1993). Low infla- - - . tuon is a corollary of fiscal prudence: East Asian governments never had Aveiag;' K to rely heavily on the inflation t because teir deficits were within fi- * Eco nonre tt96iAOi nanceable limits. ,.- HPAt. i. . r . 75 . Several East Asian governments have made fornal commitments to Hong .ong1' 8.8 low inflation that constrained their options fr acivism. Historically, tjQA Korei~iRe~of .; 1^2.2 Athe most important of these was the commitment to a fixed exdhange &z^4J, l -I i , $3 rate; however, aU but Hong Kong have since abandoned this mech- nism. Other self-imposed constraints on fiscal policies and borrowing r c -; v induding balanced budget laws and various institutional checks, remain --.. 6 > :;. in place and have generaly contributed to fiscal discipline and low in- - M-% lie- ............. :. - . fladion. (We discuss rthese in chapter 4.) In general, HPAE govenments .A ~ ~ .; >^;have been strong enough to alter public spending and foreign borrow- 2005l|wzX, - ing as needed, although in Thailand this has been a continuous struggle * i s .>^7^ ^ . . (Warr and Nadhiprabha 1993). -1.1 Reasons fbr the commitment to low inflation vary widely but are - - -: rooted in the recent economic history of each of the HPAEs. In Indonesia and Taiwan, China (and to a lesser extent Korea), aversion to inflation MiK& ld^y^ l grew out-of traumatic inflationary spirals tiat accompanied economic and politcal crises. In Malaysia and Singapore, the success of colonial-era i __*>'1120Z: 0.$g11i;ES fiscal conservatsm seems to have helped shape postcolonial policies. In no. MACRQ ECON~ Figure 3.1 Revenues from Money Creation as a Percentage of GDP.- Examples from East Asia and Other Selected Economies (Percent) 4 12 - ~ I iMalayhia -a ;:.~~~~~~~~~~~~~I I: I . ~~~~~Thafland 10 IIa- Rep. of Komea - : -I 4 :: .. 1970 1972 1974 1976 1978 198 2982 1984 1986 ±988 12- - ZaireM Arge.tnna - 190 22 194 1976 1378f- 1980 198.2- M2fy1984 1986 N ie Reene iw moe c:ato Ji a pecng of G' -s de afined asroo oaIl 10 - - o t o - GD-- 6-uce World^ Ban dat a. j, s 8 2 fro de ansfrg incrase goenmn spending.# A | , -- -2_ _ __ _ _ --_ ; 1970 1972 1974 1976 1978 1980 1982 is"4 1986 Nofte. Rev:nucs from moncy crmcaon as a pucrcntagep of GDP is ds fi ned as ra2do of nominal O nge in hirgu-powerd mony mo nominal GDP. S iwcff Weridl Bankc damc : Thailand, the traclition of responsible fiscal policies dates frm t:he niine- : teenth century, whien a strosng currency helped thffe kingdom retain ifts in- v dependence. In Hong Kong, coloniial rWfie has insffJated the government ; fr om demands for-increased govemmrent spending.- - . 0 One result.of low tO moderate i&fltion'rates pardicula-rly welco'me tO : business is stable real interest rates. Figure 3.2 shows real interest rates in V .- . -: - : . - - - . . - - ~~~~~~~~~~~III ::- ;4S 'MIRACLE; Korea, Malaysia, and Thailand, compared with Argentina, Ghana, and Mexico. As with money creation, the contrast is rcmarktble. In the East Asian cases, low inflation and flexible financial policies kept real interest rates within a narrow range. For the comparators, the combination of nominal interest rate controls with high and unstable inflation was deadly. wild gyrations in real interst rates created severe uncertainty for investorsm Figure 3.2 Real Intemst Rates Examples from East Asia and Other Selected Economies Real Interest rate (percent) 20 - ~ ~ ~ I -70 J j Rep.ofKorea 1 I C * i I - Malandu: SO-80._R.~~~ . ,~~~1~ Malaysia- 1970 1972 1974 197E 1978 1980 296 1984 2988 20- 7 - 40 -30 o4 -A-. -70 "- ' Ghana -80 j !- '- - +_ -1 Argentina X -- . ' -. 400 i7[7 _] _= !_ _ 400 19-70 1972 1974 1976 197 2S80 1 i982 1984 1986 Nte. Real intercst rates are defined as die deposit rate deflated by the consumer puce index. &-wuree: World Bank data. . 11-2., --. MACROECQ -M Keeptmg Extenal Debt under ControL Of the seven developing HPAES, only Indonesia, Korea, Malaysia, and Thailand have public or publicly guar- anteed foreign debt. The governments of the others-Hong Kong, Sin- gapore, and Taiwan, China-have not borrowed abroad. None of the four with foreign debt has faced a aisis, in the sense of having to reschedule debt, but sharp incrases in debt have led to rapid adjust- ment. In some economies during some periods-for example, Korea in .1980-85, Malaysia in 198248, and Indonesia since 1987--debt-GNP ratios have been quite high compared with other indebted economies (see table 3.3). As with fiscal deficits, however, fivorable feedback from other policies enabled the HPAE debtors to sustain higher external debt to GDP han other economies. High levels of exports meant that foreign exchange was readily available to service the foreign debt. Similarly, high growth implied that retrs on borrowed capital were sufficient to pay the interest. Korea's successful handling of a very high foreign debt illustrates these trends. Beginning in the early 1970s, Korea borrowed heavily to fi- nance private sector investment and build up foreign exchange reserves. By 1984 Korea's foreign debt was fourth largest in the world; by 1985 it equaled more than half its GNP. Yet because of its high export-GNP rauio and rapid overall growth, Korea neverlost creditwordliness. From 1986 the govermment pursued an actve debr-reducton policr, drawing on Table 3.3 Intional lndebtedness >Akioftvnal 4g-." ;t -~S ' Ra,. ofor deb to ,xen .- WStFrnom'rgo Pea pt 199 d Pk U' -:9 HPE POW -- - x0 hdonesia ¢qv478r 690 s66A -455 2 ~~MaI~jysxa ~865 47 1384 54 47"- - 390 177 T98- *, t 'A Dc';-. A..4. -,'."- ' * '''0 '' 0h te;nme - - 38 -1762II * ,: - joCr'.; * -`3408' ; ; ;f- ;-t' r4+44;-- 7:2 - 4: r. . 026B0 * ~~-.-is~,t- -'!r-sA .. -., tx "3 AaNt 1V RAC burgeoning international reserves generated by exports to make pay- ments ahead of schedule; by 1990 the debrtCNP ratio was down-to 14 percent (In contrast, when Mexico faced scvere problems with its cred- itors in 1982, it had a much /owerdebt to GNP ratio than Korea in 1984 but a much higher debt to export ratio.) Keeping the Exchange Rate in Une. The HPAEs avoided the severe appreci- ation that beset Sub-Saharan Africa and Latin America (see cable 3.4). In contrast to such economies as Bolivia and Ghana, the East Asian economies did not cling to a given nominal exchange rare (or inade- quare rare of nominal depreciation) in the face of continuing inflation but depreciated when necessary, someumes quite sharply. Fiscal pru- dence prevented the cxcessive demand pressures that appreciated the real exchange rate in such economies as C6te d'Ivoire and Nigeria. The evolution of exchange rate regimes in the HPAES has been broadly similar. Hong Kong, Malaysia, and Singapore pegged their curencies to the British pound during the Breon Woods period, then floated them in 1973 or 1974. The Taiwan, China, doUlar was pegged to the U.S. dol- Table 3.4 Average Appriation Index, 1976-85 .- I ;t : ndx Pe;rcent -.-aamgerankr>: . . f . . f(highier vau means. ( . O: si.an10 most.- .Ecoonomy .. mare app recr ap.-p. .e at0a&O .r.s . : H Kong - * *Indonesia. 98 25 Ko reo, Rep. of 1 41 Malaysia ~~~88 1 Singapore- 87; I- Oce - . ;;Slidiend e-co "nomiei-s ,- "' ''' 0 ' ' ' ' ' ~~ ~~~~~~0; ,- Bolivia 4' 11 , '- ,, ... .8 . - D- . 90 4a * ..- -- .2- J--"; MACRORC larfrom 1960 to 1973, then appredated twice, and was floated in 1979. Thailand had the longest fixed rate regimc the baht was fixed to the dol- lar during 1954-84, with a single small devaluation in 1981. In 1984 the baht was devalued and floated as part of an adjustment program. In- donesia fixed the rupiahto the U.S. dollar from 1971 to 1978. Because inflation was higher in Indonesia than the United Staes, the rupiah ap- predated vis-a-vis the dollar, necessitating mnajor devaluations in 1978 (5 1 percent), 1983 (38 percent), and 1986 (45 percent), after which In- donesia shifted to a managed floar. Korea, too, tried to tie its currency to i-the dollar but resorted to four major devaluations between 1961 and the start of a managed flnat in 1980. Most moved from long-term fixed rate regimes, to fixed-but- : adjustable rate regimes with occasional steep devaluations, to managed .* -: floating rate regimes. Hong Kong, the single exception, reintroduced a fixed linked) rate regime in 1983 in the face of fierce speculation against the currency prompted by political uncertainty. Under the man- aged floating regines that began in the early 1980s, policymakers no longrer set rates but attempt to influence them at the margin, generally to move in parallel ro the U.S. dollar Because the United States has been the region's major export market, maintaining a stable and at times slighdy undervalued exchange rate vis-a-vis the U.S. dollar has assisted exporters. The HPAES' success at maintaining stablc exchange rates is apparent in figure 3.3, which contrasts the remarkable stability of real exchange rates since 1970 in Korea, Malaysia, and Thailand with rhe severe exhnge rate instability in Argentina, Peru, and Sri Lanka. Argentina repeatedly attempted to use the exchange rate as a nominal anchor against high in- flation (fr example, in 1973-74 and 1980-81). But failure to keep other macroeconomic fundamentals in line led to the collapse of the real exchange rate and sharp real devaluations. In contrast, the East Asian economies' pragmatic macroecononic management eambled them to avoid swings of the real exchange rate, even in the face of major external shocks (see box 3.1). Responding Quickly to Macroeconomic Shocks --The HPAEs rapid response to macroeconomic shocks has been greatly . dflitated by two characreristics First, by limiting distortions and tighdty supervsing banks, governments reduced the spillover from the "5 Fiupre 33 Exmples of Real Exchange Rate variabiryy in East Asia and Other Selected Economies Real exchange rate (percent) 300 1 Rep oKorea- 25D 0- 200 | SD- __ | I__ I_ I___ __I., - __ I I ; I' : I i . 0 1970 1972 1974 1976 1978 I980 1982 1984 1986 1988 300-- --Arg ntarM 300- ------- - --e-i.o mo 1 t _ 7 X A Tr. . Z. I MO - I 100 - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ I 10 -5- 1 1 i ! i - I o I : I , I ! i t , . , i 1970 1972 1974 1976 1978 1980 2982 1984 1986 i988 I.. Inder of ral xchange rce: I980=100 cal depreciaton is down. Sot t:World Bankdam. real sector into the financial sector that in other economies exacerbated fiscal woes. Second, flexdble labor and capital markets enabled the rcal sector to react quiddy to government initiatives, seting off new growth cycles tha eased the recessionary impaa of stabilization measures. Thus the HPAEs recovered quickly from rnacroeconomnic shocks. Box 3.2 de- scribes two more typical developing economies, Cc-e d'Ivoire and Mex- Wia MACRAC ROECNO: gik"I!xL~R -I U J5tc zur &q -, i" '-#- 7 c ' ?YvsK:r~~ dafiii g'% " k c yx;rvn m; rm'n * dXc~~~~~6aecord4icsi.&acrh. aia hcdo . .~~~~~~~~~~~~~~A x1B uq`c1ie.aRAsianiconj o-n' bbiisl'btt` ~i~ * :t-1~~~~~~~~ 10 t 1" "'4t.>-iJ 0 r*T. 'a,d wini-omm j t, - * -;-- tgl g$> >0 b *t> t-;gTl, * -F' .; K>e,;efiile(nu t io*eeldvomgm Uroole'Eithti CZ>- - b&AkfrAhe cnsxilngwod4sing -lci that fMaulet taocltaket'imEely action in rs,'l;lponsewr; to uhshcsanc-v ,^f- thinaUe.Weinri or is a mn n boxa3& anca frgr ttSF ath-: term.asOft2dcbrnzoi.EhcEastArsiaccn mi*>>icshavebeoinaica wstas unfa- voabe .,as cfa-vscin ohocsf dvlo96u-t.aggl -Bld-ow we exa-- m4e for caes in whidi athe HPAESinrsonde succdi, fd uya to.ea variety of maCrOe:oneomic sods:Indonesia, ore,h, Siengapoenloand Thand.. _ j, =tr * 2,; g *_ . *&e.mtvt'w>.ttiS.................-i> . l- E' Nh, .; ; + < -,=- j* -i -13eeoiOmlecs rheismdy4bun&m ihakd&Rod&M-> *,te>;t,r<,., ,,3ii'-g\;-,..)-sVl,ast Asia has bee +lucky7 rather vorale sthose ficing othernadevl,toping zeonomis. Below e exam-4 * n ou ae in whc th HPE respode sucesll to $ aaieyo macroeconomic shomcks: Indonesiai,snc Kra,Singi;Dapore,and Thailand. 0 2.eex14f-;RS,.-~*S.-4,~-tk;k.te-fSKe-t.w,WNit..hej %-"7J ''-'~~~~~~~~-.'i'-:a~~~~~~n~;' :1 . ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 4 0' - . ..~~ ~ .. _t-j + _V *. -- ,,,',PO :fifl _~ r _ - , : t . .......................................... - A i o no fd s V . i O-MalwAfl&q8w4d4x,O Aursou1s4 ipA o -o . - ;DeB errcostlsXnomy w>W~~~~~~~~~ein;iug Oil Pries hi.s Indonesias. From 1982c §itls'l to 1986 Inonsi *fuced - tMcreaeq-puD^ltMndg- rarpily worsenuing ters oqf traecuedrmriyb decli%ni;ngi t t~~~~~~~~rcs Strtn in 1983,4> thej! _ jgovernment, r esponde.di with av remarkabl .. . Uestlrsrlvs~~omplrehelnsie anF-d vsucesuladjustmenstpro2gDram.Idealuedthe Din- .~~~~~~~~~~il in 198 awntd 1986>= an cur exedtus minl by recedln . remmentSemMewolskomeffmeasurem oDmff teachieveen fosqrom 98 i tillPustated bthe ollow *-. . -~~~~~~~n calculoytaEnViamSoaetion fo ;Littlean othes orthaocomng). The exces of ~~~~~~~~~iprso goos adnnactoservices overannon-ol texots f*i:3ls frormta¢b "8~~~~~~~~~E X y tf 5 g t 1 XR2SOZSFU.uas=g ~~~~~ .~~~~~~~~~Ddnn OF aalcs inilntacil90 adarLsIndonesa. rom.ut1982 to 196,Iadronosi faed >erStalfizaonx DastalpriXeoa.Srng n18,tegvranenTaa-nit lreOspondkled.u with a.eaml S S robinssnaett)eltSmitocapiml-intcnsive poDjeC OIhened t eue urn account deficit .~ ~ ~ ~ ~ ~~~~~1 ca'lto (fo Li' anc otes foromn) h xeso - ; ~~~~~~~~~imports- of si scdvices F ove 192no A96 exportsi facedfi - - . rapidly w~~~~~~~9orsligtrso td lslpiaiyb elnn i : rosSarigi 93 hegvrmn epoddw8 l MACROECO 15 percent of GDP i 44 to 7 percent in 1988. This shift of 8 percent OfGDP measures the extent by which absorption had to be reduced-the ects--e of increased principal repayments of long-term debt, higher in- terest payments, and reduced income from oil and gas imports. This massive resource shift, also associated with trade liberalization, was brought about without increased inflation. *-.~ M - Adjustment was not painless. Growth initially fell due to lower ex- -- : port incomes and a tight 1983 budget; by 1985 the economy had - slipped into a recession, with only 1 percent growth. Even so the gov- ernment pushed ahead with the adjustment process, which actually ac- celerated in 1986. In the later half of the 1 980s, perseverance paid off in a boom of manufactured exports that pushed the export growth rate to 25 percent a year for the period 1985-91. Since then Indonesia has chalked up average annual growth of about 7.1 percent. Respodimgto M economic Crisis in Korea. In 1979, Korea encountered a variety of problems that threatened to undercut the 1970s' impressive gro wthu Rising oil prices battered Korea's termns of trade, the world re- cession dampened exporc demand, and high interest rates boosted debt service costs. Korea was not unique in these troubles, of course; these were the same woes that led to debt cises in many economnies outside of East Asia. Korea had plenty of specific problems besides. Real appreciation dur- ing the 1974-79 fixed exchange rate regine had made exports less com- petitive, the rice crop had filed, and the assassination of President Park * ~-Chung-Hee had exacerbated political uncertainty (Coffins and Park 1989). Korea had a few structural advantages, however Unlike economies that quidyfel into debt crises, Korea was not running large andgrow- ing budget deficits. And while private savings had dropped due to de- cdining output and incomes, investment remained high. More important, Korea responded quicldy to its troubles with an aggressive January 1980 stabilizaton package backed by UmF standby credits. The government ended the fixed exchangre rate regime, devalued the won by 17 percent, and dghtened monetary and fiscal policy Things got worse before thry got better In 1980, output fell 5 per- Centr, inHfation soared to more than 25 percent, and the current account deficit approached 9 percent of GDP. The strong medicine was pardy re- sponsible for the economys worsning symptoms: the devaluation spurred inflation, while tighter aggregate demand polcies excerbated :i9g * ~~~~~the drop in output. Even so, and despite political outraies over Kords rising foreign debt, it continued foreign borrowing throughout the cri- sis, thus maintaning- high investment levels. 'Within two years, the medicine had begun to rake the desired effect. * ~~In 1982 inflation dropped to 7 percent and in 1983 to 3.4 percent. The :!current account defict fei to 2 percent of rot in 1983. Overall, the gov:edr nes proumpt and effective response to a potential crisis stresngthened the conomy, preparing it for rapid growth in the 1980s. Mjusting n to in Sho i Thailand. Thaland only partaly adlusted to the first oil shock and in the late 1970s engaged in a mild private and public spending boom Th then mcmeh second oil shock and the rise in world interest rates. By 1980-81, thet cornlidated public sector deficit was 7 percent of GDP, nearly half of which -was the deficit of nonfinan- dal public enterprises. The current account deficit was also about 7 per- ctentBecause foreign bo rro ing had been moderate-the debt-GDP radto was only 35 percent in 1982-Thailand was not &cing a dbt crn- sis and continued to borrow. Even so, the new government that rook - : wover in 1980 perceived that macroeconomic adjustment was needed. Monetry policy options were lifmited by the fixed xchange rate and the relatively open capital marke. The government therfci re took the alter- native path, fiscl contraction, movig gradualy but consistenty dudrg the next several years to cut expenditures and boost revenues Policymakers steeply cur deficits of the nonfinancial public enter- prises, then gradualy reduced the central government defici. As a rt- suit, the consolidated govemrnent deficit decined from 8 percent of GDP in 198142 to 1.6 percent in 1986-87, when adjustment was es- sentially complete. Meanwhile, steeper tax rates and tougher colection efTorts boosted cepntal gavernent ax revenue f'rom 13 percent of GD? in 1982 to 16 percent in 1988. The amdjustment process was facilitated by a 1984 devaluation. Thai gradualism was possible because foreign borowing had been mode:ra and the economy did not yet fatce a crsis. But grdualism was not h-esitancn conservative fiscal polices were consistent and were sus- tained into the late 1980s amid an exporct and foreign investment boom. Sin 19878Tailand has becen accumulating foeign excamnge re- serves, and the governm ent has rcgularly recorded a fiscal surplus. - ThaCocing Policy Failwes h Snpor. Singapore escaped the 1980-81 world recession wath secacely adip in its robust growth. But in 1985 the ;- . : - -- ~~~~ecnomy hesanqconev eficlpiiswreosset fan werosu -economy encountered a sudden and severe recession growth fel hom 120. MACKtE~9 .83 perce-nt: in 1984 to 41.6 percent in 1985. Reasons for thi~s unusual episode included government encouragement of high wages between 1979 and 1981, which was intended to speed the decline of labor- intensive production but backfired by eroding competitiveness, and the apprecation of the Singapore dollar's trade-'weighted exchange rare. These were eamcerbated by a sharp decline in public investmentL due to the simultaneous completion of several major projects. Private invest- ment, which had been declining gradually for several years, also dropped in 1985, apparently because of flling external demnand resulting from Singaprs decining comperitiveness Recognizing the problems, the government devised an integrated policy package that reversed the high wage policy to restore Singapore's competitivenes and stimulated domestic demand. The govermecnt also cut the emnployers compulsory contribution to the Central ProvW dent Fund from 25 percent of wages to 10 percent, reduced corporate taxes, and introduced accelerated depredation. Income taxes were cut andt development expenditure was boosted by 21I perrcent. These policy repnses and a fortuitous depred'ation in the Singapore dollar, due to N -: the global depredation of the U.S. dollar, contributed to a rapid reco- My-c In 1986 growth recovered to 2 percent, and by 1987 it had reached 9.5 percent, a level maintained through 1990. How Macreconic Stabiriy Contnb-- to Gmowth It cannot be a coincidence that all of these seven economies have had exceptionally high growth by world standards, and all have had unusual success managing their maa.oeconomnics over the long run. Al but In- donesia and Korea have also been long-term low-inflation economies, while Indoncsia and Korea fal into the moderately low-inflation ate- goesvLow or modate inflation for lbong periods provides a fivorable envirorinmentfor growth. Thaere are four main reasons why high inflation is likely to be adverse For growth. and why rhe conservatve policaes followed in the HPAs are likely to have been favorable for growth. Fsrl economies that are not fuly adjusted to a given rate ofinflation usually suffer from relative price distortionspcaused byr inflation Nominal interm rates are often con- aolled, and hence real interest rates become negative and volatle; de- preciatons of the exchange rate lag behind inflation, so that rel appreciatons and exche rate variability rsult. As wc satw, the grel in- GompeiitiveneSS and sumulated domestic demand. The governmenr . alsocut th emploers' cmpulsoy contibutio tO th CenrMProvi Pen 7--- 7- MAL terest rate and the real exchange rate were unusually stable in most HPAEs, compared with other developing economies, which meant that these crucial relative prices were more e&fctive at guiding resource allo- cation (a topic to which we shall return in chapter 6). Second, real tax collections lag inflation, because collections are based on nominal incomes of an earlier year (the Tanzi cffect), and public util- ity prices are not raised in line with inflation. For both reasons the fiscal problem is intensified by inflation, and public savings may be reduced. Public savings have been an important component of the unusually high levels of total savings in the HPAEs, compared with other low- and middle-income economies (see chapter 5). Third, high inflation is inevitably unstble. There is uncertainty about future rates of inflation, and this both reduces the efficiency of in- vestment and discourage it. If the inflation rate were high and stable, there would, in theory, be no problem on this account. But in reality the higher the inflation, the more likely are measures to reduce it. These in turn have a contractonary impact on private investment in the short run. Finally, high and variable inflation imposes substantial institutional costs in many economies. During periods of price volatility, scarce man- agerial resources in the economy are drawn into financial as opposed to real sector management, as was the case in Latin America in the 1980s. How irportant to growth is maaoeconomic stbility? Cross-economy, econometric studies generally find that higher inflation reduces growth (Fischer 1993). But the relationship is not robust for small changes; an economy with a slightly lower-than-average inflation rare for a longer perod does not necessarily have a somewhat higher growth rate. For ex- ample, Thailand's long-term growth rate is well below that of Korea, even though Koreis average inflation rate has been higheL Furthermow, there are many economies that have, at various times, had low inflation rates and low growth. The most important case is that of India. There are also cases of high inflation associated with high growth. The most important example is Brazil from 1968 to 1980. More recenty, Turke/s growth rate from 1981 to 1990 averaged 5.4 percent, while its inflation rate averaged 46 percent (ittle and others, forthcoming). Low to mod- crate inflation may be a necessary condition for growth, but it dearly is not sufficent. Relatively cautious fiscal and fbreign borrwing policies meant that serious debt crises were avoided, which reduced the stopgo pattern of crisis and response that characterized many developing cconomies in the In MACROACCjj_~. 1980s. As noted above, there have been problems, notably in Korea and Malaysia, but they were dealt with swiftly. The Indonesian and, even more, Thai reactions were more gradual but also effective. Avoiding crises and the need for rescheduling meant that creditworthiness was maintained, and it was easier to borrow in the shorc term and to avoid very deep cuts, especially in invesmenL Sudden reductions in aggregate demand and in investment compelled by debt crises were major causes of the sharp declines in growth rates in many of the heavily indebted economies of Latin America and elsewhere. Although macroeconomic stability and prompt responses to macro- economic shocks were not the whole story of the HPAES' success, these factors created a basis firom which policies intended to affect the real economy-the supply side-could be launched in an environment of stable real interest and exchange rates. We now tum to the most broadly shared of these supply side policy initiatives: creating an export push. Creating an Export Push GOVERMENTS OF THE HPAEs HAVE ENCOURAGED exports by fostering a supportive macroeconomic climate and by providing suitable microeconomic incentives. Macro- economic stability helped exports by easing the liberaization of re- straints on tade and by facilitating realistic and, in some cases, undervalued exchange rates. We discuss these issues below. Few general- izations can be made about microeconomic incentives, however, since these economies differed in the degree and selectivity of promotion, and each economypassed through several stages. Because of tis and because ecport promotion has played such an important role in the East Asian mirade, we condude this chapter with a survey of the evolution of export-push polices in all seven developing HPAEs (these arc summa- rized in table 3.5). Appendix 3.1 givcs brief policy histories of each economy in the fiorm of timelines. Macmecononic Stability Facilitated omre Open Economies The success of the HPAES rests pardy on what they have done and partly on what they have not done. One thing they have not done is to n3 - IRA~~~~CLUE Table 3.5 Phases in H PAETmade Regimes .~Nad&z anlimad 'War and construe- Mket-led develop- Land reformn anid Natural-resouirce- guidcddevclpmcnc, do'a,1950-O0 met .150-70 reconstruction, ae prs 194846' ~~~~~~~ ~ ~~~~~~1949-.521900 Impori-substinicing industrialirzatiLon-, 1953-57 J AF:K~~~~~~~~~~~Eport promotion 1958-72 - ZTilting.twr ~ incur 1967 73 tiC .;;J*-Export promotion. Fvoring import sb tSelectiv6 inter:vendion>- 2andirnPorts subst-~ Indi'srial. con-solida- sio 197180: Omlandcommoity dioi4hii:heavy 2nd stontut7-o5 . no -boomIn,974-81 cciakuwe growth, 1973-080 * ~~Functional abdlbcraizaon, AdjstmntE to aex- 1809 .- .': ~~~~~~~Refborm ad epr shodci` 85~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~icntvs -1980- Hihtecnlgn present Derc~~~u1adon:and Adjw~~~mciz-andh- moeniztion, oriU.nrarion,T 19iiod-ors * outward ..d cAIimno, A_ - CMLEMUOn, - ,11986-9 present. impose general import restrictions to redress balance of payments deficits. Hong Kong, MaIayrsi-a, Singapore, Thailand. and Taiwan, China, had no cause to -impose such restrictions, since their current ac- count balances neve-r faced serious long-term deficits. ndonesa and Korea. with more troubled current: accounts, might have imposed re- strictions but dlid nor. The benefit of avoiding import restrictosi iely understood. Such measures strve enterprises of imported inputs and are particularly hard on exporters, who tend to depend heaviy on i.mported materials 124 MACRO-EC;~~ and capital goods. But easing import restrictions without accommiodat- ing maacrconomic or exchange rate policy can womsen current account problems, as consumers and firms buy imported goods not previously available. The trade regimes of the eight H-PAEs have differed widely, but each has gradually liberalized without incurring a serious current ac- count deficit. Often trade liberalization has been part of a policy pack- age chat included devaluation (usually to cushion the blow to import -substitution industries), exchange rate unification, fiscal reform, and foreign aid or concessional loanis to offset a tempomarily deteriorating current account. The dose link between successfuil macroeconomic policies and trade liberalization can be seen in the experiences of Indonesia, Korea, and Taiwan, China. Taiwan, China, took its first big step toward trade liber- alization in 1958 with a policy package chat included a 25 percent de- valuation, a unified exchange rate, export incentives, and the widespread remnoval of quantitative import restrictions. In Korea, devaluations and trade liberalization have gone hand in hand, beginning with a dtrastkc de- valuation in 1961 (Collins and Park 1989). Indonesia began trade liber- alization in the late 1960s, together with the stabilization efforts chat marked the start of the Suharto era. The new government abolished comprehensive import licensing and in 1970 united and devalued the exchange rate. After an inward-looking period in the mid.-1970s, trade liberalization resumed in the 1980s and has since proceedecd together with an export boom. Exchange Rates and Exports As We have seen, several HiPAE governmentrs used exchange rate poli- cies to off-set the adverse impact of trade liberalizations on producers of import substitutes. A few went beyond this objctive, however, and used * deliberately undervalued exchange rates to assist exporters. In these in- stinces, exchange rate policy and the fiscal and monetary tools to carry it out became a part of an overal export-push strteg Taiwan, China, avis the most notable example of this, but Korea and Indonesia also delib- ecrately undrvalued their currencies to boost exports. We briefly discuss agl three belowo The very large current surpluses that Taiwan, China, ran in the 1980s, especally from 1984 to 1987 (when the surpluses 125 CL E averaged 16 percent of GDP, with an etraordinary peak of 20 per- cent in 1986), resulted from government efforts to manage the ex- change rates What would have happened if the New Taiwan, China, dollar had been allowed to appreciate more rapidly? Exports would have become less competitive, reduced export growth would probably have had a deflationary effect, and this in rum could have reduced savings. Alternatively-and more realistically-thc poten- tial deflationary effect would have been offset by increased public expenditure, leading to a budget deficir. * Korea used exchange rate protection from 1986 to 1989 when it ran a current account surplus (which peaked at 8 percent of GDP in 1988). A desire to protect the export industries was certainly a fictor in Korean exdcange rate policy, but tie main concern was to reduce the debt ratio and build up reserves to avoid repeating the close brush with a foreign debt crisis in 1984-85. m The Indonesian devaluation of 1978 can be classificd laelyas an- ticipatory exchange rate protection. No immediate balance of pay- ments problem existed, as the adverse effect of the previous real appreciation of the rupiah on non-oil exports had been offset by the rise in the quantity and value of oil and natural gas exporcs. Rather, the aim of devaluation was to encourage non-oil exports and slow import growth. By 1982, when the balance of payments had sharply deteriorated, the wisdom of the 1978 devaluation was clear. One can see a fairly clear relationship between devaluations and ex- port growth in the 1980s. Taiwan, China's, real exchange rate relative to the United States depreciated sharply from 1980 to 1985 (the period of general dollar appreciation), and the result can be dearly scen in an ex- port boom to the United States. The Korean real exchange rate was kept fairly stble by numerous nominal devaluations during the period of the first great Korean export boom, 1963-72. The effects on exports of real devaluation from 1982 to 1988 can also be seen very dearly. The Malaysian real exchange rate steadily depreciated from 1987 to 1990, and this must have been one factor in rapid export growth-. The effects of Thailand's real devaluation in 1984-88 on export growth were quite dramatic, even though there were also external fiators explaining export growth, notably the availability of capitl fiom Japan and Taiwan, China, for developing export indlustries. From 1986 to 1989 the dollar value of exports rose 12 percent. All three of Indonesims 126 MAC ROE CO NO M ICr devaluations had dear effects on exporcs of manufiactures, and this was particularly true of the 1986 devaluation. From 1986 to 1988 the vol- Iume of exports of manufactures rose 80 percent (from a Fairly low base), and growth continued right through 1990. We now turn to our survey of the evolution of aporc-push strategies in the seven developing HPAEs. 'Korea Pushes Exports and Industrial * Korea's development has passed through three smges and is currently in the midst of a fourth Unlike several of the larger economies in East X Asia, which evolved from protectionist, inward-looking trade regimes toward relatively open economies, Korea did not have a sufficiendy large :: domestic population to contemplate a strategy odtcr than export-led de- - velopment. Its performance may be described as forced growdt, because it did not stem from exploitation of natural resources, an influx of labor, flows of speculative capital, or the adoption of new means of produc- tion. Rather, growth resulted from a systematic program of importing raw materials and intermediate goods for processing and export with added value. War, P oanstrclion, and Land Refonn (1956O-01. Korea had a relaively well-developed infrascructre at the end of World War II, but the parti- tioning of the peninsula by U.S. and Soviet forces severed economic links between the heavily Midustrial north and agricultural south. The war took a heavy toll, taking 1.5 million lives and destroying two-thirds of the south's industrial capaciy. With a poor natural resource base and 'one ofthe world's highest population densities, the south was almost en- tirely dependent on U.S. aid after the war For all its devastation, the war may have helped to prepare Korea for an industrial takeoff by loosening a rigid social structure, opening the way for fundamental changes in oudook. While developmenr efforts in tie 1950s induded several false srans, progress was made in reconstruction, including the restoration of transportaton and communication networks. The government also completed a land reform program that had stalled before the war. Export Takeoff 1961-73). Under President Park, aggressive promotion of exports was combined with-classic import protection at home. Ko- rean policyakers nmainLiained dose control over trade, exchange, and - 1financial policy, as wel as aspeccs of industrial decisionmaking. In con- trast to other controlled economies, they used these instruments to pur- 127, L~~~~~~A4 41 ,. C, E- sue the primary objective of export growth. The trade regime was biased in fiavor of exports as a whole but essentially neutral with respect to dte composition of exports. Even so, the first instruments of export promotion were highly dis- cretionary. Exporters were supported with multiple exchange rates, di- rect cash payments (see table 3.6), permission to retain foreign exchange earnings to import restricted commodities, and permission to borrow in foreign currcncies. This system not only avoided hampering exporters with restrictions on capital and intermediate inputs for their own use, but it also gave access to he firavomble exchange rates determined by scarcity rents in heavily protected domestic markets. Even as discre- tionary incentives were gradually replaced by more automatic instru- ments, exporters received significant exemptions from import controls. Table 3.6 Effective Exchange Rates for Imports and Exports, Republic of Korea ' ' ' '' ' ' S' '''' ^ '#ren'ccs due '"~ ' ~' ' to: - -~ Ecnge- p'remiums phs: ':D'treet :.ax ax a n.' ; 4 f - ~d6r tcashX ji. m aterb-- -,--' -,"-, of2&i¼mt'i l ;!l: l.F.Sdcparunrnjstwe4* .PM.wodsbMwoucra'ovcnsaSo &lSCca4O :ioo . * .5 i.-1OI6W4lKLQJon s-Cot r 2t m.~ oi -EsSw-S^h jwtiglrttm&ask*Glitbaej&fl an a~- a Moi~~~~~~~~~_S ___ - .Stot F,ritiEgt,,fE,,.,^WK drafited in 97 recogiJ,zed thait the comp_9;:lexity of the>>z=> economy ws E rcouldnororZfoeveceedingwma the urg;ovrnments anagement capEacitie.h plDan's emphsi . |~~~~~~~~~o indictive planningand a greatr role *the maret srss4-wtrg-|gs#Qz;btamB s entually R$kZsEasl@rteE;s SadiNosindustres, scupport for h develokmopmcent f echolgyand the promo- don of competitiont.ti ME- Z>I&A1 M I RAC L E exchange rare policy adjustments. It also initiated comprehensive finan- cial and tax reforms, and made major improvements in customs, ports, and shipping. As we discuss elsewhere in this chapter, the macroeconomic measures successfiuly restored financial stability. The current account deficit de- dined, and inflation was brought to below 5 percent, from 7.6 percent in 1982. But trade and industrial policies gradually became even, more inward oriented and subject to government intervention. In addition, the adjustment resulted in slower growth of output and incomes, re- duced public and private investment, low rates of capacity utilization, and the emergence of financial problems among industrial enterprises. Moreover, although textiles and plywood both emerged as important exports during this period, overall manufacturing growth slowed in the face of weak non-oil exports and weak domestic demand. By 1985, a plethora of decrees had brought a wide range of products under different forms of import control. Overall, the production regime favored import substitution rlatidve to exports. As a result, domestic re- sources were drawn into relatively inefficient capital-intensive activities supplying the domestic marker- This disadvantaged the more labor-. intensive, efficient downstream producers and reduced export growth. Derigflation and Outward Orietaton (1986P--. In 1986, the economy suf- fered further shocks, as declining oil and commodity prices led to a steep dererioration in the terns of trade and a jump in the debt-service matio. This time, however, in addition to successful efforrs to restore macm- economic stability, the goverrnent launched a program of broad trade and regulatory reforms that opened the way for a decisive shlfif o export Figum 3.4 Effective RatOs promotion. Exporc-oriented deregulation has gathered momentum since 1986-8 7, culmninating in a series of major reforms in trde policy, Import-:X ... Impoflb 2sminvestment licensing and transport regulatons. The first step toward trade reforrn bad actualy been taken in 1985 Tramnston with steep tariff curs. In 1986, pushed forward by extrnal shocks, the _ l I process gathered momentum. Between May and Ocaober, tde govern- :Exportpush-dI 1 1--+,i ment announced a package of exporc incentives, a major devaluation, -- - o 20 40 6o so 100 and the first of several programs to simplify imprt and export proce- Effatve rtedf potectJon dures. Major exporters were given unresticted, duty-free access to im- (percent) * Mlaysia 31 Indonesia Thabnd porcs (see figure-3.4). Later the government deregulared domestic and -cSalieb, Xcah. and Mcyanadmn foreign private investment, and for the first time permirted the private -mcr 507zSo&k, Yeah and Mcpnathan (1993); Bbarracharyaand Pangx (1992w); sector to invest in power, telecommunications, ports, and roads. It also Brinmbc (1993). firher deregulated the financial sector 138 'M ACRO E 'CON The economy responded rapidly as manufacturing output, exports, and investment all increased. Growth accelerated from 1988 onward as a resurgence in domestic demand accompanied the continued growth of non-oil exports. Plywood and textiles remained the largest manufactured export items, but other manu&ctured exports such ats shoes, apparel, and electronics began to grow very rapidly. Foreign in- vestment projects approved by the govcrnment increased tenfold be- tween 1986 and 1991, to more than $8 billion, and domestic investmcnt rose by a similar proportion, to $30 billion. Perhaps morc important, 70 percent of the foreign invesrment approved was export oriented, compared with 38 percent in 1986. Since 1986, GDP growth has averaged 9.3 percent annually, the highest sustained growth rate since independence. Indonesia continues to encounter occasional difficulties. In 1990 and 1991, the economy experienced an import surge and deceleradon of ex- port growth due to increased aggregate demand prompted by a prior easing of monetary policies. This led to a widening current account deficit and a surge in foreign borrowing. As with previous macro- economic cdistortions, the government responded by tightening mone- tary policy and curbing borrowing. Meanwhile, the obvious success of export-push policies can be expected to generate momentum for firTher export-oriented reflrm. If this is the case, future swings toward protec- tionism, should they occur, are likely to be less pronounced and less damaging than those of the past. Thailand's Shift fom Resource-Based Exports to Manufacturing Likc Indonesia and Malaysia, Thailand historically exported primary and agricultural products derived from its rich natural resources. Early trade policies emphasized taxation of the resource-based rent associated with the production of these exportable commodities. Trade was heavily controlled. Rice exports, Thailand's major commodity export, were con- trolled by a state marketing monopoly, export taxes on other commodi- ties were heavy, and a multiple exchange rate regime discouraged export producton (Jitsuchon 1991). Natural-Reso.rce-Based Exports 11955-70). In 1955 the exchange rate was unified, and the state marketing monopoly on rice exports was abol- ished. Although these policy reforms encouraged natural-resource-based anrd agricultural exports, Thailand maintained substantial import pro- '.39 ffl ]\ ,, ,, .RWCLE tection for specific industries. As late as the mid-1960s, however, overall levels of effective protection to industry were modest by developing- economy standards. This was especially notable in consumer goods manufacruring, where nominal tariffs were in the range of 25-30 per- cent ad valorem. Tariffs on machinery and intermediate inputs were in the range of 15-20 percent. FavoiingImportSubsfition (1971401. In the 1970s Thailamd, following the import-substitution strategies favored by many other developing economics, raised tarifli on consumer goods to a rangc of 30-55 per- cent. Capital and intermediate goods continued to enter at low duty rates, leading to an increase in the effective protection to value added in import-substituting industries and to dedines in the eff-ective protection to agriculture and traditional exports (figure 3.4). The bias against agri- cultural and export production increased. Textiles, pharmaceuticals, and automobile assembly were particularly favored. At times, vehide imports were banned, which resulred in limit- less protection for domestic manuficurers But there was a quid pro quo. Domestic content requirements were established to promote upstream suppliers ofcomponents and parts. These nontariffbarriers to competing imports of components and parts raised producton costs in the assembly sector while subsidizing the produce- of inputs. Consumers were in ef- fect taxed to promote an intermediate goods sector. R-aion and Export IncentWes (19804. As was noted in our discussion of macroecononic polic, the second oil shock exposed weaknesses in the Thai economy that were the result of the import-substitution policies of the 1970s. In 1981 Thailand's trade poliqc shifted explicitly in the di- rection of export promotion. Remaining export taxes were reduced, and as described above, the exchange rate was devalued and subsequently moved to a managed floaL The government also began reducing pro- tection of local industries and making tariff lower and more uniform. The maximnum duty rate was reduced fiom 100 to 60 percent, and the average rate was reduced. The tariff reductions were alrnost immediately offser, however, by the government's decision to impose a tariff sur- charge to raise revenue for macroeconomic stabization objectves. Ef- fective prntection remained high into dte mid-1980s at 52 percent of value added, higher thaL other EastAsian economies such as Korea (28 percent) and Malaysia (23 percent). During these early stages of the export-push period, the govemment Board of Investment (BOI) played an important role in promoting cx- 140 MACROECO?rw1 . .~~~~~~~~~~~~~~~~~~~~~~~~. port growth. The BOI, which had been established in 1960 to supervise the application of investment incentives, had long been a tool for im- plementing the import-substimtion strategy. One of its key roles was to distribute (and often increase) the incentives for firms operating in pro- tected domestic market niches through waivers of tariflis, import sur- charges, and occasionally bans on competing imports.5 The shift to export promotion brought a different role to the BOI. A - major study of the exporc-promotion system was carried out under a uNDP-World Bank technical assistance project. The study recom- mended radical reforms in the way in which the BOI did business. It ad- vocated using more automatic, sectoraily based incentives (in contrast to - the firm-specific discretional incentives of the past two decades) and shifting the target of incentives to labor-intensive, export-oriented, geo- graphically dispersed activities. In 1983 the BOI announced new promo- * tional criteria that f&vored, among other objectives, exports and labor intensity. Box 3.4 describes the impact of this shift on direct tariff in- vestment. Other incentives for exporters induded tax exemptions and rebates, reductions in electricity tariffs, automartic access to credit, mar- keting assistance, and promotion of tracing companies. The govern- ment also stramined customs procedures and abolished unnecessaty regtuations to expedite export shipments.6 Automatic and concessionary credit was another major element in the export-push strategy. The Bank of Thailand had taditionally ex- tended refinancing flaciiities (RFFs), through the commercial banks, to key economic sectors.7 These RFFs were mostly used to rediscount cx- * - port bills of large exporters handling traditional producs As part of the export incenuves package, the Bank of Thailand revised its rediscount irules to focus more explicidy on small, nontraditional exporters and on other productive activities. Many of the export-push incentives were specifically designed to off- set remaining distortions from Thailand's import-substitution era. Tax rebates, duty-free imports, and export-processing zones correced par- tially for existing distortions. In this sense, Thailand's experience is sim- ilar to that of the northeastem HPABs; the initial surge of export promotion tookplace in the context ofa domestic market that remained modctey protected from the rigors of foreign competiton. Thailand has recently begun a more broadly based liberalization of imports. In -1990 tarifs on capitl goodswere reduced, and in 1991 large reductions werc made in tariffs on automobiles and computers. .. -,J - , g f : ; -_ ii~~~~LE' . ~ ~ ~ ~ ~ ~ ~ ~ . .g . . -~ ~ ~ ~ ~ ~~ie .,L . E-or s -.~~~~~~~~~~~~~f wS_ .n . '', M '.ape'roa9AesS>'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'f',-f ' ''.~~~~~~~~~~~~~~~~~ ' -" - - D X = ~~~i ar =CU roa es :~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ NT :~ -: g :~~~~~~~~~ f - ~~~~~~~~~~nstet sm W-96 ne :--: : ::~~~~~~~~~~~~~~~~~~~~~~~~~i3 ! :g :- ;~~~~~~~~~~~~~~~~~~~~le - l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~a - u - ~~~~~ - ~~~ - The -outcome of these pol~~~~~~~~~~~~~~iy a laacB 96* X - - : - manufacoJres reprcsented 30.6 pelt of a gowing volume of That ex- - - pora;. Lading secors inclu ed d Woin, ~IO W f oj- : - - -. elrr, ad integrarcd is E Oirect foei n ilVtmetpl la ao .~~~~~~~~~P A .. . . . X - 0 - ; - * - :xS 5 ; .; 0 - .: - 0-- - - 0 MACR0O J~~ Table'3.7 Distribution of Man.factnriug, Thailand (Percentage of gross domestic product) Senor' ,97. '1986 : - 0 0 v3: ;n.uMtA C R1O36 _ Hev inutis3.9 42.6: 3. High-skill labor-intensive idusre 9.7 11.1 .13.3 Traditional light inutis19.2 201 .23.'3 * Total. . - 100 100 . 100 mTab e 3 Banko ofTi2n aflTiland The Elements of a Successful Export Push These summaries of the evolution of export push reveal the plethora of policies consistent with export promotion. (China's recent policies fa.- voring manufactured exports are outlined in box 3.5.) AssistaLnce to ax- ports varied over rime and across economies and induded p..ferential .financing, promotion stubsidies, tax incentives, subsidized inf1auc- tLre, and foreign investment incentives. Moreover, at any given time and in any particular economy, the magnitude of the incentives varied across activities. Some governments favored nontraditional exports, some favored particular types of manufacmured exportsh Even so, there are significant commonalities among these ponlCis. Each contributed to one or more of the four elements of a successfl export pushi accss for exporters to imports at world prices; access for exporters to lon- and short-term financing-, goverment sistancn in penetrating markets; and flexibility in policy implementation. Access to Iviports at Worm Pricer When imports are restricted or rtaed, exporters are: disadvantaged vis--vis exporters from other cono e who do not face similarly expensive imports. To combine import pro- tection with export promotion, these ill effctcs must be mitigated. HPAE governments, have found myriade ways to grant exporters access to ir- - ports atworld ipories free trade zones, exppot-processing zones, nd : p sorts-tewrld prinacer in bonmd aitn npntrin aed warehouses, luty drawbacks, or exemptions from tiffs. However, the higher thelevel of irmport-substitution protecton, the greatcer the level wof adminisrative cormiptenc needed to cseate countrvailing imcentves for exporters. Perha because of these difficulties and the dedining 0 ovrnen:shn foud yia wystogan epotesapssst img benefits of import-substitution protection in successful export-oriented economies, each of these economies has gradually liberalized its trade '-43 = X _5 -tCL~CE ... .~~~~~~~~~~~~~~~~pt, g. , ar, L .* G .. ~ ~ ~ ~ ~ M ..~ ." <._ ... . . ~ ~ ~ ~ ~ ~ ~ ~ . .~~~~~~~~~~~~~~~~~~~~~~~~~~~ , MC - pogamtoenux aces o reir oce a sbizex apries,.FButagion-,.,. - : 0 we oberve i ipesvartyofhye of creit j ntnvru 0 - - shorr-tan), the degre oSsubsidizaon9 ( !;aMantS; a, as subsi : f 0 dizedraresX, theselecizviy all QqporT ve9u 5age , ex ortatirres T~~~~~~~~~~~~~~~~~~~ ~ -a Lgo-. .~~~~~~~~~~~~~~~~~~~~~A- - T_ ..~~~~~~~~~~~~~ .r, .5 .V. . :.: . , .:: - ;: .- ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~I MACRO.ECRO:E Mabrket Penebaiom. NearLy all governments recognized the difficulty expoers Fce in cracking fbreign markers but again chose various means to encourage exporters to overcome these hurdles. Some directly subsi- dized export activity (direct income tax incentives), some subsidized marker penectration (through exporter associations), some subsidized small and medium-size exporters to offier their difficulties in marker penetration, and some promoted the creation of international trading companies. fleibfit. Pragnatism and a commitment to results mark the HPAE governments' approach to export policy: Policy flexibility proved impor- tant because hitting the right strategy is not easy, fir three reasons. The right strategy depends on the circumstances. It changes as the economy changes. And it is not always obvious. Making mistakes is not necessar- ily bad, as long as the mistakes are reversed. The dassic example of pol- icy flexibility is Koreals reversal of the HCI drive. After a large commitment to promoting heavy and chemical industries4 the Korean government nevertheless changed course when circumscances changed and the policy threatened exports. But Korea was not unique in this re- gard. In each of the lPA.Es, policies evolved over rime and often required major revLsions. In chapter 1 we showed the impact ofthe HAEs' export-push sategy on the rapid growth of manufactured exports. Another outcome of the export push can be seen in their changing composition of exports (see figure 3.5). With the exception of Indonesia (whose exporcs of refined petroleum products are induded under chemicals) all of the HPAEs dra- matically increased the share of exports in machinery and equipment at the expense of such traditional manufiacmres as food products and tex- tiles. In 1991 the share of rnachinery exports in total manufacturd ex- ports for the developing HPAEs was about double that for other developing economies. This shift clearly shows the response of exporters to changing market opportunities, rising labor force skills, increasing technoogical capabiliy, and changing comparative costs. What is surprising, however, is the extent to which textiles and ap parel remained important exports in the face of shifting comparative ad- vantage. As late as 1991 most of the developing HPAEs had shares of textdles in manufacured exports which cxceeded the average for other developing economies, suggesting that these sectors retained substantial cost competitiveness despite rapid capical deepening and rising real wages (a phenomonon to which we shall return in chapter 6). :-45 IE~~~~~~~~~~~ MA CR- O The successful East Asian economies have had much greater macro- economic stability and much more rapid export growth than most other developing economies. They achieved macroeconomic stability by ad- hering to macroeconomic fundamentals-particularly by keeping fiscal deficits within the linmits of prudent financing-and by rapidly and ef- fectively correcting major macroeconomic imbalances that emerged. They achieved rapid export growth by fostering a favorable environ- ment through macroeconomic stability and by applying on a trial-and- error basis a vast array of market interventions and microeconomic incentives. In contrast to the many less successfiul economies that have clung to filled policies, the HPAEs have pragmatically chosen and as- sessed policies according to their impact on macroeconomic stability and export growth. They have adapted policies as necessary, kept those that worked, and dropped those that failed or outlived their usefilness 'Why did the East Asian economies choose macroeconomic stability and exporrs as their policy yardsticks? And why have they been more succsfil than other economies in selecting, implementng, and alter- ing policies as necessary to attain these goals? We mrn to these questions of political economy in the next chapter. 3-47 R'A CLE Appendix 3.1 Economic and Ploflitcal rimermes Indonesia OpemdanatsubhJJisa':- M.~ ~ ~~nnhcfup mita ho & b m n m _ _ _ _ _ _ _ _ _ _ mawiled ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~Ol dm Ci Wed ccanoiny.. a. dkea,mtecuainl Rdninrpaaaim McPtsUniicatone. -fr'nlto.'--totim atpinduainnmmd Th~Nw ndp in. -f rank gugjsajfraden Wailvs!m cak- . - r.hghmafi firms ewzuedwt HjyCIfDAhtuIKI Sq 01 . -minm ins -LM 1New bcip - ntu aio asI&uiamoat iila c=7- Su&,wont z.kdqaSns,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. 7. Gicderiy "~"_... 148 MACRO 4 it i i- - - Rqmn. 'S~~~~~~~~~~~~~~F .h1 .c J;.Dfdb . .rtat& cpu! Sore t~~~~~ 4(** 4 ~~~~~~~~~~~~~-__ __ __ _ Applbw mun-of * ~~~~ *. aduap ~~~~~~~ Duicamny&mns~ EWuaJn kauga -Oc .DcpIaa -iiil0 .--. . .:PN Smwod 6hd ai sar bb-l ,;-b Fa t h- ' appmv.4mpa; '.- ;elto m-pr ~~~~~~n , ; .,p9& ;9?, ,,.E,--.-.ica :cauua ,had;, . rus,* ; t SB S lBS. 1982 ' 185 298 2 0 8 9 - : 9 '149 ::::~~~~~~'W -- t L Malaysia t,. -_ ;' 1 o - ! _ . , - ' -. ; E' XFin; Ma/ape ' . RmMaj' P-a -; -d -d Flu' -|-.: :Invutcu Guarimeu REr;Fads''\ ' Wea ' 'eT - ''H a' ElbI- f -'''t In ra 'a 'za A ., -; ,,hiu..m- - : . V '- .>::-* ; ~ rn . .- -. ; , - -: t * - --.lnancc .;Me:.i C.nnhzmo d c - n: P-o -- ...;.i: AeMof T : F uIn t = . Ai. ..s-- .~~~~ -it.hl,duc.rf TaiFhvki .- ; nduth Dcv r*m t. PmThdc F- - . .UIfllf - : .,,ar -,' '., . --," t ;--T F95 19:7'2 *e §-! 1 -' .as -f-l _ak '. CoaiI .- .edpam Atikd 'Z,Aa --uc-y:- -, - _. 155 1;:' 1.:; - 2.95 286 26. ;36 .56 2.;6 2.6 _39 27 8h - .~~~~¾ :~ ~ ~ ~ aidnfrminMlyCnrn . . -. puawersii.linccihispendcin 150 MACROE!cONOI'd Jlae f'aasuanu 4 mm w?Lntae flundd A H~ ~~~~~~~~~~~~Hwedpma d rA ~ ~ 7mteaa& fmAapaL ~ w?apAL ik.; Caaveuaan -~tw -.-- Ccep0 mdon.has 6dm; Pummlga9onc .9 g -95 flatten CoalIuonAa hakswete, DFMdapt ate ZntaAa~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ SzittqfSis~~arhz4wteutitym k rwmpato a cnpi alirsics~ ~ ~ ~ ~ ~ mqtpmoD - lctor Fmna , Dndh p mn Praadnfpn)- aun Avl&mmmac=:-cdtdldnrc RelasionsAct kbca~hsrt' Ddpsasll emda.A toa rsc apusrake f Epl'e q byp(uu imInanics .lard indusrs indusif Vzmz * A Cud adlcldorrr ad~~~~~~~~~~~~~~ asnr tarn - ~ ~ ~ ~ ~ ~ in cir dvilnnm Z 152~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. A 'I p 1 .1 ijr Ii I: S.  4 5. I N U E. . .. . .  .2   Liii r U Fgfl  j iI jS *j. * H* I I: * lB * K * i i.irr a * up . ' r I ¶ L I" U r I, '(U I * - N - I Id I I JJ'L[ ill i-Ii ** I a C L E Tainan, China 141gb ma aflntmt ~~~~~~~~Tbhir liur-Yac; Ema,nonc aauvlapdcpnr Indprn i D mv"nt Kmr Pi'lt *' . -cnt t Stwn fa I of~~~~~~~~~~~~~~~ [WI oa ~ e * law- ~~~~~~isarml Smax far SWiPCle panacfmcf- Nto 'CLWO' LaNd to LiRnClp meat nd Tnug.IJiar herpm Ddnr ~~~~~~~~~~cadlazao lan lWborm 'InniCal RateL .7. .- 1845 - 1550 1952 1953 1965 1571568 1359 290 86 54 185-19 16 * -. .93'164 '54 MACROE A C. . ~ ~ 1 . Rauanachimn pn.ductviycatcr vith . ,;s s ; -,?a nma,nr,aklbrc . 7 -. - ---- - - - , RL. Urnlimdon ;[7'< ; ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~f f f d ; Dewlcpmmr a Foreign FAmU6k,cn& SAkbmcnt .d;1972: 197n - =, -ofIndustr[al O~Iiutr &ah - ^ ¼' - - Estahlubman &uU'nihmc, < * -; ' - t; ' 0 i-, - ,, 7 - :-, 0 T.-; S - f u ,: " laboriaws a 'x - . d zial .,N e l.:-:, Wdncbu c - - : infmu,-c . , d '5en;- -d' 'J;D';; - ,.,,'N D ,' T. -'','.or .' s&.a. _______ pu4n *-.' .d .; d''','' Ins e D -l Initkial- .. ..I 010 -. . ~ . WoddEatak . . oppm-IIn . aiwan. Liwnadotpl f .... -.1 . * .. m7i~~~~~~~~~~~~~~~~~~~~~~5 Nat 0 '' is, i1ce - - m--a-r"- ,,,me%'to Notes 1. Bradford is among the first to categorize a combina- 4. The Promotion of Investments Act of 1986 pro- tion of functional and sclective interventions that result in vided incentives for investments in agriculture, industry. the effective exchange rare for exports exceeding that for and tourism. imports as an export push. By this definition the South- east Asian NiEs would not qtalifl (Bradford 1986, 1990). 5. The BOi's policies of selective intervention to pro- We we ds e term somewhat more broadly ro indude sus- more firms and specific sectors are discussed more fully in tained movement toward parity of incentives between ex- chapter 6. port and import substitutes, combined with institutional supporr for exporters. 6. Onc impormtn example of these efforts was the es- tablishment of export-processing zones (Ens) and bonded 2. The analysis in this section draws on Easterly and warehouses. Drawing on Korea's and Taiwan, China's, ecx- Vchmidt-Hebbel (Ibrthcoming. The data on consoli- perience, Thailand used the EPs to streamline temporary dated public deficits, as well as the resr of the data in this admission of inputs for exporters in the zones. The zones section except where otherwise indicated, are from the also offered lower-cost elecrricity, frcedom of immigration same source. Consolidated public deficits, though less for managers and technicians, and. common-facility widely available than central government deficits, are a bonded warehouses to reduce customs and storage costs much more reliable indicator of fiscal management be- for enterprises located in the zonc. In general these incen- cause they indude operating deficits of public enterprises tives also favored large and often foreign-owned firms. that have played a critical role in some macroeconomic cases. 7. A firm wishing to obtain access to the scheme issued a promissory note, which was discoLnted by its commer- 3. Although the mean level of effective protecrion was cial bank and firther rectiscounted by the central bank low, variance among sectors was high. both at subsidized interest rates. .56 R 4 An Institution Basis for Shared Grot N EACH H-PAE EXCEPT JAPAN, NEW LEADERS PACED AN URGENT need to establish. their plitical viability before the economic takeoff. The Republic of Korea was threatened by invasion fr-om the North; Taiwan, China, fr-om China; and Thailand, from Viet Nam andt Cambodia. In Indonesia, Malaysia, Singapore, and Thailand, leaders faced formidable commnunist threats. In * addition, leaders in Indonesia, Korea, and Taiwan, China, having taken power, neededt to prove their ability to govern. Leaders in Malaysia and Singapore had to contend with ethnic diversity and attendant questions of political representation. Even in Japan, where the competition was less immediate, leaders had to earn public confidence after the debacle of World War [I. In all cases then, leaders desperately needed to answer *a basic question: why should they lead and not others? Whatever strategythe leaders of theHPAE governments selected to an- swer the basic challenge of legitimnacy, they included a pindpoIe ofsbazred growth. The leaders hoped chat rapid, widely shared improvements in economic welfare would bring legtimacy- Of course, feiw political leaders anywhere would reject, on principle, either the desirability of gwrowh or that the benefits of growth should be shared. 'W1hat distinguished the HPAEe' leadershp was thec extent to which, they. adoptedt specific institutional mechanismns tailored to these goals, and that workced. In this chapter we briefly review the history of the adoption of an explicit commitmnent to shared growth. We then dis- * cuss the main institutional mechanisms adopted to achieve this goal and how they were areated and mainmainedh * Wealth-sharing programs designed to indude non-elites in eco- * ~~nomiuc growth '157 4~~~~~~ ~C £ KlA CLE * A cadre of economic technocrats insulated from narrow political pressures o InsEirutions and mechanisms to share information and win the support of business elites. Achieving Legitimacy through Shared Growth T tHE PRINCIPLE OF SHARED GROVfH ASSUMED DIFFERENT forns in each HPAE It was most explicit in Malaysia. In response to racial riots in 1969, three dominant parties, each representing a major ethnic group, widened their coalition to indude other parties under a national firont. The coalition's goal was to share wealth more eq- uitaMly without stifling growh The coalition's solution was the New Economic Policy (NEP), an attempt to improve the lot of the Bumiput- era, the largest but poorest ehinic group. The coalition's own early de- scriprion of the NEP is quite explicir: The NEP has as its overriding objective the promotion of national unnity through the two pronged strategy oF. (i) eradicating poverty by raising the income levels and increasing employrnent opportunities for all Malaysians, irrespective of race and (ii) accelerating the process of restrucruring Malaysian society to correct economic imbalance, so as to reduce and eventually eliminate the identification of race with economic function.... The efforts to attain these objectives were, in -turn, to be undertaken in the context of rapid structural change and expansion of the economy so as to ensure that no particular group ex- periences any loss or feels any sense of deprivation. [Salleh, Yeah, and Meyanathan 1993] The principle of shared growth is also explicit in Indo- .esia, where President Suharto advocates "economic democracr. For example, in 1990 President Suharto himself strongly endorsed research results of the Indonesian Economists Association stressing the importance of sharing the benefits of growth. 'I hope that IsEfs findings on the sys- tem of Economiic Democracy receives as m.uch public exposure as pos- sible and that all levels of society become involved in the discussion of this vital issue. This will enable us to reach national consensus as we strive forward toward the full implementation of Economic Democ- : 58 ::- -: A N..:I NsTn' racy" (isEa 1990). Economic democracy advocates reliance on the free market to stimulate growth. But it also explicitly acknowledges the market system's inadequacies in distribution and actively seeks to re- dress them. In Taiwan, Cbina, commitment to shared growth grew out of the Taiwanese authorities' searching analysis of their failure in mainland China The leaders' five-point diagnosis included three equity issues: agricultural tenants had rebelled against exploitation while the party continued to identify with the landlords; labor unions had run out of control; and government was beholden to vested interests (Wade 1990). These realizations, combined with pressure from the United States to undertake social and economic restructuring, led the Kuomintang to adopt shared growth as the formula for its rule in Taiwan, China. Unlike the Taiwanese authorities, Korea's post-World War II leader- ship lacked a synchronized view ofgrowth and distribution. Even so, the sequence of policies it pursued, and the flattening of wealth resulting fiom the Korean War effectively produced egalitarian growth. Pardy by accident and pardy by trial and error, the leadership gradually adopted the principle of shared growth. In both postwar Japan and post-independence Singapore, growth with equity was the leaders' primary concern. In Japan, the war's de- struction of landed wealth actually helped the leadership to implement its redistributive goals. In Singapore lack of a significant rural sector and the absence of a landlord elite also made pursuing equity goals easier. But neither set of leaders was complacent; both adopted effective con- sultative arrangements between labor, government, and business.' Chapter 1 presents evidence that the leaders of the HPAEs have been unusually successfud in achieving shared growrh. Growth-rate and income-distribution performance measures show that the HPAEs signifi- cantly outperformed other low- and middle-income economies. How did they do this? If adopting the principles were sufficient, certainly achievement wouid be widespread. The answer must partially lie in the insttutions and mechanisms and in how they worked. To win the support of non-elites, the leaders of the I4PAEs introduced mechanisms that drastically increased oppormtnitues to share the benefits of growth. These mechanisms varied from economy to economy but in- cluded education (in all the HPAEs); land reform (in Japan, Korea, and Taiwan, China); support for small and medium-size industries (Hong Kong, Japan, Korea, and Taiwan, China); and government provision of '159 such basic amenities as housing and public health services (Hong Kong and Singapore). Nearly all HPAE governments walked a delicate line re- garding labor by limiting the power of unions and intervening to check labor radicalism, while at the same time encouraging a cooperative di- mate in which labor was rewarded for increases in producriviy. These wealth-sharing measures have differed from the typical redis- tributdve approach of most developing economies. Instead of granting direct income transfes or subsidizing specfic commodities (for exam- pIe, food or fuel), HPAE leaders have favored mechanisms that increase opportunities for upward mobiity. The frequent result is that indivlidu- als and families, provided the opportnity and convinced that efforts will be rewarded, srudy more, work harder, and save more. Universal Education Education is arguably the most important of these wealth-sharing, op- portunity-creating mechanisms. The provision ofuniversal primary edu- cation and wide access to secondary and higher education contributed substantially to opportunities for upward mobility. This mobiity in turn mirigated the feeling of non-elites that society is unjust and made them more accepting of the market-oriented policies needed to foster growth. In Korea, for example, the government invested heavily in the expan- sion of education at the primary and secondary levels soon afier the wan In rhe mid-I 960s, Park Chung Hee's export-promotion drive boosted demand for educated labor, providJing jobs for the first crop of graduates to benefit from the postwar education push. Equitable Land Holding and Land Refonn Theory and empirical evidence suggest that widespread ownership of land not only improves equity but also improves land prodlccrMry (Berry and Cline 1979). All the lPA.Es with subsEantial agrarian sectors have widespread land holding, resulting from either taditional owner- ship patterns (Indonesia and Thailand) or land reform (Japan, Korea, and Taiwan, China). In Malaysia, corporate-owned plantations have dominated agriculture since the colonial era. But with a relatively small dpopulaton and ample land, it has avoided equity problems common-to other developing economies with highly unequal land distributions. Hong Kong and Singapore have almost no agricultural seroc. x6o AN Korea and Taiwan, China, began land reform under broadly simnilar circumstances. In both cases, authoritarian governments &cing a com- munist threat were dependent on the assistance of the United States, whose advisers urged them to adopt more egalitarian land holding. In Taiwan, China, rhe government seized land from the landlords, com- pensating them with shares in state enterprises. It then sold d-ie land to the tillers at fitvorable credit terms and fivorable prices. The government then helpcd the tillers upgrade production for domestic and export mar- kets. The program worked, economically and politically. Land reform helped Taiwan, China, achieve one of the world's most equitable income distributions (Kuo 1976). Political stability benefited in two ways. Newly Ianded farmers, focused on boosting production, had little inter- est in radical activities. Former landlords, as new shareholders in state enterprises, had a vested interest in the success of the Taiwanese author- iies economic program. In Korea, land reform occurred in two stages. The first, initiated by U.S. forces in 1947, distributed the land confiscated from the Japanese at the end of Wurid War II to the tillers and put a ceiling on rents of other land. The second, begun in 1950 and completed after the Korean War, was undertaken by the Korean government after a lengthy debate in the legislatum The government took over landlord properties, paid the latter nominal compensation, and distributed the land to 900,000 renants, effectively eliminatng tenancy. Euuterpises Small and Medium-Size Are Beautiful Just as numerous small land holdings improved equityand efficiency, the HP'AEs benefited fiom a profusion of small and medium-size enter- prises (SMEs). The large number of SMEs generally reflected market forces rather than government intervention. But several of these economies supported SMEs with preferential credits and specific support services. Rapid growth of labor-intensive manufacturing in these firms absorbed large numbers of workers, reducing unemployment and at- tracting rural labor As firms shifted to more sophistcated production, efficiengr rose and workers' real incomes increased. Support for SMEs has been most explicit and successful in Taiwan, China. As shown in table 4.1, SMEs comprise at least 90 percent of en- - r-erpnsesin each secor. Not surprisingly, the SMEs also dominate the export sector, producing about 60 percent of the total value of exports r16i - g,sji~~srA1N I IAC CL E Table 4.1 SmaUl and Medium-Size Businesses in Taiwan, China --. ;e - ,.-; - -. -; ..--,, Smell .d medim--e ;-u.,.nt -s - . -i - -;f -- - 0E u~0--:- -: :-Tou - --tal- - -.Percentageof Cue.ory nipX X res Number0 : a -ni vi - rts ^Agri culur, fishing, livstock, forestry 3,256 3,171 97.39 ..Mning and:quarlying -1,390- - .1,360 97.84; MItanu-;ting . -157,965, 155,263 98.29:. ;Ee c,gis,nd r gm a . . .: 1t.. . 86- 168. 90.32 Constructon -. , 28,419 26,456 93.09. Commerces: 489,864 - 475,106 .96.99 Tmansporr,-srorngc,and ommunimcations 36,897 . 35,818 - 97.08-; Funancc,.i surancerel estatebusiness services 30,534 28263 92.56: ; --.:.Comxruunit$, social, and personal servic .69,550 69,229 ..99.54 -'; Total 0 =- :': 0 . . -- --: - ,: : 818,061- -794,834 ,97.16 Stmrn nTai China (1992b). (see tabIe 4.2). Other HPAEs have also encouraged small and medium- size industries. Japan has direaed enormous financial resources toward dcveloping small and medium-size enterprises. Public financial insti- rutions have allocated an average of 10 percent of lending toward SMES. During the rapid growth period of the 1 950s, as much as 30 per- cent of their total lending for fixed investments went to SMEs. The var- ious government-supported direcred-credit programs have proven particularly helpful during times of transition and rapid change (Itoh and Urama 1993). Although there are mixed evaluations of the success of these programs, one thing is dear. The SME sector has become an important cornerstone of Japan's economy. In 1989, SMEs accounted for about 52 percent of both manuficturing value added and sales, and their share of employment in various manufacturing subsectors Table 4.2 The Export Value of Smali and Medium-Size Businesses in Taiwaan, China (millons of U.S. dogrs) .,.i .,. .:: Auqon. -V -e~tkr.f g . Po 'M2d .436.89 J 64 -2.'...L.', ;,23525 - 42 674 - 3--- '-'4 ' ,: ~ OSrTIIJ hn 9b. . .~. _ - -' , .ni& _ i6z~~~~~~~~~~~~~~~~~~~~~~~~~~6 ranged from a low of 41 percent in transport machinery to a high of 100 percent in silverware. Korean development has been largely driven by the expansion of con- glomerates, the so-called chacbols But beginning in the early 1980s, the SME sector began to grow rapidly.2 SMES' share in total manufacturing employment rose from 37.6 percenr in 1976 to 51.2 percent in 1988; while at the same time the SME share in mnanufacturing value added rose from 23.7 percent to 34.9 percent (Kim and Nugent 1993). Korea es- tablished an extensive support system for SMEs. And, as in Japan, finan- cial support systems, such as export financing and credit guaree programs. have becn d-ie most imporcant. Housing: Successfully Targting Low-lncome Households Two HPAEs, Hong Kong and Singapore, intervened heavily in hous- img markers to win the support and cooperation of non-elites. By pro- riding low-cost housing for the majority of residents, both programs have helped to decrease inequality and minimize social unrest, thus pro- viding< the long-term stability attractive to investors. Moreover, the mas- sive construction effort created lobs when both economies faced higb unmemploymenrt subsequendy, the wide availability of low-cost housing for workers helped to hold down wage demands, subsidizing labor- intensive manufacturing. In Hong Kong, which has generally followed a laissez-faire approach to the economy, mass housing programs were a iesponse to a massive in- flux of refiigees and migrants from China. The rapid increase in Hong Kong's population from 600,000 in 1945 to 2.4 million in 1950 spawned slum areas, high unemploymenr, and poverry. These led to so- cial disturbances, which culminated in riots in 1967. To difflase the ten- sion and improve living conditions of the general public, in 1972 the govemment launched a Thst- track public housing program that has since expanded to include the construction of entre new towns outsidc the city propet By 1987, more than 40 percent of the population lived in public housing. Like Hong Kong, Singapore experienced a dramatic influx of mi- -grants in the late 1950s and early 1960s and faced similarproblems. Its expulsion fiom the Malayan Federation made matters worse. The gov- ermuent responded by creating the Housing Development Board (HDB) in 1960 to provide public housing to low-income f2milies. As these -63 -FIAA CL E needs were met, the HDB has turned to development of middle-income public housing and self-sufficient new towns similar to those in Hong Kong. Today 80 percent of the population lives in public housing, and more than 90 percent of the families in public housing own their units. While success in Hong Kong and Singapore has been aided by small ter- ritories and populations, the experience of Indonesia suggests that pub- lic housing carefully targeted coward low-income groups may have benefits in larger economies as wel (see box 4.1). The Labor Trade-Off: Cooperative Unions Get a Bigger Slice In Japan, Korea, Singapore, and Taiwan, China (and to a lesser extent Malaysia), governments restrucured the labor sector to suppress radical ac- -. '-V.- -A \,\4NC PIN# GECONONESf-EU-'E N,:i_OPULt-; MM. "F J ~ ~ ~ .3>titY$~ t d l y,W ojcalits RPe%mcn9do JU,SF c Urox iiGwn~~Skj&tiEboy, *~.~FL t ~L !A? ia.tj. 1t -, 1 new.urbanj ,gos aalHrnlaaf.ok unablc os afford, . .. . ,,peratlvece; isarbcle uron, lu; d.:-oLo6. i.>rohow nri r schaSuetck&^le conca4ctconianrsbdI'nolidal . NI- -4 &tr;$ - .$QyiTOi'eSpottiid2*¢it& ,dii,i potntil cisr;+f_,';i s,the lndoneslan 0E e 'i=rt%v(i shCa zi 0~~~~~~~~~~~~~' . .2 praflvhezoeinxa e4!zng.aa-WpdcaaunOp;pflennca oaraW-X;4:: .~~i a., 14PT 'jA'A ThdlW1I&wnroSfi mmp~~~~g V, mmnnns tt O' q own 4paymentt, .rp4 atM rE onrir.2~, 5L~; - .. .rcen . 4zh4overnnenthshepd,h ooen3 rop enu~1auz b rl __ 4' ,- ;OR¼S.. 01 1onap l mcwohWidin-ahd.s ao0O&n1nttl&n4 1c64 AN IN4S-TQ -* tivity in an effirt to ensure political stability. Governments abolished trade- based labor unions and pushed the creation ofcompany- or enterprse-based unions. Management and company union representatives were then re- quired to joindy formulate and implement work-related poliies. Labor movements in Indonesia and Thailand, while not subjected to systematic restructuring, were nonetheless routinely suppressed at the first sign of radicalism, primarily because of the governments' fear of communism. In Hong Kong, conversely, union radicalism has been held in check by a steady supply of new labor from China and by the in- fluence of Beijing, which has encouraged moderation in the unions under its Hong Kong umbrella organization. Governments in several Asian NIEs, particularly in Malaysia and Singapore, have modeled their systems of labor relations after Japan, where labor radicalism was quickly suppressed after World War It and labor-management consultations were gradually institutionalized. In Japan, these consultations have led workers to expect that adjustments in their wages will correspond to trends in the national economy despite the unions' lack of political weight. 'Labor has been compensated for its decreased political role through wage policies tied to increases in pro- ductivity" (Johnson 1982, p. 151). Like the government in posrwar Japan, the government in Singapore sharply lirmited union autonomy soon after independence due to fears of political instability Lacking natural resources following the separation from Malaysia, Singapore courted foreign investment in labor-intensive manufacuring by suppressing independent unions and assuring in- vestors industrial peace. To meet workers' rising expectations, the gov- ernmnent in 1969 instructed the National Trade Union Congress (NrUC), the officially recognized labor umbrella organizaton, to suggest new directions for the union movement. The NTUC, which is in effect a branch of the ruling People's Action Parry (PAP), made two recorimmendations: govemment, management, and labor should work together to avoid antagonistic wage negotations; and workers should organize into cooperatives with a stake in the econ- omy. To firrher cooperation, the NTUC established the trpartite National Wages Council (Nwc). Each year, the NWC reviews wage and economic trends before advising the government on wage adjustment guidelines. Although recommendations are officially nonbinding, the govemment usualy endorses them wholesale, and business and labor use them as wage settlement benchmarks. The NWCs effectiveness was evident during I65 . , 1, ,t77;tS>ri C L E the 1985-86 recession, when it persuaded labor to accept wage cuts of about 12 percent. The rAP also reorganized the NTUc itself, transforming it from a negotiating agent to a provider of goods and services. The NTUC has since established a number of worker cooperatives, ranging from dhe Fairprice Supermarket Cooperative, with more than 240,000 members, to the Comfort Taxi Service, with 5,500 members (see box 4.2). i-ONQO5tTHEE1R. FIRCOOPE ATVESORAGANIZD*.BYSINGAPORE'S K LW. ,>NtwOACoA't bf.lrodany, byOf&r( fridhbYii ari a I I ..... t f,ruung_. -r eforp1crncuorVrne a arry rJo 80.ebrrelxelw:pa 6N-ii!'Ode4if&1I te .Tdfoms;. iAti4niz oflisr int tt-. lcenseds n m,drNeriiBccause there west limirbon ZihcnumberErg4$-;$ ;$S-E wre&nn owners .were-ab1le to hA %L&AY.=u'.&O Ar'hAgsLen.fl O &sute htAusvresorred-ro dnyangiotr * q,qfgoods41n1.p !h 'j -s;- -ac,a.l 'kasb'Vid'e6B'6&f iservitlones, |cd h rt eranve tovbay4larg&numbers.ofwasxn dnL1c Uii dehWdiilcdt ,s e z6b-d_ sout__ro admzn, tamnve5 , . ;~~~cas.D ;.si tmhmr ~~~1miUtw Xonda; toso&.faealtrncensei unDedarUG. id.@ 2 *~~~~ g 30 X123 ,T.4 ,5r S 25R"-iTr7:aCPVSi:*' rtt''''';>r*CA -RX 'SgBgWtt f- I;" .,i.. 2 M os D;9iS!si i 3'r-7.aWef . ,5i2&~o Zrix ,.Kt,> r,=LTz.eiC nd.soton uceSE oarEs;;2ancasc. ,C , N 6 A ;~~~~~~~~~~~~~g . .:. -~ ~~~~~~~~~~~nf .wn . Ei .ii - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - .~ _' : -:" .. - ..- :s~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I Other economies have forestalled development of large and politically powerful unions with less formal arrangements. In Taiwan, China, for example, close managemcnt-labor contact in small and medium-size en- terprises reduced workers' interest in union representation (see table 4.3). Moreover, the emphasis on labor-intensive manufacturing generated rapid increases in the demand for labor. Consequendy, as figure 4.1 il- lustrates, from 1952 to 1979, inflation-adjusted basic wages increased more than sixfold in manufacturing and more than fourfold in agricul- ture. The total increase in compensation has actually been higher, since basic wages are supplemented twice annually with profit-sharing bonuses. Insulating the Economic Technocracy IMPLEMENTING SHARED GROWTH TO SOLVE THE PROBLEM OF political legitimacy requires sharing but it also requires growth. We now discuss HPAE institutional traits that have been critical to achiev- ing both these goals. Foremost among them is technocratic insulation- the ability of economic technocrats to formulate and implement policies in kceeping with politicaUly formulated national goals with a minimum of lobbying for special fivors fiom politicians and interest groups. Without i, technocrats in the high-perforningAsian economies would have been unable to introduce and sustain rational economic policies, and some viral wealth-sharing mechanisms would have been neutralized soon after their inception, as was land reform in the Philippines (see box 4.3). Table 4.3 Labor Union Activit by Scale of Business in Taiwan, China (pnenetz ,p- *-3,--es No . -i-:;. -o jo.itX6 Nor renponre 4:i30-49 f 4 .-. 9: 06.M 008 9 29S . .85. . .; -f67- So28 015 ,a 51:05 .44 . .51 0 , -2 - .t 19.45 :X05 . 6.26i_' r T' ..37 r'- Sourer Taiwan, clnnChial Bure of rheTii _Proviial Government, Annual tSt ~ & '980-84 : 167 Figure 4.1 Index of Real Wags 1952-79 Percent (IndEx: 1952;100) I I ' ~~~~~~~~~~Murnsfacturlng 300 o;#{~~~~~~~ ii. I l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ :.: - 400 - glAtr 200- .:: 0~~~~~~~~~~ Fue41Inv of RelWgsI927 p1952 1955 1955 1M1964 1967 1970 1973 1976 1979 SPrcent , Kuo, and Ranis (198 1, p.2). Direct insulaton Mechanisms Hong Kong offers the cearest case of insulation, since the colonial government, which has been strongly committed to free marker forces, need ior depend on the political support ofprivate interests to maintain power But insulaion does not necessarily diminish as democratic insti- tutions proliferar. Indeed, the technocracy in authoritarian societies -may be utterly without insulation, while those in some democracies are heavily insulated from outside pressures. In Japan, insulation chacacterizes not only the economic technocracy but nearly the entire bureaucracy. Japanese bureaucrats draft laws in consultation with the policy committees of the ruling liberal Democra- tic Party and private sector representatives; the diet (parliament) merely ratifies what the bureaucracy has prepared. Because the bureaucracy has independent power, it can often ignore pressure from individual actors in the private sector. Moreover, rhe bureaucracy can achieve policy goals through administrative guidance-nonbinding recommendations. Guidelines ale enforced with generous incentives-such as licenses and ixG 0 ~~~~~~~~~~~~~~A N- -INS . ~ ~~~~~~ ~ . . k t.di i h .u' . .____.~~~~~~~~~~~~~~~~I ::~~~~~~~~~~~~~~~~~~~~~~I: ::: 4W " -: EP,Pr u G 1;ivestex n6nLY:Ji CVt'i: 9 - L c 3 utyt~~~~;g ,psuN'_O ,i dgm :~~~~~~~~~~~~~~~~~~~~O C e . u~~~~~~~~~t)r,, V ormmg c Ico, 0 -ws C 0:00000~~~~~~~~~ICR 3986 :P3p -~~~~~~~~~~~~' a an rcotan aswep 0.''0 .~~~~~~~~~~~~~~~~~~~~~r .f ' .wLtS~~~~~~~~~~~~~~~~~~~I - 3 f : ' -:~~~~~~~~~~~~~~~~~~7 } ,-. :g_D . - D~~~~~ok , X ; 0~~~~~~~~~~~~~~~~~~~~S . f , . : . .t: = f f - : -- - . D6d .t- : D r r ;0~~~c-un0000.;:0-0--0I6.: At 'R' CL foreign exchange allocations-and the implicit threat to withhold them from companies that refuse to cooperate. One of the distinct features of the b'ueleary process in all the HPAEs is the primacy given to the bureaucracy over the legislature in drafting laws. Though the.legislature must approve the laws, the bureaucracy studies, analyzes, and drafts the bills. It has considerable control over the agenda and can use this to minimize political pressure. IThejapanese bureaucracy is further protecred fiom political pressure by the Natonal Personnel Authority, an independent body that sets the bu- reaucray's pay scales and promotdon policies, administers civil service exams, and makes most appointments. Japan's prime minister names only his ministers and, except in a few caes, one of the two vice-nunisters in each ministry; the National Personnel Authority is responsible for the resrt3 As in Japan, the bureaucracy in Korea has a long tradition of inde- pendent policymaking. This was boosted in 1961, when army General Park Chung Hee seized power and reorganized the govemment to pro- mote rapid development. He createa a powerful Economic Planning Board that has had broad budgetary authoriry. Acdministraive control over the banking system has given the bureaucracy strong leverage over big business, which relies on government-backed financing to fund rapid expansion.4 The partnership has generaly been uneven, with the bureaucracy dominating the relationship. Whe few of the HPAEs have such thoroughly insulated bureaucracies as Japan and Korea, each boasts at least a small, ideologically consistent technocratic core that answers directly to the top leaders and therefore has some independence from the legislature and other sources of politi- cal pressure. Indonesia and Thailand, where bureaucratic traditions more dosely resemble thse in other developing economies, are instruc- tive on this score. Inflation Control and Debt Management in Thailand and lIndonesia For historical reasons, Thailand and Indonesia have been pro- roundly concerned with low inflation and debt management. Thailand is unique in Southeast Asia for being the only economy in that region not to have been colonized, in fact one of the few in the.developing world. One reason for this remarkable alchievement was the openness of the Thai Kingdom to trade with foreigners. Another was the monar- chy's concern with avoiding indebtedness to foreigners. And a third, - 170- A IN.r:a.n.s-ANNm. and perhaps the most important, was the establishment of a bureau- cracy in 1892 that enabled the monarchy to centralize political control over the kingdom's many regions. Historically, then, the bureaucracy in Thailand has always had a powerful influence over public administra- tion and policymaking. But it was-not until 1957, when Gcncral Sarit took over the reins of government, chat bureaucratic reforms were in- troduced to address macroeconomic imbalances. Since then, keeping inflation low and efficientdy managing debt have been among the gov- ernment's key concerns. These same issues have been extremely important to the Indonesian government since President Suharto took power in 1967. The previous Guided Dem'ocracy regime of President Sukano had let goverrnent spending run out of control, leading to annual inflation rates of 1,000 percenic and other serious economic problems induding a rapidly dete- riorating infratructure. Suharto's New Order govemment ascended to power with an explicit conmitment to fight inflation and develop in- fratructure, particularly in the rural areas. This led to passage of a bal- anced budget law and to financial controls that addressed infrastructure d development. To achieve the rwin objectives of low inflation and prudent debt management, both governments created mechanisms tO insulate their -: .economic redcnocracies. In Thailand, the government's Budget Bureau has tight control of the budget drafting process and has maintained a stable exchange rate and low inflation. To draft the budget, the Budget Bureau consults with the National Economic and Social Development Board about proposed public investments and with the Finance Min- istry about expected revenues. It then deternines together with the Cen- t tral Bank how much deficit financing the economy can tolerate without increasing inflation. Having detcrniined aggregate allowable expendi- ture, it calculates how much each government agency may spend and :forwards a broad outline of proposed expenditures to the cabinet. The cabinet may propose changes to the outline before forwarding it to a parliamentary budget committee for evaluation. But since neither the cabinet nor the committee receives budget details, the bureau's pro- posals usually survive with few alterations. After approval by the com- mnittee, the budget outline is submitted to the endiic parliament fior routine passage. Throughout, the bureau draws support from budgetary laws that -'-'lUriiit the government deficit to a small percentage of the year's total ex- '7' V , MRCLE penditures and that cap the percentage of the budget that can be spent servcing the foreign public debt (the current cap is 13 percent). Parlia- mentary rules offer a fiirther unusual guarantee against runaway deficit spending nonrninisrerial membe of parliament on the budget com- muitte may propose- only minimial changes and in any event may nor re- vise the budget upward. The results have been impressive (see figure 4.2). Except for the diruptions caused by the 1972 and 1979 oil shocks, the annual inflation rate has been dlose to 5 percent. Few. developing eoinoies can lay dlaim to such an achievemet and even more ad- winced HPA.Rs such as Hong, Kon-g and Korea have had higher rates. This demonstrates how a small technocratc core, insulated from politics, can set a positive tone for an entire economy. In Indonesia, too, technocrats keep a tight rein on the budget. Under Suharto, a balanced budget has been the cornerstone of government 6i- nancial policy: In 1967, the legislature approve-d a law limiting expendi- ares to domestic. revenues plus foreign assistnce. Since then, he finance ministry has institutionalized a review that requires each min- istry to justify prposed expenditures on a line-by-line bAsis. In addi- Figure 4.2 Thailand Consumer Price Index (percentage change from previous year) Percentag change 30 -- 20-- 10 -i' k.ii---- 4j If spend ~ing nminisr Iemer of priaet onIh ugtc 1955 1958 1961 1964 1867 1970 1973 1976 1979 1952 1955585 1 &unr Innualtional Moneraty Fund da.el r72 S~~~~~~~~~~~~~~~~~~~~~~ N ST dion, parliamentary rules restrict the legislature's discussion of the bud- get to broad policy issues. Armed with thesc instruments, the finance ministry has established a macroeconomic environment favorable to growth. Inflation has dropped since the 1960s and has generally re- mained low and stable (see figure 4.3). Indirect lusulation Mechanisns Some HPAEs have enhanced bureaucratic insulation indiiecdy and gen- erally inadvertently. For example, wealth-sharing mechanisms reduce the potential gain to interest groups of intervening in policymaking and im- plementation, since some share of the firuits of growth has already been as- suled. In addition, when wealth-sharing mechanisms reduce the size of oroduction units, as with land reform and support for small and medium- * . size industries, the marginal cost of organizing and lobbying increases. 'With the benefits of lobbying down, and the costs up, interest groups tend to be more willing to leave the policy process to the technocrats. Reozganization of labor from industrywide unions into company unions, along with labor federation provision of goods and services, has Figure 4.3 Indoneia Consumer Pnce Index (perentage changefim previous year) Percentage change - 0 f -'-N- 30 --rv-- 20 -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - . ; 0 I I I A | ~~~~~~~~~~~I 1 j I - ~01970 1972 1974 1976 197t1 "SO 1982 lS84 1986 1.98S 1990 0~~~~~~~ ~ ~~~~~~ f - - I 40 l- 5 eItional Monc4y Fund daa . . 0 - :n h . _ '. _ -17_ 197, 0 197 2.97 197 29: =8 ;8 .98 196 819 ;our.neranaa Mocx Fun darn00. , similarly reduced the marginal benefit and increased the marginal cost of collective action. Thus, in contrasr with wokers in many other developing economies, workers in the HPAEs arc more lihkly to refrain from work stOp- pages and other disruptions and fiom lobbying the government for man- dated wage increases. Because employers faced fewer demands fiom labor they, too, have had less incentive to press demands on the reclnocracy. Of course a powerfil bureaucracy insulated fiom external pressure can also be dangerous. Indeed, the description ofjapan's bureaucracy- an independent force that controls generous incentives and issues vague guidelines-is nor unlike that of bureaucracies in ocher economies that engage in corruption and malfeasance on a vast scale. In Japan, itselE some ministries are perceived to be captured by the interests they regu- late. These ministries enjoy less power and prestige than those that resist politicization.5 Even well-intentioned bureaucrats may become counter- productive or lose sight of the goals they are supposed to pursue if bu- reaucratic insulation means that they are not held accountable for their decisions. How have the HPAEs kept insulated bureaucracies honest and responsive to legitimate private concerns? The next two scctions address these perennial problems of bureaucracr. honesty and responsiveness. Building a Reputable Civnl Service While insulation of the techmocracy may be necessary, it is hardly suf- ficient-in the long term. To sustain growth, a bureaucracy must have the competence to formulatE effective policies and the integrity coiimple- ment them fairly. The more policymakers attempt to fine tune the econ- omy, the greater the need for competence and honesry. Among the HwAEs, Hong Kong, Japan, Korea, Singapore, and Taiwan, China, have been successfiLl in building relatively competent and honest bureaucra- aes. The Southeast Asian NiEs, Indonesia, Malaysia, and Thailand, have gradually introduced measures to upgrade theirs, with Malaysia in the forefront of the process, but they still clearly lag behind more industrial- ized economies. Despite the stress on culture in much popular writng about the Asian miracle, the most successflil bureaucracies have not relied on cul- ture alone. HPAE bureaucracies have employed numero us mechanisms to increase the appeal of a public service career, thereby heightening com- petition and improving the pool of a1plicants. The overall principles of these mechanisms, readily applicable to any sociecy are: - '74 ANN I NSTUT * Recrauitment and promotion must be merit based and highly competuve. * Total compensation, induding pay, perks, and prestige, must be competitive with the private sector. * Those who makc it to the top should be amply rewardecd Merit-Based Recnuibnent and Pmmotion. Recruitment in Japan revolves around highly competitive cmil service examinations administered each year by the National PersonneI Authority. For the higher-level exams, the rate'of success has been less than 8 percent. Indeed, in 1980 the av- erage rate of success was 2.7 percent (Kim 1988). Promotion within the bureaucracy is based on a combination of seniority and a host of perfor- mnance indicators that dif&r across the ministries. Because the number of personnel in the bureaucracy is fixed by law, compention for promo- tion can be intense (see Pempel 1984)!' Korea also relies on exams and performance but has supplemented thcse filkers by,,recruitingheavilyamongacademics with advanced degrees. (Box 4A describes the turnaronund in the Korean civil service in the 1960s.) This has resulted in a prolifcration of research instiutes associated with ministries. Since die Economic Plamning Board established the Korea Development lnstiute in 1971, most other ministries have fol- lowed SUiL Most institute researchers have a PhD., many from a re- spected foreign university (im and Leipziger 1993). Like Korea, Taiwan, China, recruits heavily from academnia, primarily to offset weaknesses in the civil service exam system. Academics are recruited from major univer- sities, usually for fixed periods ranging from three to six years. In Malaysia, civil service exam results aresubject to affirnmative acion guidelines meant to increase the number of Malays in gaovement. Be- cause this reduces the pool of eligible applicants, it probably hinders de- velopment of the bureaucracy Incentie-Based Compensation. In bureaucracies, as in nearly everything else, you get whar you pay for. Table 4.4 suggs that relative pay is sig- - nificandy berter in the Four Tigers than in other economies. Relative pay in Mvialaysia and Thailand is about the same as the average for other low- and middle-in&me econo-mies but is brill signEficantly higher than -in the Philippines, which is widely perceived to have one of the weakest bureaucracies in SoutheastAsia. In general, the more fa-vorably the total public sector compensation package compares to compensation in the private sector, the better the quality of the bureaucramcy '175 2 ~~~~~~~~~~~~~~~~~, L~~~ Wotsurp'sne 4ingapre,uowhidii 'widely' peroivd' tohv> he ....~~~~~~~~~~~~~~~~~~~~~~~~~~~ 'e.:. bCS 'fie ontlybaejmlayofacputHministercvdin ingapordranws d:o .~~ ~ ~ ~ ~~ ~ ~ ~ .~ .4 tyJQ.y~i*S ~ t trta r'A)l4A' JapanF.t anfiS Korca alsomke^6 a, dcieatem; - effor*;t >Jto<> match the p rivate sector I Japn,o£teNationaul Pcrsnne icAuthmForityHo annuay sur y Korea basc sldsaelwrthan salnre.otarein lare pivaslrbed - copnis bu W-pine aflownrs hlp to narrowr th;e ; *iffcen- Tain acon of salary plus* bonuss in thlet pvteo seaorzan sahlary plus hatl towne in ~d. . ;1 tNrnsnorecdRBErmwt1 an oinon thepulic sectorwforsentry-level, m_id-l n posi_-io dif- gin's mtost compentsan uaprihtabreacracyb pmays itasn iburacr b7estT thly base salayoaflmister iSngorerang S$22, 100 ;3iF to S$7,2 (aou US$i3,812'l_e9 to s US$1,390) whl ..a n Japan ndKreaalo ak dliert effort tomathte privaste s seconn Jeapan th@e Nationalt Per,.sonnel kAu*thoit annuallyS*,-,{F.4 surves thu dso o mpanis toderin :job pay rates$ (Kimo 1988 p.) In Koea bs p;saries arcloer than saares in large pr,;iFitvate cmpaies.x ofi salar plus onuss n te riae sector amnd alrypls a llowncsin -~ ~ ~ ~ ~ ~~~~~ne in totasr nly compnsatione appear to biel smlerie tOhane in most 176-go' otcmern n pgtbracaqspy t ueurt A 1NSTwt:ANWpNS Table 4.4 ICP Estimates of Per Capita GDP and Raftio of Public to Private Sector Salaries, 1992 (penent) Seir leve 'Mddee GDPpacai &ononq/regran A B~A _B-A ;Ennyluas HPA&; ,. f.4i. -a), ';f - i ; *; apo * r; - n4:b 114. 1150 107-'. - . 7. 920 :Korea, Re.. of -.,':o - 98 . 8 8 .3 822 83.6 -710 ; Ta a China M ;65.2 *.; 60.3 63.5 - 65.8 : 6 0.0 7,954 0LMa1-aysia 40 33 343 -5001 - 5oo Thailand. 0 f; :- . : X . 47.G.6-; . 0 -2- i7.50 . -; - . 33.6 -43.0 9 4418.0 0 - .Philipie i277 L2r;4:;.3 25.0 325 62. 232 ;Tind-anToa, 63.5b - _. 63.5 76'9 7 779 8,51 ;0 ° ~~2tmgu ; - - 3 007 1P 3 7.1 60.0 :---06 A- ;~aLua 2416 24 25.0 2 4 - 4 . ,3 SnbSsjz70Afia 6-g, Somi- a ;. 11.2=, -' - 2 122r- 3 5 ;wCili toth UN4. Ineraoa ComparZion Pr e. ,. reth to ' i- ernational ! ;. - a. -LEsciae indude alloaneos and boiisetDtarnibrpriarteseor comnpmsdoui ae ased in asurveyof impranis with 500or more;- 2. i~ '. lAor on smalrilab, average compdranontn u iconomige te o (o public sector. In other HPEs public sector base salaries are consistendy lower than :-their privare sector counterparts (table 4.4).7 W7hile figures were not available for Indonesia, anecdotal evidence suggests tat te private-- : :tpublic wage gap tre is the largest among HIAEs. A stdent graduating - from an Idonesian university can expect to earn t o to four times as much in the private sector as in a government job. Nor surprisingly, c frsle Som the technocratic core, Indonesias bureaucracy is perhaps the - j least sell regarded among HPAES. In economies where public sector wages are at least decent, prestige, c job:seciny, and other advantages of public employment will persuade - - - - : : - -:: r~~~~~~~~~~~~77 cas cll reare amn H -=AE-s. se y, and other 2civa oDf pu plymn wil peruad V ItALE some talented individuals to forgo higher earnings in the private sector. The trick for governments is to hit on a combination that will attact competent individuals to the ci-vil service. Prestige is of course partly a function of culture; thus East Asian cultures that regard dvii servants highy have an advantage in building bureaucracies. Confurcian thought, with its veneration ofscholars and pre- occupation with written tests, especially civi service examns, remains a pow- erful force in Hong Kong, Japan, Korea, Singapore, and Taiwan, China. Not surprisingy, these societies have produced strong bureaucraces. But bureaucracy building need not be culturally constrained. Merit- based screening with economic incentives to attract a large pool of applicants can produce high status for civil servants in any society. Pro- vided that the exams are sufficiently competitive, individuals who pass them will be relat-ively rare. Thus passin the examn or winning a com- petitive divi service promotion becomes similar to being inducted into a hail of fame; it signals the public that an indLividual is special. 'While the value of prestige cannot normally be quantified, the monetary value accorded to some Thai royal decorations, described in box 4.5, suggests that the value of such pre-stige can be very high indeed. job security can also offset slightly lower pay. In most HPAE bureau- cracies, dismissa is unlikely unless the bureaucrat commits a serious mistake. Security of tenure translates into lower income variability, which in turn provides incentives for public- employees to accept a lower salary. Furthermore, in many economies a public employee can look for- ward to a retirement pension, a benefit not normally available in the pri- vate sector except in large corporations. A Clear Path to the Top. A final requirement for a successfuil bureaucracy is a well-defined, competitive career path with a substantial prize for those who make itto the top. In Japan and some other E*AEs, retrement comes early, and the rewards to a successfUl bureaucrat are substantial, extending beyond the pay; perks, and prestige to indude a golden parchute. The Japanese have a special word for it- amakuckr4 which means "descet from Heaven"' and connotes a lucrative job in a public or private corpora-. tion or occasionally election to political office. Similar practices exis in Korea, Malaysia, Singapore, and Taiwan, China, in differing degrees. Unlik the much-criticized revolving-door, which whisks bureaucrats in other economides fiom the pubfic sector to the private sector and back again, descent from Heaven is a one-way tp. Moreover, rethring bu- reaucrats in Japan do not choose their sinecures but are assigned them 178 0~ ~ ~ ~ ~ ~ ~~~~ ~A IN.ST-W A NijN ~~~~~ ,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - B The Iu1e.f i &A, NXA nL~ : IIIGHLYCOMPErITJV& MESBASED iR:ACkAUnENrTO` A BUW- -auaa creaeswuciors Ilentrusns%lenlowingimem-w.?- , g E6~~~at fieai o r . | - r - ftheel - cr the witha'L, r -~~~~~~~~~~~ruCG-2~eP roc=on Dbutmntiad-pr MouShaMW no noiv;Enr Be--'-4 3'- ra;cmtomsfo:grslcne unnv,zes;2ix cra~tcsanz espnt de4Ff*1, iWZIi& rbiidMdZl2& ~ 6n,isui, -. 1,4u2 ~ ~ ~ ~ - i` :2O to int.scl cotxmpedarive vimnmcnt idii vatue beisreweeninstrvies,i faos Whl thcaw re is n denying whohatv firns welcome etredseniceor bu- reuras partyo b strs~eopsilngthe pusfultiesih gzovenmersternno aafti naueo h ueacaymasta the rtirees LiPare ofe except l mana le iteia vy e e CompetemncBmet~ids nbrityi gofclmgi by a acommite wthatincthei cminitrycThese faturs, combined wihoanesy Ci ompetitivniscourgesenvironmesny witein anbtworyof e ministries, guard againrs firsq o iffern apst-rticrement jobs cin etis hae for specia faos Whil there is no deyn tha firs elom rtiedseiob crrcntueo h bureZaXi oaucrac mens that the reatiree 'i,sg ar ofteni2/ exceptonal mnagers v \alubl in theivr own t9ightF,.4^S-t>> Research group (Outside lcturers) ( Subcommittee) Otdelcue; {eedbacM (Briefings. subcomnittees' reports) .~~~~~~~~~~~~~~ -|Deliberativecouncil Publec r Ationr - Public Aions (Report) ---- -* - Explanatory mngs - Lectures -Others S&urn-. Ono (1992). : X . ' - . -~~~~~ In Korea, from the mid-1960s to the early 1980s, goverment and private sector relations were also dlose and cooperative, although some in the p'rivate sector argue that goverment was coo strong handed and dictatorial.10 Goverment and business leaders met oftn and regularly although less formdlly than in Japan. Government solicited businesses' views and included them as a critical policy component. Until the early 1980s, the most imporgnt communication channels were monthly rexport-promoton meetings, at which the president of Korea himself presided over discussions berween the economic ministers andictop business leaders. Businessmen expressed their views about mar- ets, regulations, and potential plans. The president: then instructed spe- cific ministers to attend to each important issue; at the nt meetng, * mnUnisters dlivered a progress report on their assigned tasks. These meetings were supplerm ented by goverment-initiated dis-of sion groups involving company managersi middbe-level government of- -ficials, and experts or scholars. As in Japan, there were basicaluy two types: fumnctional (for example, taxation) and sectoral (usually focused on a particular industry). In both, bureaucrats invited incerested partes and cxperts to present their views. The focus was on information gath- - eren bureaucrats made the final decision. E vernmsonoicitied oin- -cor poate al viewpoints in policies. 'With Koreais gradual democratization since the raid-1980s, irlations between goverinmet and business have become more distant and the meetings less fprequent. Recently, however, as the conomy has slowed, there is a growing consensus that monthly meetings should resume. 'Similar consultative mechnisms appear to be evolving in Malaysia. From independence until 1970, the government focused on infrastruc- tueand generally Left the private sector to fimction freely. From 1972 t'o 1985, as the goverment responded to the racial drots-of 1969 wich the N ewv Ecovrnomic Polncyd relations between govenmentand bness tended to be contentous, in part because of the requirement that asl en- terprises have at least 30 percent Malay participation. Since 1986 business-government relations have gradualy imprOved, culminating in 1991 ind he New Development Policy- (ND), which has struck a more Ne cn balance between growth and distributon. For example, uner c venalac beCe rwhaddsrbnn o xml, under the NEP, all firms with more than M$250,000 (US$104,167) capitlization --were required to comply with ethnidcy-based shareholding rgulations. Under the NDP, the cut-Off point ha been raised to M$2.5 milion (US$1.04 rmillion). 183 The NDP specifically seeks to emulate Japan's government-business consultative mechanisms, a determinat-ion reflected in Primec Minister Mahathir Bin Mohamad's 'Look East Policy." Formal consultations began in 199 1, when the government invited capt-ains of industry to dis- cuss the national budget with finance officials in a much publicized budget dialogue. In thessame year Mahathir aeaaed the pinnace of the conulttiv stuctre, the Malaysian Business Council, which'i chaired S~~~~~~~~~~~cni taiv stmcFu I by cth pruie minister and includes more than sixty members from in- dustry, labor, and govermennt. Thailand has been slower to develop consultative mechanisms. In 1983; the govenmment established the National Joint Pubfic and Private Consulkative Committe (NJPPCC) with the prime munister as chairman and the head of the plnning agency as searetary general. Othmer mem- bers are top government officials, particularly from the economic agen- aies, and nine private sector represenmives dmawn from the Thai Chamber of Commerce, the Association of Thai Industries, and the Thai BankersAssociation. Provincial consultative comrnitees have been established to supplement the NJPPCC (Samudavanija 1992, p. 26). Views about the effectveness of the committee differ widely. Public-private consultation is most cxplicit in Singapore. The private sector paricipates in policymaking in many ways. For example, private citizens serve as directors on government statutory boards and as mem- bers ofad hoc government advisoy commirtees. In both capacities, they review policies and programs and make recommendations for official considemraion. In addirion, the government regularly invites chambers of commerce, trade associations, and professional societies to subniit their views on specific issues (Chia 1992, p- 11). The Singapore ents efforts to manage dhe economy are most. fully apparent in the National Wages Council, which indudes rp-resenta- tives fiom govenmen; business, and labor Because of this tripartite struc- tumr, the Nwc fulfills several coordination funcuons, simultaneously fiurthering the govenment's guidance of business and of labor, as wel as business-labor cooperation. Having rejected more typical denocratic ifrms as too adversaial, the PAP has also etablished a wide variety of citizen asso- ciatiorLs to provide channels of communicanon with the government Formal mechanisms for business-government interface are almost en- tirely lackingin Taiwan, China. Instead, coordination is handled tirough large public entrrise that provide basic inputs to manu £cturers and through enterprise owned by the Khomintang the dominant political 184 AN NST .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- ----- - --- parry.-1 Indeed, public enterprises, broadly defined, account for a larger share of GDP in Taiwan, China, than in Korea or even India, where cen- rual planning plays a key role. These public, quasi-public, and party- owned firms give the government a high degree of control over large-scale sensitive investments and a means to influence the generally smaller firms that oomprise the private sector. Over the years, the large role of the pub- lic sector has become less obvious, as some parry and govemment firms were sold to cooperative private interests. Coordination is also facilitated by middle- and lower-level officials of the Industrial Development Board, who make routine, infornal visits to thousands of small and medium-size enterprises within their jurisdictions. Like Taiwan, China, Indonesia lacks the formal government-business coordinating links of other HPABs; this may be pardy explained by the weakness of its bureaucracy. Coordination appears to be handled by in- fbrmal networks linking senior officials with major enterprises.'2 Little is known about these networks, but they appear to have produced a high degree of cooperation in the economy. In contrast to the deliberation councils, where rules are more transparent, these networks are more sus- ceptible to captLre by participants and to corruption. Deliberation councils natually reflect the history and culture of the society in which they operate Even so, the experience in the HPA:Es sug- gests that their applicability is not limited, as some analysts may suggest, to the Conf&cian cultures of Northeast Asia. Because the operation of deliberation councils has been so importanr in winning the support of business elites, we now look in greater derail at the operation of these councils in two very different economies, Japan and Malaysia. Japan's Councils: e Foreunnes The concept of deliberation councils in Japan is most closely associated with Mm. Because of m1Im's role in developing and implementing the economys industrial policy, the public-private consultative process was most dearly manifested in the councils attached to it. Government-business collaboration has long been a hallmark of the Japanese economy. The zaibauu-Japan's large corporate holding groups-were dosely linked with the military-bureaucratic complex of the 1920s and 1930s and were a key element in building the economy's industrial and military base. At that time, the influence of big business matched and sometimes exceeded that of the bureaucracy. The U.S. occupation forces, ideniliing the zabatiu as a key element in the build-up ofJapan7s wartime machiney tried to limit their power, 185 = - . i QSI.YMiI R-A C L £ in part by conferring more power on the politically independent bu- reaucracy. Ironically, this not only strengthened the bureaucracy- particularly MITT, which became the center of developmcnt policy formulation-but also faciliraced the resurrection of industrial holding groups, organized this time around banks (the so-called kid'mu system). These new groups becamc the main. players in the government-initiated consultation procss. The principal instrument of the bureauicracy in managing this process has been administrative guidance, the use of threats and rewards to induce privatc parties to follow or implement measures suggested by the bureaucracy. Administrative guidance does not depend on any law or regulation. Rather, it employs eidter threats of adverse consequences should a private party refuse to conform to a 'suggestion or recominen- dation" or, alternatively, rewards dhat can be granted should the party concede,I MITE is well kcnown for having employed this with great skill. Until the mid-1960s, administrative guidance was a powerfil4 tool for Mm bureaucrats. Mm had firm control over the allocation of foreign ex- change and over importation through the Foreign Exchange and Trade Control Law. But intermational pressure, combined with the need to maintin an export-based economy, led the government to gradually lib- eralize the rade and investment regime. The process began in the mid- 1 960s and continued till 1979, when the Foreign Exchange and Control Law was dismanded and the Foreign Capital Iav liberalized. Of course this meant that MmT was losing key components of administraive guid- ance. By the late 1970s, guidance had taken a distinctiely different form-mergers of industies spiced with financial support from the government-and coercion declined. Ealaysia's Councls: mnulalingJapan and Korea. Malaysia's experience with public-private cooperation through ddiberation councils has paricular rel- evance for developing econqomies. UnlikeJapan or Korea, Malaysia is a mul- tiechnic society, and unlike Singapore it has a eladvely large population and land area. Thus Malaysia is in many ways more tpical ofa devloping econ- omy than are te high-perfomiing economies of Nordteast Asia. When the Malaysian govermment introduced foirmal public-private cooperation in 1983, it embodied-the concept in the slogan "Malaysia Inc.,' an overt attempt m emulate Japan Inc., partcularly the dose rela- rionship between Japanesec ministries (especially )and big business. In Malaysia, officials have applied this concept to their quest for growth with equiit, a particularly important goal in a multiethnic so6ey- -I86 :.; AN INS' Before 1983, Malaysia had some consultative panels within min- istries that dealt with specialized issues and were concerned mainly with cutting red tape. Many had litde private sector participation; those that did, however, produced concrece results. Since then, more deliberation councils have been formed, dealing with broader and more complicated issues (the budget and trade, for instance). One of the most ambitious, the Malaysian Business Council, which was formed in 1991, resembles Koreas highly successful Export Promotion Council. Although it is too early to know how effective these new councils will be, thev would be hamstrung without an efficient and rcpurable civil service bureaucracy. The government recognizes this and is trying to in- crease the bureaucracy's efficiency, eliminate corruption, and educate * . civil servants on the value of cooperating with the private sector. Why Deberaton Councils Assist Growth From an economic standpoint, deliberatve councils facilitate infor- mation transmission- They enable the bureaucracy to gadter infbrrna- tion about world markets, technology trends, and the impact of regulations domestically and abroad, synthesize the information into an action plan, and communicate the plan back to the private sector. Self- interested behavior is more or less constrained by the repeated nature of the collaboration. This in part helps establish credibility-private sector participants believe cheating and reneging are less likely. Politically, these councils serve as proto-democratic institutions, providing direct channels for big business, labor, and academnia to the seat of power. Fi- nally, because the rules that govern an industry are effectively esablished - rwithin the council, every member is assured that the rules canolt be al- tered arbitrarily. With dear rules established, members can concentare on market competiton and not worry about others tring to curry spe- cial &ivors from the government. In effecr, their share of industry profits is determined more through competition than through rent-seeking. In this sense, a deliberation council can be viewed as a wealth-sharing mechanism. : . T . , - -. Studies of the successful Asian economies have fiequently empha- sized the authoritarian nature of their leadership. Leaders of these soCd- - - eties are described as benevolent dict;ators who, in contrast to most other- - - 1 0 0 X H :-: 0 - - : - . I~~~~~~~~~~~~~~~~~J87 - - - ~~~~CL E dictators, uscd their authority to musde businessmen, workers, and bu- reaucrats into a single-mLinded pursuit of development. This chapter of- fers a different interpretation. Leaders need the active cooperation of their peoplc to stimulate and sustain growth. Heavy-handed treatment might subjugate the ruled, but it would also increase the perceived risk of exprpriation, driving away capital. Lack of capital would slow growth, in turn driving away skilled workers and entrepreneurs. Thus, even authoritarian leaders cannot rely on coercion alone Rather, all leaders need a strategy that will induce people to participate willingly in growth. Whether by design or accident, the leaders of the HPAEs have to various degrees devised mechanisms thar effectively deliver the promise of shared growth. Although policies to promote growth are generally well known, they tend to be either difficulr to initiate and/or to sustain over time. Growth makes those at the top of the economic pyramid uncertain about the fiu- ntre of their wealth. It introduces changes that are likely to open avenues for fellow elites and new competitors to contest control of their e- nomic fiefdoms Elite groups therefore generally find it in their interest to try and maintain the status quo unless they are assured that they will benefit from rapid growth. Furthermore, the initial stages of growth often require that the general population make sacrifices. Hence, even if growh can be initiated it may be interrupted by social and political un- res; unless its fruits are shard by the non-elites. The task of initiating and susmining growth thus presents a very dif- ficult coordination problem. Lcaders must persuade elire groups ro share the growth dividend. They must demonstrate to non-elites thar part of the dividend will indeed accrue to them. Governments must create acedible means to persuade everyone that their plans are nor all rhetoric. The governments in the HPAEs all established mechanisms that opera- rionalized te princple of shared growth. The NE in Malaysia, deliber- ation councils in Japan, public housing in Hong Kong and Singapore, and small enterprise promotion in Korea and Taiwan, China-all helped to demonstate the reality of shared growth. Finally, govemments must foster the conditions in which the bureau- cracy, or at least a dedicated technocratic core, can design and implement effective policies. To varing degrees, and with a variety of mechanisms, the HPAEs have performed these tasks unusually welL In the next two chapters, we examine the policies:and instruments that HrAE govern- ments used in their effort to rialize the promise of shared growth. ; I88 Notes 1. For example, the National Wages Council in Singa- 8. The criteria selected are (a) the transpare-ncy of the pore has helped coordinate wage rises with increasing pro- government, (b) the effectiveness of antitrust laws, (c) the ductivity. In Japan, the establishment of the Japan security of property andt person, (d) the faimirss in the ad- Produtctivity Center helpedi institutionalize labor- ministration of justice, (e) the extent of improper prac- management consutrations. tices in the public sector, and (1) the extent to whiich lobbying influences government decisionnmaking. The 2. In fact, the smE. sector declined only when the HCI rankinigs for each criterion are based on responses of busi- push was initiated in the mid-1970s. ness executives stationed in each of the economies in the survey to questions pertaining to the criteri-on. A Borda 3. In those cases where the prime minister appoints score was computed for each country by summing up its both vice ministers, he is constrained by custom to ap- rankings for each of the criteria. The overall rank score for point a career bureaucrat from the ministry to the post of a group of economies is the avenage of their Borda scores. admninistrative vic minister. 9. Industrial associations have been important partici- 4. The debt-equity ratio has been inordinately high in panes in the consultative process. In many cases, these as- Korea. Official statistics indicate that this is in the range sociations are empowered by law to represent their 310-80, although the 'true' figures are said to be lower, sectors. T'hus, the business community has strong repre- at about 160-80 (Wade 1990). Nevertheless, this is still sentation in a council. high compared with the ratio in the United States and the United Kcingdom, where the range is 50-90. 10. Though dhe channels of communication are still operational, the democratization of the polity has led to 5. Okimoto (1989) ranks the power and prestige of some chilling of relations. The country is currently under- ministries as the inverse. of their degree of policicization. going rapid political change that has affected the business- * The heavily politicized construction ministry ranks lowest govemment: interface. There is a recognition in many and the relatively nonpolicicized Mmu the highest. quarters, including business, fir the need to reestablish good relations but under a new form of collaboration. * 6. Seniority bears heavily on promotion decisions. It is That form is likely to reflect a more even balance between assumed that, once one has successfulfly hurdled the diEf- the tWO SCectr. cult entrance exams, she or he is capable of doing (effi- * ciendly and eff-ectively) what is askedi of her or him. This is I11. It is also widely known that the relatively feiw large not -to say that performance does not count. The pyra- private businesses have dlose links to the party. midlike structure of the bureaucr-acy guarantees some c * gree ofcmeiin A neigds ok oehrt 12. In fuct, inte'rviews revealed that even. some midsizc ensure that its members prosper.. ..Who becomes a bu- firms have established relations of this type. reau chief, a director-general, or ultimately the (adminis- trative) vice minister is a source of intense competition 13. Johnson (1 982) gives an interesting example: "The among the classes in a ministry" (Johnson 1982, p. 62). city of Musashino in the suburbs of Tokyo issued a policy statemnent sayving that housing contractors who built large 7. Finance-related agencies in Thailand, for examnple, projects had to cooperate in provicding or helping to buy the Bank of Thailand and the Budget Bureau, have their land on which to build elemnentary schools. Whena. con- own personnel and recruitmenit programs. There, salaries tracor ignored thes guidelines, die city capped the water are said to be about 30 percent higher across the boardt and sewage lines he had built for the project with con- than in the rest of the public sector. crete. The contractor took die city to court, but the court upheld the city" (p. 266). 189 - - w~R S Strategies for Rapid Accumulation CCUMULATION OF PRODUCrIVE ASSETS IS THE foundation of economic growth. It is therefore noE surprising that, as shown in chapter 1, the high- performing Asian economies (HPAEs) accumulated both physical and human capital much more rapidly and consistently than other economies and that ac- cumulation accounts for a large portion of their superior performance. More interesting is the question of how the HPAEs were able to achieve these rapid rates of accumulation and, in particular, to what extent policies were responsible. This chapter examines the HPAEs' superior performance in three accumulation-related areas: human capital accu- mulation, financial savings, and investnent. For each area, explana- nions fill into three broad categories: - Conditions that were not the direct result-of accumulation policy : Policy fundamentals: areas in which nearly all governmenrs inter- vene, but in which HPAE governments acted more effectively i ActivisE policies: selective interventions underraken in some HPAEs to increase accumulation. Rapid economic growth and the attendant changes in economic structure not directly related to accumulation policies were important factors in increasing all dtree forms of accumuladion. For example, the HPAEs' rapid sLift from high birth rates to low birth rates increased the re- sources per child potcntially available for education at home and in rhe dassroorn. At the same time, the rapid deceleration of population growth increased household saving, in turn creating resources for investment. :n addition, the HPAES performed well in selectng and implementing . ..fuindamental economic policies related to accumulation. Every govern- fundam.nml econon ry gover9 :~~~~~~~~~~~~~~~~~~~~~~~~~~g ment spends on education, but the HPAEs spent their money more wisely, emphasizing universal primnary and, later, secondary' education. Similarly, every governmen tcries to control inflation and secure prop- erty rights, and most regulate banking institutions to protect savers against default, all of which encourage savings and investment. The HPAEs didt these things better than the rest. Moreover, as the HPAFs de- veloped, their government:s strengthened the legal and regulatory firame- works fir bond andt security markets to facilitace investment. Besides fostering Favor0ble general condidions and getting policy fun- damenrals right, some HPABs have deliberately intervened in marker to solve specific coordlination problems related to accumulation. These in- terventions aimecd at boosting savings and investment rates are the most controversial of the HPAEs accumulation policies. Generally, interven- tions have been more comrmon and more successful in the three northern HPAEs-Japan, the Republic of Korea, and Taiwan, China-al of which have strong govermnnmetl institutions. With some notmble exceptions, attempts to replicate such efforts among Southeast Asian newly industri- aiigeconomieis have not been successful. Atrthe outset we must acknowledge a fundamental methodological diffickty. Since we know the HnPAFs bad rapid accumulation (and to some extent were chosen as a group on that basis), we also knoyw, even before examining their policies, that they wer not inconsistent with rapid accumulation. The challenge is to separate policies.that were neu- tral or perhaps negative but not sufficiently negative to hinder accumu- lation significantly from those policies that actually faLcilitated rapid accumulation. We do thtis, noting that our findings are based on antalytic and empirical judgmecnts rather tha-n derived from statistical models. Explaining East Asia's Hi'gh Human Capital Fonnation XCEPTr FOR IN THAE AND, THE QUANTITY OF BASIC EDUCATION provided to boys and girls of school aehas been consistently iEtiigierin the HPAEs tha in economiies with similar levels of in- come. After having achieved universal primary schooling, thereby elim- inating the gap between boys and girls at the primary level a decade or .mren searle than most, the HPAEs expanded secondary educa- 192.- er rapidly - aantdfut l fwikecorg aig n netet h STRATEGIES FOR RAIMDACPIDA tion and were particularly effcctive in reducing gender gaps at that level. What accounts for this extraordinary performance? We focus first on three enabling factors; high income growth, early demographic transi- dions, and more equal income distributions. Each of these greatly in- creased the resources available for education. We then shift focus to two policy variables: the overall budgemry commitment to education and the distribution of the education budget. We seek to show that the allo- * cation of public resources to primary and secondary education was the major determining factor in East Asias successful educational strategies. Rapid Economic Growth Rapid economic growth is the first factor that accounts fbr East Asia's extraorlinary quanrity of education. For example, during the decade 1965-75, CDP growth averaged 6.7 percent in Malaysia and 4.1 percent in Argentina. This implies that over the decade a conscant share of GDP allocated to education would have doubled the resources available for education in Malaysia, while in Argentina they would have increased by - less than half Rapid growth also creares jobs, increases real wages, and raises the rate of return on labor force skills, thereby increasing the deo mand for education (Ing, Anderson, and Wang 1993). Thus, rapid growth operates on both the supply side of the market for educational services, by increasing the potendal resources available for education, and on the demand side. ; The rising real wages thar are a product of rapid gwdicould have a dampening effect on educational opportunities, if the relative wages of teachers rise. Fortunately, rapid accumulation of hurnan capital facilitates subsequent educational expansion by increasing the potential supply of eachers, thereby reducing the premium teachers command.' The wages of educated workers, including teachers, tend to rise at a slower rate than average wages. Therefore, the impact of rising wages on the cost of edu- cation and hence on enrollments is weakened.2 For example, per pupil operatng costs at the primary level average roughly 13 percent of per capita GNP in Indonesia and Malaysia but more than double that, 29 per- cent, in Sub-Saharan Africa. In Indonesia and Malaysia the annual earn- ings of primary teachers are roughly 2.4 times per capita GNP. In Sub-Saharan Africa, however, where human capital is relatively scarce, that ratio is much higherA-The pattern is similar at the secondary leveL. 'I93 ---RAC_L E Declining Popubtion Growth East Asia preceded other developing regions in moving through the de- mographic uansition, and the education sector was a major beneficiary. One important outcome ofa deceleration of the rate of population growth is a deline in the rate of growth of the school-age population. During the 1980s, the growth rare of the population age 6-11 years was very low in East Asia-so low that the absolute number of schoolchildren in Korea, Singapore, and Thailand actually declined-but was phenomenally high in Sub-Saharan Africa. As a result, from 1965 to 1989 the share of the pop- ulation up to 14 years old in Singapore fell from 44 to 24 percent, while in Kenya it rose from 47 to 51 percent. Similar contrast are evident between other HPAES and Sub-Saharan African economies (see table 5.1). When the school-age population is growing rapidly, as in Kenya or Pakistan, rising expendirures on basic education are needed just to keep enrollment rates constant. With declining or slowly growing school-age populatons, however, similar increases in expenditures can go for more schooling or better quality. Alternatively, expenditures as a proportion of GDP can be cut while current standards are maintained (see table A5.1, in appendixA5.1). Table 5.1 Size and Growth of School-Age Populabon of tota pp ., (rnl *.FIPA&E2.1: ong-Ko 402 1: 0 -- f-.- -rcaRPf - - 143r4o- i'o n i iio:' 26¾--r:'0.7. -0,.,3' p"t.of." ' ' 4 - _- 37 ' 0.2 d' ' . '.4-4 1~~~~~~~~~~~~~~~~~~~6 a 2 9' . ...-' ... 1Banglaesh .X' ' ' ..'-.'43 . . * I94 - 0 . ; Brazl t, 4 35 120 - ' "1- ...:,..:a' 'C: '.47 0 ''., T- 47s51_3 4 igrn.~~~~~~ ~~. , 46 ~~~~~~ 9~11 r! 7 t '194. STRATEGIES FOR. R-A:P-D` I Accordingly, declining, stagnant, or slowly growing school-age popu- lations have allowed substantial increases in per pupil expenditure in all of the HPAEs. Table 5.2 shows the magnitudc of the savings on expendi- ture on basic education in East Asia due to the decline in the proportion of the population of school age.5 Consider the most striking example: if the share of the school-age population in 1988 were as large in Korea as in Kenya, the Korean govemment would have had to spcnd 5.6 percent Table 5.2 Percentage of CNP Allocated to Education Saved Due to Lower Fertility Rates t f 9 ;\; _ i X~- dne torowthm;s#e Expendtre '.on ; rhuol-aepopnltdon #t basic eduation: - -ere tower an. a -a prcenage : Economyp- - :. -; = :of GNP: . fi!ny' Mcio P X.n , Ho,gKon' 19 75 2.0 12 1.0 1.0 - 7 t180-1 1.7 1.5 L 1. 2 ;1975 . 4.2 4. 38 3 . . 1 98849 .*- - - - 2.8 :4.8 8 3.9 w;~~~~~~~~ ..w -..; l9-; 1975 13 0~~ ~ ~~~ ~~. 6 04 04- 1988-89: 28 . . %8 :14 20 1980-81 ~~~4:4 11 3. 0 4'0 ;1988-8 4.40 16 04 0 * ''J1975 .21 ; 08 20 "_J ',' ''05i ''i 5 :ijS" < t'' '! '.' -9 8 268 1 0.3 ' 08 4t,'-,'^ .iNnc' ofcul lan o. [Cur on basic ed an an asapccara2e aof ) ' Fr. percencddference between ErAsiznan4orhcrccononucs in scihoolWsg popadi) z s S * q ;~~r ;2,-,; j:u;f . t *, , 1;-th.iii 'n2u',_ -+ ,L{'; (cxpcn mme on Cai ied on =a a; p ;oFGN)- an c f addal itht would havte been rqe h p gw e m rp , . s. 0- C- -d . . t - - - r- . A4 %Th%isanndImrrofd1Saviui LTvuuut1accoarcstuiooup.½ Thris ra ndxue n asic dlihearo,usso(iouy",brn"~ ':95 .~~~~~~~~C C of GNP on education instead of the 2.8 percent actually spent to achiieve Korea's high enrollment rates. This suggests that lower fcrtility rates saved Korea 2.8 percent of GNP. Conversely, had Pakistan's school-age popula- tion grown at the relatively low Korean rate, the govcrnment of Pakistan could have increased enrollment rates by as much as 50 percent. Equality in Distributon of Income The more equal an economy's distribution of income, the higher pri- mary and secondary enrollments tend to be. In a crosa-section of more than eighty economies, there is a strong and statistically significant neg- ative correlation between basic education enrollment rates and the level of income inequality, as measured by the Gini coefficient (Clarke 1992). As we discuss below, educational expansion can have an equalizing im- pact, so the causality in this relationship could run from enrollment ratcs to the distribution of income. However, there are a variety of rea- sons why low income inequaliry, which we docurment in chapter 1, might have been a third factor co:ntributing to high enrollment rates in East Asia. Due to greater income equality, the income of the poorest 20 percent of the population in East Asian economies is higher than in economies with the same average income but greater inequality (see table A5.4, at the end of the chapter). Because of budgetary constraints (and capital market imperfections that preclude their borrowing to invest in educa- non), very poor households are unable to invest in their children even when the returns are high: the pressing need to subsist crowds out high- return investments. This is less likely to occur in East Asia because the incomes of the poor tend to be further above subsistence level. Measures of the income elasticity of the enrollment raio derived fiom cross-economy education expenditure equations are 0.31 and 0.43 at the primary and secondary levels (Schultz 1988). This implies that in East Asia the positive effect on school enrollments of lower inequality can be quite large. For example, given an income elasticity of basic edu- cation enrollments of 0.40, if the distribution of income was as equal in Brazil as in Malaysia, enrollments among poor Brazilian children would be roughly 33 percent higher. Moreover, the income elasticity of de- mand for basic schooling among poor families is likely to be substan- tially higher than average and may exceed 1.0. In this case, enrollments' among poor Brazilian children would be 80 percent higher. . . . -96.- - STRATEGIES FORR-R A P Coordination Probems that Hinder Human Capital Fomation Numerous studies show that a person's carnings are higher the higher their human capital, as measured by their education and 'iealth status. As shown in chapter 1, the same is true for nadons; we know that invest- ments in education contribute significandty to economic growth. But neither of these points necessarily justifies public provicion of education. Indeed, if the private returns art as high zs social returns on these invest- ments, individuals and households are likely to make adequate (from an economywide point of view) investments in human capital on their own. However, two kinds of problems generate a gap between private and so- cial retumrns, which-if not countered by government action-will result in fnilies investing less in their children's education than is in society's interesL First, there may be filures in the capital market and in informa- tion, discussed above, that reduce parents ability or interest in investing. Second, educational investments have positive externalities, which imply that families thar invest in education are not the sole beneficiaries.6 The difficulty of borrowing to send children to school affects the poor especially Creditors cannot easily stake a fuiture claim on embod- ied human capital (as they can for other types of collateral). Even poor &milies who might be willing to borrow, because schooling has high pri- ratc returns, usually cannot The poor are also likely to be less aware of fixture returns on education-and therefore invest less in their children's schooling than would make sense even from a strictly private point of view.7 These market faulres in principle suggesc making loans available and improving information about future retuns. A simpler, and more common, alternative is for government to reduce the direct costs of schooling by making public schooling available and free. The second class of marker failure-externalities-has more imme- diate implications for what type and level of schooling the government should subsidize and for whit groups. There are several sources of exter- nalities. An educated person may increase not only her household's in- come but also the income of other households, because her ideas and innovations spill over to others. Similarly, education may have social benefits not directly compensated monetarily, for example, a reducton in the spread of contagious diseases. In these and other cases, coordin a- non probleiims ast, and governments acting on society's behalf have rea- son to choose a higher level of education than families acting alone. I97 . --MIRAC L E Figure 5.1 Returns on Inveshnent Policy responses generally consist of making schooling available and in Education by Level (Latest Year) free and in some economies compulsory. The difference between social Ratreof rtuwm(percent) and private returns from educauton is probably higher at tle primary Pri_ 1 and secondary levels than at the university level (see figu 5.1). Many :Primary i ---c- ndia1 --I-i positive spillovers come from literacy acquired at lower levels of school- SondarJ Kr fhing, while the retu:s from training at the university level are almost - Hi4her I ! i fuly captured by te higher income of university graduates. Vocational 05 ID L training 2may also havc high social payoffs, if it improves worker pro- X 0~~~' 5 10 15- 20 25 30 3!i 3 SocIal return U Pilvato ductivity not only for the trainee but also for her co-workers. More im- *; Social retrns N Nvaf mt portantly, evidence suggests that vocational training is most Soww Pswchaopoulos (1993). cost-effective if trainees have a solid base of primary and secondary edu- cation. All of this argues for universal primary arid broadly based sec- ondary education as a means to improve economic efficiency and income distribution. Table 5.3 Public Expend4tnre on Education as a Percentage of GNP Policies that Promoted Human Capital Formation if ;E con>msrnyI +. 5 .- -1960 196.i 9 ! f Higher shares of national income devoted to education cannot fiuly explain the larger accumulation ofhuman capitl in the HwAES. In both oegg :~5 Kong2.8 ;1960 and 1989, public expenditure on education as a percenrage of GNP Korc p.oE 2.0 ]46 - was not much higher in East Asia than elsewhere (see table 5.3). In 1960 ;:,--Smngipoin -2.8 34 ?g1ayi 7~ '.9 pV W - the share was 2.2 percent for all developing economies, 2.4 percent for T ¾ J.'x',tfg>R; 23.", '.3]9.2. Sub-Saharan Africa, and 2.5 percent for East Askia During the three z lndocs 2z3 0.9 decades that followed, the governments of East Asia markedly increased *~~~~~~~' ovr nts- ogVIg f! Eas.............t2T- -J *; *ZAvctnX ;.sc ,.1 2i.5{ Si ;. nthe share of national output they invested in formal educaion, but so -- - ; ' '_t_AX'l *'ii;i governments in other developing regions. In 1989 the in Sub ;;rg sa 1;9 q r -l Saharan Africa, 4.1 percent, was higher than the East Asian share, 3.7 PakfisraaV$1 'I 26{- .-- J bSs percent, which barely exceeded the averge share for al developing -- ttdc4cs:oPcdr ; t economies, 3.6 percent8 \ ; , 1 .7.'. 3.1S'St; E fl \ 9'n 5|; >Nor wer initial conditions, for example the colonial legarcy, decisive. ; ' - Et -6 J 4 V's 41/ W e Korea did have much hier enrollment mtes in 1950 thn most d0Zn~ * oFcDnms' j^['.''Lli{,.rM What proportion of these gaps is due to initial conditions? At the pri- .98 --f E g .- . ... -.. 0 -. STRATEGIES FOR RID7ArCCU-M4 mary level, Indonesia increased its enrollmenr rate by nearly 80 percent- age points since 1950, while Pakistan managed an increase of only 34 percentage points, implying that inost of the current gap is explained by the pace of increase rather than initial conditions. For secondary school- ing, Pakistan's enrollment rate in 1950 was actually higher than Indone- sias; all ofde current gap is explained by the rates of increase during the past thit-seven years Prmawy and Secoudary Education. The allocation of public expenditure between -basic and higher education is the major public policy factor that accounts fbr East Asias extraordinary performance with regard to the quantiry of basic education provided. The share of public expendi- ture on education allocated to basic education has been consistently higher in EastAsia than clsewherm: Korea and Venezuela provide an ex- treme example that nicely illustrs the point. Table 5A indicares that in 1985 Venezuela allocated 43 percent of its education budget to higher education;, by conwzst, in the same year Korea allocated only 10 percenc of its budget to higher education. Public expenditure on educa- tion as a percentage of GNP was actually higher in Venezuela (4.3) than Table 5.4 Allocation of Education Budgets, 1985 f~~~~Pbi 'K> . r;j6Z.et.t . Pub&-, Peretge of Percenag f k ~~~~~~~~~* ; L;;te>ni8s. cn. cio* s --g. ; r ' ; - on eoncain on'! - .- .t' -* ~- as , ;- educanon sit - - Mn 6 . ,9; n ' * t-- ...i --. -- . .. .. ;E A:LdcaoonzyGW o N- Siid,cinollnt bsc ducanon i:980. . -Indones23: 72.. .0 " , 10.3, 0f iKoeaRepo 0 *W- -'~~~~~~~I" 7 .2' tw 1vIalay~~~~~ia 79 2-~. - 14.6~ 749 ~ Singapore 5.0 32, 30.7: 4 . - .Taln .t3. . 2..10 .8. 'F. .4. .. - '- . -1 - --: ; . e-3 ;E-310 : t- . .- .,Lltf -atv.Sosc Qu-. 't t L' a6E a9d4.3i fl95.46,iiuhatrthe per - , eniag ofthe eu4oi buidgealarcmbic?Iuconwns_81 m 9S48; and 80, ~tini985-8 .- dc: ng .adalu`p* St e it -c - Mg(* 'I P-d -- W5Ofli '99. £ in Korea (3.0). After subtracting the share going to higher education, however, public expenditur available for basic education as a percent- age of GNP was considerably higher in Korea (2.5) than in Venezuela (1.3). Box 5.1 shows how Indonesia's emphasis on primary education, contrasted with Bolivia's relative negIlect of primary schooling, is re- flected in rural educational opportunities. The share of public funds allocated to tertiary education in East Asia has tended to be low, averaging roughly 15 percent during the past three decades.9 In Latin America die share has been roughly 24 percent.'0 In South Asia, the share is dose to the Latin American leveL This had been the case in Sub-Saharan. Africa as well, but in recent years the share has dedinedrto East Asian levels. By giving priority to expanding the primary and secondary basesof dhe educatoa pyramid, East Asian governments have stimulated the de- mand for higher education, while relying to a large extent on the private sector to satisfyr that demand. Eni all developing regions the probability of going to university is miarkedly higher for secondary school graduates from high- thian from low-income famirnls. Typically, in low- and middle- income economies government subsidies of university education are nor related to need, implying that they benefit fimilies wi 'h relatively high in- comes that could aflord to pay fees closer to the actual cost of schooling. At the same time, in many economies, Brazil and Kenya being no- table examples, low public funding of secondary education results in poorly qualified children fr-om low-income backgrounds being forced into the private sector or entirely out of the education system. Because of the higher conce-ntration on basic education in East Asia, public finds for education are more likely to benefit children of low-income famnilies who might otherwise have difficulty remaining in schooL Vocational Twining. Human resources and the trainingr to upgrade them have been important to the HI'Ans successful export drives, despite the high degree of labor intensity in their manufactured exports. High- level skills are essential for such manufacturing-related aaiviries as mnan- agement and entrepreneurship, information -technology, finance, markceting, accounting, and law. Moreover, adaptive innovations on the shop floor, which are responsible for a major share of productivity in. manufacruring, demand both higher- and lower-level skills. But while vocational taining is widlreonzda imotant such traing is rarely cost-efficient when provided in the school systems. Firms prefer to do their own trairnin pardy because many skills are firm-spedific.. -200 STRATEGIES FO.R. k-Ai -Wl Us N 31 SZ Ed Mud n :p quesm--l- gg 0 . ou 0 0 to .,H, citi0ii"bf- propoipomp Y, $61 -5-i PC-= -r. n. ZOM ulca'don- ud et,' 00 Pff pq*" 90 r w=ntpmbamr- ucauonz,-,F-,,"j,,. C uc2uoq diiii2b M-ffiki6 0,-percmrp -,-PFF 2=enta o Bft a c;-,MC:rcmamclcri,,-,, CUOn. arsozaLs= R -'zcu -m y,'one in C O j'an jopPrO 7 o non, ZhQ? 1987' :ax .11;2pcm gr ;",O "ea ):PM m 's, iham liEfic,ar-kniv P C Li C=O caiiNwduj &-Co zm owcvcr Lti SCIOLOO 201 Few studies have been made of training in the HPAES or ocher devel- oping economies, in part because of the difficulty of defining tamining. One lage survey of 48,00 0 manufliLccring firms in Taiwan, China, con- duded chat firm-level trining rised productvity by encouraging the efficient use of technolog (Aw and Tan 1993). The study also found chat returns from training were higher in industries with well-educaced workers in an environment of rapid technological change. Another Tai- wan, China, study with a smaller survey sample and stricter definition of training found that only 24 percent of Sins provided trining and only 7 percent of technicians received it. As would be expected, training was positively affected by the level of technology and negatively afficted by labor turnover and high opportunity costs of training (San 1990). Other studies in the HPAaEs have reached similar conclusions. Enter- prise and preemployment training have produced social rates of reTur of 20 percent in Maaysia. In-plant taining for welders in Korean ship- building had a socal rate of return of 28 percent, higber than in non- firm. training institutions (Middlecton and others, forthcoming). On-the-job ntrainig was an tremely systemainc and powerful ingredi- entb in the rapid growth ofjapanese companieas" (onishni 1989). In some instances, govern-mnt efforts to pomote taining have gone awry According to one study, Koreis 1974 Special Law for Vocational Training, which required firms to provide six months' training in am proved schemes, discouraged firm-level training; firms considered the period too long and opted to pay a fine instead (Kim 1987; van Adams 1989). More narrow governmentefforts to forecast occupationdemand or provide tnained workers to fill anticipated jobs have also generally not been successfil. One exception is Singapore's use of training to promote the inforna- ton technology sector. Singapore has achieved world leadership in information-related services thrugh a concerted program thar involved educational insintitons (specializng in business and engineering software training), training subsidies to schools and office workers, computeriza- don of the dvil service, and establishment of TxadeNer, an interatonal infornation network (Hon 1992). This success illustrates the importnce of a govenmmenes ability to foresee a major trend and coordinate comple- menary privare invesctnents. At the same dime, businesses must stand rready to take advantage of the comprehensive support that the gvern- nment ptovides. Other East Asian economies have had imilar, if less spec-. tacular, successes in telecommunications (Mody and others 1992). zoz STRATEGIES FOR RAP Has training been more effective or widespread among the HPAEs than other developing economies? The evidence is too limited for a dear conclusion. However, ir is dear chat retunms from training are aug- mented by emnomic growth and consequent job creation and are there- fore higher in the HPAEs. Moreover, the HPAES' extensive participation in international markets aided the success of their training pmgrams. In- tremational competition encourages finls to train workers and increases the number of skill-intensive jobs, thereby ensuring that new skills do nor erode. Training, then, appears to function like many other aspects of development in East Asia. It contributes to rapid, susained growth but does not, in imself make such growth possible. Why East Asia's Enducational Poicies Wirked Ln most of the ecorarnies of East Asia, public investments in educa- tdon were not only larger than elsewhere in absolute rerms-they were also betten They responded more appropriately to coordination filures in the market for education. Emphasis on universal, high-quality pri- maryeducation had importantpayoffs both for economic efficiency and for equiy. The excess demand for secondary and tertiarv education, generated by rapid ateainment of universal primary education, was met largey by a combination of expansion ofa public secondary system with mentocratic entrance requirements and a self-financed private system. This stands in stark contrast to many other low- and middle-income economies, which have stressed public subsidies to university education. Explaining East Asia's High Savings Rates : - A5S IS THE CASE WITH EAST ASIA'S RAPID HUMAN CAPITAL formation, the region's high savings rates are in part an outcome of rapid growth and rapid demographic transitions. Govern- ment policy also encouraged (and sometmes compelled) increased sav- ings through a variety of means. The most basic of these was the maintenance of macroeconomic stabiliry. In addition, the HPAEs have addressed savings-related coordinaion problems with numerous mea- sures designed specifically to boost savings rates. Thcse range from pol- icy fundamentals (such as rguatory supevision of bnks, winch 203 0j IRACLE address an absence of deposit insurance markers), to targeted interven- tions (such as restrictions on consumer credit and forced savings plans). Growth and Demographics: The Effect on Savings Many popular efforts to explain the Easr Asian miracle have invoked cultural factors to explain high savings rates and used those high savings rates to explain high rates of investment and growth. Neither element of such explanations has received much empirical support. Indeed, studies of the income-savings relationship in a broad cross section of economies indicate that while income and savings growth are highly correlated, in- comes often have risen befbre savings rates rather than after, suggesting that growth drives savings rather than the other way around. Recent econometric studies support the idea that rapid income growth boosts savings rates as households acquire resources fister than they increase consumption (Carroll, Weil, and Summers 1993). According to this view, East Asia's high savings rates since the 1960s are partily an out- come of high growth rates rather than a cause. We have attempted our own tests of the savings-growth relationship for the HPAEs (described in greater detail in aDpendix 5.1 and its corre- sponding table A5.2). Income growth has been a remarkably good pre- dictor of increased savings rates in Indonesia, Japan, Korea, Thailand, and Taiwan, China, but savings has not been a good predictor of growth. Results are mixed for Hong Kong and Malaysia, and causation might run either way. In Singapore income changes were not a signifi- cant fctor in the speccacular rise in savings rates in the 1970s and 1980s, consistentwith the view that demographic fiactors and, to a lesser extent, the policies of the Central Provident Fund, a forced savings pro- gramn, determined the rise (Monetary Authority of Singapore 1991). Rapid changes in rhe size and age composition of households may have cont-ributed to rising savings rates for many of the same reasons that rapid demographic transitions fostered human capital formation: as birth rates fell, the dependency ratio-that is, the ratio of non-woring- age people to working-age people-decreased. Theories of savings based on the life cycle of increasing then decreasing income relative to con- sumption predict that societies with a high proportion of prime-age workers will save more than those with higher proportions of young or old people (Ando and Modigliani 1963; Modigliani 1970). Of course, the benefit of filling bir rates evntually will be reversed as the popu- 204 STRATEGIES FOR At lation ages and the old age dependency ratio rises, an effect seen in stud- ies of European economies and Japan (see Blumenthal 1970; Horioka 1990). For the HPAES other than Japan, there is litle evidence of the rela- tionship between demographic changes and household savings. The Singapore study mentioned above showed that most of the increased savings rate fiom 19 to 46 percent betwi. :n 1970 to 1989 was attribut- able to dedining dependency ratios (Monetary Authority of Singapore 1991). In Taiwan, China, where household savings rates are generally very high, households headed by prime-age workers tend to save more, possibly reflecting in part preparation for old age. (Bur savings in older households are also remarkably high and stable, contrary to the view of the life cyde savings hypothesis [Deaton and Paxon 1992].) Demo- ,raphic factors may also interact with corpwt"c behavior to increase sav- ings. In Japan the tradition of employees working beyond retirement age at reduced wages appears to contribute significantly to higher savings rates (Horioka 1990). Policy Fundamentals and High Saings Rates In addition to benefiting from higher growth and demographic shifts conducive to savings, East Asian governments selected, ofien through tial and error, a broad spectum of policies rhat encouraged savings. Two fundamentals that facilitated high levels of savings were the man- agement of macroeconomic policy and the making of budgetary choices. The Role of Real Interest Rates. Foremost among macroeconomic fmnda- mentals was the effort to control inflation discussed in chapter 3. Be- cause high inflation rates also tend to be volatile, real interest rates are often negative and unpredictable, discouragng domestic financial sav- ings. Moreover, since unanticipared high inflation erxdes the ral value offinancial assets, volatility of inflation ,increases the riskassociatedwitth holding, them. Conversely, low to moderate inflation, partculady asta- ble rates, encourages financial savings. Thus the HPAEs' generally supe- rior record at avoiding long-term fiscal deficits and restaiing inflation gave people the-confidence to save (Neal 1990). Table 5.5 documents average real interest rats in the uAEs com- pared with other low- and middle-income economies. Real interest rares on deposits in the HPAEs were either positive or nildly negative. In con- 2.05 MC LE Table 5.5 Average Real Intkeres Rates on Deposits, Selected Economies - *. . r~~~~~~~~~~~~~~nwiu&wrd MJPA& AIcegtsx ' ~~~~~~1973-2. -tJ . . 3 * Tan.Iiq,~~~~~d 1973-90 AS3 * MmhI& . 177 1 .2.47 - ', 19,7-91 ~~~~~~~2.43 i Taiwan C. 94-3 .8 7* '- ThIml 1177-got 4.4197 -. Naqj.dab ~~~~1976-92 04 .'3 NcpI - - . 117449 ~-3.6 55 M-ils 1791443$3'.4 1935-91 ~~~~~J314 94.9 Un4m7 3976-9' ~~~~ ~~~~~~~~~~LA9 Ism, EWF -. -. ~~~~~~1976-9 -63m- Taft,1 ~~~1976-1-5 83 At.tqc . , -~~~~~~~~~~~521 334~~19 Am.- '. 144 -243 * , l - '*aIa~~~~~~~~17 I . -: :I!-: ;w -.9 21033j uas mAi 12793-q 5 Zsra ore 1Sgonsshw -5an ecnmc*it1agIeadertso average ae er 1. 13.1 prciinTASseS an G2aeret nZaba.Ra intercst ra4e weeas' evlOS e as 2o6~~~~~~~~~~-l 109 3 3 STRATEGIES FOR 'AlI' measured by the standard deviation) in other low- and middle-income . economies than in the HPAEs. While the standard deviation was 3.5 in the HPAEs, dose to the OECD economies at 2.8, on average it was 40 in Latin America, 14 in Sub-Saharan Africa, and 6 in South Asia. Malaysia and Singapore have maintained the most consistendy positive and stable real interest rates on deposits, paralleling interest race trends in the - -..United States. Japan maintained negative, but highly stable, real interest races on deposits until 1983 and has had very low btut stable positive real miterest rates since that time. Korea, especially Indonesia, and Taiwan, China, have had much greater volatility of real interest rates than the :other HPi-Es but stl substantially less than in many other low- and -middle-income economies. Several studies have documented the positive association between real interest rates and the growth of savings deposits and broad money aggregates (Fry 1988; Lanyi and Saracoglu 1983; Gelb 1989). This is especialy evident for changes from highly negative to positive real rates of interest. For example, when Taiwan, China, raised real interest rates on bank deposits from a negative 300 percent in 1949 to about 8.5 per- cent in 1953, the ratio of time and savings deposits to money stock rose from 2 to 34 percent in three years (Chiu 1992). Indonesia and Korea - achieved similarly dramatic increases in financial savings after stabiliz- ing inflation and shifting from negatve to positive real interest rates. More recendy, Argentina, Chile, Mexico, and Pakistan have done the same. The Role of Pubic Savings One direct way that any government can ad- dress the concen that aggregate savings are too low is to generate public sector savings through a combination of tax and expenditure policies. Although in theory forward-looking private savers should increase con- sumption (and reduce savings) to match any increase in public savings (because their future obligations are reduced), in practice households are faced with liquidity constraints and are not able to act on expectaions. Thus total savings may rise with an increase in public savings (Summers 1985). Empirical evidence shows that government savings does not fuly crowd out private savings. Furdtermore, the method of raising public savings matter.s: on average, increasing public savings via reduced ex- penditures is more effective than raising taxation. For a sample of thir- teen developing economies, a transitory increase of a dollar in public savings, miade through a cut in expenditures, reduces private savings by --only 1-50 cenrs. If the dollar increase in public saving is achieved 207 UI4MXLCLE through a current period tax increas, private savings dedlines by 48-65 cents (Corbo and Schmidt-Hebbcl 199 1). Most HPAEs have maintained high public savings compared with other low- and middle-incomececonomies througi. -ombination of tax policy and expenditure restraint. The extent to waich HPAE govern- ments have restrained current expenditures relative to other developing and industrial economies is shown in figure 5.2. which gives average public consumption as a percentage of GDP between 1970 and 1988 for a sample of 111 economies. Indonesia, Japan, Korea, Singapore, and Thailand are in the bottom third of the distribution of all economies in terms of the share of public consumption, with public consumption shares below 15 percent of GDP. Malaysia and Taiwan, China, are ecx- ceptions, faling in the middle of the distribution. In Malaysia the high levels of public consumption reflected the explicit redistribucional ob- jectives of the New Economic Policy As a group, the avcrage share of public consumption in the HPAEs is below both the OECD economies and all other regional groupings of low- and middle-income economies. In contrast to the situation in most developing economies, in the HPAEs, rates of public and private savings have been high and growing. For example, Singapore's public savings rose from 5.5 percent of GDP in 197480 to 18.5 percent in 1981-90, and private savings rose from 22.6 to 24 percent. Malaysia's public savings increased from an average 3.2 percent of GDP in 1961-80 to 10.3 percent in 1981-90, while its private savings increased slightly. Th2iland's public savings was high in the 1980s (87-15 percent of GDP), but declined as a proportion of GDP in the mid-1980s, when the privare savings rate rose dramatically (see table 5.6). While in some other economies (for example, Pakistan, Zimbabwe, and Venezuela) public savings was relatively high and accounted for a major part of national savings, generally public savings in low- and middle-income economies was substantially lower than in the HPAEs and had decdined. For example, public savings in the Philippines de- dined from an average of 10.4 percent of GDP in 1980-83 to 1.4 per- cent in 1984-87, leading to much lower national savings rates. Both public and private savings dcdined in the mid-l 980s in Argentina and Chile. Brazils public savings dropped from 8.3 percent of GDP in 1980 to 3.2 percent for the period 1983-85.11 Some Sub-Saharan African economies (for example, Ghana) had negative public savings m the 1980s, offseting increases in private savings. 208 STRATEGIES FOR RAI3D Figure 5.2 Averag Public Consumption as a Percentage of GDP Taiwan, Ch Ina -~~~~~~~~~~~~~~~~~~~- I Singapore Rep. of Korea EHPAEs Oethe eownomies o,10 20 30 40 Average public contsumptlon (percentage of GDP) . Soe _, --orl dB2nk6 - 209 !&4iIRfACLE Table 5.6 Public and Private Savings of HPAEs and Other Selected Developing Economies (pereL'ntage of CGP) u -ble ; saings V i4o6 J-apa, ,* ')0 z | g 95 5-70 ' . S 0 t ^6.2 17.2- -19714=0 4.6 '201 1 6 12 15 0 ; i . R :|?8 ~~191480 3 f.;:: .2;X 0f: . 187'; -198190. 103 191i Singapore . *. . - 4;---o 1981-90'b 1-ft 3- >;- ;:' - I83 - .-- . . --i24.0 1980-B5 . 14.3 * .- -j198 8.6. 14 Oird deSpigeconemi -Argendna~S.Kw'lii -. . ,t''-'.'' _ -' s_ --;--.1980-85* ..... . 5 - I4.2 -19804871, 1980-8. 235 0 :: : 985647 j- 0 6 1980-84 3.9 2~~~~~~1.6 *19858 7.9 06 *'~~~~ 1980-84~~ 1364 -~~~ 1985~~~~-187 7. Csti Rica198-8 7. 7 P'ia9i5-8704 161 61", ,., i.o,9.O - I9 .4-. , . 1. .s '.--'-,0.;;'--. 1690 97- - '.=-' 'Philippines~~~~~~~~~~~~~~~~~~J .2IO: ~ ~ ~ ~ ~ ~ '1503 0 t I STPTRATEGIES FORJLAiPPD& Coorditon Problems that Hinder Savings Savrings are afficted by several types. of coordination problems in- duced by features of financial markets. Below, we briefly describe each of these coordination problems and the policy responses of HPAE govern- * ---ments+The range ofthe HPAES' policrresponses to theses-rual or incip- -ient marker fhilures range fiom standard remedies that the HPAEs may ...have implemented better, such as the prudential regulation of banks, to more acavLst interventions, such as- limitng competition or regulating tinerest rare spreads. Abse--.,nf Deposit-Insurance People will only deposit funds with a bank or other financial institution if they are reasonably confidenr that '.the insirution will retrn the. funds as promised. In theory, private -firs. or eVen the banks themselves could. sell insurance to depositors, in practice,theydo nor (These points are discussed at greater length in . 'Stigiz -1993a) T.he reasons are straightforward. Assuming a pae firm-didLoffer deposit insurance, depositors have no more reason to tru:st.an insurer than they do the bank: either or both could faiL In- deed, -a firm that insured several banks would risk nearly simultaneous losses during a general economic downturn or-a bank run. The risk of such losses m-akes aprivare insurers promise to deliver.even-more sus- pec than. the banks. Thus, withour government effors to guarantee that banks will be. able to honor their liabilities, potential depositors .would rend-tosave less and keep a higherproportion of theirsavings in nonfinancial assets.. Governments have. a second-motivation for insuring deposits.- Fman- .ial.hlures. of banks have a strong negative- excernality; when some bank fail, depositors lose confidence even in. sound. financial institu- dons, as the history of bank runs attests- Govemrnments therefore ty to preventrovert bank 6ailures-, fiequendy acting explicitly or implicidy as the-insurer oflascresorr through the central bank. :High rnsacti.CostsPrivate banks seek large-deposits and those thar- are relatvely inexpensive to sevice theg therefore compete for the de- pits of -corporations andtaffluent urban -dwellers and. make lite or-no efforrro attact ithe savings -of workers, poor people,_ andLresidents-of snulLi owna andshe: countyside In. addition,- banks may- rcfiise such ;marg inall po eposits by settng minimumin deposit lewls- The:pracceaise.the transactons costs forsmall savers ancdiscour-- age themfruomn aking,financial saving. | , t 0 ;g , . . . . . .~ ~ ~ ~ ~ ~ ~~111 I`- LE Isolaton and Insufficient Savings. Some economists have also suggested that individuals acting on their own and their flimilys behalf may save less than would be desirable from the point of view of society as a whole. This is because they fail to consider the impact of their savings decisions on others. If savings decisions could be coordinated, each individual would save more and aggregate savings would rise (Dasgupra, Marglin, and Sen 1972). In such circumstances social returns on savings exceed private retums, and there may be a role for government efliorts to in- crease savings. Policies that Promoted Savings Government responses to these coordination problems are varied. All governments take steps to alleviate the lack of deposit insurance markets and to promore stble financial institutions. These generally indude prudential regulation of savings institutions and protection of deposi- tors from bank defaults. In general, the developing HPAEs have per- formed these fundamentals better than many other developing economies. East Asian governments have artached great inmportance to maintaining saver' confidence in financial institutions. Often their policies have gone beyond prudential rgulations to include protectng banks fiom competition to increase the financial strength of banking in- stitutions.12 Sometimes this has been done at the expense of long-nm efficiency. When necessary, all EPAR govenments have bailed out troubled fi- nancial institutions through financial and managenent assistance or mergers with stronger banks (see table A5.5, at the end of the chapter). This is not to saythat the iEns have not experienced banking crises. In- deed, some financial instability is probably unavoidable in economies that are expanding as rapidly as those in East Asia Nonetheless, the crises that have occurred, for exanpie, in the wake of the 1 980s' real es- tate speculation boom in Hong Kongi Japan, Korea, Malaysia, and Thailand, have generally been brief and modest, particularly in compar- ison with the protracted and disruptive banking crises in Bolivia, Chile, Ghana, the Philippines, and Turkey during the same period. Regu;llions to Proane Bank Solvency. Central banks and departments of 'finance in the HPAEs appear to have been generally more successful in su- pervising commercal banks, which have reported a rdatively low pro- portion of nonperforming loans in thcir portfolios. There are some 2.17 STRATEGIES FOR RAPI major exceptions: Indonesia in the late 1980s, Korea in the early 1980s, and Malaysias Bank Bumiputera in the 1970s. Generally, however, non- performing loans have been less of a problem in the HPAEs than in many developing economics. For example, one study of twelve Latin Ameri- can economies found that nonperforming loans accounted for nearly 14 percent of total loans in 1987, while the proportion of bad loans sur- passed 20 percent in Argentina, Honduras, Uruguay, and Venezuela (Morris .ndothers 1990). In Pakistan, required provisions for bad debts constituted 14 percent of tota pordolios in 1989 (an internal World Bank report). In Sub-Saharan Africa, many banks have failed, and 25-40 percent of bank portfolios have had to be written off. While macroeconomic instability and low growth are partly responsible for this poor performance, inadequate prudential regulation also contributed. The more effective prudential regulation of East Asian banks is ap- parent in the ease with which most of them adopted the international capital adequacy requirements set by the Bank for International Settle- ments (Bis) to ensure that banks do not take on inappropriate levels of risks. Hong Kong, Japan, and Singapore began to stengthen prudential reguation in the 1970s, and Malaysia, Thailand, and Taiwan, China, oi-llowed suit in the 198Os. All IPAEs have since adopred the BIS capital adequacy reo'4rements. Indonesia, the last to do so, has had some diffi- culty achieving BIS capital requirement levels because it introduced proper prudential regulations rather late.13 By conast, in Lain America only Ecuador adhered to BEIS standar4 iwile Colombia was in the process of adoptng the standards (Morris and others 1990). Moreover, some Latin American economies have not established minimum capital requirements for new banks. In Sub-Saharan Africa, the net worth positions ofbanldngstyerms siarply deteriorated in the 1980s, and adopting BIS standards wold entail subsantial rcapialization. Most East Asian economies continue to rcly on dose contact between supervisors and banks to encourage prudence. This routine, often daily interacton enables regulators personally to assess the riskiness of a bankA portfblio. This pragmatic approach is evident in their attitude toward col- lateral requir~ments& To reduce the risk of borrower defaults, regulators -encuraged banks to require substantial collateral. But because the collat- era. was usually real estate, banks became increasingly exposed in the -1980s to the risk of asset price declines.4 Face with potenial insolven- -ies due to heavy real estate lending, regulators in Hong Kong, Japan, Thailand, and Taiwan, China, have limited the use of real estate as collat- 2.13 LE eral and sought to reduce loan concentration by limiting lending to re- lated parties. Enforcemnenr has been difficult, however, because companies are dosely heldl and dlisdosure rules are geCnerally weak. Moreoverinm sev- erl economies, notably Indonesia, Japan, and Thaiad, banks and firms rend to have inrerlikng ownership. Many East Asian financial systems appear to be moving to ara mixed system of enfbrce-menr, in which the highly structured exposure rules that characterize international supervision are backed by tradi- tional moral suasion in which bank regulators use their control of branch licensing, rediscounts, and other regulations as leverage to ensure cooperation. The Bank of Japan, for example, has used constant inter- action between city bankcs and the supervisory authorities, coupled with its tight control of branch licensingi to encourage prudent behavior. Malaysia's bank supervisors operate in a similar flhshion. Coping witf Distessed Financial Iustitutioas. In spite of generaly good pmudential regulation, there have been financial crises in Eas Asia brought about by a mix of macroeconomic shocks, insufficient pruden- tial regulation, and seculative lending (see table A5.5, at the endt of the- chapter). However, goverrnments rapid and effTctivc response has been able to maintain confidence by acting as implicit insurer without exces- sive fiscal costs. Rapid growth and spiraling real estate prices have often tempted financial instittions into property and sometimes stock spec- ulation. This has happened repeatedly in Hong Kong (in 1965, 1982, and 1986), but it has also been a problem in Malaysia, Thailandt, and, most recently, Japan. By the late 1980s, financial institutions in each of these economies were facing diffliclties, as nonperformning loans rose due -to the economaic downturn. Korea encountered similar difficulties but for somewhat difkirenr rea- sons, among them the heavy indebtedness of firmns that had been bene- ficiaries of government-directed credit. 'When exenldemand fell in the 1980s, many firms had large excess capacity and poor -debt service capability. Among them were gvrmecnt priority projects in shipping; overseas construction, texties, and machinery. Because the firms were highly leverged, they were financially vulnerable to changes in demandt and interest rates. 'When fnnildistres has occurred, HPAE governments have come to the rescue_ Table A5.5 describes the origins of and responses to some of the significant banking crises in the HPAEs. Whil only Korea andt Tab- Wn, China, have estOablished explit deposit isurance (both in the 2-4 STRATEGIES FOR RiAPVT early 1980s), they and other HPAEs have for decades implicitly insured deposits by stepping in whenever necessary to prevent bank failures. In some cases, especially in Japan, stronger banks were pressed to take over banks in financial trouble. When mergers have been impractical, East Asian governments have served as lender of last resort. Only Hong Kong and Thailand have liquidated insolvent financial institutions. Yet iecause of rougher prudential supervision and better econornic growth, the costs of bank rescues in East Asia have been lower than in many other economies. During the 1980s the estimated fiscal cost of bailouts in the HPAEs was 0.5-1 percent of GDP, compared with 2.3 percent in the United States, 2.5 percent in Turkey, 3 percent in the Philippines, and 26 percent in Chile. Puntein of Banksfrom Compedfioo. Nearly all governments regulate the acation of new financial institutions and the entry of foreign financial institutions to try to ensure thar newcomers are solvent and will remain so. But many East Asian governments have gone fiirrher, limiting com- petition and protecting existing banks both fiom foreign banks and from new domestic competitors. With the notable exceptions of Hong Kong, Indonesia, and Singapore which welcome the opening ofnew fi- nancial insritutions provided they meet capialization and other regula- tory requirements, most HPAEs have had rigid restricions on entry. When expansion has been pennitted, it has mostly been through the li- censing of new branches of exsting banks. Despite limits on enny, rhe banking sectors in the HPAEs are no more concentrated than in otier de- veloping economies or OECD economies (see table A5.6, at the end of the chapter). This means that banking sysrems differ litde in structure among economies, although the onrgins of their relatively high interna- tional degree of concentration may differ, as we shall discuss below. Why were entry restrictions used rather than high net worth stan- dards to encourage prudent behavior? Entry restrictions give govern- ments a more discretio, -.y tool to influence the behavior of banks than net worth requiremnentr ' tovernments may exercise discretion in decid- ing which banks shouldexand (through more branches) and when. Banks have an incentive to respond to govenments' actions lest they forgo the additional profits from asset growth. Net worth requirements do not provide goveriments this avenue of influence. Have these restrictions on compention contributed to increased fi- nancial savings? The answer is ambiguous. Larger, more profitable banks have more to lose and so a greater incentive to maintain their rep- .. . . . . .~~~~~~~~~~~~~~~~~~1 utations by prudent lendling. Moreover, restricting competition mn- creases profits, and higher profits increase the financial strengthi of the banking system (provided banks do not distribute profits to share- holders). These effcts would raise saver confidence and hence increase savings. Yet, in most economnies regulators have argued that the financial sectoes tendency toward concentration is so great that govermnments must intervene to maintain competition within the banldng sector, pre- venting monopolistic practices that squeeze dlients, reduce efficiency and innovation, and hence, lower savings.' Therefore, whethe-r controlling competition results in &avorablc or unfaivorable outcomes for savings depe-nds on whether the increased sol- vency comes at the expense of efficiency. The evidence among devel- oped economies supports the view that financial systems with fewer, larger banks outperform systems with many smaler banks (Vitrs 1991). But in many developening econoes financial markets are domi- natal by a few large, veq inefficient banks whose size is bad less on economies of scale thn on restrictions of bank licenses and othecr limits on comp'etinon. The impact of limited competiton on banking efficiengc vres across economies in East Asia. Japanes commercal banks, for example, are regarded as efficient, despite enty restictions, of the pres- sure of competition among existing banks (Sakaklbara and others 1982). Thee is no dominant commercial bank in Japany The top five ban are roughly equal mn size, and they accunt for about 40 percent of total commercial bank assets (see table A5.6). In comparison to large German banks, Japanese banks tend to have lower lending margins (spreads), although smal German banks have lower spreads than small Japanese banks. The Japanese banks also have lower lending margins tha equivalent bankcs in the United States but approximately equal rates of return on cquity (Ueda 1992). Gross nterest miargins as wellas net operating costs (as a percentage Of total aSSetS)of banks in KoreaI, Malaysia, and Thailand were in line with oECD economies and substan- tialy lower than othecr developig economics notably Liberiof Sri Lanka, and Turkeys and selected economies in Latin America (Hanson and Rocha 1986). Conversely, studies of the banking system in Indone- sia and Korea find that protecion has led to less ianovautve pracbtces (n- ternal World Bank reports). Thus the evidence does not condusively suprt the efficacy opl - = ner ~~~~~~~~~~~~~~~~~spo p rtn cot(as prcetg of tt t)of Be aces to imit entry. CWhere ted,encies toward large nancial institutions z16. STRATEGIES FOR I _ are strong, govrnment cfforcs to limit entry will mainly reduce efficiency and lower innovation. This is especially true of Mstrctions in some economies on foreign banks, which tend to be the leaders in innovation. Regilalg Spreads. Except for Hong Kong and Singapore, ali HPAEs have at times simultaneously regulated deposit and lending rates and con- sequendy the spreads of financial institutions. If spreads are controlled, banks will earn limited rents despice protection from competiion. In gen- aal, lending spreads of the HPABs have been comparable to many OECD economies and lower than most other devdoping economies, indicaung that controls on intermediation magins were generally effcidve. Figure 5.3 shows that avage terest spread of dte iHPAEs (about 3 percent) is lower than aveg spreads in other developing regions, which range from 4.0 to neary 10 percet. Maren in Indonesia, Thailand, and Taiwan, China, were about 3-4 pcent in the 1 980s. These margi ar compara- ble to those in the United States (2 percent) and Japan (3 prcenr). In Korea and Malaysia they were generally igher at about 4.5 percent (see tableA5.7, at the end ofthe chapter). Thus, while restricingcompetition grenerated some rents for banks, a combination ofcompettion among in- cumbents, regulation of spreads, and evenntal liberalization of entry re- Fwue 5.3 Spreads in Financial lnternediatioi, 1979-89 P-cntge pent spread - I I r @ _ l / 14~~~~~~~~~~~~~ - ---' t -, ! fl---l --- -/- .- . . \ ; . , ~~ ~~~~~~~~~~i I o - 0- , . ' ,-m'- 1- 0 ~~ ~ K . ;(t,~ . 2- 1 1-' 1979 1930 1981 1982 1983 1984 1985 1986 1987 2988 1989 a. Indudes countnres in Sub-Saharan Africa and Nonh AfricL S:ournranamdonal MoncrccyFund dan. 2;7 ..-A GL EL stricdions made banks in East Asia rlaively efficient compared widt those in other low- and middle-income economies. Regulated spreads are not always binding, however, because banks can circumvent them by charging fees or requiring compensating bal- ances. Banks in Japan and Korea, for example, charged compensating balances. There is some evidence that these practices were sufficient to circumvent the regulations in Japan (Ueda 1992). Intermediation in- come of commercial banks in Korea has been higher than the simple dif- ference between lending and deposit rates. In contrast, regulations on dcposit and lending races have been binding in the state-owned com-- mercial banks of Indonesia and Taiwan, China. In Thailand, commer- cial bank lending rates tended to bunch around the ceiling rate, and interest rate spreads increased slighdy after the mid-1980s due to a drop in deposit rates and the relatively unchanged lending rate ceilings (an in- ternal World Bank report). Creatng Postal Saiugs lnsoboans In addition to effective prudential su- pervision and regulation of entry and spreads, all of which encouraged savings by ensuring stable banks, Japan, Korea, Malaysia, Singapore, and Taiwan, China, established government-run postal savings systems to attract small savers. Postal savings systems offrecd small savers greater security and lower action costs than the private sector and were therefore particularly effectivc in attracting to the formal financial sector the savings of low-income and mral households Japan established the region's first postal savin progam in 1875, with the explicit goal of fostering savings of rual dwelers and people with low to moderate incomes in the cities and towns. Undl then, such people were effectively exduded from the financial ssten, which lacked rural networks and discouraged small depositors by requiring high minimum balances or paying very low interest races on small deposits. The Japanese government heavily promoted postal saving among low-income households and made the interest income on small postal savin deposits tax free. Similar institutions with the same savings mobilization goals were es- tablished in Korea, Malaysia, Singapore, and Taiwan, China. Like Japan, both Korea and Taiwan, China, have granted tx- exempt status to the interest income from postal savings during long periods. In Tai- wan, China, postal savings offices account for abour a third of all finan- cial institution offices, and the postal savings service has longer business hours than other financial institutions (Shea, fbrchconiing). In Malaysia and Singapore, where postal savings have also accounted for a large pro- ,.. - - :.... .: STRATEGIES FOR. RAPD portion of domestic deposits, the governments separated management of the savings system from the post office in the early 1970s, when the proportion of postal savings in total savings dedined, cvidendy because postal employees were not enthusiastic promoters of savings. Even so, the savings system continued to utilize the post offices as a deposit- taking branch network. Postal savings systems can be an effective way of mobilizing house- hold savings, provided that governments take care thar the cost of ad- ministerng the systems does nor outstrip the benefits. Geneally postal savings systems piggyback on the mail-delivery infrastrucumre, thus min- ining overhead and fixed costs. In the early years in Japan, the system often operated out of the houses of wealthy landowners that were al- ready serving as postal branches. As postal savings expanded and the ad- ministrative burden rose, posta officals demanded wage increases. Overall, however, the postal systems cost of accepting deposits has been much less than that of the private banks (Mukai 198 1). Fored Savins Besides the above measures, some HPAEs have tried to compel savings through several measures, including mandatory pension schemes and restictons on consumption and borrowing for consump- tion. Because these measures consttute more active interventions in markers, their efficacy, even in East Asia, is open to question. Restric- tons on consumer choices, induding the basic consumption-savings decision, have welfare costs. Moreover, similar efforts in other economies have been spectacular failures. Examples indude the Wide- spread deprivation and massive waste associated with forced savings in the now-defunct command economies. Three economies in East Asia, Japan, Malaysia and Singapore, have well-developed, mandatory pension plans The impact of these plans on aggregate savings depends on the degree to which they subsdtute forvol- untary savings. Evidence of the impact of pension funds in Japan and Singapore, the only two economies where the issue has been studied, is incondusive. One study of savings in Japan found pension funds had no significant impact on tocal savings, although they did channel household savings into financial asset thereby contributing to financial deepening (Delde 1988). Anoher found a small negative effc (Nogui 1985). This SU that the Japanese pension fund squeezed out a portion of volunary savings Conversely, there is evidence that Singapores Cenal Provident Fund boosted aggregate savings by about 4 percent of GDP dur- ing-the 1970s and 1980s (MonetaryAuthority of Singapore 1991). .219 * jILA~~~CLE The lack of consumer aedit to purase housing, constmer durables, and other consumer items may have induced increased household savings in somc East Asian economies. Bank regulators in Japan, Korea, and Tai- wan, China, rsriaetd credit available for consumer durables purchases. Maki (1993) offers evidence that the rapid incrase in savings in Japan after World War II was driven by the need to acquire consumer dumables, and he shows the same pattern existed in Korea and Taiwan, China. As household incomes and the demand for consumcr dumbles inacased, sav- ings as a proportion of income rosc rapidly. Once the excess demand for consumer dumbles was mem savings rates smbilized and even declined. Have Pordes Increased Savings? How effective were policies in the HPAEs aimed ar increasing savings? Favorable growth conditions and good policy fundamentals account for most of East Asias superior savings performance. Maoeconomic sta- bility combined with solid prudenrial regulation of financial institutions created an environment in which savings were secure and retained their value. Limits to competition among banks and the willinpess of gov- enments to rescue financial institutions in trouble had costs-reduced efficiency in the former case and fiscal costs in the latter-but these were moderate and may have been compensated for by the increase in savings engendered by public confidence in banks. Postal savings had the virtue of being both secure and accessible instruments for small savers. Gov- ernments in Japan, Korea, and Taiwan, China, as wel as Malaysia and Singapore, operated these instimtions efficiently, tapping a nontradi- tional source of financial savings at rdatively low adminisa tive coss Have measure that required consumers to save more than theywould have voluntarily increased economic wvelfare? The answer is ambiguous. Restictions on consumer credit raised savings rates tempoarily in Japan but at a cost in terms of consumer wdfar Forced savings through pen- sion schemes have not been a major ficaor in increasing saving rates in Eassia, except perhaps in Singapore, but public savings are an impor- tant and complementary source of high aggregare savings. To jusr in- trening in households' savings deasions, social returns on the incread savings must exceed the privatc oppomruity cos detenined by the rate of return on investment. This would be the case, for example, if total fic- tor productivity growth were positive and a significant component of overall growth, either as a onsquence of dynamic scale economies or of 220 STRATEGIES FOR RAPD X i.* cxtnalities associated with investment. The social return also must ac- ceed the private return to jusfyr intervening in households' savings deci- sions. In Japan (during its rapid growth period), Korea, and Taiwan, China, there is cvidence that economywide productivity growth has been positive and rates of return to capital have been good. In Singapore, how- ever, evidence suggests that aggregae total fctor productivity growth is negative and rates of return to capital are low (Young 1992). Singapore may have compelled its consumers to save too much. Explaining East Asia's High Investment Rates Q JNE OF THE MOST REMARKABLE ASPECTS OF THE HPAEs IS their unusually high rates of private investment. Private invest- ment in the IIPAEs averaged about 7 percentage points of GDP higher than other low- and middle-income economies (see chapter 1). As with East Asia's high rates of human capital formation and savings, invsment has been fostered by macroeconomic stability and rapid eco- nomic growth. Growth, savings, and investment interacted in a viruous cirde as high investment initially spured growth, which resulted in in- creased savings to support continued high rates of invesnnent. In addi- ton, two general conditions have been particularly important to nvesntmen. Thes-secu property rights and complementary public invesrnenr in infrastructure-are key elements of the market-friendly institutional environment discussed in chapter 4. Security of property rights is so obvious that it might easily be over- looked. Where foreign or domestic private investors fear complete or partial expropriation of their investments, private investment initative is naturally reduced. Thus secure property rights embodied in the legal framework were a key to high investment rates. In addition, formal and informal mechanisms leading to enforcement of contracts were an im- portant contributing factor. Data also suggest that the developing HPAE governments have done a better job providing infrructure themselves and, at times, creating the conditions for the private sector tO provide it. Limited econometric evi- dence suggests that these investnents, at least in the cases of Korea'and - Tiwan, China, generate very high rates of return (an intemal World :Bank report). Easterly and Rebelo (1993) present cross-econonly esti- . .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~22 CLE mates that infrastucture investment has a high payoff for economic growth. These investments havc bccn highly complenenary to private investment, especially in export-oriented manu&cauring activities. Table 5.7 shows some indication of infrastructural development for the developing HPAES and selected low- and middle-income economies. Even ifwe set aside urbanized Hong Kong and Singapore, the HPAEs' per capita electricity generating capacity is equal to or exceeds that of economies at similar levels of per capit incomc. Similarly, from 1975 to 1985 the five developing HPAEs with nontrivial mral sectors increased telephone service much faster than other economies with roughly the same population per phone in 1975. For example, in Malaysia the num- ber of telephones per person roughly quadmrpled, while Brazil, Mauri- Table 5.7 Direct Provision of infrastructure, Selected Years -- ;-W ta r FB!N cenw f. Paetge of w.c4 e odaivr roads e, padca ; telhone' Ec; rsonomy^ 0-i- 1971- *1986 _.1975 -1985: g RPA Fs - Hong-Kong- . 100 - 10 0 160.6 -18.3 s ^^. . ..... ... . 025 62.2 77 . 4 '205 coe.Rep. of -14 .2 542 332 . 5621-- 2- 5 . $a1 .yuia ' . 85 80 2818 42 12 S ;-i n , '= *r,=, ,'-. i 95 8 182.3 3- 13t3; Thailand ' 67 396 --18.2, 174.2 13 6 W 49: ,~~~~~~~~~~~~~~~~~~~~--8 -:351'7 < C-' - : ; .... : ~~~->-: -:11.6 ,...; 12.4. 16 i95.-- -;309.0, :.......... 24 ¾ 167 -.C'Ivoi-e .0.0 7.9 1' 98.6: '- 1 Gn . ..528: 79.7 India L 34W ~~48.0- 209 89.5 '352, -20 muridu 0 935 M'-'co ,60.0 3 21 KU W-- -hilippine. - - - :-- - ..12 5 - 1 .7 . Si tnlt 66 1P7o-mlj FL 7~~~~~~~~~~~~~~~! =2.2 STRATEGIES FORRK R.A%W tius, and Peru had smaller gains. Although increases in infiastructure are clearly correlated with income growth, recent econometric evidence suggests causality may run from infrastructure spending to growth (Easterly and Rebelo 1993; Canning and Nry 1993). Coordination Pmrblems that Hinder Inrestment Beyond establishing ivorable general conditions for investment, most HPAE governments have employed a variety of policies designed specifi- caly to address coordination problems that may hinder investment. These problems are derived fiom shortcomings in markets for informa- tion and in markets for sharing risk both types of problems exist in all economies, but the problems and the consequences tend to be worse in developing economies. For example, such developing-economy informa- don problems as poor accounting standaris and a dearth of merchant banks and other institutions to monitor corporate performance mean that bond and equity markets are often weak or entirely absen.'7 Simi- larly, while market mechanisms for sharing risk are limited in all economies-firms in developing and industial economies often cannot buy insurance for the most serious risks they face-the consequences are more severe in developing economies in which lower wealth levels make households and firms more vulnerable. Furthermore, the transforma- dons inherent in development, such as the creation ofnew firms and new industries and the absorption of new technologies, heighten uncertainty. Polcies to Pmoute Inmeshment HPAE responses to these problems can be grouped into three broad categories FLrst, as these economies have ma- tured, policynmakers have gradually increased their atrteion to the creation and improvenent of bond and equity markets. Hownrer, while these mar- kets are playing an increasingly important roe they were not generay a key factor in mobizing investment during the HPAEs economic mkeoffi. Second, many Hwss created development banks to ease constraints in long-tern capital markets, as wel as specializd institutions to provide fi- nancig to agriculmre and small- and medium-size enterprises. These c- forts-appear, on balance, mt have been more succcssfi4 in the HPAEs than in other developing economies, for reasons we shal explore Fmally, HPAE gov- ernments have encourged investment with a wide array of other mecha- nisms designed to inacase the atrcdveness of private investment.. The remainder of this chapter describes specific policies in each of these thiee categories-supporc for bond and equity markers, establish- 223 ment of decvelopment banks, and mechanisms to increasethe attraative- ness of private investment-and attempts to evaluate whether each was effective in increasing investment. We devote particular attention to the third category, whiich includes several policies that involve a high degree of intervention in markets. At the simplest level, many HPALE boosted corporations' retained earnings by adopting low corporate taxes. So'me also reduced the cost of investment by adjusting tax, tariff, and exchanige rate policies to hold down the relative price of capital goods. Several gov- ermecnts, particularly japan, Korea, and Taiwan, China, tried to reduce risk for investors by shouldering some of it themselves or spreading it across other firms. The northern-tier FIPAE-s also stanld out for their ef- forts to increase domestic investment by restricting capital flows, Finally, several HPAEs, again particularly the northiern tier but also Malaysia, sub- sidized investment financing by slightly suppressing the interest rates on savings depo'sits and corporate borrowing. This last mechanism, which we term mild financial repression, plays a k-ey role in capital allocaidon, the subject of the next chapter. Bend amd EqLit Markets, Developed after the Takeoff. Despite the phenom- enal growth ofAsian co:poatons, bond and equity markets have played a relatively small role in Asian finance. As table 5.8 shows, bonds on average between 1970 and 1990 accounted for less than 1 0 percent of the net financing of corporations for the five HPAEs for which dataexist This is higher than the share of bond finanding-in the oECD economies, however Ironically, pofides that contributed tco some of the HPAEs success in developing other financial market institutions may have been inimical to the early development of a corporate bond market. In Japan, for ex- ample, the long-ternm credit banks govermnment-protected monopoly on the issue of long-ternm debentures discouraged the development of bondi markets.'8TeTa government created the same monopoly for its pri- vate long-term credit banks. In addition, tb"~ lack of government securi- ties in most HPAEs, a direct outgrowth of their success at minimizing fiscal deficits, has deprived embryonic bond markets of the most widely utilize-d risk-free benchmark rate. This has impeded corporate bondt ism- sues, since markets find it difficult to determine simultaneously the risk- free, rate and the risk premium associated wiiA a specific corporate bond. In response, Hong Kong has issued bonds, re.-n though the government did not need the financing, to establish a risk-flee benchmark for the corpomte bond market. Malaysia and Sina ar now considering following suint. .22.4 STRATEGIES FOWIRAP-r"DA& -table 5.8. Net Financing Soumes of Nonfinancial Corporations in East Asia and Selected Other Developing Economies I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~'S57D.75 Z'~~~.6 $245 LI14 Malajiula. 126-Si I $61 36.125 L 14 .- . 'I 0,W~3N ~ 423 137 ~~ 1¶32-W 293 ~~~~~311 IjaV . ~~ 3a~1(tnpI2p.frmr&uu)..' - -3. 4 971-1113 t1 I L-) 7.7 1564-US' -~~~~ 26.0 .7 I&U LO 4 1m971-$ .0 '23j.- 47 -- L32 ~ Ca.uu3.I97-Z5~~~~~ 80. 5 '0i 10d !P jan23.2.&) 1. LuSti. v~74;- 23 2 7 8 ' T..wi b.CaWuW.K m.a1dua~~~~~~~~~u~~~y.,~~~ _:; immu' a w0.A,i~ -VNIIA RamilkWm B.ubumiawamuiEsu Wu hl.1z9h2.~ Iua a hitoicAym been small.Makfhzds appeamrs t b changing- as the Asian- economiecs mature. Asian corporations with their headquarters in Japan, Korea, Taiwan, China, and othe-r I?PAE aeincreasingly issuing. corporate bonds, drawing on the 'strength -of their worldwvide reputations and access to international financial mrar- ks'19 Recognizing- the value -of a healthy bondmarket, Japan has *eased lalws thar discriminared agamnst corporate bonds. Similarly- Thai- land, a.nd -rec'endy. Indonesia. hav strengthened rating and reg ulatory agendes, -while -Hong ,Kong has refornictl its . regulatory- structures.: Malaysiahas in troduced acrotebond rating-agency. Thus,_ whlile. 'corporatebonds have -notibeen? imnportant -in Asia's takeoff,: theyr are - n !14r - ow .omorumavntaaeealorne hstrowCr,,>.mdlnon; xbat'flpC c-..s% rte- {g ' E aji&ht w*kw4ti4<_4$SF-> 4nordi#Pnd fr^½Mstoa ts)boni.tacnnve~idrssi. rte 'lmiberonibo£anM mictaves:ant .~~~ .rDooe>Doe ite: baafpmeERm-@edlafsicrtoqsnutoC*Ddae - 'zi*:zem ,J 49.W4.,. 0,fl ' tl 5 f1 e ^D¢ 4$i.9-= I .;-i .~~~~~~~~R~p . XtnEt s: .i&l i .-a:- S ; f --kii N........................Wj { 4swrs emosd(firsJ. .lE:reaLtdepoitr atesas iw itres-onldt -.ec: oemihan a6ouvzRcDrcenq-, .t.s - > ,",snewep.zSerm 68.;e. &1t 'aifer- irg;ont dw mtnznllts-isl-mmedoucti;. r ' on+ A * _ w a c.en I_ . Id « - S.. t '+!F 4 .%>l- # {-- -tuv--tr>> s rv;s.> t, '-tF H: : U - 9 ] t ! ),.- rv-S , S 7 \*.M1 s- X$Sssrtmus y>dforeacchan^ ~~~~~~contos.ce ponsdw7e.cmpabl Ronn sann= mxtl!oruenv cuuenw,te oryaetd::-.. dA acul- z:pRrC=sOn,,wo eavinten ded:ir- .-i-f : f^we*,a ts^ r.4 £} -v>l-t-) g4i aiS^- ii4tz1- wt1E -;Si -;;f4 -ltiH .Mi*navnolX -icmiemltGsaco cosraire-mti=ur.a: - f~~~~~~~~~~~~~~~~~~~~~~ _N 2 P i ¢ itrFt..4->>,'gsSw4'St;t ,-; -tU -J;.>a)>iJ--.i j>S;<, 6ti-;5f ~S- t',S:- i ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~,r v =tr2aonnPnrllgno-vnvscLitg.de6tea sr-eecv4ec,.ak-o | . 4 F l- = oSaN ,f ,4 J t: ' _fi< . - ' r St-, , *-; 0 2* s e Xs > ; ; vE;uldshor-rem;tfinancrqprn=svsitrmalsmm4ia;UkI, b pem6d t -vat.FSeracser. : . I g ; ;,= a ' Srsge wer fNtf .w--- ¢\*1.tPGK> -O f mof- . . S.7;1vt ; mE~~~~~~~~~~~~~~~~~-m "Proodx tssXadMeneeRnf>Bcue.i zn- S'6Stae>-en f u bz matH.fRftAS;sf-he;Ho S-,R,;wiftW:2.S'sruXs .{2;;7t>iv1X;742-- it^.=x < 4.- *l -. -r-9. Pj r ~~~~~~~~~~~~~~~~~~- a.-::' , A.,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ': . : - . ' e~~~rapi finfancial Av ind twirdl bankt r eguatossquee th fo dcepoit. Svrate | r - 0 - - - stuleding rate sp=,d, ehansth rhatidie that resuts pId into depositorsla ;.~. - f' . 0 -7-' : .0 : households) wercpaed on to borrw s rorld ations). 8).e' ' . - 1, ' '' . ' : - ' ;: - - '-. Nonethesltni WSuring tleatlo in raest racsae posauedo efetos'ofmow- '; --' S X" -- . X . - ' . ers, finiandial repression with geneitie. mess dmanct frae :ayic SevAeralc d-': .-~~~~~~~~~~~~~~~~tde hav arue tha thrd th o.- ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~m at reslt is infiin and has . .. -f0 -~ ~ ~ ~ ~ ~ ~ ~ ~~eav cosqu-e fo grwt (Wrl Bank 1989: Fr 198. 0 . :-- : - - , j z X ,, thsaye, e.msancs.-stngin2da.r pes :on !'paricuarl *,~ ~ ~ ~ ~ ~~~~~~~~~~~hs resutin in' ne''4rtv 'ea -' e,'-t'o mod--,- 0,' ' -:f ,'' .- . :0:,-': :.: - STRATEGIES FOR P.A P,r. Evn widhoutfinancial repression, banks do not allocate loans to the high- est bidders. Since bidders for credit offrer promises rater than payment, banks must evaluate the bidders and decide which promises are most likely to be kept. Since rationing already takes place in financial systems, the additional impact of moderate financial repression can be relatively small. Indeed, credit rationing may actually be good for growth, if the mechanism used to select borrowers generates incentives to perform. For -emmple, Japan, Korea, and Taiwan, China, allocated subsidized credit on the basis of past performance, usually with exports as the criterion of suc- cess. Since the value of subsidized capital was high, firms perceived high marginal returns from greater effort and competed accordingly. We ex- plore this finction of credit allocation in detail in chapter 6. Did moderate financial repression hinder growth? We address this question by examining the relationships between income growth, real interest rates, and inflation using two statistical methods and two data sets: a sample of nineteen economies that includes a subser with positive real interest rates, and time-series data for Korea and Taiwan, China. (Details are in appendix 5.2). In the nineteen-economies sample, infla- ton rates are far more important in explaining growti than real interest rates. i the subset of economies with positive real interest rates, the in- wrtest rate is statistically insignificant and actually has a negative effect on gowth when inflation is included. Similar tescs can be made using time-series data for Korea and Tai- wan, China We use the difference berween the curb rate and the insti- tutional interest rate as a measure of financial repression. Its coefficient is statistically insignificant. In Korea, inflation and the curb market rate (our proxy for the equilibrium interest rare) significantly explain income growth. In Taiwan, China, none of the variables is significandy related to trhe growth rate. While these tests are not condcusive, they support the view that moderate fnancial repression did not necessarily hinder growth in Korea and Taiwan, China. Which were the factors that contributed most to the HPAEs7 extra- ordinary record of accumulation over the past thirty years? Judgments are difficult, given our incomplete understanding of the costs and bene- fits of all the policies used. But on balance the common elements across the eight emnomies tend to favor the importance of doing the finda- -mentals well-encouraging stable and predictable macroeconomic envi- ., = , . . .~~~~~~~~~~~~~~~~~4 .~~~~~~~~~~dr '. ronments, universal primary (and later secondary) education, sound and solvent financial insctations, secure property rights and comple- mentary public investments in infrastructure, and low relative prices of investment goods. Efforts to improve the institutional framework for capital market development came later in the process and were not re- sponsible fbr takeoff. In some cases well-functioning development banks were a positive but not a determining factor. More selective interventions-forced savings, tax policies to promote (sometimes very specific) investnents, sharing risk, restricting capital outflow, and repressing interest rates also appear to have succeeded in some HPAEs, especiallyJapan, Korea, Singapore, and Taiwan, China But the potential costs of these more selectve interventions if misapplied can be very high in terms of consumer welfare, and strong institutional capability is necessary. They would not have succeeded without the im- portant monitoring and disciplinary roles performed by the banks and public institurions of these economies. Where other East Asian economies have lacked this capability-in Indonesia, Malaysia, and Thailand-ef%ors at selective interventions to promote rapid accumula- tion have been generally unsuccessful. Appendix 5.1: Granger Causality Tests for Savings Rates and Growth Rates EXISTENCE OF A REIATIONSHIP BETWEEN VARlABLES proves neither the existence of causality nor the direction of in- fluence. Granger-type tests are ofien used to test for causality be- tween two variables, in this case the growth rate of real cD per capita i) and the gross saving rate (s). Tests were done for regression result fir the eight East Asian economies and the United States. Vector auto- regressions were performed on equations 5.1 and 5.2: (5.1) g=X g',- + s,+,i k-I k- (5.2) s= S*-k + gr- k-I k-I 242 STRATEGIES FORItAI I '. Then F-rests were performed to see whether the lagged savings rates are jointly significant (equation 5.1) and whether the lagged growth rates are jointly significant (equation 5.2). Up to five years of lagged rates are included. The results show a causal relationship from ODP growth rates to savings rates for most East Asian economies (see table AS. 1). The Granger definition exploits time-series relationships to identify causality. By this definition, g causes s if the prediction of the current value of s (s) is strengthened by usingpast values ofgt. Thus, gcauses s if c2a(s1 IA,) < cr2 (rsA%-gt ) where A, = {A*ik< t- 1I denotes the infor- mation set that includes Ak for all past k up to and including t-1, and a2 (st 1A) is the variance of s, about the best predictor by A-,. To defline causality in a time-series model involving two variables, we make two simplifjing assumptions. First, that the set A, indudes s, and g only, and not a third variable. Second, that tansformations 7Tand T exist, such that ST Tss, and G,= Tg, are a pair of linear, covariance- smtionary dine series, and that S, and G6 preserve the causality relation- ship of s,andg,. Ordinary or seasonal differencing is an example of such :ransfonnations. Causality can be defined by using several representations of the time- series model for the independent and dependent variables. The autore- Table A5.1 Public Expenditure per Student on Primary and Secondary Education (US. dolirs) -~ ______ -. 1975:18 28 ,cmy -i , ' d P - Seconday Primac - y ;C -.Hog'Y . 1923.-3 180.5 8.15581 810.2* 'co ra,Rep.of-' 90 147 512 32 ' 3898 352.5 ,;j 09699 1'aaysi=5. 572.5 PT ~~~~~~~193'3. 314.7 -8340O t0xnanana .176 -67i i49 1881 2 9 ~~ 'Brazil 76.9 1~~~~~~~~~79.1-' 15.4-22. t~ ~ ~~2. 1r2 a e I 778.0.' 0j 4 175 30.9 ~~~~ ~~~~~~~~27 9 62.3 3. - I _-'3 . 7 India 53 713 ~~~~~~~~5038 `6 M† Mexicot. ~23& 566 1210 312 ' 10. 875' )3 . . ,.., fl wwevtAunat.wl c tea, 'UNFUd--. . ' '. ta. 243 Table A5.2 Tests of Bivariate Granger Causality between the Real Per Capit GOP Growth Rate and the Gmoss Savings Rate Si ~~~~Sig *~~~~~~~ ~ ~~~~~~~ *2 S; S 4 .m .Si NS~~~~S 3;.KS. . . g 0 ~~~~~~~~~~~~~~~- , 1S SBal ;-i -,*-tX ; ~~~~~~~~~~~~tjas S - 8 MSN Z ,;t-fI.. . .. g N'S Ng~~~~~~~a f ~~~~~~~~~~~~~~~- S, , - ' Kqq, -S-- NS MS Mg~~~~~~S MS ~~~~Si ..~~~~~~~ MS' , -~. -Si Mdapua1- I Si . S .i -41rs .2 PM- .~~~~~~~S . , KS 16 ,~ .~~~~~~~ -, 11 ' 115_ ; ;. -S 3 la S 4 In.i - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~2 MS ' ' : Si K~~~~~~~~~~Si, Ti wuyo-u "S .- ,,- - . . - - :7 - j . ; ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ luhLgO8 '' -; KS . gX - md &" 244~~~~~~~~~~ STRATEGIES FO EGERAYIF gressive representation was selected fbr this study. In this representation : and by Granger's definition, ifg does not cause s,, lagged values ofgcan- nor add to the predictive accuracy of sp, given past values of s. Using the autoregressrve representations, s. can be regressed on its own past values and on past values ofgto rest whether g, causes s, by resting whether the coefficients of the lagged values ofgare significandy different fiom zero (Granger 1969). The results indicate that past values ofgare a very good - predictor of s, but that the reverse is not true (see table A5.2). Appendix 5.2: Technical Note on the Relationship between interest Rates and Growth P; = REVIOUS STUDIES, NOTABLY BY FRY (1984, 1988) AiND GELB (1989), have tested the relationship between real interest rates (R) and the rate of GDP growti (G) using the foilowing equation: (5.3) G= consrant + -IRu+ Gelb's equations also indude a shift variable to account for structual changes in growth rates following the breakdowvn of the Bretton Woods monetary system in 1973. Both Fry- and Gedb apply ordinary least squares (OLS) estimation to pooled cross-economy dine-sries data Both studies show a consistendy positve and statistcally significant relation- ship between economic growth and the real rate of interest But equation 5.3 has major shortcomings. Fr, it fahils to disinguish between small and large degrees of financial system repression (the dif- ference between the equilibrium rate of interest and the controlled or in- stitutional rate). High negative real interest rates have severe negative .ff&ets on the economy and may have driven the results of equation 5.3, whereas moderate financial repression may have benign or positive ef- fects that are not captred by the equation. These difirences can be ex- plored by spliming the sample in two according to degree of financial system repression, putting economies with negatve real rates of interest in one group and those with positive real rares in another Second, financial regression is not the only reason for high negative in- terest rates. Other economic problems can afifct interest rates as well. Many ofthe economies in the sample with high negative interest rates also 245 had high and often volatile inflation rates (Murdock and Srigiirz 1993). Introducing inflaton to the equation, this corrects for the rce of inflation in analyzing whether financial repression has a negative effectr on growth: (5.4) constant fPi ,R+ J321NF + :. Third, investment opportuities also aff*ect real interest races. A &ll in the real instimutional rate of interest does not mean a change in financial repression if the equilibrium rate falls equivalently because of lack of in- vestmenit opportunities. This study, thercfore, introduces better proxies for the degree of financial repression in the single economy regressions: the difference between the curb marker rates and the banking system's interest rates. (These are available fMr Korea and Taiwan, China, but: not fMr the other'high-performing Asian economiies.) Fourth, OLS estimiates couldt be biased when the regressors are corre- lated with the error term, as in equation 5.3. If an economy experiences a sudden increase in productivity, for example, the demnand for credit will rise, and inflation is likely to tu'rn out to be lower than expected- Both ef- fects will result in a higher-than-expected real interest rate. Thus, two- stage least square estimation (2sLS) is used in this study. (In the event, the results for OLS and 2SLS suggest thar the condlusions in this study would not have been substantially changed had oLs estimates been used.) Using pooled clint-series data for twenty economnies, this studly esti- mated equation 5.3 applying a two-stage least squarcs estimation to both thLe complete sample and to the subsample ofeconomies with pos- itive average rates of interest. Avenage GNP growth rates were used as proxies for the growth rate, an1d real lending races (from the Inter- national Financial Statistics' database) as proxies for real interest rates. T'he 25L estimation induded the previous year's interest rate as the in- strumental variable for the real interest rate~, a year variable to account for any sectoral trend, and two measures of an index of relative income because standard growth models predict that poorer ecoonomies have higher equiibriumn growth rates. All regressions indlude dummy vari- ables for each economy. Table A5.3 reports the results for the imact of the real interest rare. and the inflation rate on the GNP growth rare. FoIhCsml of twenty economnies the real interest ratechas astatisticallysignificant and positive impact on growth (equation 5.3). But when inflation is induded (eiqua don A),the cefficient of the real ineetrare is no longer statistically significant, while the negative coefficient on the rate of inflation is. For 246 STRATEGIES FOR, Table AS.3 Parameter Estmates for Interest Rates and Growth , '; aJrnsr&-vru.' Rea hien A, 'T, tuft* nn.. . ,F- Rea *cur = *b.,, ~~.L. rw&ratr'- rate ~~~ rat interest rnf 'Whole sample (2S):- - . - ,.- -'.'-' P1=2 92 ; ;7s . . r .;. . ; Ekiuado .Rn 1 . . - . . 0.148 5LAJIA tT-.>g - -., ; . . Equaiontl ' ' -0-: & Equarioni).. 0.391~~~0. t i Equaon 2 ;'w- - -. 0.059 - -~-::~;0363u- t Eqarin3 -- - . A389 0 x0 438a M6 E a C~1;perat lewd.,' WattTrto- sSt ae Icr............,;><,. -~ Ieastst regesons were estimated o n s hmp b of w- i otlie suhsampl.ofcaniameoLswason the nm :l ' -: - on - . . ..-* - . - 2; >: < - ~ .;:-;frA. ;:. f!:.y*:............... ,.. r .. , . . - ...... :- swarcw&iBank dima the subsample of economies with positive real interest rates, the coeffi- cdent of the real interest rate is not statistially significant cxcept when inflation is induded. Then the coefficient of the real interest rate is sig.- nficat and ngative, suggesting that lower real interest rates may have had a positive impact on growth. The coefficient on inflation remains negtive and significanL 'Table A5.3 also reports estimares of a time-series analysis of the rela- tionship of interest rates, inflaton, and growth in Korea. A third equa- tion includes the,real curb rate as a proxy for the equilibrium interest race. For equation 5.3, the coefficient ofthe real interest rate is positive, 'large, and statistically highly significant. When the real curb rate is in- dudced, the real interest rate is no longer significant, but the real curb rate is. When the infladoti variable is added, the eal curb rate rerains highly significant, but the real interest rate and the inflation rare do not. 'When the same tests are perfonned on time-series infboration for Tai- wan, China (not reported in tale A5.3), the esdmated coefficients have the same sign and similar manitudes as those for Korea, but the standard errors are much larger and none of ti coeffiacnts is statistically significant. 747 fI. F--ACLE= . - Table AIA IcomeDo Rzobution GNPper shareof . arof: tope :: - -..i- cap ib omm 2 :top 20. botm20 Ec--fonomy, ;- .f(perytr) -t of pferentof pertof ; :pon~f;= =2 1: I96590 Ysrr eholkoouse hold hovsseholds : .HogIong- 6.;62 1980' --- 5.4A -47.0 - 8.7 -'Indonesia) 4.50 0>i1976' 6.6 . 49.4-' 74 -Japa 4.r 9AEV - f i1~ ; 1969` 7.9 41.0 5.2 f - *0 -XX1979.. ' 8.7 ;375 431 - Korea,R ep. of .- 71 1976.- '5.7 453 7.95 4Maaysia"' . .0' - 1973:- 3.5 6.1; --16.03 = R - . . 0;198T'0.:4G6 ' .51.2 R E11.1 - tq- . -gapo 6.5: 91982683 5.1. 9 .-.6- Taiwan, China. 9- 1976 95- 35.0 -3.68: h Thailand- 4.4 - 1975-76 5.6 :49.8- 8.89 * - : , - . .I - 00--- *. f.OzhtAid PhilPEippines000: 1.i3 19e g70-71 - 03.7 .;53.9 14.6 *-0 --0 -;; : . - - ;- -l985b- f51985 5 .48.0' 8.7 Brad! 3.. .... ; 3 1972 - 2.0 6 3i3 1983' 2.47 626 26.0. -. taiRica ... - .:. 1.4 '1;:-971. 3;3'.......... - 54.8 - -16.61- 4- .. ;. } :t. ~~~~~~~~~-.; - 8c ...... 33 . 54 8--S ;:--1 .51- 1986" ~..3 -5.6.5 -Mco 2.8 1977 Ž2.9 57.7- - 19.9 p~. * 0.2 1I72 ;L- i-9 -0. 2 . 11 -I 1- { - - ^: --:- - - - . 00 98546w 44< - :5139 179 . V&elzuea - '_!. -10 0 -.-1970 T .- 3.0 -:. ;.:-54.0 . .- .................... :;.*18.0 {; 7197' ;76 0~~~~~~~~~~~0 i - --^ Boiswana. =' :i -- -4- --- 19856 `2'.: 5 0 00- Cbce'Ivi-re 05 198- ;8-T 5 0 527 10 5 5 Gh-.na ',-. -14-. 19s8j9' j rt dia 1-9 '19645 6.7' 48 -9 7.2 1975-76 7 .: . 49:4 7 --- 248---:- .- : ^ . , . - , : X , 0 - . . - . - : -.n s04.. 314.... ± a.-~ ~ ~ ~ ~ ~ ~ ~~~t'.nerprcznmoc -.. r'| -s.-u 1 i '' ' . "' ''- ; i '. q ' '- ' . '-,: '~.D a~ r .o bou'e.i"k.c'p.ndz'ut. : STRATEGIES FOR RA:PtDiW Table A5S5 Natuire, Causes, and Resolufion of Banking System Crises Naturnfnere. Nature ofbiot .:'Enfisrj ::s Causes or re : T-TnnHon n a . .-- Nine deposi:-aing c- (a) rg exposurc to real The govement rem reg- -1982-83 . estateiled . . e lending f-raud ulatoryand audiuing system and - and mismanagement. liquidated troubled DTCs. * andw wak prudential t* regation. Hihinternational g;or&., inh teiiga m or:. 1rg* itrst t- res 1{o4 Kong, Eight banks, including a maJor large-exposure to real The gormentookoverthe : 1;8-86 :: -international ankwere in. estatelendingand spillover lagebankcandintrduced ne 'fnancial difficultie . effers fiom 1983 crii in cludingtopexecu- since che banks o-;ccL- t-v -sewn&de from tce Jargeie o rn- [7stwi .. . ' ;..- - - ' - DTCs aS subsidiaries. - me. nrial ban.kThose no: taken ovcr by thgovernrmenrreaivedcrdt - b. .... . -- - , : ;- - --:-. tr from toher com rcialnk Japan. ACenini Bank repor has esrti (a) Excive Encourage rnergerioFweaker ; 11; ...- -l991 mated the size of problun loans --- real estate lending, corn-1 banksb with hthier ones, of the top 21 banks o be. prisi about 90 p- r o b provide.- betwc _ 3.5-4.8 percentof centofthe bid loansof d bki systrem:ase,.S. .n and a epdeclinein tO e ofaount.to'be wrt- ' ral esae prices ban-s- ten off2s about 1 1-5 percent -nc Ina-ecjuare - No-~ lon to be banking system a. transu t we .s;r- P- ---re: ale to dalinstidon,and dthecost o, - r exposuredititugh de&eoftbsie * .. 4, - - . - ' . Ins to nonbanic among commercia bank.; - ;. , -a-lure of one deposit- -k;ng (a) Preoinant causesaThe goverinrrescud -24 7 '1985-88 ~~~cooperatie ini8cuec ru n speculationi, isletcooperatives and con-~ 4. { 88.rins: n`32(if35o) r. d : sldar In addituo;n 4! (o38) bnsstOCks iic npM'.Te eta and4 (of 47) i' naiicecompa . Bank = Iedfe ,ic -. ,.gee' (b) Deterioration rtemsia- nieterealsoi inacial dis- l i OXi iiit of , s b , r '~~~I -' ' '. :i' .. .. ~ ~ ~ s,. .butOO~~ .1C - . ..trbek~ X . - ,- ' . . . 4 ~~~ ~ ~~.. . .-...Z9 - -d, . SI"' ':ononur - . - 1' .i44~~~.;'C~~ijrr. M i.., .n7.' ,.- u.. i ae .luo opt (&4kivŽirnnre -.'- i" n.r-'$20 4..h,4.to ~ J8' 4r y0'"C-: - OUr1Uust:-&rnpan _S n N Z L'ooeratLVeS ,L' i -J, 'S44. ,,.. ,-4 249 L E Table ASS (coawiaWed) Economy NantresJiitttn.iI C Aauee fbaib ut - - ThailandL- Thegovernment'st ofb- (a) Fraudand_speculation - The government liqtuidaed 24 -1983-87 out of 50 finance companies on real,estare and finance companis and-meriged w..as estiinatedtUS$190 mi- exch agerae c--ansac another 9, andthe Centra-: lion or 0.48percent ofGNP. - ons. Bank cook overthe other 17 : 0 .X 0f: - -ii - - -; - - t 3 u; 9 ; - - ; - 0 45 .an -t sold thmto ncw mvsots -*-- 0 , Five commercial banks () EHigh concenaon of ands tinvestors accounting for 24 percent of- unscuredinsider - i n o banks) commercia bank assct were in loans. - It bouiht some shares ofrou- - financial-difficulties in -. (c Hit te bled bank ' -- *a 1 - '' < < ' >> 5 0 0"~ 000 t9B6-;87 . - g 0t, - " - - inestrae - ksocaeda-Ioarid to_ provid emnergec loans to rubd -The Rind would be finlanced by contribui-: tions frdmn commercial-banks.- ~- . i 0 - 0- - 0- -; - :0~; - . - Mat-uri-y .6 of Trobe -m-00-orriom t Gl' ;;' fChide,' :- ' X 7';EntLrc mortgage banking sys- Matitymismatchesof .Troubled mowwcre' 0 - X -1976 - 0 - - - - te was insolvent.n assets and liabilities as liqudat -' ~~~~~~~~~~~~~interestrates wosc andi - ~~~~~~~~~~~~inadequate Capital * ~~~~~~~~~~~~~requirements-,. -Chile,- Severn! financial insdiutons (or -a) Severe. r i . n1981. thegoverrncmnt bqu - *: ' :.':198b-83- . :-; t: :iabout 60pcrcentofrhefinan-, ,'1981driasi c ase dat ed:8insticunon&sIn:1983:it - .0 .-.. ;.:al ni'sln 1porolio) m nterest rts,adX rook ovceS (liqui da*ng-3nd w insolvencii-n 1988;ceniral decline i terms of- rccapitzng5).n The:Ceir -5banlholdingsofbadconcr- trade due to :ihe col-. Bank also reqs-- pe G &iiloutrct coppero ec e rate.-, cost -was est dat US$6bil- scheme tomee-±rcgnkabik lion, or 26.2 percent of GDP.r ( p ' tesd interest rate sub ltoadsupervisowo er4utpwp o ea and r2Mpaniisider and speculative lendling raie nc 'Ghana Thlicnetwot ofthe banking -.Hug foreinecag 0 7 . . .0 !, _, ' ~~~~~~~~~C ,.-,, -. 0, '''t ,';30>-.Maa -1988 system wax negativsiat ,-lose and ahppor-, -- , *c&r of restructuriii as 7 'rionirononiperfrrnng- * US-$300 million ornealy6 o 2ans 1 hr ng (a) Collapse 1in he pic;of -Government- ' 5 heavylosses.~ ~ ~ ~ ~A oi. wie o Oam oans. resieur n,' ipuin enig '~tew. capta Encourage drpe:and o L.... ......I. ; . . . . . .- - _'50 STRATEGIES FOR RAPI Economy ~~Nature offlwwal Naftre of Lalikt -1-; Eco ov . ; dt-ress cause: or rescue .~*-PhiUippinas,--- -- :. Among smaller finanacal insti-- (a).Inrcrascs in oil prices . Insolvent isttutions weas i- 5;- :1981-87 wituuons, comprising 7.6 percent - nd international inter- liquidated and depositors paid o- deposit money bank assets,w esi rats, coupled with off by ie deposit insurance 182 iilai In a deterioration' i he -. In,a~~~~~~~~ddizio an cc corporatio. ' erment intcrvened in thc two tenms of tradc. largest p bli -banks. - f Th 2 publicbankswere. private banks, accountingF .Ilb) High concentration oforestu-c-tr: in 1986, and their. more than two thirds of the. i lending to relate , par- adporios wcr put und raa asset. s .-.-ofde|itnolney ban. tienes and inadequate : epate agency under over -- ;;:-.By-1986, centl b:.lcassis- -- piud al rigon t;meat mMnaeent. Five private e.- :;: - -- :-.tance to financial nstirudons and--:- suevso, . bak reaie under......... direct.. j r- . --i4, - c !ene'X ~~6-.~ tral-banlk supervision;: am9-:outto 19.1 bmiion pesos ; E !. y:. . . Sor 3 percent-of GDP.i i;Turkey.- A- :- Afinancialcrsis eruptedwith: ~Inadequate prudential The government recapitalized- '- - .'.198245-8 -' 5 '-:',. the collapsc ofsal brokers, = superision.---- . the--a banuks an'd isisted onthe-ir 5. .-.-8- -adfeba weres dara ~ :- op-- - - erriaires opucrurt, re+ xc=ured. .- * on its prudential regulaons. ' ,Uwe .> Strs . 'Of appodirel 4,500 .- (a)Predominat &wewa Simia approach tora oth .21982-8 7. bank, 1,575 er rporen~ spcltv rea esat S&L FedraDeoi Isrc-). ;N ; E >- -'-troubl in 1987. tec en 1982 - loans:;; - nce :Corporation l riquidated~ - and 1988, e ou 424 banks banks (a d od sp Berien 1 D988, about ( X i57s9 -89; ..:-'.: 1,5O6 (ofES,147) S&Jwre 2 lœ ~~ 98Ireiducedrhev aluc -;ur;;IeC&porons.ised .- ; f al crisis uptea wid- -n T :i - 175 liqudated. n -:ofanks'aserand in goemn-e A meried 'andfi ia b1~! upcr themers'nd.rnew'ana2ei- 4> i . 1 9 w _- -- 1 89 over60 0 (o ne fift oif mcesineetrob er liaoso noe -!;4 -. i -- '. S&Ls) were solvent Toed paid t depostors ,. ... e0 S L a acquiredw - ibi-i,sl9 ~US-$13 bilion r22 (b Dergultion ofS ,s, -t rb -7 of >'thertJ5Ss -'1. ~~~~~-:.M' fiwxra.ouc'acdequateat a pru x!srctn&- - ' of 'D?. '. '' - costs'OLcLa.ofA.Lupec,rtsoFGN ..4..~., .VV~'p aennas supervision; let 'I additionlg' up Q f . , . . _ ,~ i: _lrg , b=., . 6 Sincel9i epe ~ ~ ~ ~ ~~"rceve ,, 0 .4-- . , S&. to le_h ' ' ' - _ .- , ~S' 4A 'f* _ ' _ . ' _ _ ' : __ . _ ,_ .A., -:' ~~ ~~ . ~ ~ ~ ~ 'to poo7r-q,ueprtc~ alr ecer- - r. --A. ~~~~~ (~~~~~~iarioi1987.sB KiCwewn198 (19)lindz CO'.rbn (96;Seg(18A9)Sii&ae . ~~~~ inic vi 1~~~~~~~~~~~~~~~~~~~~totspJ !ei2 ge ioanC.Q- ~il~jii15o Table A5.6 Comparison of Ba nkingSystem ConcetainRto ka ses r ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ &rp .' ,. Tnw _.e .;g -tuSI -. L E * g f ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ mi Hmlshnet ~~~193 75 77.7 .2 Indaish 28-54 59 690.10 lapi ' 2 40 37 1.- KcrsRep.J 41 50 '51 _ d qi l n 1 5 94 2 : 5 6 6 1 D I O P : . 0 0 . . 0 , 0 1959 60 6 7. 0i-., : >; ~~~~~~~~~~~~~hil.,nd . 359f8 9 56 66 72 t, 0 j t la .; s; ] s -i s ;= > . - - . s ;.^r - S @ - - - - .- < - - -. - = r . . z~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. .. . . .. . i ;J1980 *',i-i#t.. ' .................... l'-' '......... '; , '1;'; ; -; -3, 21. .. r199 09' I--1 S Silg9O8|e to -- 0S @ 0 3e2 -7.0 0 r ,s; a 1 ' _ L _N . r ."5 ;~~~~~~~~~~~~~~~ 6 X .. @V i or:ptnsnnfre. .l °fl°,c.,. x,,ua,o.lpoxasedr lGo e s.v . .;- <,>3.g,c 'f, 'MWkm~ ; -m'. '' -to ,. . ra rtu7 'tn t DcbuaP.ace- (US$54 i-:)k haEraw$ 0;.. . . .................. ; 15*41= D,adepeet. BakefMaLquh 821 sata .huu ! hp t beta-1usdm pre:,: - - M s'.- b*r * (383fpm r"T"'ad W" ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~mylwdncg iu a .e .' "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~w (U M F. . .. . 1 - .! ; . 1 - -'ps, :w aucef,u DZI~hhy Saa,Dwqscfa f6ai,na Ouse aena,iacpehfehtd cMa&aashqm na _~ ~ ~ ~ ~ ~~ ~~ ~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i.anuctl&nl ... 254 -. - .. - ; , . . .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ST RATE GI ES FORRWD a notAum 56Na'.".- bnal.pnwant nkdkdw -17MpcLIpa:Dad.xemowwt mnouvkt Phllpphim (inw) . UISI WUDNW Caa,g--st aayTciamr salmawn 4 youn fixwmldtus Hai,3 Ealcmwmaai ma 25dluftolankga.safaw, * *. Ml~~~~~~~~~~~~~naCr) -. - . . ~~~~~~~~~~~~~~~~~~~~~~agmriaa ftfiwm. Fwre io fhlsi5p).Amam 1ka(vn$ . - . . taandeatnamI . &Wtuat.,. L U.Jm.. 'ra'. Cia... ...lwx.~ flakdC.ma,wdcttaaa. .. I -. - Psunatatd - .. ~~~~~Damu-lag rum tht ptwmsnt. M.&Sns- toharm spn Indian] Rime. ~~~~~- 21.1 raiKon bAixtaIwna Ova.sbmmPt-gbaii.ouwbg - -isllh- s.leantam line. - Capstans: Jlbadand ~(USSLIS ils.) hAlfpnaee fa pnaar'AaSlak - fiwrdx sA aee nmt& nxd.dbamd mAman Back fhcAgcdswc (MAO ~45JW KWon 54 pm-- Drpsu : PiLbSzaa (W1i5tPSmaxlonwund- aghhinAxn m 0(W) G Na'orSbn .easmatn m Haaxid2781 snaenntw ix ai t 13tmx aotnaan d9=it dndaif.d,136 (n bnvarFan&nsn lelmnaina Rantm (199011. nc.Ceainul lash ripen Exap nImpaMllk a Japa bmtlwa kandblnkLTuixwaLnz h Eaptut-ap Bank f,V bases mixs dhc DdJchi Kangac Dank. lad. Malayt, teat] banksa.iaa.Dn fllnAadRps 33Twhfpa.Ca ai u.uo Iqi Bthtapnbahndoknet Japa Cntlnmmod Kot.spaeqtaa(nip am TDank 52ldr. w mi lee N.re hl~C"lbakum T a Novtes 1. See chapter 6 below for a mor'e derailed discussion 3. Of course GNP per capita is much higher in East Asia of the relationship between educational expansion, the than in Sub-Sabaran Africa. Because of the bias that re- * growth of labor demand, and the educational structure of sults from using exchange rates rather than measures of wages. purchasing power parity, we do nor attempt to weigh the effect on costs of higher GNP per capita against the effect * 2. From aross-economny school expenditure equations of lower teachers wages relative to GNP~ per capita. esinared by Schultz (19 88), the elasticity of enrollments with respect to the relative price of teachers is -0.80 and 4. the relationship between the abundance ofthumnan --0.70 at -the primary and secondary levels. capital and reachier pay is not firm. In Korea, ~where CRACL E human capital is most abundant, the earnings of primary 1O; Given the smaller size of the basic education age teachers are roughly 5 times per capita GNP. However, be- cohort in East Asia than in Latin Amcrica (see bdow in cause pupil-rcacher ratios are higher in Korea, operating che text), this difFerence underestimates the gap between costs as a perccntage ofper capita GNP are, at 16.5 percent, the nwo regions in the strength of dte public sectors com- not much higher than in Indonesia and Malaysia, sug- mitment to basic education. gesting an important role for policy both in determining teachers' vage and in controlling costs. Sec Tan and 11. Howevcr, since the mid-1980s, public savings in Mingat (1992), pp. 28-34. Argentina, Brazil, Chile, Colombia, Costa Rica (since 1983), and Mcxico (in 1987) have incrcased in response 5. Implicit in the calculations is the assumption that to fiscal adjusrments, leading national savings to rise (an expenditure per pupil would remain the same even if the intcrnal World Bank report). growth of the school-age population was faster. 12. In Japan, by managing excess competition, indus- 6. There may also be problems on the supply side-in- trial policy may have also assisted ir the stability of finan- adequate capital markets or entrepreneuri deficiences- cial instiutions. Japan's recession cartels, fir example, may that limit the private provision of educational services, have avoided large and systematic bankruptcies, which Psacharopoulos, Jimenez, and Tan (1986) present esti- surely helped the solvency of financial institutions. mnates of private and social retrns from education. See Psacharopoulos (1993) for more recenrestimares. Behrmn 13. The danger of casing intcrest rates and entry con- (1993) cites the many critcisms of hese estates, indud- trois without simultancously strengthening prudential ingtheir failurc to acoount forabiityand forqualiy ofCed- regulations and capital adequacy requirements has been ucation. Neither he nor others, however, question the evidenr in Indonesia, where low capital requirements for existence of some gap between private and social returns. entry and poor enforcement of prudential regulations led to mounting concern about insolvenr banks. 7. In fact, it is difficult to distinguish cmpirically b-i tween the e1*lrEs of nor being able to borrow and the ef- 14. In several economies, the risk was aggravated by fects of poor inbrmnation on the demand for schooling- the banks' tendency to speculate in securities. Table A5.4 shows that the major source of financial disruptions in 8. Government expenditure on education, exprssed as Hong Kong, Malaysia, Thailand, and, recently, Japan was a percentage of GrNP, was used as an explanatory variable real estate and securities speculation. in a cross-economy regreSSion in which expected years of schooling of the school-age cohort (essentially an aggre- 15. Stiglirz (1993a) notes that in economies in which gateof enrollment rates) is the independent variable. For governments do nor take an active pro-competitive a sample of fifteen Asian and Latin American economies, stance, there are a limited number of firms in the banking the expenditure variable was insignificant. See Tan and sector. Furthermore, even when there are a large number Mingat (1992). of banls, e&ctive competirion (if viewed from the per- spective of the receiver of the financial service) is likely to 9. In Korea and Taiwan, China, the share of public ex- be limited, This is because information pertinent to a sin- penduimre on education allocated to higher education has gle borrower is specialized, and each of tie major cate- increased during the last decade or so for two reasons. On gories of loans represents a diffcrent product markerk the one hand; universal and near universal enrollmrient There are differences in the kind of infbrmation that is rates have been achieved at the primary and siondary lev- needed across product categories and few economies of - , respeccively. On the other hand, the increase has been scope across the categories. Patterns of specialization consistenr with the shift in the strucrure of production among financial institurions confirm this view. - and exports to more technologically sophisticated and skill-intensive products, and die consequent increase in 16. Data on quality are extremely limited, but avail- the demand for engineers and other skilled workers. able measurements suggest that infiastructure in the -256 STRATEGIES FOR R_ HPAEs may be of better quality than in other developing enterprises in Indonesia and Malaysia have reported diffi- economies. Table 5.7 shows the percentage of roads that culies because of a growing volume of arrears. -are paved in a large selection of low- and middle-income economies. 23. For example, in the 1950s and 1960s, the ratio of write-offs and bad loans to rotal assets of JDB was lower 17. There is a large literature not only documenting than private commercial banks ODaIJERI 1993). The De- the relative unimportance of equiry markets as a source of velopment Bank of Singapore has evolvcd inro Singa- finance but also explaining why this is so, both in terms of pore's largest private commercial bank, largely on the the evidence concerning the negative effect of share issucs basis of its excellent financial performance (Gordon on share price and in theoretical terms. See Stiglicz 1983). Yaron (1992) suggests that traditional financial in- (1993a). dicators may not be the best indicator of peformance of development banks, since they have been established with 18. Long-term credit banks, but not city banks, issued development objectives. Hc mcasures instead the degree coupon debenturcs of arnaturityof fiveyears, and this has of depcndence of developmcnt banks on subsidies from been centrml to the segmentation of short- and long-term the government. With some exceptions (for example, finance (Packer 1992). Malaysia's Bank Bumiputera, Indonesia's Bapindo), the HPAEs' development banks are self-reliant. 19. Korea's bond market has been expanding more than other East Asian economies, except for Japan. But 24. There is remarkable umiformity in the pattern of most of these bonds arc government guaranteed. Firms investmenr outcomes exceeding investment efforts among have begun to issue bonds, because governmenc has the five HPAEs in the sample. Malaysia is perhaps the least begun to control the debt-equiy ratios of large corpora- representativc in the sense that during the 1970s (and up tions (an internal World Bank report). Many of the cor- to 1982) the nominal investment share cxceeded the real porate bond issues may be simply a form of bank investment share. All five economies, however, had real financing that carries a government guarantec. investmenr shares that exceeded the nominal share be- tween 1982 and 1985. Korea continued to show a declin- 20. This is known as the price discovery finction of ing relative price of investment goods throughout the stock markets. 1980s, while Thailand had a major increase in relative capital goods prices in 1988-89. 21. More substantial long-term lending to industry has been carried out by other development banks (for ex- 25. De Long and Summers (1991) obtain a similar re- ample, Eximbank, the Small Business Finance Corpora- sult for the relative price of equipment (to the GDP defla- non, and private long-term credit banks). tor). They predict the real equipment price for the sample of sixty-two economies as a function of relative 1980 GDP 22. While the experince offEastAsian economics with per worker. Of the six HPAEs in-their sample (Singapore development banks has been generally positive, there also and Taiwan, China, are excluded), Indonesia and have been failed development banks. In Japan (Recon- Malaysia lie dose to but above their predicted values; struction Finance Bank in 1952) and Thailand (Industrial Hong Kong. Japan, Korea, ancd Thailand are all signifi- Bank in 1959), insolvent development banks were closed. cantly below their predicted values. Real equipment prices Development banks lending to srnall and medium-size are unusually low in these economies. - . . .~~~~~~~~~~~~~~~~~~5 0; a E R 6 Using Resources Efficiendy: E Relyiing on Markets and Exports T IS APPARENT THAT THE HIGH-PERFORMING ASIAN economies were unusually successhfl in strategies to achieve rapid accumulation. But high levels of physical and human cap- ital formation are not a guarantee of economic success, as the modem economic history of the former Soviet Union and East- em Europe sadly attests. Resources, once accumulated, need to be allocated to high-yielding actirities. This chapter first considers the strategies, explicit and implicir, that the HPAEs used to achieve efficient allocation inte areas: the labor market, the capital market; and inter- national trade. It then turns to industrial policy-deliberate, government- sponsored interventions to alter industrial structurewhich was an important feature of public policy in some HpAES, and asks whether these efforts have resulted in superior economic performance. Finally, the chapter considers the role ofmanufacured exports in promoting the demand for skilled labor and the acquisition and mastery of best- practice technology. Explaining East Asia's Efficient Resource Use T 0 t|HE HIGH-PERFORMING ASIAN ECONOMES HAVEEFFICIENTLY absorbed unprecedented rates of growth of physical and human capitai for thirty years. As we showed in chapter 1, all of the HPAEs had total factor productivity growdt that exceeded that of more than 70 percent of developing economies between 1960 and 1989, and some-japan, the Republic of Korea, Hong Kong, Malaysia. and Tai- -wan, China-had among the highest rates of TFP growth in the world 259 - 0 - _EE [+ ~C L E This means that fictors, once accumulated in the HPAES, contributed much more to output growth than would be predicted on the basis of the average performance of a broad sample of economies, both develop- ing and industrial. -How have the East Asian economies succeeded in using resources ef- ficiently? Good fundamentals again tell much of the story. Price distor- dions, while present, were limited, and markets were allowed to work HPAE governments have relied primarily on market mechanisms to guide allocative decisions in both labor and capital markets. Labor markets in the HPAEs have been remarkably flee of the interventions that restrict labor mobility and create small, high-wage elites in other developing re- gions. Whle capital markets have been frequently controlled, resric- tions on deposit and lending races have generally distorted markets less than in most other low- and middle-income economies (as we demon- strated in chapter 5). Government efforts to direct credit have generally been undertaken within the framework of moderate interest rare subsi- dies, and financal institutions have been given the final decision on the creditworthiness of investment projects. This has meant that loans sup- plied through directed credit have been monitored and generally have been repaid. The relative prices of capital and labor have broadly re- flccred their relative scarcities. The HPAEs have also used the international economy more effectvely than many other low- and middle-income economies. Each of the HPAEs, except Hong Kong, went through an early phase of protection of impo rt-subsituting industries. But these policies were later modified by reductions in import controls and tariffs, combined with strong incen- tives to export (see chapter 3). This opened up much of the modem economy to international compeiition and introduced world prices as performance standards, not only for exports but also for tie intermedi- atc and capital goods used in export production. As a result, domestic prices for traded goods in the HPAEs are more doseJy aligned with inter- national prices than in many other developing economies. Apart from these findamnentals, however, most IIPAEs seletvely in- tervened in financial markets and used a combination of policies to pro- mote thc development of specific industries. Most HPAES halve directed credit to priority sectors and sometimes to specific firms. Where these programs have been effective in -allocating resources to high-yielding investrnents-for example, automatic credit for exporters-irt was be- cause allocation rules were performance-based, and borrowers were ac- :26 USING REESOURCiS tively monitored; in short, effective contests were created. These condi- -tions have been more- common in the three northern HPAEs that have strong governmental institutions. * 0 Indonesia, Japan, Korea, Malaysia, Singapore, and Taiwan, China, have tried to use targeted industrial. policy to promote the development of specific knowledge- and capital-intensive industries. We test the rela- tionship between industrial policy and productivity-based catching up and conclude that industrial policy has generally nor been successful in changing industrial structure or raising productivity, What, then, has contributed to the iPAEs' apparendy superior performance in adopting and mastering international best-practice technologies? We argue that the combination of compedtive discipline and well-firictioning factor markets with a pro-export orientation-the export-push strategy- employed by all eight of these econormies was responsible for their supe- rior productivity performance. Manufacturet exports accelerated the acquisition and mastery of in- ternational best-practice technologies in highly imperfsc international technology markets. Hih and rising levels of htiman capital in the HPAES, especialy the superior educationd atminment and cognitive skl of the industrial labor force, helped to ensure that these new processes and equipment were used and adapted effectively. Thus export orienta- tion and high human capitl formed a virtuous circle; exports raised the returns from education, and educationi raised the retuns from exporting. UsUsng the Market Labor Markets in East Asia 5 xT rILE HUMAN CAPITAL ACCUMULATION MAY BE A NECES- condition for sustained rapid growth of output and wages, it certainly is not a suffiaent condition. The Philip- pines, the republics of the former Soviet Union, and Sri Lanka are promi- ; nent examples of economies that have had high school-enrollment rates but low rates of growth of per capita income and of wages. Utilizing humnan capital in activities that yield high returns on the prior invest- ment in education and training is as unportant to growdh as the accu- mulation of human capitaL We have fiequently looked to coordination turilures in markets to understand why public policy might be used, to help markets function - - - - ; 0 ~~~~~~~~~~~~~~26t - ,'',' e!'kCLE better. In the case of labor markets, however, coordination failures are limited, and information, although not perfectly symmetric between workers and employers, is better than in capital markets.' Extenalities, as we showed in the previous chapter, primarily alTce investment deci- sions in human capital, not interactions in the labor marker. Thus, there is a presumption tiat, ifleftto themselves, labor markets would perform their coordinating (allocative) role quite well. Two conditions must be fulfilled for a growing supply of educated labor to be utilized in high-retum activities. First, rhere must be rapid growth of labor demand relative to supply and, in particular, of demand for skilled labor. Second, the labor market must perform well. It must be efficient, flexible4 and responsive to changing conditions so as to ensure that workers are employed in jobs in which their skills are most produc- tively utilized. If these two conditions are not filfilled, there is a risk that rapid educatonal expansion may result in labor market problems dtat erode the benefits of human capitml accumulation and have a seriously adverse impact on the growth of output and wages. In our assessment of why human capital in Fast Asia appears to have been efficiendy utilized and why changes in labor market outcomes have been so fuvorable, we focus on two facors: the dynamics of labor supply and demand and labor market performance. We also examine the inter- actions between these two factors. Then we consider whether there are any common denominators in thae policies of East Asian governments that have contributed to the efficient utilization of human capital. Labor Supply and Demand in the HPAEs Shifts in the aggregate supply and demand for labor in East Asia com- pared with other developing regions have led to more rapidly rising real wages (see figure 6.1). East Asia's more rapid rate of increase of wages, following an initial period during which wages did nor increase, . is the reslt of slower growth of supply and more rapid growth of demand for labor.2 For a given rate of growth of labor demand, the slower the rate of increase of labor supply, the bigger the increase in wages likely to result from competitive pressures in the labor market. One measure of the dampening influence of labor force growth on the rare of increase of real wages was obtained for a cross section of twenty-two economies in the 1970s; a 1 percent increase in labor force growth resulted in a 1 percent decline in real wage growth (Gregory and Lal 1977). z6z USING RESO£5 RC Figure 6.1 Increase hi Real Earnings Japan f ; ;- [ Rep. of Komea __ l Tawan,nChina 0 7= -Z I . Hong Kong | Singapore . Indonesia } - - - Thailand I - } Latin Amneuca and Certhean F- mi 1 - 1SrO.:- SubSahIranAbica -'A- It - U d980-90 t I -t 0 1 2 3 4 5 6 7 8 9 10 11 Increase In real eamings (p'e.nt) Nwc ldIdoc for Taiw2n, Chin= 1979 = 100. Other Asian cconomics arr Bangladcsh, India. Pakistan, and the Philppines. SouwreWorld Bank dacvi Taiwan, China, National Stistcs. Labor Force Growth. During their period of industrializarion in the nine- teenth century, the populations of the currendy high-income economies grew, on average, ar an annual race of only 0.8 percent.3 Today, the pop- ulation of Sub-Saharan Africa is growing at roughly four times ca rate; the populations of Latin America and South Asia are growing at romughly three times that race. Only in East Asia have populaton growth rares, and the share of the population less than fourteen ycars old, declined to levels approaciing those that prevailed in the high-income cconomies.4 We have already seen how the early denographic transition in East Asia markedly reduced the rate of growth of the school-age population to levels well below those in other developing regions, thereby easing the financial burden of maintaining education enrollment rates. The early demographic transition also reduced, with a lag, the rate of growth of new entrants into the East Asian labor force. The annual rate of labor force growth during the 1980s was 2.6 percent in Sub-Saharan Africa and Latin America and 2.2 percent in Soudt Asia. In East Asia the rate was 1.8 percent (see table 6.1). The labor force growth rare in Singapore is only 0.8 percent, about the same as that experienced by the high- income economies in the nineteenth century. In Sub-Saharan Africa the rate of labor force growth continues to increase. In Est Asia, despite in- creases in the particpation rates of women, labor force growt rates have been declining much faster than in Latin America or South Asia. 263 RACL E Table 6.1 Labor Force Growth Rates Ecenomy,vfonegjn. J--.90-85 1985-2000 H; , s;- 2.5 1.8 Indonesia: . - ~ -; ; - - 2.4 - 2.2 t'rKc, Rep. of. 2.7 1.-9 Malaysia - 2.9 2:6 ..Sinore 1.9 0.8 Thailand -; .- 0 - 2.5- 1.7 thAsria 2.2 2.2 L4 in Am6icai bhea-n 2.8 2.6 ST--S . m2ranAfr. : 2.3 2.6:. S are WorldBank(1987b). Growth of Labor Demand. Not only has the growth of labor supply been slower in East Asia than in other developing regions, but the growth of labor demand has been faster. For the period 1960-90, the rates of growth of wage employment in manu&cruring construction, and ser- vices have tended to be substantially higher in East Asia than in Sub- Saharan Africa, Laiin America, or South Asia. This is not because gowth of employment has been unusually slow esewher Widt the ex- ception of the 1980s, during which output and employment stgnared because of the macroeconomic adjustments made necessary by debt- service problems, the pace of industrial sector employment growth in Sub-Saharan Africa, Latin Arnerica, and South Asia was more than dou- ble the pace in the currently high-income economies during the perod 1880-1900 (see table A6.4, at the end of the chapter). Rather, the in- terregional differcnce reflects the unusually rapid growth of employ- ment in East Asia. Even as it has been growing rapidly, labor demand in East Asia has become increasingly skill-intensive. As a share of wage employment, white collar and technical employment increased steadily during the. 1970s and the 1980s; in Korea, for example, it rose from 29 percent in 1980 co 36 percent in 1990, and in Taiwan, China, from 32 to 40 per- cent during the same period (see table 6.2). The pace of change in the occupational structure of employment is lower in other developing re- gions. The change in the occupational composition of labor demand in the HPAEs reflected increases in the abundance of educated labor, and consequent declines in its relative price, as well as changes in compara- 264 USING RESO-UE Table 6.2 Professional and Technical, Administrative and Managerial, Clerical, and Sales Occupations as a Percentage of Total Employment e H i Kora,JRp of Shngapore Tariano Chbina 1980 29.0: - 42 31.8 .1981; - 292 30.8 -42.8 - 32.7 0 S-; 198200: f - . 30.6: 0 : .--0 *;: . -;-44.2 33.1 1983 32.1 - 45.6 33.1 -1984 32 - 46.5 33.7 1 l985*. :-- 34.3ta;; - - - |--.46.5 34.2 1986. 34.3.3 379 46.2- - 34.3; - 1987 2 34.1 - 47.0 35.1 : - 1 98 8:. . : . 34.6 - .-X 46.1 37.2 ,1l989 ~0  0 - 35A4 ~> V - - ? - -t-46.8 38.5 ;1990_ 36.2 . 60.0 X 39.8 * £ / r'-_Notvanilablc; - aToml employment for Hone Kong indudes Linmployed who have previously bald K Iidudes serviceworke -X : Somiwfe::ields (1993). tive advantage. East Asian exporters shifted into more technologically sophisticated, skill-intensive goods as rapidly rising wages of unskilled labor eroded international competitiveness in labor-intensive manufac- tured goods. What accounts for the extraordinary pace at which labor demand, a-nd in particular the demand for skilld labor, has expanded in East Asia? Since the demand for labor is derived from the demand for output, the explanation must be sought, in the first instance, in those factors that contributed to East Asia's rapid output growth: high rates of savings and investment and high rates of return on investment B at it is also im- portant to consider the extent to which the path ofgrowth of output has been more labor demanding in East Asia than in other developing re- gions. As we shall see, the dynamism of agriculture and export-push stratcgies contributed to rapid increases in the demand for labor and to rising skill-intensity of labor demand. Letting Markets Woric Public Policy in the Labor Mardt The term "labor market?' generally refers to the complex of interac- tions between employees and employers. The labor market thus bears the responsibility for melding the changing needs of the economys pro- 265 iR_AC LCE ductive apparatus with the changing skills and preferences of labor force participants (Berry and Sabot 1981). In the course of economic devel- opment, workers-most of whom are following educational, training, and vocational paths different from preceding generations-tend to shift toward more dynamic sectors, from self-employmenr to wage- employment, and toward higher-skilled occupations.5 An efficient, flex- ible, and responsive labor market allows rapid matching of skills to needs, improving resource allocation across firms and industries and thereby contributing to growth. While in most developing economies, labor markets have tended to unction rather well, where they have not, the costs have at times been high.7 In East Asia, more than elsewhere, governments resisted the tempta- tion to intervene in the labor market to counter outcomes unpalatable in the short run or to particular groups (for example, wages below where workers believed they ought to be, educated unemployment, or the em- ployment of educated workers in lower-level occupations). A relatively high level of efficiency in the allocation of labor was achieved by allow- ing wages and employment to be determined largely by the interaction of those supplying and those demanding labor services, rather than by government legislation, public sector leadership, or union pressure How WePl Did East Asian Labor Markets Perform? In East Asia, wages were pulled up by increases in the demand for labor, whereas elsewhere there was a greater tendency for wages to be pushed up artificially. Earnings growth was determined more by the growth of the economy as a whole hn by growth in any particular sec- tor (Fields 1992). In economies with highly segmented labor markets, the opposite is the case. East Asia's Flexible Labor Markets. By not allowing the price of labor in some sectors to rise well above what workers could earn elsewhere in the economy, most HiPAEs avoided the creation of a high-wage labor elite. For example, there is evidence for Taiwan, China, of a remarkably inte- grated labor market; there is no significant correlation between the growth rate of earnings and the growth of output within sectors. Manu- &acuring wages for unskilled labor are only 20 percent higher than agri- Lultural wages. By contrast, in Colombia and Jamaica, where labor markers are highly segmented, workers with the same ievel -of skill earn nonagricultural wages that exceed agricultmral wages by 150 percent 266 USING -R (Fields 1992). Still larger intersecroral wage gaps have becn observed in East Sub-Saharan Africa (Ranis 1992). Another noable feature of the -structure of wages in East Asia, as illustrated by Korea and Taiwan, China, is the modest size of the gap between skilled and unskilled wages in the nonagriculturil sector (see figure 6.2; also see Ranis 1992). The small gap did not result from minimum wage legislation pushing up un- skilled wages. Rather the growth of demand for unskilled labor, in com- bination with marked increases in the abundance of educated workers, compressed the occupational structure of wages. One benefit of the compressed wage structure in the HPAEs was that it reduced the incentive for educated workers to conduct a lcngthy search for a relatively high-wage job, and thereby remain unemployed, rather than fill a job slot at a lower occupational level. Conversely, a seg- mented labor market may provide an incentive to workers in low-in- come employment, and to the unemployed, to lobby the government to provide more high-wage jobs than justified by the derived demand for labor.8 Thus, the share of the public sector in the increase in total wage employment in recent decades is a crude indicator of the magnitude of public sector surplus labor, and hence labor market segmentation.9 As table 6.3 shows, in all the Sub-Saharan Afrcan, Latin American, and South Asian economies for which data were available, employment in the public sector grew more rapidly than wage employment in the pri- vate sector. The median share of the public sector in the increase in total employment was between 71 and 87 percent.'0 With the exception of Malaysia, the experience of tl-e East Asian - economies is markedly different. Omitting Malaysia, the share of the public sector in the increase in total wage employment in those Et Asian economies for which data were available was less than half the me- di-an share for other developing economies. The limited growth ofpub- lic employment in these East Asian economies is evidence that public surplus employment has been kept in check and stronger evidence that the scope Sor lobbying for make-work jobs has been limited. The need to maintain international compeitiveness helped to limit growth of V public employment because importers who purchased intermediate in- puts from public enterprises lobbied for low prices. Moreover, govern- ments striving for macroeconomic stability stricdy limited the size of the public enterprise deficits that they would finance. - In- Malaysia the share of the public sector in the increase in tota wage employment is as high as in other developing regions. The increase in 267 1Figum 6.2 Industrial Wage Gaps Wage a Taiwan, Ch._a Mean monthly earings (thouK;mds of 1955 NT S) -~~~~~~~~~~~~~~~~~~~~~~~~~~~~t ' ' . ' ..~~~~~~~~~~~~~~ L .... T- O. IF= F4 to-I A -l rea walp :~2~ - : ; - - ~~~~~~0.5- - - . I | ' ]I I . U UnakflledlIndustulal real wags4 - °°- 5 ! s X t f 1 11 ~~~~~~~~~~-MeradusLnduilrsalwags 0.0 1953 1955 l9S7 1959 1961 1963 1965 2967 1969 2.971 Wags gap, Rep. of Korea Moan monthly eamings (thousands of 1975 won) 40 1 10 - *I Unskilld odustrial real wa _________ I V -.- Average Industrbl reawuge 1963 1965 1967 1969 1971 1973 1975 1977 1979 Sounrs: Kuo (1983); Taiwan. China (various years, 1988); Korea (1978); Kim (1990). 26-: USING-RESOUIRCE Table 6.3 Cmwth of Wage Emnployme.t in the Public and Private Sectors, Selected DeveloPing Economies __ __ __ __ __ _ __ __ __ p b tsector In inrase p/otl &onorny/~on Peno Public rkwre Toal wageemplyment LnAmerica aznd C"bantbe: * - : ~~~19733 1.4 00. 0.3 92 CostaRica ' 1~~973-43. :6 , 28353 Panama *':, :1963-82 . ~7.3 ,.,18 2. 45 Peru ~~~~1970-44 * 6.61- 061.4 Trinidad` :1970-84 4 12 ,195 Venezuela ;1967-82 ~~~~~51 3Ž4 '. 37 27 Middle ase and Not nfipc Egjp . , . 1~~966-76f 23-. . 103 n*nj.I tA -c Gh~~~~ina ~~196048-734 -5.93-4. ,Keny 196381 64.2. . 67 Tanzania ~~~196246. 6.1' --3.8' 1.619 LZambia -1966-080 7.2 -r0943 K India .~~~~~~1960-80 42213.2 7 SnI nka ~~~197183" 8.0 09393 ? Taiwan,China . ' -196.5-85 23- 6 65501 ~~Thailand .: ~~~~ 1963-83 ~ 6. ..7 33, I' .~~~~~~~~~~~~~~~~~~1 Unwciglired.mC"ean" bl'emlyetwsntarsostolbyninud by2. la0or mar2- 6 publennat0. ahr,pbic employmentEwswns aresons tolbyn nue bynlabor cma- ployment opporrtunities for the historically disadvantaged ethnic Malay Majority. The Bumriputera program can be seen as an aggressive aflirma- tiye action policy that resulted in considerable public surplus labor (Free- man 1992). In contrast to many gvrmnts in similar circumstances, however, when the macroeconozmc costs of public employment began 'to mount theMalasian government, manifesting the adaptability that sewed governmnents so well -throughout the region, changed direction.- Why Were Eas Asian Labor Markts More Niedble? Rapid growth was ake -fcror in making wages more flexible in the HPAEs. Workers are generally more willing to accept flexibility of wages around a rapidly rising trend bc&ause downward adjuLstment implies a: slower rate oinrease rte than an absolu:te decline in reai earnuings, (Shiah and Mathur -1992). This 2.69 is manifested in wage-setting practi ces that tie compenation to enterprise performnance, another example of how the extraordinary hibor marker dy- nainics in Eas Asia contributed to superior labor marker perforrnance. But some HPAE governiA.. "ts also intervened in labor markets, pri- manily for political reasons, to suppress the activities of industry- or economywide unions and to ensure that wage bargains were set at the enterprise level. We have suggested chat the rapid growth of wages con- tributed to the explanation of why, in contrast to odchr regions, wages were not maintained above market-dclaring levels in E-ast Asia. An alter- native, or perhaps complementary explanation is that governments, in collaboration with employer repressed laibor markets. Wage-seuingpractices One manifestation of the link between rapid growth and wage flexibility is the apparent willigness of workers, espe- diaily in the northeastern economies of Asia, to accept wage structures that include a substantial element of-performance-based bonuses. This' practice links compensation to enterprise, rather than wholly individua perfbrm2nce. It is comnmon for current cash compensation to have two comnponents-a base wage and a bonus (Shah and Mathur 1992). To the extent th at "special cash payments" are more responsive to changing economic circumstances than are base wages, thiis arrangement could help account for the greater flexibility of wages in East Asia than else- where. During the period 1958-83 in Japan, econometric evide-nce, shows that rthe ratio of bonus payments to base wage-s varies positively and significantly with such indicators of aggregate economic conditions as cnrporate profits. By contrast, there was no significant relationship between base pay and corporate profits (Freeman and Weitzman 1987). Nevertheless, the impact of changes in bonuses on coral compensa- tion appears to be marginal. Whlile bonus payments as a proportion of total current compensation (15 percent or more in Japan and Korea) or of corporate pois(40 percent or more) rend to be large, the elasticity of bonus payments with respect to profits is small. In Korea, most bonus payments have become contractually linked to the base wage, and in Japan the implicit link of bonuses to base wages appears to be strong (Shah and Mathur 1992). How rep ressed uor H-PAE la bor ma rkeex? Repression in the labor mar- ket can rake two forms: dlirec restraint on wage levels or growth (wage represion) or restraint on organized labor (labor repression) with indi- rect restaint on, wages. Wage repression was dlearly nor a necessary ingedintin -the East Asian recipe for economic sucesThgorn 7-7 -- USING RESO mernts of Hong Kong, Japan, Korea, and Taiwan, China,- did not repress wages, ant yet in these economics output, exports, and ultimately wages all increased at rates well above the international norm. These economies were adept at adjusting to changing comparative advantage as labor abundance gave way to labor scarciy. At the other extreme, the govcrnmnent of Singapore repressed wages, at times quite severely, and the consequent costs of labor shortages appear to ha've been high, lead- i ing the government to abandon. the policy (Fields 1992; Freeman 1992). Several HPAE governments have reprcssed labor organizations. Korea, Singaporc, and Taiwan, China, suppressed indepcndent unions.11 Malaysia has independent unions but highly restrictivc labor laws. Hong Kong and Japan do not intervene in labor relations. H-ow- ever, ir Japan most unions are company based. In Korea and Taiwan, China, real wages increased at roughly the same pace as real GDP. Nevertheless labor relations in Korea have been deteriorating in the 1990s, so that labor relations problems serious enough to threaten economic growth have arisen, partly in reaction to repression of unions (Freemanl 1992). In Singapore during 1972-79, r real wages increased at less than one-quarrer the rate of increase of real GDP. This divergence was the consequence of the governmenes aggres- sive, and quite open, intervention on the side of wage repression in the market for labor12 The minister of labor was quoted in the press as say- ing, "It is clear that an essential clement in our new strategy must be a tighter grip on wage increases.... If we donor tquicldy and willingly change to low gear on the wages front, we shall fisrher discourage in- vestrnent and aggravate the unemployment problem" (reprinted in Fields 1992). * . EIfwages fall well below market-dcearing levels and employers are pre- vented by the government from using warge competition to bid up the price of labor, chronic shortages of labor are likely to be the result. This is what occurred in Singapore, with the result that the rate of growth of output, as well as the rate of improvement of living standards, declined.. In 1979 the government recognized the costliness of this corsraint and * changed course: a policy of "wag,e correction" was adopted with the ac- plicit aim of alleviating labor shortages and shifting the economy away from labor-intensive and toward capital- and skill-intensive industries. Although the high wage policy was abandoned in 1981, in the 1980s ..the grovth rate of wages accelerated:and matched the growth rate of SGDP (Fields 1992).- 271 *CL Hong Kong represses neither wages nor. unions. With respecr to the labor market, the govemment of Hong Kong is among the least inter- ventionist in the world. Despite labor laws that are favorable to union formation and action, union membership in private sector fims is low; unions engage in relatively litde collective bargaining and are generally considered to be "weak." Perhaps because wages were rising rapidly in Hong Kong, workers did not fcel the need for collective action. Hong Kong demonstrates that East Asian economies need not aggressively re- press unions to have a weak union movement, suggesting, for example, that in Korca there may have been frw benefits to offset the political and economic costs imposed by labor repression (Freeman 1992). Japan's expenence with labor unions is another counterexample to the belief that East Asian success required goverrnment repression of labor. The Japanese example may be the more telling of the two because in Japan, in contrast to Hong Kong, union membership is high and unions are, at least potentialy, quite strong. Yet unions did not extract economic benefits for their members at the expense of econornic growth (Freeman 1992). Fow Did Efficient Labor Markets CoutnIbute to Growth? The benefits of maintaining wages at or marginally below, rather han well above, marker-cleawing levels were numerous and substantial.13 FiSt, because inflated wages were avoided, the profits of firns and, most likely, aggregate savigs rates were higher. As a consequence, retained earing accountecd fior a higher proportion of investinent finance than orherwise would have been the case, redung reliance orn underdeveloped capital markets. The probable net effects, therefore, were higher levels of invest- ment, greater compeitiveness in international markets, and faster rates of growth of output, employment, and ultdmately eanings.14 Superior labor market performance in East Asia thus contributed to sustaining the extrordinary dynamics of labor demand we documented above. Moreover, by not allowing the price of labor to become inflated relative to the price of capital, wage restraint encouraged the use of more labor-intensive technology. Ar a means of avoiding excessive capital in- tensity, this was particularly important, given that government interven- tions tended to lower the price of capitaL Efficient use of skilled labor also contributed to the high levels of publicsector savings in most of the HPAEs. If the govenment capitulates to demands for make-workjobs the subsidies required to finance the te- .272 USING RESOURO suliting gap between the wage and the marginal product of labor are likely to divert scarce savings fiom productive investments, lowering the rate of capital accumulation and stcadily eroding the grwth potential of the economy.15 For a time, Malaysia was an East Asian illustration of this point, but there are manry more examples outside the region where governments have been slower to rake corrective action. If the government provides make-work jobs, squandering the pro- : ducdive potential of human capital may be another consequence. In the : worst case, which unfortunately is quite common, workers in whom a substantial educational investment has been made are paid the prevail- ing (relatively) high public secror wage to do jobs in which their mar- ginal productivity is zero or even negauve.16 In this way creation of makei-work jobs impedes the process of hunan capital deepening that is cen-tral to education's contibution to economic growth. Had the labor market been allowed to functon without intervention, these same workcrs would have entered lower-paying occupations, and the consc- quent mcrease in the educational level in those occupations would have contributed to an increase in labor producdviry. Becausewages were responsive to changes in the demand for labor, ad- Justent to external maaoeconornic shocks, such as those induced by the oil crises of the 1970s, was ofien quicker and less painful in the HPAEs (Mazwrndar, forthcoming). Ifthepriceoflabordoesnotadjustquick-lyand smoodily to macroeconomic shocks or to secular changes in the macro- economic environment, then either the quantity of labor must adjust or the rates of output growth or inflation are liely to. In relaton to fluctua- uions in aggregate demand, employment stability tends to be much greater in East Asia than in other rgions, so that East Asian econonies genelly adjust to downturns without laying offworkers (Shah and Marhur 1992). Assisting the Market Fnancial Markets and AllocaioDn -T-L MARKETSSPERFORM THEIR ALLOCATIVE FUNCTION IN -ways: they aggregate savings, allocate funds to competing a_J uneat, and monitor performanc Without financial intermediation, firms would have to rely on retained earnings to fi- nance investment, so marginal retums to investment would diverge 2173 . ..'., J:RACLE Markedly, especially in the short run. Thus the policies descrilbed in chapter 5 that contributed to the deepening of financial sectors also contributed to better capital allocation. In addition to private capital markets governments in East Asia, like governmecnts in many developing economies, have attempted to allocate credit directly to high-priority activities. Elsewhere, directed-credit programs have been catastrophic failures. In some HPA-Es-parricu.larly Japan, Korea, and Taiwan, China-these programs have caused less damage to capital allocation and may have been beneficial. Our examination of capital market performance begins by docu- menting the highi rare of financial deepening in the H-PAEs. We then describe the nature and magnitude of directed-credit programs and as- sess their performance. We show that, where these programs appear to have had the most beneficial results, they supplemented fundamen- tally sound capita market policie-s with improved selection and mon- itoring. Interest rate subsidies were small, loans were repaid, and strict performance criteria were used to ration fumds. These preconditions for efThctive directed credit were present in only a few of the HPAEs. Where they were absent, directed-credtit programs largely failed. East Asia'ls Record of Growing Financial Depth The growth of financial assets in East Asia has been ve-ry rapid- faisrer than the growth in aggregate savings-leading t usata grwth in the financial depth of East Asian economies. Figure 6.3 compares East Asia's ratio of M2 (rime plus demand deposits) to GDP, which is often used to measure the depthi of the financial sector, with the M2IGDP ratio for a group of low- and miiddle-income economies and the OECD econom-ies. The relatively greater financial depthi in East Asia far surpasses most developing economies and compares weli even with the OECD economnies.. A broader meaure of financial depth, the ratio of financial liabilities (M3) to GDP?, shows a similar trend. Korea is unusual in that its ratio of M2 to GDP? is low compared with many developing economies at similar levels of per capita m- come, and these ratios have inacrased more slowly than Korea's income growth would predicc. The largest incremental change was evident after the inte-rest rate adjustments in 1965. However, M2 fails to cap- ture the presence of a vast nonbank financial sector. The rapid growth of income in Korea since the 1970s has been accompaniedi by large in- :74 USING RESOI5 Figure 6.3 Financial Dept in IPAEs, Other Developing Economies, and Selected OECD Economies Etrope and Maidde East I .~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ I Lai HAmaEs 190:7 Sub-Saharan Afrks OECD economiesI_________________ 0 20 40 60 so 100 M2/GDP (percent) Not latin America includes Argendina Barbados. Bolivia, Brazil. Ch_c. Columbia, Costa Rict, Ecuador, El Salvador, Cuatemala, jamaica. Panamna, Trinidad and Tobago, Uruguay. and Venezuela. Earopc and Middle East indudes Algeria, Egypt, Greece. Morocco, Portugal. Syria. Tunisia, and Yugoslavia. Sub-Saharan Africa includes Cameroon, Congo. C6te dt'lvoire, dhe Gambia, Gharn, Kenya. Madagascar, Madriania, Nigeria. Senegal, Sudan, Tan.a.ia, Togo, Uganda. and Zirbabwe. OECD economies include Atriza, Belgium, Canada, Denmar Finland,l Germany, Ireland, Netherlands, Norway, Spain, Swecden and Switzerland. NoweM2IGDP = [currency + demand deposits,+ time deposits, + savings deposits (except large certificates of dcposit)j as a percentage ofgross domesdtc product. Sawrrr Intemnatonal Monetary Fund data. creases in die assets of' nonbank financial instirutions (King and Levine 1993). Fnancial assets in industrial and developing economies, incling East Asian economies, are held largely in deposits with banks and other financial institutions and to a very limitedt extent in government or pi vate securities (World Bank 1989c). The public policies described in chapter 5 that have- emphasized the development of banks and other savings institutions may partly explain this composition, but there also are market explanations. Equity and bondt markets develop slowly, espe.- daiy in emerging markets. In East Asia, the combinaton of rpid growth of financial assets in the banking system and rising savings dut- igthe past two decades has meant' that bankcs have played an maceas- .Mg-~~~~~~ CP inly i an role in capital allocation. 2.75 A C L EACt Financial Markets, Coordinaion Failures, and the Role of Government The lack of perfect information that characterizes all markets is most serious in capita markets, where acquiring information to ficilitate bet- ter selection and monitoring of projects is a key function. Imnperfecr in- formation has several consequences for allocative efflciency. As we noted in chapter 5, bond and equity markets are slow to develop. This places increased emphasis on the role of banks in selecring borrowers and mon- itoring their performance. But credit markets are not like auction mar- kets or markets for ordinary commodities. Because banks cannot discern perfecdy the probability of default, they cannot simply aUlocate credit to the highest bidders-those who promise to pay the highest in- terest rates. Rdther, they must choose among agents willing to pay the going rate those that seem most likely to repay the loan. This allocativc process is, in effect, a form of rationing. Moreover, it gives banks some control over entrepreneurs, who must maintain their good credit stand- mg in order to renew loans. Financal institutions evolve in response to these market limitations. For example, bank-dient arrangements respond to incomplete in$onma- don through relationship banking. Loan contacts are characterized by more than lust interest rates. The namre and amount of collateral be- come key. Yet despite efficiency-enhancing marker responses, capital markerS remain imperfct. Market power by banks may raise inrermedi- ation margins, resulting in monopoly rents for the banking sector Banks also face incentives to misallocate resources; underfunded banks rend to finance excessively risky projes. Social returns on investment may differ markedly from the private remrns to investors, and social perceptions of the riskiness of investments may differ substantially from pnvate assessments. For the above reasons, most governments are reluctant to rely entirely on private markets to allocare capital. Two types of government involve- ment are therefore common: first, efforts to reduce misallocation of re- sources by financial institutions; and second, eflorrs to control banking spreads ro limit monopoly rents. While these measures are by no means unique to East Asia, the HPAEs have been unusually active in exercising them. In addition, most HPAs have gone a step further and attempted.. to direct credit to specific sectors, industries, and even firms. We discuss this more complex and controversial topic in a separate section below. -76 U S I N G R E S OUlCM Reducing Bad ImuestmentChoices. Portfolio requirements, either to avoid excessive lending in some areas or to increase lending in others, direcdy influence the allocation of credit by banks. In chapter 5 we argued that East Asian govemmcnts bave emphasized the safety of banks through both regulations that monitor and enhance the solvency of banlks and measures that avoid financial disruptions. These have positive effects on resource allocation. Solvent banks have fewer incentives than under- funded ones to take excessive risks in lending. Solvrent banks can also lend more. Portfolio requirements to enhance prudent lending-such as limits on lending to real estate projects- -also may guide credit away from some activities to others with higher-perceived social returns. In general, restrictions on lending have favored producers. For example, East Asian governments have often restricted lending to real estate and other spec- ulative activities. These governments are concerned nor only that such lending increases the riskiness of banks' portfolios but also that it diverts lending away from manufacturing. Thailand, for example, does not specifj how much of their portfolio banks should lend to industry, but it provides lending guidelines that favor lending to industry and agri- culture and limit lending to other activities, induding real ,-r e. and securities. Most BAEs were assisied in their efforts to influence lending deci- sions because they directly held a high proportion of savings deposits. Table 6.4 shows that in Indonesia, Korea, until recendy, and Taiwan, China, very large shares of total deposits of the financial system were held in publicly owned commercial banks. In addition dte governments ofjapan, Malaysia, Singapore, and Taiwan, China, have discretion over a large (and growing) prop ,rtion of total deposits iri postal savings and/or provident and insurance fi^nds. Some governments have also been able to complement those resources with budgetary allocations or borrowing fiom abroad. Japan and Singapore have the most well-defined programs fior allo- cating s;-vings held by public finanicial institutions. The financial inst- tutions created under the Fiscal Investment and Loan Program (FuLs) of Japan allocate finds originating from the governments postal savings, national welfare deposits, and national pension deposits. The FILa ab- sorbed 20-40 percent of household savings during the rapid growth pe- riod. The Development Fund of Singapore, whose resources originate primarily from the provident funds, lends direcdy to the governrmcnes 277 Table 6.4 Propotin of Total Deposits with Government Financial Instutions in East Asia (percentage oftotaldeposits) m S ^ ~~- - XjiOQ~~~~~~et-r .,- -, 7M E -. ; :0; 1 1 c3 - - - - Wi.h .:mmcw or Wi.h IM it eI P . .3- pus ed sings pwvidearinsturance aem rid n Economy/year insa*adem fiends s~Pecielixzdi banks' Thud 1971-76 N' NA >ing-sy - N-. NA ; - : . 8.79.9a 7 ,9 8 1-8 5 N A N A . 7 1. 7 1. 7 1960 -6 5F-A ' 61.3 461.3 - -10-. . 108, 4. NA. 151 197 14t3} '.0 NA. 22-.3;-':r. _ . .19 -. - 0 22,N.37 . ~ ~ ~ ~ . 1971-75' Noa 80.7 855 1981-85 ~~~~~~ ~ ~~2.2 NA61 _1971-75- N' 44308½ 4 1976'-80 36 24.7 .~~~~~~~~~~8 ..02. 1906-89 ~ ~ ~ itede 15 30cora4 prdn beavo an3nhne 1ovnq9 e ---I 0-~~~~I stradbtutory boardsin and goernment coprateos. ndesprad Kreauete 198145:~ ~ ~ ~ efiieg 57rsueealoainb dcigtevlrl9o netel Mhal aysa and Thailand hyagven guided thei saig.Tate-owned comerieal banks ofinanen priority projmbiiets. cilsngsb rmougi duvenc thoeef fcmptto mn banks and inceas thecen averageins.- n~ ~ ~ ~ ~~~~sra bewe borwn.n edn aes ihsrasrdc h 2- t78 -: .~~~~~~~~~2 .2 . . : -0 - ; ;. - - - =.: - - --~~~~~~~~~~I USING. RESOI1 - With pvernsmen t or Vit? pensoi 4J pbi * * ~~~~~~postalsavings ~provident awntrance commercdialad *SingPapre 1971-75 2.1- 21.6 -23.7 1976409. 30.0 39.4 19814~5 -1.0- 34.6 456' 1986-90 ~~~~~12.7 34.9- 47.6 -.1971-75- 10.8 A67978.7 1976-80 ~~~~~-13.81 NA. 63.6 77.4 t18-5 - 92N.53.5 727 -:1986-90 1 9.4 NA 52.4 71.8- .Thailand -1971-75 .12.7 N.1.1 -13.8 1976-80 10. NA. . 13.7 * -1981-85 8.2 NA1. - - *1986:190 7.6 NA- 2.2 . . . a.Th fguxre P6rpotd is fbr 1978-80. c Te figure -for-Taiwan, China, are bed on the deposits of all COMmercial banks and'accneunl vrsiae eas ny1 - f he16 commnrdalbans& are gov&nmncnr owned. Nvtlds,xgocnnrwndbanks acon for the mjrshare of!commer-, cilbank deposits. SomorrlndoesizIndonesian Financial -Sratiirk (Vadlouis issues). Japan: jDDIFJI (1993); ooi Staijsnal(aiu er) Kore= fronmndcWSraisdt Yew-book (ari,ous years). MaasaQuartrl(yBaliab (199O)-. Singaporc~ Yearbook ofSaiac (1991); Eoomic ~a.uSoraiSwszi 1:160-2,195).Taiwan, China: Sraisrca Yebrnoak (variou vears). Thind& Qnarurly Bulletin(vrosise) Except for Hong Kong and Singapore, all HPAES at Various times have simultaneously reguated deposit and lending rates and consequently the spreads of financia institutions. As a result, rents derived from im- perfrct competition have been modest. Our evidence led us to condude -in chapter 5 that, in comparison with other developing economies, bankcing system spreads were relat-ively low in East Asia. and have de- clined as entry restrictions and interest races have been liberalized (see table 5.7). Given the rapidi growth of dhe financial sysems in the HrAs and their limited intermediation margins, we conclude that regulators -have done a good job of balancing the need to limit competiton in. order to enhance bank solvency with the need to limit imperfect ornm- 279 ~~~~~~~A L E- petition renrs. HPAE regulators did not always get it right, however. It is notable that, in Indonesia and Thailand, regulated intermediation mar- gins were substantially higher than those after liberalizarion of entiy and interest rates. Using Directed Credit Like many governments, those in East Asia have gone beyond the in- direcr guidance described above to target credit direcdy at priority acriv- ities. The categories to which they directed crduit have differed litde from other developing economies. However, HBPAEs, particularly Japan and Korea, have uscd unusual mechanisms that have helped to increase loan repayment rates and overall success of directed credit. They have limited the size of credir subsidies, and they have applied stringent stan- dards to the selection and imlplementation of projecrs. Criteria forSeecting Dreeed-Cree. Tagts. There are three broad types of directed-credit interventions. First, governments can direa credit to specific firms or ind ries. Second, they can direct credit on the basis of broad functional criteria, such as promotion of exports or small and. medium-size enterprises. T ird. they can direct credit to accomplish so- cial objectives, such as mass housing or redistributon of assets among ethnic groups. HimAE governments have engaged in all three. Ta?eting firms or indutdes. Japan cuinng its postwar reconstruction and Korea during the 1970s directed substantial credit to specific sectors and firms, mostly in heavy and chemical industries (Has) (see table 6.5). Japan's FILP accounted for about a third of new equipment lending in the 1950s. Between 1973 and 1981, Korea's policy loans were about 60 percent of the total loans of its deposit money banks. Most ofJapan's priority lending targets through the early 1960s were industries associ- ated with large optimum scales and increasing reurns to scale; so were Korea's during the HCi period. Indonesia and Malaysia, conversely, had disappointing experiences with credit interventions targeted to specific industries in the 1970s and early 1980s and abandoned the schemes in favor of more functionally directed credit Hong Kong and Thailand have not been active in using credit instruments to push selected industnes. In contrast with many other develop'ing economnies, where directed credit often went to public and quasi-public enterprises, direcred credit in the BPAEs has generally gone to the private sector. Consider the cases o80 USING RtESOURGWE,j Table 6.5 Pmoportion of Loans Accounted for by Government's Policy Loans, Japan and Republic of Korea Japan Republic~~~~~~~~6 fKa:reat 'Pa/C P k'ons as a' -: o a Pol.1 lons sare f ne'.7-- sa or 1955 J3~~~.2. 32. .. .- I96 '':. -22. , iV 4965. 09 . .SE5.9uS: ~~, 1970 ~~~~~~9.8 -13.8 47.41 "11.~~~~~~~~~~~J771: 1975 W 106' :1.75.390 1980 1 141 ,~~~w'.u½ 176, 59.78" 19.4 8:12'.~~~~ ~~~~~~~~~* ..;,..e i Table6. loanspertain wi Lows Adeounder by iavemmentad Imn'Policgram (FIlP) Outstanding Rloancre f2 finanKial .,W ............ : ~ I' ' t; .. ', w : '' ................ ' .,1 7 , oF *~~~ PoS loan indude loans0 t udrJ some kind oft exii ;rrna ;&I prg;m tht, lon eted&emnali term .D,>,,0 ,;; inuerestzat eandaada8llt or suppor;tend b'' ch Cenra Bn's auomr 'nicws&tdnglar h au of al lan and d- c-unts -ob privasector made by &- '::'; j'Ban of ICor&de-'o' * fi:'ir -',ey b- ', ', cd iwo _ ,_,_ _ va7&bni rh6~~~~~~~~~dveometinttuin ~2U'Foreig 'rmd.nn o icudkey al- lansfo foa- n i se b,- dt d a d loa : ;te banI a Srnc:JpaCiDqE ( ;Wr;-;rfX 19).0 i-r uS-.rn;n-oth: (192. 8 :, - ' - - :;rl)w of Trkeyand exio. Lans y te bakng system to public enterprises in Turkey'in 1975 were 28 percent of bank assets, and to the public ad- ministration, 19 percent; those proportions rose to 24.9 and 35.9 per- -cent, repectively, in 1980. In Mxico,- the largest development bank provided about 75 percent of its industrial loans to fbur state-owned en-. terprises. UJhfortunatelyv, t-hese public enterprises often did not perform well economically or fina-ncially. In East Asia, Indonesia, Malaysia, and Singapore have directed credit to state enterprises, but the proporions of total acrdit were not p tversistently Ehigh the parastatals tended to perfbrm better financially, and thae intLerest rate subsidies were nor large (except in Indonesia).171Korea was an exception; its lending to the public steel * plant, Posco, was substantia but is widely considered a succes story. Ta.rgeting eorerne and small and medium-size enteTiprises. The broad- est functional-targeting of credit in the HPAEs has been to exporte-rs. All ---an econmic c193)ep o S-goes 1 9t* Easturke an ecnmisexcept Hong Kong and Sigpem have subsidized export credits, often through the central bank redis onSt te (see table 6.6). Exporters from all industries' have had access. Small and medium-size-firms, which often had to rely on curb markets ibund ax- port financing one of their few ready sources of fnral finaneow. Box 6.1 rAa, Maltznus onpagd284) -f -- Sngaore avedircte creit o ste nterriss, ut he popotios o .AC SKAN MIRACLE Table 6.6 Characteristics of the Export Credit Schemes in East Asia Hong Kong OfQers export credit insurance with a slight subsidy through the Hong Kong Export Credit Insuance Corporation (HKECIQ. Japan Between 1946 and 1972, the Bank ofjapan (BO,I redis- countcd axport bills at low. rates of interest (I to 2 percentage points lower than that applicable to general trade bills). During this period, the HOy directed 10 percent of its loans to export finance (which financed 50 percent of toal exporr- related bank Icnding). Kara, Rep. of Offers concessional rediscounting of prreshipment credit based on Ictters ofcredit (tcs). The differential betwecn loan rates of this cxport credit and general loan rates was between 3 and 20 percentage po;nts until 1982, when this difFerence was eliminated. In 1984 the dififrcntia was restored ar 1.5 percentage points. The Korean Exporr-Inport Bank provides longer-term post- shipment finance. Provides export credit insurance, but this has been used spar- ingly (by about 3 percent of exports in 1980). Malaysia Since 197i7, the Export Credit Refinancing (ECR) faciliry has rediscountd export bills fbr postshipmcnt financing Since 1979, the ECR refinances peshipmenr export acedits at subsidized rates for cligible exports, based on lecters of crediL A domestic leter of credir was also introduced to enible the financing of indircct expores. Singapore In 1975 the government established a rediscount facility for exports and re-exports. The Export Credit Insurance Corporation provides insurance for exports by Singapore. Taiwan, Cbina Offers firms short-tcrm preshipment loans based on LCs. After products are shipped, firms are cligible for loans based on LCs. documents agait acceptance of payment, and ship- ment documenrs. In addition, the Central Bank provides a special discount rate to designated domestic banks for export loans, usually at 1 percentage point lower than the usual dis- count rate The oeporr loan rate was about3 to 5 prcentarV,- points below the minimumn intercsr rate for secured loas between 1970 and 1980, and it dropped to I tO 2 percentage point after 1980. Thaand The Bank ofThailand rediscoun:s at subsidized rates 50 per- cent ofthe value of Icnding to exports by comrmeraal banks. Commercial banks lend on the basis of LCs. 28z USING RESOURCS Box 61 Directed -Cref'it Work.s- - r ,Bed as aSupplemen, - MANY ECONOMIES HAVE mID TO USEDIRECD CREDIT TOE-: - counge small and mediuni-size en erriss;, few have succeeded. The -contrast betwcen an unsuccessfiulland ulumately unsusinuiable pro- i.. ,.. ..=;.tE,.,;-n'- in K -a , ; gram in J-Idonesia and a successhfl program in Koreatshows why. * .-f--' In Indoncsia which itdl 'th_e arly. 198 s hada heavily reprcssed financial system similar tothatof' niany, developing economies, heav-, '..y subsidized credit dixectcd to enterpriseir. pnvdmnent-d&ignared - categones often went awry. Cr&diz allocautions wre divcn- by politicalf = as well as conomic considerations. Repayment rates were as iow as 53: percent im one program supporredby 'the World BanL Meanwhile,. porentialy aedrcwdrth medum-ze firms that lacked the politcal .. d R - ' ; ,m a um' ; i V -,P. q.R. d-,--,!j,:- -:influence ofireenterprises and the speci.al cret access cnjoyed by .. smallfs were firzen-'ourtof h finanial systeCM..-- PartyP to re=edj thesewproblemts the goveniment began .o liberal-' "'-izz he fiianial systmm in te -early 1980s, permitringa rpid increase in thenuiiiber of'banls, 'iralizin intert -d. dimann - cdirccted-credit programs. Aa result, medium-:size firms'access to iii' .nance has evidently improved, although te situation may have wors- died somwar for vey small firms l-that lack- ur established :commercial tack record. . In contrast, -Korea's ect caei t-progrmn -or small.. nd ; - : . .= .. -.~h;--.;.S Me- -M., - er- medium-size. enterprises has been compaible from the outset wlm'-a- a disci&pnec- miarket-responsiveFminal: isysen.B ks; whlchflave ''..had ia large degr¶e of autonomyin:choossingborrowers,,have pref-r-ed. f '' rinwith i 'estiblisheduack'record: of 122 firms su2rvd, only20 :.,drew.onKflnand inaik&t rcsouraes during rhei.start-Tuhje Kiorean.- -., program raiieed hrms withournor acesothe banking sector To- keent&rpdsesgoi Mep apMrv idcdfoSorldng,aswdl s- iutvesm5nt finance, \xth working cirzal far eexp,r&avafable on par- .:cuailawly -libenl tem&,Repayment.rates ii Kbrea hare beeAin.-excess-.. npn;rnr . .. -.'. . .- : Perhaps most important -to dme success &4I{e KoreanP program, Jn,:. -. teresr.iates subsidies have ben modest- wtcompason wih thoseT in .:I. n*edonesia and other developzg-e'onoiiest h'hdy;epies-ed fi- nancal systems Despite the KoreanD gov.emnenr's perveitnrern rt > e ysr - , { ., , , c _',pi,' lx, - a.. -. ,; .Fs.,,. ,.!], r,0I. 2n ti7;nfiacamar its;'- dleedcr dr:f4o'r.srn4all =n iumsz _.. -4h 0-- - - _ - ' r * Prlss- 2:rCm SU- c. a *II.MIRACLE contrasts directed-credir efforts for small and medium-size enterprises in Indonesia, where they fiiled and were discontinued, and Korea, where they have been more successful. As we argued in chapter 2, the size of the subsidy was probably less relevant to firms, especially to small and medium-size exporting firms, than was the automatic access. Allocation was stricdy based on export performance, and commercial banks moni- tored performance, since they only collecred repayments on realized ex- port sales; governments, through customs, had the means to verilfr performance. This created a simple contest for export finance. There are other more limited examples of functional directed credir. Indonesia, Korea, Malaysia, Tlailand, and Taiwan, China, have re- quired banks to extend a proportion of their loans specifically to small and mnediun-scale enterprises. Small and medium-size enterprise lend- ing through specialized development banks is a growing proportion of Japan's directed credit There are also attempts at facilitaring crecli ac- cess of technology-inrensive producers, especially through Taiwan, China's, Bank of Communications. Targeting social objecaieL. Some IuAEs have used directed-credit pro- grams to achieve social objectives. As we noted in chapter 5, Japan has directed credit to facilitare the adjustment of declining industries. Malaysia and Indonesia have directed credit to promote income redis- tribution among ethnic groups. Even Hong Kong and Singapore, which have few other directed-credit programs, provide subsidized credit for mass housing. KeepingSubsidies to Directed CmditSmalL Compared nith most other de- veloping economies, the subsidies for directed-credit programs in most HPAEs were small. In part this was due to the way in which directed- credit programs were financed. The credit programs of Japan and Tai- wan, China, were financed by postal deposits, whose repayment set a limit on subsidies. Singapore also ensured that the Development Fund financed by the pension funds was self-financing. It was also the case thar, with the exception of the HMc period in Korea (1973-81), the more prudent macroeconomic and interest rare policies of all HPAEs meant that real interest rates to borrowers of directed credit were generally pos- itive, if not at market-clearing levels.'3 Measuring effective subsidies to borrowers across economies is diffi- cult, since it requires estimates of equilibrium interest rates for each type of credit in each economy. Table 6.7 contrasts real interest rates on di- rected credit for the HPAES with a sample of other low- and middle- 7,84 USING RESOURCES*'EflI Table 6.7 Real Interest Rates on Directed Credit, Selected HPAEs and Oiler Developing Economies (percent) Economy Direcedit Nonireted credt FIPAESI Indonesia, 1981-83, liquidity credits -1.7-4.0 Japan, 1951-60 0.5-3.0 3.1-4.6 Korea, Rep. of, 197040, industry -2.7 2.9 cxporrs -6.7 2.9 Taiwan, China, 1980-89, industry 1.9-3.9 4.6 1984-85, exports- 1.5 4.6 QOter developing economies Brazil, 1987 -23.5 Colombia, 1981-87, industry 1.5 13.5 India, 1992 - -2.5-4.0 7.0 Mexico, 198748 -24.0 6.0 Turkey, 1981, indusay -4.0-15.0 13.9 1980-89, cxports -14.0 13.0 ,,- _No: mIvilable.: .iurerJDB/JER (1993); World Bank d2t:L income economies. Directed-credit programs in the HPAEs had higher real interest rates than in other developing region&s Given capital's high international mobility, we conclude that interest rate subsidies in the HPAEs were lower than in other regions.'9 The HPAEs generally have smaller shares of their credit systems subject to government direction than other developing economics. For exam- pie, in the early 1980s, roughly tiree-fourths of loans from Turk-ey's fi- nancial system were made at government directive. In Pakistan directed credit as a proportion of nongovemment credit was 54-68 percent in the late 1980s and I990s (an internal World Bank report). In Brazil, about 80 percent of the total average stock of credit was directed in 1986-87. Colombia's directed credit to indusay in 1984-87 was on av- erage 56 percent of total credir. By contrast, in EastAsia, Korea's policy loans in the HCI period were more than 50 percent of total credit, and Japan's never exceeded 15 percent of total credit. Indonesia briefly had 48 percent of its total loans in priority liquidity credits in 1982, but it reduced these rapidly following the liberalizations of 1983 and 1991. Moreover, dircted credit is becoming less importanL WhiLe earlier efforts of govremments in East Asia concentrated on developing the 285 ' S!WNMIRACLE banking system and guiding its lending decisions, more recent policies have been directed at deregularing financial markets: Japan started in the 1970s and the others in the 1980s. The fearures of deregulation are sim- ilar: interest rates are gradually deregulated1 directed credit is reduced, and entry restricrions are slowly eased (see table A6.5, at the end of the chapter). The majority of HPAEs deregulated their interest rates during the 1980s, jointly with efforts to liberalize their economies and their fi- nancial markets. The deregulation of entry has been much more lim- ited, however, except in Indonesia. This process in the East Asian economies has been gradual and was meant to minimize disruption in the manufacturing sector. Institutional Mechanisms for Selection and Monitomng. In contrast ro most other economies that have rried but fiiled to use directed credit, Japan and Korea appear to have strong institutional capacity for project de- sign, appraisal, and monitoring. Because of this, direcced credit has gen- erally gone to projects that are creditworthy and viable. More importan-and in sharp contrast to many developing economies- funds have usually been utilized for the purpose for which they were al- located. The result has been a high level of loan repayments and a correspondingly low level of loan losses. In Japan, officials working within government sectoral guidelines operated much like loan officers in private commercial banks, scruti- nizing projecr proposals and the firms involved to ascertain their abil- iry to repay. Once a loan was approved, close cooperation between development and commercial banks ensured continuous monitoring of the performance of the borrower and enabled development banks to take early action if loan repayment was in arrears. Coupled with Japan's overall economic success, effltive pre-appraisal and monitoring re- suited in very low loan losses. The Japan Development Bank, the main government lender, experienced write-offs of a mere 0.09 percent of av- erage loans outstanding in 1951-55 and just 0.01 percent in 1956-65, lower than comnmercial and trust banks that focused on short-termn lending and had a more diversified portfolio. The JDB's superior per- formance may be partly explained by its greater reliance on collateral security, particularly equipment purchased with the loans. Fear of los- ing collateralized equipment gave borrowers a strong incentive to repay their loans. In Korea, the government, banks, and industrial firms worked to- gether and shared responsibility for the success of directed-credit pro- 286 USING RESOURCESO jects. Various institutional arrangements such as monthly axport pro- motion meetings and monthly briefings on ecoonomic trends, discussed in chapter 4, supporred this close consultation (Cho and Hellmann 1993). Enforcement of government goals--primarily the imperative to export-was achieved by rewarding successful firms with continued credit support while reducing credit to less successful firms. Such con- tinual and pragmatic review and adjustment of policy implementation reduced, but did not always eliminate, the moral-hazard costs of gov- ernment intervention. Other HPAEs fiall along a spectrum between Taiwan, China, which has utilized mechanisms similar to Japan and Korea and has had similar suc- cess, arnd Indonesia, which was plagued by the problems discussed below and has largely discontinued directed credit. In contrast with Northeast Asia, credit allocation decisions in many developing economies were often motivated by political and noneco- nomic considerations. Projects were designed with conflicting objec- tives and lacked proper evaluation. Moreover, the very large rents that could be obtained from subsidized credit were a strong incentive to cor- rupt practices. In smaHer economies, credit allocation to specific sectors became credit to individual firms, because of the limited number of companies involved. Consequendy, projects were subject to govern- ment-monopoly negotiations and were prone to poor appraisal and dis- bursements without proper documentation. Credit allocation to large and politically powerful firms or state-owned enterprises were charac- terized by weak monitoring and follow-up on debc recovery, producing high loan losses and eventally large-scale failure of directed-credit programs. How Did Financial Market Policies Contribute to Growth? Capical m;arket policy fundamentals are undoubtedly the major ac- tor explaining the contribution that capital markets have made to the 6f- ficient use of resources in the HPAEs. Rapid deepening of financial markets, combined with prudent restrictions on lending by banks for real estate and stock speculation, are probably the most important con- tributions of East Asia's financial policies to growth. Recent cross- economy evidence shows that the average level of financial development is strongly correlated with the average rate of per capita income growth, both before and after contuolling for initial conditions (King and Levine 287 3iR A C LqR"A`ThMIRACLE 1993). Moreover, the level of financial development imporEandy pre- dicts economic growth.20 What about attempts by government to direct the allocation of credit? Here tie record is mixed. NWe conclude that credit directed to exporters was a positive factor in accelerating growth. All East Asian economies provided automatc, generous access to credit (or credit guarantees) to exporters with varying degrees of subsidies. Japan, Korea, and Taiwan, China, went a step further they used exports as a criterion to determine whether priority firms, regardless of the market for the actual priority investmenr, continued to receive credit.21 Other more specific directed-credit programs had both positive and negative results. In general, the economies with strong civil services and highly professional public financial institurions-Japan, Korea, Singapore, and Taiwan, China-apparendy had some success in implementing credit policies to promote priority industries and firms. Key to this suc- cess was the use of economic performrance criteria and the monitonng of firms by banks and bureaucrats. T hese traits were evident in an extensive study of Japan!s policy- based lending in the immediate postvar period (DB/JERI 1993). The study found thar interest rate subsidies were no more than 2-3 per- centage points and that effecive monitoring kept loan losses very low. Moreover, direct,. :redit was a minor aspect of the overall financial system, accounting for less than 10 percent of total loans. The study identified three characteristics ofJapan's policy-based lending: reliance on the private sector; the dose correspondence between national eco- nomic plans and directed credit; and respect for the managerial inde- pendence of the government financial institutions responsible for channeling funds to the targeted sectors by selecting individual firms according to purely economic criteria. By comparison, Korea's use of directed credit during its economic takeoff in the 1960s and 1970s was more extensive, and interest rate subsidies were larger. As in Japan, the governmnent in Korea consulted dosely with businesses and banks, attempted to monitor perfbrmance dosely, and relied on a competitive business environment to provide an objective measure of performance. Accordingly, policy-based lending probably contributed to the rapid expansion of Korearn indusy, partic- ularly manuFactured exports (Cho and Kim 1993). Most of the directed lending to priority industries in Korea translated into investments in equipment. Despite Korea's severe adjustment problems due to the HCI 288 USING REsot1RCE`S,r drive during the early 1980s, rapid growth resumed easily in the mid- 1980s, in part because of increasing utilization of the HCI investments, which were in physical capital, as demand increased. However, the cost of extensive and highly subsidized directed credits was substantial-and likely to increase as development proceeded. Di- rected credits burdened banks widt nonperforming loans, interfered with the financial system's efficient operation, hindered the growth of fi- nancial savings, and sometimes threatened macroeconomic stability. Korea was able to weather these difficulties partly because its closed cap- ital market at the time enabled the government to suppress interest rates on deposits, thereby funding the subsidies for loans to corporations. In addition, relatively easy credit in international capital markets and Korea's strong relationship with two key lenders, Japan and the United States, helped to ensure the availability of foreign capital, which was ei- ther fimneled into directed credit or went to bail out troubled enter- prises unable to repay subsidized loans. Today, boch the greater openness of capital markets in most developing economies and the heightened competition in international capital markets suggest tiat financial re- pression and directed credits are now less viable options for developing economies.22 To argue tlat directed-credit programs in general reached their tar- get groups and that loans were effectively monitored does not establish that they contributed positively to growth. To do this, microeconomic evidence is needed to establish two facts: government lending pro- moted investment and did not simply displace investment from other sources, and the social returns on the incremental investment ex- ceeded private returns, justifying the targeting of specific sectors. Such evidence is rare-even for the northern-tier HPAEs-but two recent studies attempt to address these questions for a sample of firms in Japan.Z3 A careful microeconometric study of Japan's industrial credit pro- grams in general machinery, electrical madcinery, and precision instru- ments during the 1980s-sectors selected because of their high potential for externalities due to innovation and learning-reaches the following conclusions: * Government lending had a large, positive, and statistically signifi- cant impact on both gross and net investment; the three-year cu- mulative effect of a one-dollar increase in debt financing from 289 MRACSN MIRACE government sources was more than two dollars of additional gross investments in fLxed capital assers. * Government lenders withdrew credit once firms had earned access to private credit markets and, in turn, extended credit to new bor- rowers rather than preexisting borrowers, consistent with (though certainly niot concdusively establishing) the daim that credit was being allocated according to objective lending standards, rather than on thc basis of political considerations. * The characteristics of firms receiving increasing government credir indicatc that assistance was targered toward firms most likely to grow, invest, and generate rechnological externalities; firms receiv- ing credit had research and development investment shares in out- put approximately twice those of finms not receiving increasing government financing. The authors conclude: "Possibly because of the institutional structu of these lending programs, it appears that economic logic rather than political capture motivates policy-based lending decisions. Govemment credit is withdrawn quicldy from seasoned firms, is targeted toward growing, R&D intensive firms, and produces substancial increases in investment and access to private credit. The fact that these results are measured with data from the 1980s leads one to expect even larger ef- fects for earlier periods? (Calomiris and Himmelberg 1993). Limited evidence fiom the 1960s tends to confirm Calomiris and Himmelberg's conjectare. Horiuchi and Sui (1992) examine the role of the Japan Development Bank's directed crcdit in promoting private in- vestment during the 1960s, when JDB was the major policy-based insti- tution lending to medium-size and large firms. Their sample consisted of 477 medium-size firms listed on the Tokyo Stock Exchange in 1965. Among their results are: * JDB lending resulted in a net increase in new investmenL * JDB loans were accompanied by incremental lending from private banks, suggestng that JDB's loans did not crowd out private bank lending. * JDBs loans to firms preceded the firms' increased borrowing fiom private banks and improved the access to credit of firmns that were not artached to private "main banks" (often firms affiliated with a keiretsu bank). 290 USING RESOURCESj = Thus, directed credit appears to have acted as a signaling and insur- ance mechanism. By selecting among borrowers, the JDB encouraged complementary lending from private financial institutions with a conse- quent net increase in investment. Moreover, policy-based lending was most successfiul wlhen it was directed at nontraditional borrowers. The JDB's loans stimulated greater lending by private banks to firms not affil- iated with main banks but did not have the same impact on firms affiliated with main banks. Another study of credit programs in Japan and Korea based on an analysis of the objectives and outcomes of the programs adds an addi- tional element, noting the positive role of coordination achieved through govrmment lending programs. During the 1950s the bulk of directed credit in Japan was given to "basic industries"-electricity, iron and steel, and coal mining-that were characterized by increasing re- turns to scale and that produced basic inputs for other parts of the econ- omy. Once these basic industries were developed, Japan subsequendy promoted industries whose expected spilover effects on the economy were large, for example, machine toolS (JDBIJERI 1993). During the 1970s, Korea channeled most ofics directed credit for fixed investments in manufacruring to the HCI industries, among them such producer goods industries as steel and machinery. Using a framework not dissim- ilar to the contest framework oudined in chapter 2, the study reaches a similar conclusion: the coordinating role of government in Japan and Korea has diminished over time, and hence the effectiveness and desir- ability of using credit programs to improve coordination among private firms are diminishing as well (Cho and Hellmann 1993). But many other directed-credit programs, even in the HPAEs, did not mect their objectives. In fact, the track record of most directed-credit programs in Indonesia, Malaysia, and Thailand was so poor that their governments, demonstrating the pragmatic flexibility tihat we have ar- gued characterizes economic policymaking throughout the HPAEs, aban- doned many of them. Limited differentials between directed credit and market interest rates and smaller volumes of directed credit relative to the size of the financial system meant that financial market disrortions in the HPAEs, even from unsuccessful attempts to direct credit, were not generally large. In addition, the focus of HPAE governments on provid- ing credit for the private sector meant that, in gcneral, loss-making pub- lic enterprises were not priority lending targets, unlike in other developing econonies. Z91 ANL~MIRACLE Using the International Market: Trade and Industrial Policy W HILE THE LINK BETWEEN INTERNATIONAL TRADE AND economic growth is widely accepted, the precise nature of the relationship is controversial. Early writing on trade pol- icy and growth stressed the benefits of neutral incentives between pro- duction for the domestic market and production for export. With neutral incentives, it was argued, resources would flow to sectors in which the economy was most internatonally competitive (Corden 1971), and gains in technical efficiency (or "X-efficiency") would result from a more competitive environment (Nishimizu and Page 1991). Studies in the late 1960s and early 1970s convincingly demonstrated the extreme bias in incentive structures against exports and agriculture in most developing econonmies and the resulting high costs of import- substituting industrialization.4 Recent theoretical work has argued that, where market power, economies of scale, learning, or externalities are significant, departures from neutral incentive regimes (low and uni- form rariffs) may improve economic performance (Krugman 1986). This literature, however, unlike the earlier carefiully doctumented studies of the costs of protection, provides very litde evidence conceming the importance of these factors or the potential benefits to grwwh of depar- tures from trade neutrality25 Proponents of trade neutrality and intervention both dee the records of the high-performing Asian economies as evidence supporting their views. As we noted in chapter 1, international trade is important in all the HPABs and is the factor most consistently correlated with their suc- cess. Balassa (1991), Krueger (1993), Hughes (1992), and others argue that openness to international trade, based on largely neutral incentives, was the critical factor in East Asia's rapid growth. Conversely, advocates of trade interventions, while acknowledging the importance of trade, note that incentives deriving from quantitative restrictions on imports, tariffs, and subsidies were not neutral among sectors (or firms) during their periods of rapid growth. They argue that the HPAE govemments successfully intervened to change comparative advantge (Amsden 1989; Wade 1990; Singh 1992). Industrial policy interventions, which often use trade policy instru- ments, are motivated by the belief that shifting industrial structures to- ward newer and more modern sectors increases the opportunities for 292. USING RESOURCESF capturing dynamic scale economies that result from learning. During their heavy and chemical industries programs, Japan and Korea were the most active HPAEs in promoting individual industries and sectors. Sing- apore and Taiwan, China, have also actively provided incentives for tednological upgrading. Malaysia had an HCI progrm reminiscent of Japan's and Korea's, while Indonesia has attempted to leapfrog from labor-intensive manufacturing to high-technology industries such as air- craft and electronics. How have trade and industrial policies affecred growth in East Asia? To answer this question we first examine those conditions that might justif government efforts to promote specific sectors. We then describe the trade and industrial policy regimes in the HPAEs and compare their domestic relative prices with international prices to determine the extent of interventions. Finally, we evaluate rwo approaches to increasing pro- ductiviry. We find, on the one hand, that government efforts to promote specific industries generally did not increase economywide productivity. On the other hand, the evidence shows that broad government support for exports was a highly efFective way of enhancing absorption of inter- national best-practice technology, thus boosting productivity and out- put growth. Market Failures, Trade, and Industrial Policies Why should governments inreriere with the level playing field cre- ated by the international market? Our ditcussion of coordination fail- ures in chapter 2 touched on many of the reasons, but it may be useful to review some of the arguments specifically as they apply to trade and industrial policy526 Among the many reasons for discrepancies between socal and private returns cited earlier, four stand out: X Interdendent investments and economies oficale. Increasing returns fom scale and capital market imperfections may mean that investments that could be internationally competitive at optimal scales will not be undertaken. This is especially true with large, interdependent projects for which optimum scale depends on simultaneous investnent in up- stream and downstream industries. The larger the indivisibilities and re- turns from scale, dhe more likely that private initiatives will be absent. a Straegic negotiations. In negotiatons with other economies and for- eign companies, governments can alter the nature of the marker envi- ronment by intervention. The outcome of any bargaining problem 293 RACR"MM!IRA C L r depends on the strength of competition on both sides. By coordinating thc actions of buycrs of technology and trying to increase competition amnong sellers, governments can appropriate more of the surplus associ- ated withl the transfer of technology than they otherwise could (Stiglitz 1993a). e Pecuinay externalities. Pecuniary external economies arise if, as the size of a competitive industry increases, the long-run supply curve faills (Marshallian real externalities). Such gains in productivity are attribut- able to economies of scope in the use of specialized equipment and greater specialization of individual skills. When economics are small, currcnt prices may not convey adequate information about prospective lower costs of production through larger plant size (Scitovsky 1953; Chenery 1959).27 Externalities can also arise from the interaction be- tween suppliers and buyers about the design or production of a product leading to a better or cheaper good than is available internationally. In this case, the source of the externality is the nontradability of some types of inputs or knowledge-otherwise the improved method or product could be obtained from international suppliers. * Learning. Externalities related to learning have traditionally been identified as important sources of mark-et fai]ures in developing economies. When firms gain knowledge of production from other firms without incurring cosrs, real externalities are present.28 Because of in- complete appropriability of knowledge, individual firms may spend less on obtaining production knowledge than is socially optimal.39 Exter- nalities due to learning may also be conferred on other firms by the first entrant. These indude the demonstration that the sector is physically and economically feasible and the leakage of information on technology and marketing (Pack and Westphal 1986; Rob 1990). Economists initially responded to these market failures with argu- ments supporting the protection of so-called infant industries. Protec- don was seen as necessary so that firms could gain the experience needed to lower costs and become viable.30 Industrial policy advocates take the aTgument a step fiuther. Because they associate learning with capital- and knowledge-intensive industries, they advocate a rapid shift in industrial structure toward these activities, even if they are not intema- tionally competitive at present levels of scale, knowledge, and fiactor prices. In this view, the short-mn aflocative costs of establishing intema- tionally uncompetitive industries will be outweighed by the long-run benefits of rapid productivity change in the promoted and linked sec- 294 USING RESOURCE RCE tors. But as tlc many inf.int industries that have never grown up amply demonstrate, prorection does not enisure that the promised learning and economies ofscale actually materialize. Trade Policies in the HPAEs Most IHPAES began industrialization with a protectionist oricntation and ihave %radually moved toward increasingly free trade. Along the way (as we showecd in chapter 3), they often tappcd some of the efficiency- generating benefits of international competition through mixed trade regimes: they granted exporters duty-free imports oF capital and intcr- mediate goods while continuing to protect consumer goods. Export prices were set in the international market and were often substantially less than current marginal or average costs.31 Losses on axpart produc- tion offset profits in the protected market, while competition in the in- ternational market ensured that the firm would not suffer from loss of cost discipline. More recently, all the HPIAEs have reduced their protec- dion of import-substituting industries. Below we describe the evolving patterns of protection in the HPAEs, with the exceptions of Hong Kong and Singapore, which adopted es- sentially free trade stances early in their development. Japan. Japan pursued an early import-substituting industrialization strategy, similar in many dimensions to those of Argentina, India, and ocher less successfid economies. As late as 1968, effective rates of pro- tection (ERPs) in Japan were still quite high and exhibited the cascaded pattern from raw materials (low) to consumer products (high) that is rypical of most developing economies (see table 6.8). Unlike many import-substituting economies, however, there was surprisingly high protection of machinery (final producer goods), confirming other re- search indicating that the Japanese atuthorities engaged in an intensive effort to develop this sector. Eiu levels in the machinery sector were re- duced during the 1970s, only after it was evident from export perfor- mance that the sector had become internationally competitive (Itoh and Kiyono 1988). Quite high levels of protection were afforded sectors such as iron and steel and nonferrous metals as late as 1970. Protection in capital-intensive sectors such as pulp and paper and chemicals also re- mained high, to say nothing of the remarkably high levels in textiles.32 Korea. There is considerable evidence, summarized in Pack and Westphal (1986), that Korea selectively protected sectors that the gov- 295 A RACLE Table 6.8 Effective Rates of Protection in Japan (percent) By type afgoork 1963 19 8 _ _-NRP Type -A . -P. NARP ERP 1 1968 Raw matcrials 3.1 0.8 3.9 0.9 -2.3 -3.0 Producer goods -13.7 29.6 15.2 22.3 15.9 7.1 Intermediate producer grods 12.3 28.0 14.1 21.7 15.7 7.6 Final producer goods 15.9 32.3 16.9 23.2 15.4 6.3 Consumer goods . 21.6 44.6- 23.6 . .35.8 . 23.0 12.2 B yindry'.r Iad*m1 :*1963 1968 1972 Manufacturin 32.3 24.2 14.4 Tattiles 54.3 282 .18.6 Spinning 27.1 12.5 I.0 Weaving -44.6 30.5 15.5 Products 72.8 32.8 22.4 Wood products 14.0 25.6 16.1 Pperandpulp 9.7- 18.0 11.0 Publishing -16.7 1.0 . -0.9 Learher and rubber products . 30.9 21.8 12.3 Chernicals 33.4 17.7 8.8 Perroleum and coil products 19.5 14.5 . 7.1 Nonmeallic minerals 22.2 15.7 8.1 Iron andsterl 30.1 30.0. 17.1 Nonrfrous metals 30.4 34.1 22.1 Metal products 13.8 -19.9 9.9 Machinery 36.7 20.0 7.7. Genemal machinery 23.0 14.5 8.7 Electrical machinery 30.9 16.5 -5.4- Transportequipment- 61-5 31.0 9.2- Precision instruments - 34.9 22.9 IG.4 a. NRP=Nonial Rate of Pro tcdon. b. -ERP=Efficdvc Rarc of Protecion. c. ERP based on simple averages of riffraes-- Sovw Fortop panel, Itoh mnd Kiyono (1988); fior bottom pancl,'Shouda (1982). erinent hoped to promote. Protection consisted of both tarifis and nontariff barriers. Table 6.9 presents a summary measure that combines the effect of nominal tariifs and non:ariff barriers.33 The figures span the period 1966-85, and thus indude the early efforcs at industrial pro- motion, the HCI drive, and the subsequent liberlization. The 1966 fig- ures show a relatively protectionist stance that becomes somewhat morc 2.96 USING RESOURC E SE . Table 6.9 Overall Degrees of Import Liberalization by Major Industry for the Republic of Korea, Selected Years (ConsoEdating Both QRs and Tariffs) (percent) Mgjor indunrLy £966 1970 1975 1980 1983 1985 I. Printary Industry 56420 5.5 55.1 58.8 65.3 71.2 * II. Food, beverags, and tobacco 30.0 32A 38.4 49.6 49.6 64.1 - III. -Textiles, clothing, and lcathcr products 34.7 .36.6 43.8 74.0 74.3 87.8 IV. Wood and its produas 24.2 64.0 69.1 76.5 82.7 92.4 V. Paper, paper products, and printing and publishing 39.5 54.7 54.5 78.4 88.8 90.8 VI. Chernicals, perroleurm, coal, rubber, and plastic products 47.9 57.1 58.4 65.9 75.2 80.0 VII. Nonmetallic mineral products 41i. 77.0 76.8 89.1 89.2 91.7 VIII. Basic metal products 44.7 73.3 74.8 86.2 90.1 92.0 IDC Meral protucts, machinery, and eqwupment 41.0 59.3 55.4 63.6 69.7 77.8 X. Other manufiacturing 31.8 39.5 38.7 58.2 65.4 76.1 All manufacturing (II-X) 37.5 47.7 52.6 66.8 71.3 79.8 Lighr industry (II-VX) 33.7 38.2 43.7 62.3 63.8 76.8 Heavy and chemical industry (VI-DC) 44.5 62.5 61.2 70.5 76.5 81.7 All-industry avcrage (I-X) 39.6 50.8 52.3 65.6 70.4 78.5 Near: This table givcs the degrees of import liberalization by major indusuy these are obcained by averngig scrmal dam weighted by the current price valuc of dom:sric producton for respcrive yer Sowcan:Sakong (1993). liberal by 1975. Nevertheless, even by 1983, when Korea's success had become an established fact, most sectors were still protected by some combination of tariff-s and nontariff barriers. While Korea utilized a va- riety of instruments, especially export targets and rebates, to ensure that exporters faccd international prices for their tradeable inputs, there was considerable protection of goods sold on the domestic markec. Taiwan, China. Wade (1990) provides a careful description of the ex- tent of intervention by authorities of Taiwan, China, in product mar- kets. The pattern of protection is not dissimilar from that of Korea. As late as 1972, a significant percentage of items were subjecc to nontariff barriers, and two-thirds of potential imports were subject to nondinal tariffs in excess of 30 percent. As late as 1980, more than 40 percent of imports received nominal protection in excess of 31 percent Indonesia, Malaysia, and Thailand. As we described in chapter 3, Indone- sia, Malaysia, and Thailand all had import-substitution regimes that while modest by international standards, nevertheless favored produc- tion of manufhactured goods for the domestic marker at the expense of 2.97 MURN!iL IRACLE agriculture and exports. Malaysia was notable for low, if variable, pro- tection of import substitutes, while protection levels in Indonesia and Thailand were higher. All three economies began export-push trade strategies during their periods of protection of the domestic market. Tables 6.10, 6.11, and 6.12 provide recent data on the structure of ef- fective protection for these three economies. Effective prorection rates are declining but in Indonesia and Thailand they remain sufficiently high to result in some anti-export bias. The free trade regime for ex- porters in these economies partially offsets the structure of protection. The Closeness of East Asian Domestic Prices to Intemational Prices Notwidtstanding the protection thar exists in all the HPhEs except Hong Kong and Singapore, domestic prices in these economies are closer to international prices than in other developing regions. Two bod- ies of evidence lead us to this conclusion. First, nominal tariff rates ad- Table 6.10 Selected Effecive Rates of Protecticn, Malaysia, Selected Years (percent) Sector 1969 1979 1987 Food prnducts 83 88 Bev:rages 20 38 -22 Tobaco 125 - -26 Textiles - 58 15 Apparel 400 45 6 Wood products 33 38 82 Furniure 40 84 43 Paper and paper products 140 66 29 Printing and publishing - 26 -9 Industrial chemicals 230 32 12 1Peroleum refineries - 22 5 Rubber products 60 129 14 Plastic products 265 312 163 Iron and sted 84 63 289 Fabricated metal products 40 26. 30 Nonelectrical machinery 1,600 89 19 Electrical machinery 130 4 12 Transport equipmenr 185 59 65 -Not available. Note The eff&cive mce of prorection is the percentge by which value added at domestic pnices excereds valuc addcd at world prics. SoweSalleh, Yeah. and Meymnathan (1993).- 298 USING RESOURCE-SC Table 6.11 Selected Effecve Rates of Protection, Indonesia, Selected Years percent) Seaor 1975 1987 1990 Food, beverages,andtobacco 336.2 122 124 Textiles, cloth, and foonvear 231.8 102 35 Wood products -12 25 33 Paper and paper products 873 31 20 Printing and publishing - - Industrial chemicals 28.4 14 13 Petroleum refineries - -1 -I Rubber and plastic products 426 57 48 Ironaandstcel 18.2 13 10 Fabricated meaul products - - - * Nonelectrical machinery - - Electrical machinery - - Transport equipment Non-oil : -1 1 - Cement 63.6 60.2 53.6 Manu&crmnrng Total (eclduding oil sector) 74.1 68 59 -Nor ailablo Nan The cflicrfie rare of protection is the pecennge by wcb value added at domestic prices exceeds valuC added arworld prices Sowr.cBhmacharyaaand Pngestu (1993). jusred for the presence of nontariff barriers are lower in the HPAES than in most other developing economies. Second, comparisons of real GNP across a broad range of economies indicate that domestic relative prices for tradable goods in the H-iAEs are doser to international prices than in other regions. Table 6.13 presents one of the few systematic attempts to compare nominal taiff rates for a broad range of developing economies (Erzan and others 1989).34 The authors have also atrempted to adjust for the eflc of quantitative restrictions on levels ofprotecion. (We have added comparable data for Taiwan, China, and have computed the average nominal tariffi for Korea, Malaysia, Thailand, and Taiwan, China.35) Nominal tariffs in the HPAEs are lower than for any other grouping of developing economies except the island economies of the Caribbean and the oil states of West Asia. The difference between Latin America (albeit before its recent trade liberalizations) and the HPAEs is striking. Thus while these economies favored production of import substitutes, they did so less than most other developing economies. 299 W N MIRACLE Table 6.12 Selected Effective Rates of Protection, Thailand, Selected Years (percent) Sept Marc/i Oct Nov. April Secor 1971 1974 1978 1981 1983 13984 1984 1985 Averages All seccors 87.2 18.6 70.2 - - - - - Exduding food, beverages, and tobacco 44.2 45.9 90.3 - - - - - Manfacmuring - - - 7.4 67.4 65.3 57.0 66.3 Textikl prmducts - - - 248.5 144A 127.8 108.0 118.4 Leacherppmducs - - - 151.3 199.7 199.0 167.0 152.7 Wood products - - 60.7 53.7 55.0 49.7 62.0 Paper and pulp - - - 39.5 43.0 43.0 39.0 53.5 Chemical products - - - 54.3 52.9 51.9 46.6 44.5 Rubber products - - - 48.7 46.0 46.0 39.3 42.0 Other nonmetal products - - - 99.8 102.7 102.3 89.8 108.5 Metl products - - 633 62.0 61.9 53.9 70.9 Machinery - - - 14.1 23.6 23.6 21.0 29.3 Consumer goods and motor vehicles - - - 34.9 33A 33.5 30.6 45.6 - Nor available. Noz The effective mrre ofprmtecton is the percentage by which value added at domestic prices exces value added at world prices Swxrc Brimble (1993). Table 6.13 Average Tariffs and Para-tariffs, by Region 1985 Tranip Ceatral Soudt North Saharan West Otler DeveAlpg All para-ta f Caribban America America Africa Africa Asia Asa PEAt' aglow Tarffi Economies unweighlted 16 23 34 29 32 7 36 19 26 Economies - import-weightedb 17 24 38 30 35 4 22 17 24 Tariffplus pars -tar:ffi Economics . unweighted" 18. 65 46 36 34 9 42 24 34 Economics import-weightedb 17 66 51 39 36 5 25 20 30 - a Simple averages across products and economies.- *b. Simple aveages cwss products across countries, avcegs weighted by otal impor. c Indudes Kdrca, Malaysia, and Taivan, China. d. For Taiwmn. China, 1981 dat:. now= Erzan and others (1989). 300 USING RESOURCEStURECF This is borne out by comparisons of international and domestic prices. Figure 6.4 shows an index of outward orientation based on in- ternational comparisons of price levels and price variability for the HPAEs compared with other regional groupings (Dollar 1990). The HPAEs as a group are more outward oriented than other regions; their relative prices are doser to and more consistently related to international prices. Of course, there is substantial variation among the HPAEs. As would be ax- pecred, Hong Kong, Malaysia, and Singapore rank in the top nvo deciles of the index, consistent with evidence on their stmcrures of ef- fective protection. Somewhat surprisingly, Thailand, with higher effec- tive protection, also is ranked in the top decile, while Indonesia is in the top third. Japan, Korea, and Taiwan, China, rank in the fifth and sixth deciles, below such developing-economy compararors as Brazil, India, Mexico, Pakistan, and Venezuela. This is consistent vith rhe evidence that the three northern HPAEs intervened far more frequendy and sys- tematically in their economies than the southeastem HPAEs (Dollar 1992). While any large, multi-economy effort at real price comparisons is subject to methodological and empirical criticism, the results are broadly indicative and consistentwith other evidence: EastAsias relative prices of traded goods were closer on average to international prices than other developing areas. Openness to Forin Technolo An important fctor in East Asia's successful productivity-based catching up was openness to foreign ideas and technology Governments encouraged improvemrents in technological performance by keeping Figure 6.4 Index of Outwanr Orientaion HPAEs OECD economies South Asia I Latin Ameadca and Cadbbman Mldcile East I SabSaharan Africa 0 1 2 3 4 Svnnre: Dollar (1990). 301 .SIAN'-MIRACLE several channels of international technology transfer open at all times, even though some, such as direct foreign invesunent (DFi), were re- stricted or dosed for varying periods. In contrast to most economies with imporc-substituting industrialization strategies, even whcn protec- dion was practiced with respect to the domestic marker, the search for and absorption of foreign technology was encouraged iWhile Japan and Korea set obstades to DPI, they were hospitable to licensing, though even here the Japanese attempted to ensure that they did not incur ex- cess costs. Singapore was exceptionally welcorming to direct foreign in- vesunent, and a major task of tie economic planning agency was to locate appropriate foreign investors. Malaysia has aggressively sought ex- port-oriented DFI, particularly from Japan (see box 6.2). None of the HPAEs was hostile to the establishment of local buying offices by inter- nadional purchasers, an important source of production and marketng knowledge. This selectively pennissive attitude toward the acquisition Box 6.2 -Forign I stEn BuigsExport Technology to Mablysia- : ;W.6.:--, .ri . , - : a .INmTHE 1970SAND EARLY 1980S DIRECr FOREIGN toward eecrcal- and electronic ;priductsC chemical2 investment Was weloMe in Malaysia. Gonenbbe rduc mel proucs, and policy -encourd foregn investor ye the eec pelm. More oftheinvtents were-dedicated:- : was. inhibited by- i a of i ing t- exorsiv before. ;-Dlocal partcpaton incorporate owner. Then, -.e The-benefits toMaysia of DF havibekensub- -receson of 1985-86 large fiscal antrdedefi stnal par genrang reign echange and declinig invesmt broun htan aur spuh for a ploymenr.-.Howe er, some citics -argue that for -xpos o d i eloc no-my .,pC . . . . . . . : PO Z - - . .. , . . , . = V . . .. . ..I. t--- ax cenives andrladdomesticquity strast,not atogas i neighig& bparigpaiong requzrcnmcis-or potentialinvesto, cpits of DETwoimportntreasonsfirtatare, tecleogy and-e-portan& employment thec young g :#ny of the investments LIn ~~-.,,CO~31 Wr-.-an MC MU O opporrumnes. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Malayuiidd eaiescaz&cl ofmangainet cac The new flcdilihad a draiaic confo padiyIskilld lbor,and h-qualiy supplierate- averagd U US$300 million ayear in198385,jmpe foreig inveset pj cts thmselves -h to US$2 bbi981 -. t "became rnbuied to blng. no-hoyw in iec y ''. --Som-L.@ tii&&ibJanii -- ! - f-mc d- moredynaic. n>198a qU2rerF Ere,T og 1985 tn tirie mrcnsmiocntrm nare frriiJapzi>fllo~1.. arwah, Chi&anc &crrinMlysase9io~ than$0 ilo y when thegreatest soure of-D.El hd'bcen_:W i;L u a tsi e -k gaporet nanfrlly afraed tiicafti&and bfisb&ffrtedrd hertcnooyr kceepu 302 USING RESOUPRCEts;EPfl of knowledge of international best practice was a reflection of the view that the world market for goods and services provided an opportunity, not a threat. In contrast, many other developing economies that tended to em- phasize the dangers ofopcning to world markets were equally suspicious of open policies with respect to knowledge acquisition. Suspicion of ex- ternal trade was often reflected in a mistrust of DFI and licensing. The absence of exports reinforced the suspicion. Even whiere DPI was permit- ted in inward-oriented eco nomnies, it was not viewed as providing access to i nternational best practL e b ut rathier as a source of additcional domes- tic production. Thus some economies in Latin America that were hios- tile to licensing nevertheless allowed DFI in production for the domestic market. The basic difference between these economies and Singapore, which was much more heavily dependent on DFI, is that the multina- tional corporations locating in the latter could only do so to export, b~ ~ ~ ~~ , --actmis - - er ., .o M2)Sa. - eprs W i-h& t'u n semiconductors to their assembyatvtes hr s toMlyi. The eiets aysihabulupn also no doubt that Dfl has,creaced a lot of empiloy electronics.is. Con sidemrblc NIxdrTomue AG, mea-often. fbrpoor. rua aay. women-=to, e- .wiEc recen'dy established.a US$3,'' iniIl souibfware:z- tune of 85,000 inelectronics alone in the late 1980s cetr ocreate4applications 6r, UN-aeworkism MAiyaysi- has found :that,- ainetments mature- sin>a trce oMl yiabth&r'arts --adluand ndrginl'ikae aefrtifed ml-. sotaeegnes&roe-fifth of Germany4 snre nainl hat' are fe' huh fa "foodoiose" *.called ii its Mlyineprst_hpstu;ci;. stay, and the~ poiiplovers ~created-y succesfu asmlline iinArzo'na.' vntures encourage more. investment. Fo xapc Mlasa spanproduce:a:-rang-fpout- *-~ Si c th 90,DI1 nd aysi '& rom coiorTy<~,xado6Lassett-pyepl;;ar ioatD5asters cx word'sithird argest,--producer f. semic-bondiu"cto& compdze"eiiheras. Mcomom istat scarte& i--, Tmrld' I li- Ma .n.flt.2 -garnering billion. almost US$' - in-ne, exportsin: asnl inme Mavlyahved!n jon - 19 I6.hth-laiil 980s, -_Siemhbuiltin iiMalaysiitts 's bcm i4le i-svmioinrecrsi fourt plat in he wilctto manulicture m~egach ~ Maaia -hreit .asernle;col& ilvisoi ad: Unfite rie-s). Birocincmais a&okp os oief die ornporflezfbr:itsnew clri to _seinc5ndu&rois-- vestments in dectroAics have plntffaiya,jn th&tat&-1980 Matus `ia's r~ ~uppljilocal ~eincohduccbrPa4s.A i%5Pi~&nr4' rei:ywor1d= m rke Sr-room tair,-'- Motbrolaind Hiachilii~e ~llo(r& tSei~a&has ~conchcioner n op.(-1991) 303 MiR A C LN.MIACE given the small internal market. Korea and Taiwan, China, with larger domestic markets, created a similar situation by encourging DFI in pro- duction for acport and in those sectors where substantial technology transfer could be anticipated. In contrast, the large import substitution economies, such as Argentina, Brazil, and Mexico lured investment of foreign firms by offering access to protected domestic markets. Industrial Poicies We define industrial policies, as distinct from trade policies, as gov- ermnent efforts to alter industrial structure to promote productivity- based growth. Productivity-based growth may derive from learning, technological innovation, or catching up to international best practices. All the HPAEs, except X llug Kong, have employed industrial policies as defined above. Japan and Korea had che most systematic set of policies to alter industrial structure. Efforts in Taiwan, China, were less system- atic but were nonetheless widespreacL Industrial policy in Singapore was more functionally directed at the rapid upgrading of technology by di- rect foreign investors, regardless of type of output. Indonesia, Malaysia, and Thailand have all used industrial policies but much less systemau- caly than the northeastern HPAEs. Industrial Growth and Producivit Change. HPAE industrial growth pattems diffier from the patterns in most other low- and middle-income economies in the relative size and growth rates of two important indus- trial subsectors: metal products, dectronics, and machinery; and texies and garments. Table 6.14 shows the share of value added in nine Inter- natonal Standard Industrial Classification ('sic) subsectors as a percent- age of the tota value added of manufacturing. Among the HPAEs, metal products, electronics, and machinery (isic subsector 38, or MPM) have grown unusually fast The sector's share of manufacturing value added doubled in Japan and Singapore, nearly tripled in Indonesia and Korea, and quadrupled in Malaysia. More surprising than the importance of growth in MPM, which provides vital inputs to numerous other manu- facturing subsectors, is the contnued importance of tctxiles and gar- ments even as the rapidly developing Asian economies shifted fiom labor- to capital-intensive production. Appendix 6.1 describes our method for determining the projected relative importance of specific industrial subsectors in cross-economy comparisons. As can be seen in table 6.15, both capital-intensive seccors 304 USING RESOURS O Table 6.14 Current Price Value Added as a Share of Manufacturing Value Added -m TM"& w-d X.- _ - - pA- ' ;dr Eismjf hwo~~~~~~'" ad ad...d "ad u ~a HPA& 3973 0.05 0-. ' .I gm 0.12 Ut 0.01 0. 33 gmJZ 3573 L15 .4 0333 "I 0.36 00) DA3 0.0 DA. IW 3511 0.-16 0.35 all 0.01 0.35 0.05 an0 0.34 .03 0.59 3053 IL"9 0.36 WI 0.10 0.36i 035 0.12 ON2 0.02 IL" MP a" am oi3 - e5 . -0.6 0.4 s os 1 0.46 e a -Y 196 .2 0.39 I.3l1 0 0.21 00Ms 013 013 Wu0 3.00 ISO538 0.13 0.35Om0 0.05 .3 as6 Alli 0am .0 La 1960 0.36 5 0.7 all 0.27 -O0 0L a0 0.23 J1A 3583 0.18 0.07 07 00 p.27 0.06 0.1 0.27 013 tAO0 3,319 0.05 0.0 WI3 0.05 0.20 003 0.01 0.6 "I a." 1313 e.s L137 0.33 1.04 O 0.16 -D0 0.32 .Ot 1A 3936 0.42 0- 02 0.03 0.103 0.9 DA0B 0.0 0.1I 0.01 1.03 3970 0.32 0.23 3101 0.A5 WO3 04 0.12- 0.2 O.01 0IL" 1916 0.12 0.5 001 004 0L4 05 0.10 0.9 01 .03 1970 0.24 034 0.0 0.0 0.0 0.04 Om.02 0.06 014 14" Iwo 0-17 0.33 0.5 - .06 0.19 OA 0.12 020 I0D * 3965 0.24 -04Z am 0.0 0.24 0.0m 0.011 .27 0.0 30IJ g 3970 0.37 0.12 .3 01 W0 0.03 - 4.36 - 0 '-56 1916 Gm2 0.33 0.3 -L30 0.36 0,4 0.23 0.7 am0 Lot 1970 o0e5 013 0.06- 0617 00e4 0to .314 0.l 30 19[g 0.30 0-5 103 .C0 01.15 a3 GAe ILl5 01 .00 I=7 0.31 0.8 D101 0.0 0.35 0AS 0.24 01e 0.0 1.03 * 1,34 0.~~ ~ ~~~~~ ~~28 OA05 AM3 01 0.37 0.33I 41.1 0.5 amo 3.0 (mpm) and labor-intcnsive sectors (textles) play a greter-than-predicted role in manufacturing in most of the HPAEs (excepr Indonesia and Malaysia). Besides the unexpectd prmninence oftextiles and garmencs, the aaalysi yields another surprise. Chemicals and nonmetallic miner- als, two sectors often assciated with indusErial modernizadon, ame no -large in the HPAES by interatdonal s o.ndards. Detailed sectoral growth rates of total factor productivty are available for Japan, Korea, and Taiwan, China. There are now sufficiendy long time-series dat to conclude that in these economies, T* growth has ac- counted for a substantial raction of the growth of constnt price value added in manuFacturing. Table 6.16 shows a varety of-esmates oflong- 305? AiRAC MIRACLE term rFP growth rates. Given the length of time of the observations, it seems unlikely that the measured growth rates of TFP could be attribut- able to cyclical phenomena or growing capacity utilization of initial large investments.36 Sbtres of Selective Prmnonion. Here we describe the strategies of selec- tive promotion that have evolved in the HPAEs as they developed their industrial policies. Japan. Early Japanese industrial policy aimed to encourage sectors that faced income-elastic demands in the international marker and ex- hibited economics of scale, large fixed costs, and the potential to learn from experience. This purely economic rationale was supplemented by a sense that some sectors were critical to national morale and the Table 6.15 Actual/Predicted Share of GOP Originating in Manufacturing Sectors ITC ir.. P4- r ukA N..- ;,.daV TsIt &nwqd# hnnwj. a&d a" _d ad a Mlfk Uhr sad m - rqi" gadbe &tS.. p-AKw jminaAw ruUvr uhmn6wf nuMb 0nay ter ftrn 1973 e12 10-3 lo .89 0.53 0,7 e.14 0.37 2.6 0o35 1 2s 9ISBt 0.43 IMF95 036 065 Dh67 0.14 0.17 11.6 0.3 1.2 1973 1.63 [.7 0h30 O3 IAO 0.20 0.00 0.44 0." Oda 15SC 0.75' 03' 1.42 0.6 0.54 095 Z.62 a.ll 0.01 0.57 _,~~~~~~~~~~~~~~~7 :-5 19 o n 1329 2.412 1 36 0.73 0.63 I.W- ' 423W7 9s O.98 13a52 ns a,66 0.67 0rt oil L69 0,14 e0 Kar. Rxor d 1563 0.99 17 2.0 1.04 0.94 0:73 0.52 2.0 0.20 1.04 19 0. D35 2:4 o.n1 0.64 0.56 02 1.13 276 0.32 1.26 155 0.J5 025 1.55 O." G." 0o.e5 @0A 1.03 1.43 0D4 19B1 - 0.94 0e5 33 0.7 C."9 0.70 Jo57 '.44 0.07 OS7 1973 0.31 2.14 2.40 0.49 L09 0.75 1.53 1236 0.13 IAI t9S9 0@36 11.32 0.31 0.47 t.7 0o27 oa 50 0.34 to3J 1968 0.!0 121 4.35 al 0.3 1.9 02 3.16 0.09 0.95 19: 6 .34 3.33 1.85 071 0.65 135 033 1.32 022 3.s SeM - 1I56 0.50 6.00 0.54 233 033 0.37 1so 3,74 0.32 1.10 19S6 0.56 - 2.39 (L 1.02 0.39 0.35 0.63 L51 D.09 052 Lai . AM 'listS_;a - '- 1 - 3963 0.91 I 10 1.34 1.03 Q69 0.51 1.31 218 G.AD 1.02 1785: 03 4 170 1.21 0.73 LIZ 0.48 0.11 1.92 GaG 1.09 Chile 193 1.11 3.04 1.21 1.07 oi9 0.73 L63 2.65 0.24 121 19t0 1.1 3.M6 I1 '[94 -.67 0.43 436 D.8S 4.03 1.1 1915 : 37 0.45 0.41 15 -76 0.84 37 am64 0.0 1954 _ 1.57 ' 4 0.15 t a a.70 1.04 am e l _32 - £. ?dk±cJ uhiew 90711C aku*t-l.t..p.. -306 US ING RESOURSCOU Table 6.16 Long-Tenn TFP Growth Rates by Sector TFPgrowtld rare Taiwan, Japan, C-/ina, Sector Kirea 1960-79 1966-6 Food- 7.30 -1.76 2.0' Bcverages 7.90 0.0 Tobacco 13.40 - Apparel - 1.98 10.5 Tcxctiles 10.70 0.47 7.6 Lcather 12.60 1.03 _ Shoes - 1.03 Wood 9.40 2.81 o3b Furniture 12.10 1.74 Paper 8.20 1.44 2.3c Printing 10.70 -0.18 Chemicals 13.10 3.36 3.3 Pecroleum -0.30 _3 55d O.Ob Rubber 1 1.40 1.02 63b Nonmetallic minerals 2.80 - 2.4 Basic metals - - 7.2 Iron and-steel 3.70 134 Metal products 7.60 3.41 4A Nonelecrical machinery .00 - 230 67' Elecncal machinezy 10.70 537 Electrical equipment - - 7.1 Transport equiptnent 11.20 432 2.7 Prccision instrumencts - - 11.0 Plastic products - o.92f: 0.0 Oilher manufacturing 7.50 -1.76 Average 8.8 1.2 4.6 -Not availablc. a. Food mnd bevrages. b. Rubber, petroleum, and wood products. c Paper and paper products. d. peuroluem ref and coaL- c. All machincry. f Plasdc.. Sowre Pack (1 993b). achievement of international respect37 In later years, industrial policy had a narrower technological focus, for example, promoting establish- ment of the technological base for very large-scale integrated circuits (Borrus, Tyson, and Zysman 1986). Among the sectors carefyl7 nur- tured were steel, automobiles, textiles, shipbuilding, and aluminum re- fining in the earlier years and then electronics and semiconductors in 307 A :LA?WUIRACLE later years.38 (Effective rates of protection in these industries can be seen in table 6.8.) Japanese industrial policy until the early 1970s had several strands (Yamamura 1986; Yamnamura and Yasuba 1985). The government di- rected credit to large firms and protected the domestic market to help the firms realize static and dynamic economies of scale. mmTt allocated foreign exchange to favored large firms and tried to ensure that technol- ogy payments to foreign firms were neither excessive nor duplicated, often forcing licensees to share their information with other Japanese firms. Relatively high levels of protection were used as part of a strategy to help firrns to export successfully. As late as 1978, the average of effec- tive rates of protection across the manufacturing sector was 22 per- cent?9 Such rates may have been responsible for the very low level of manufactured imports, an issue widely discussed in the scholarly litera- ture (Balassa and Noland 1988). Korea. Korea's policies were similar to Japan's with respect to selected aredit, protection, and limitation of entry into specific sectors (Padc and Westphal 1986). However, the Korean government promoted individual firms more often to rectify perceived entrepreneurial and skill deficien- des, using export performance to determine whether firms deserved con- tinued promotion. Other policies to encourage industrial growth and exports induded making direct and indirect inputs to exports available at world prices. In Korea, the selectively promoted sectors were the heavy and chemical industries: iron and sreel, metal products, machineiy, elec- tronics, and industrial chemicals. The motivation fbr these appears to have been both strategic-to increase defense capability-and economic-to shift to capital- and technology-intensive sectors in anticipation of a loss of competitive advantage in labor-intensive sectors.4t The costs of the Ha: drive are still not filly known, but they were high (see box 6.3). Al- though there are no recent esrimates of Korean rates of effective protec- tion, there is considerable anecdotal information suggesting that the government afforded these sectors relatively high protection. Taiwan, GCina. The intervention of Taiwan, China, in manufactur- ing has been similar to Japan's and Korea's, though less important quan- tiratively. Wade (1990) has documented tariffs, quantitative restrictions, and selectve credit policies, maintaining that the success of Taiwan, China, was at least partly attributable to an intensive government effort to direct the economy's sectoral evolution. This conflicts with the stan- dard neoclassical view that its development was primarily attributable to 308 USING RESOURCES.E low protection, tde availability of inputs to exporters at international prices, a conservative macroeconomic policy reflected in low inflation, and competitive factor markets (Little 1979). Unlike the situation ill Japan and Korca, in Taiwani, China, it is dif- ficult to discern a pattern of economic incentives. Wade (1990) notes that a guiding principle in the 1960s may have been the existence of . ........~ ~~~~.. Box 63.31Te HCIDrive:Costsof lidtervention THE-KOREAN C;OVERMENT'S INTERVENTION sidered the lower bound, however. Those industries .policy, most notably ihe HCI promotion, has often whose credit was squeezed were ,frced to borrow at . been evaluateCd from thc.perspective of the success or- curb marker rate. The reaL effectivetsubsidies fior rCI .. fEijure of industrial policy. !Aniorher ;inporrant ap; . - borrowers weitreherd6rcbas hrigh as3 percentof GNP. proach in evaluaing int-rentioh policy, however, the Korean goverment opted to bail our. the. Wo woud -be toestirmate the fisal/financial cost asso- strugglingheavy machiney; shipbuilding, overseas r CaCed wit intervenCiin. Govcrnment intervention in-:X-- construcrioh, and shipping indusries in' th mid- .-',.cureddiecthefosts thform ofsubsidies to strateic 1980s, nonperfbrming lo of conmrcial banks sectors through-policy loans andl tax-exemptions, es- accumud rapidly and accordny bank prof- ¶ecialfr during the..1973-79 periodof.HCI .promo- i;abi.y seriously detoed During 198-7, fr McQf nflosPA in:iec tidon. Intervention - rred idirici c t exaanplIe, the share of nonperfmii lons I to -rmn .of accuimiated nonperorming loans assets readied eInost 10 percent. Between -1985 and: resulting port:bli difficties'ofcommeicial banks.: 1988, 78 coqorations were "rationa DrC ThepornmoUooftea scows spporteda ti process, rn b. abroaod ran.oft fscalae,dFCa.ncirwa.i'srumets. iwntie-o. of principal aonc amounted iy~-oroid-='rangds deo vted- totheCninstrent. Far tr*illonwn,-orabour-I.prcentofcNxin 1985. Z .-Goiernment ids t.devotedal dtoe-tc dHC.,isector d-am Farmaging-was tihe frter issunceofsubsi- amounwird roperceatord2tc rationaiefis adt p -romotrionpmo.n aton, ations to Cen Bakdiscoun-rswhich hash:urtbthe Upfipjwo 100pr.fr:hi ea ad 50 ..balaces . ; .0 percht-irfor the` ew yas -provi eK overnmeab d r rf ;'-rate abouta two-tl owerfor strategic IndustrIs-, encesd uring the traumatic economic adj(nt nt of. beilijoeawon q-;~ drive in.197 alone,abou& 82 1l979-8; the direct costs ofS l u- e vbenuo won s o(3t percenmeofrtot sixt gb'p a.~~~~~~~ ~~1ost ' of fi ~ ~ ~ 5qt;ra;n it has taken, mre tan-a ;- XstedT ha..spredomnant .sdircefi- -. . + .idede toawith ind costs i n .th ?-.-,.&asupoirH U-wa:pr&etl pOliclon 1 . form of accuulated.nonpdomn'on an dirtedtokey-industes. n977r snp C45 'squed ba pr N 93t percent wfahe:oiidome T& tioflthe kaMng tKore governt aonce med pl .nanci I r iii dHCIi uppsorCof theiic to Insfpoic lo stlact rhplicinin&i&~~~~P oubida ib n6perceinnt domesais,ei- ,.sectr: I.l HdVseccoI'in. more than40 . of tot 1977-a wroe ati5& 75io & separate.; Da&k- ad to.z adle.t tc piceit 4 tGNP '(6kIiby aitg an nerest< troughicbuaet atyet to bedeermicos:ts.-:. Afrgen&al bankIoSii4- Ths eshdmshoulbe con- (1993 NM-stB. ank (1987) LKep Pew _;- issW_-_. L-:-, V- It... orin:.gounentannoLnce~a-memura-; .pan---- ;.-...; … - 309 ?SdRA C L EMIRACLE gaps in the domestic input-outrput table that revealed potential areas of import substitution. The autobiography of a major architect of the is- land's development supports this view (Li 1989). Although anecdotal evidence summarized by Wade demonstrates considerable intervention in many sectors, there are no studies of effective rates of protection or subsidy after che early 1970 . The more recent intentions of the gov- ernment may be discerned directly from two infrastructural effiorts, namely, the Hsinchu Scicnce Park and the Industrial Technology Re- search Institure. Both were major investments undertaken by the gov- ernment to provide the basis for a rapid shift toward higher-technology sectors. Malaysia. In the 1980s Malaysia experimented with a heavy and chemical industries drive similar in focus to those of Japan and Korea but dependent on public investment. The Look East policy adopted in 1981 was, apart from its sociocultural aspects, an explicit attempt to em- ulate the heavy industrialization efforts of Japan and Korea, which Nvere regarded as successes. The government created the Heavy Industies Cor- poration of Malaysia (HICOM) to be the vehide for the heavy industrial- izaton push. HICOM targeted a number of large-scale, capital-intensive projects for development including iron and steel, nonferrous metals. machinery and equipment, paper and paper products, and petrochel... ails. While public investments were intended to be catalytic in each of these activities, die government actively promoted joint ventures with private (usually foreign) investors. By 1988 HICOM had set up nine com- panies, employing a totai of 4,350 workers, involved in steel, cement, motor vehicle, and motorcycle engine manufacruring. The highest pro- file of these investments was the PROTON car project, a joint venture with Mirsubishi designed to produce 100,000 units per year. The ambitious targets of the HICOM program ran into the maao- econornic constraints described in chapter 3-declining terms of trade and a deteriorating fiscal position. Prompt responses to the deteriorating macroeconomic situation induded reductions in public investment in the HiCOm program. Poor management and low profitability also forced the treasury to bai out loss-making enterprises, a practice it could not sustain under the fiscal austerity of the adjustment program. With the shift to policies of privatization of public enterprises and private sector develon- ment in the late 1980s, the HICOM period had ended (see box 6.4). Indonesia and Thailand. Neither Indonesia nor Thailand has made the systematic efforts to change industrial structure characteristic of 310 US ING RESOURCESCE '-'Box6.4Ro-' gnizingea Mistke, Malasi Pratzes. -THE PUBLIC SECIOR IN MAIAYSIA WAS RELATIVE- many so-lled pivarized firms becae more pr ly snall util the late 1970s, wheni the gopertlnment - ductive and innovative. Nor can the macroeconomic .rapidly,expanded the umber of stat efirms, pary to bencfits be overlooked. in the end, tihecderal gDi'- -,;;givehMlay's opportuniues and.partlyoutofa convic- -emuent anticipates thata-tleast 15 percentofitsour- i, noniat the govea entcould s de&elopmentr. snd -in debeen trred to privafizedi M LSifuch of th& vgrowt i the early 1980s was due to firms. 'HICOM,. the Heavy and, Industrial Corporation of Malaysia's slow but steady divestiture and' the Malaysia, which esiablished nine subsidiaries 'in such, b enits it realized by 'adaping flexibly to thcir- =areasas steel, automobils, cement, ant paper. cumsrances of-each--cas provide anI examplefor .- , State :fkis, especiallEy in HCOM, performed less. m . niany oter deveopig cnos, wh privatiza- ,'wl thanexpeced nd in 1984 :the defici of the- 'tion' is an offical goal, yet much less has been p'ublic enterpnses tached 3.7 p ercentofGNP.As ris-.- achieved A derailed analysis of the compauies that ,, Dipg fiA:il cEdfictraa&la mngs of trade threat- -w:er privatized'shows consistent gains in soial we0 -ened -macroeconomic stability sthe govemenc fare. As two cases belowillte, changes m incen- begat pdvaize in 1983.iOf m orerhan 800 state tives and a broader role for priate sector interests fiMrs existing Mi th mid9Os, about a hundred had, have 'frequently been more imporrant.-lthan the been by'WO.Maire s conmecial change in. oners pe . . manuring oper s,butsome were quite big,. M A&r?Systems Berhad (MM).- -Wlhi48 .- incl=du,ding an--airline; a shipping' company,, and a percent of AS was sold to the publicin 1985,ther telecc}tnmuincatiotis fim, wa no gc in an t or de 'jur autl{orit'" * there have beeinglitches.-For instance, t public Eve so, managment gnd de t contrl over offeingshave been heavily esscbed,:suggt investrment decisions and eventulywon permission' in ; that the:government upde.*ied.i' s sha'res per- to. raise domesc- fares. Whie the farie har. Imps l6sing as much as US$150 million i, t sl of consumers, less, rationing -o - tickets ra c ur big companies.A-lso, morc atenon to promor- w r Moreoe, ile a b b- , u wr',p r; "O ';L Mocvr-1ic ALA l'i2bu b-,cn even, ',R :. mg¶5mpedrion h monopolisriasitu?aribnsvol 'in the early 18sir becamec a~ profitablta-paying Dnont-. Finally, priver&.to-h _a; 17Y corporation? ;'on-as proceededveslowl; Ii.-mg Co.tainer Temindl (z).. 'In 1985,90- while ir contintue to s day, -ie Malysia ;piblic Kr f was sold to de-izep - Sect -f or r ; X emain ., , ,, . ;; . T . . t ;, sL ibni d - -e.; --. dlS. - - to., the pia hysia. s -meant thoughte ublic secor retained control, the more.than ounigh sales. Ith. indmcude'l ashosr"of ' ;.rm that took over, many repres'nnt'' Bumi -p '-; :schems c.thh' d h ontro -' srests, ireincentvet to ear Pr6fit.nK ~~rodxicri tvi&de 6& eII&' 'ubl-PiIcsc~ ui tovrpro tiea cotrings r phec s a luc c r es, nopbz- t H-tover *as- theaiSaxontrall PRO- --1,ing container. traicinand out ofdie caf nil reion ON,taornil company inWCOM. Even ienad E o ref.ier . ah :2 ~flrn ihake no, t. .nb-e'rir the-fi-ms&vi s.s sub tanrlyimprove iworkeri,productv-- y__Fxpcrcnxe m mnagrsinat-nyaestanisinad shamtrome.' w, Ke, 'SczE,,edr ,h&,,,6nrinf taonintd r'' Sw,,t; (iu) L (1992); - kspoldec mor' toit~z rk&4he~~ wa tha ~ Slkii(199); od Bn (.1988)"'- 3"I - _A C IRACLE Japan, Korea, Malaysia, and Taiwan, China. Thailand's Board of Invest- ment (BO}) has attempred to promote various activities viewed as having the potential for technological learning, but, as we pointed out in chap- ter 3, it is difficult to discern any sysrematic effort to change industrial structure from the pattem of promotion of the BOI. Indonesia has used public investment in an effort to move roward high technology indus- tries. Indonesia's technology development efforts represent an cxtreme case of state-supported technological upgrading, but the public sector research facilities and strategic industries have fev links with the private industrial sector. One experiment with leapfrogging into high technol- ogy industries-Indonesia's attempt to create an aircraft industry-is described in box 6.5. Did Indushial Policy Increase Productiv? Have atrempts [o alter industrial structure helped accelerate produc- tivity change? We attempt to answer rhis question by addressing rwo re- lated issues in appendix 6.1: * Firsr, did industrial policy alter the sectoral configuration of in- dustries in ways that we would not predict based on facror inten- sities and changing relative factor prices? If changes in dhe seaCoral composition of output are largely rnarket-conforming, industrial policies must have failed irh at least one of their objectives: to guide industrial development along paths thar it would not take if it were guided by market forces. * Second, what were the rates of productivity change in industry in the HPAEs, and what were their sectoral patterns? If rates of pro- ctucivity change in industry are low overall or in promoted sectors, prima facie, industrial policy did not meet its productivity- enhancing objective. In both cases, our answers lead us to conclude that industrial policies were largely ineffecnive. Indusbial Polcy Only Marginaly Altered Induri Structure. The cross-econ- omy comparisons of industrial structure raise quesrions about the effi- cacy of government efforts to promote or discourage specific sectors. In Korea, for example, despite the government's extensive efforts to speed the private sector's shift from labor-intensive to capitml- and technology- intensive industries, the relatively labor-intensive tetiles and garments 312 USING RESOUCES sector was nearly three times bigger than intemational norms prediaed in 1988, a substantial increase relative to international norms from 1968. During the same period, Korea merely maintained the interna- tional norm in chemicals, a heaviLy promoted secror, while ocher heav- ily promoted sectors, basic metals and metal products and machinery, achieved only modest improvements. Similar surprises are evident in Singapore, which, like Korea, has a government that aggressively inter- Box 6.5 Indonesia's Turbulent Leap into High Technology * SECr1oRAL-. TARGETING TO ACHIEVE RAPID PRODUCTIN . change is a risly business-paTilarly when publicly funded firs are pemiedm to ily on protecred domestc mardratherdian bing subJece to international compeunon. The danges oof such a coursc * areevidentin the difficulties encountered by Idonesianai.ramfpro- - ducer PM. Industri Pes;wat Terbag Nusantaa(iv), which has absorbed $1 billion in goverment. fiud since its: esitablishient in 1979 but has yet to become internatonally compmtive or geniiney- .profitble Intheory, IPN is exectcd to use the domestic marker as a spring- board toiinternation2l sales. Te strategy is to shift-froiproducing' -.-aircraftunder licnse, tO co-design and co cton, tohsole designa -and production. In practice, vns.p cpdo meticxmruuei is tooe:- .small togenciate the economies ofscale necessayfor efflciehtaircraft. pro- 313 :AN MIRACLE vened through a variety of mechanisms to promote capital- and technology-intensive industries. Here, the importance of textiles relative to the predicted norm has increased even more sharply. The textile sec- tor went from double its predicted size in 1973 to eleven times in 1989. Over the same period, metal products and machinery declined from rwelve to fivc times its predicted size. Appendix 6.1 describes a simple rest of the impact of industrial policy on the basis of the factor proportions theory of comparativc advantage (Hecksher-Ohlin-Samuelson). We assume that given the relative labor abundance in all of the HPAEs early in their industrial- ization, the share of labor-intensive manufactures in toral manufac- turing should rise as they begin exporting. We compare statistically the pattern of growth of the industrial sector in each economy with its pattern of factor intensities. We argue that if industrial structure is derermined primarily by market forces, low wages and low capital in- tensity at the beginning of the period of active industrial policy will predict the pattern of industrial growth. If industrial policy to pro- mote capital- and knowledge-intensive sectors was important in de- termining the sectoral pattern of growth, however, this should not be the case. Indeed, the opposite should be true; high wages (a measure of human capital intensity) and high capital intensity at the end of the promotion period should predict changes in industrial structure. We have performed these tests on each of the Four Tigers plus Japan for periods relevant to their industrial promotion efforts, using two- digir isic classifications of industrial sectors (see table A6.2 in appendix 6.1). Our efforr to differen dare berween a comparative-advantage- based evolution of industrial structure versus one characterized by sig- nificant intervention is notable mainly for a number of negative results. In particular, in Korea-the economy for which significant interven- tion is best documented-during 1973480 the most rapid growdt in seac 'I shares of value added occurred in lower-wage or lower value- addeC_ per worker sectors. In Korea at the two-digit level, sectoral growth was broadly market confbrming in terms of traditional factor intensities. The results for Japan are incondusive. In Hong Kong there is no relationship between our explanatory variables and patterns of production. In Singapore, for the period 1980-89, output grew more rapidly in more capital- and knowledge-intensive sectors, supporting the view that the Singaporean authorities successfhlly intervened to en- courage increasingly capital-intensive development. But, given the 3"4 USING RESOURCES F rapid growth in the capital-labor ratio in Singapore, this result also con- forms to factor proportions theory predictions. These findings do not imply that governments were not attempt- ing to influence industrial structure. They undoubtedly were. But they suggest that, despite governmenr intentions, the manufacturing sector seems to have evolved roughly in accord with neoclassical ex- pectations; industrial growth was largely market conforming. Rates and Patterns of Product Change in Industry. There is both good news and bad news for advocates of industrial policy in the productivity performance of East Asian industry. The good news is that, on average, rates of productivity change in industry in Japan (before 1973), Korea, and Taiwan, China, which are the only economies for which we have derailed sectoral estimates of TFP growth, were high by international standards; productivity-based catching up was tak-ing place (Page 1991). The bad news is that, in general, productivity change has not been higher in promoted sectors. Japan may be an exception. Between 1960 and 1979 chemicals and the metalworking machinery complex have un- usually good TFP performance Jorgenson, Kuroda, and Nishimizu 1987). Japads industrial srructure differs from international norms in these sectors and exhibits quite high values of the share of value added in total manufacturing. These industries are those that observers usuaDy point to as having received significant government support, induding efforts to stimulate producrivity growti. A number of calculations of Tr have been carried out for Korea for a variety of periods (Dollar and Sokoloff 1990; Lim 1991). From these studies a number of patterns can be identified that are broadly consis- tent with one another. Most striking are the high values ofTFrP change in most sectors by intemrational stndards (Nishimizu and Page 1991). Al- though the Korean government selectively promoted chemicals and iron and steel (included in basic metals), the large growth in the share of iron and steel was accompanied by quite low TFP performance between 1966 and 1985; textiles and clothing conversely, had very high rates of TFp growth. The promored chemical sector, whose relative size was decreas- ing, was characterized by considerably higher-than-average TFP growth during this same period. The government in Taiwan, China, did not attemnpt to influence sec- toral evolution as strongly as the government of Korea. Nevertheless, there was more than a small efforr devoted to encouraging specific sec- tors, particularly those viewed as either capital- or technology-intensive. 315 1STN MI R A CL E The three sectors that exhibited the greatest expansion in the share of value added-apparel, electrical equipment, and metal producs-are all characterized by below-average wages, capital-labor rarios, and labor product. Moreover, there is no statistical relationship between wage or capital inrensity and productivity change at the sectoral level. In fact, the highest sectoral rates ofTFP change are recorded in textiles and apparel. Recent sectoral TFP growth rares for industry in Malaysia from 1973 to 1989 showa similar pattern ofgreatvariabilicyin cheTr growdt rates in promored sectors, from high in nonelectrical machinery and paper and paper producrs to low in iron and sreel and transport equipment. Textiles and apparel, a nonpromoted sector, are among the TFP leaders (Maisom 1992). In short, there is no apparent relationship between the sectors promoted under the HICOM drive and rapid productivity growth. Overall, the evidence that industrial policy systematically promoted secrors with high productivity change is weak InJapan there is some sup- port for the assertion that TFP growth was higher in selected seccors, while in Korea and Taiwan, China, activities that were not promoted (for cx- ample, textilcs) had TFP performance as impressive as those that were. Moreover, arrempts to determine whedter high rates of TFp growth com- bined with rapid growth of promoted sectors can plausibly explain the very high overall rates of TFP change in manufacuing yield mosdy neg- ative results (see appendi 6.1). The main reasons for manufactrings success in Japan, Korea, and Taiwan, China, lay in the high general rates ofT"P growth, including those in labor-intensive, nonpromoted sectors How Manufactured Exports Increased Productivy OES THIS MEAN THAT PUBUIC POLICY HAD NO ROLE IN THE rapid rates of TFP change found for many of the HPAEs? We be- lieve not. In chapter I we described how rapid UP change in developing economies can be a result of the ability to move quicldy doser to international best practices. We termed this "productivity- based catching up." We believe that rapid growth of cxportsi a result of the export-push policies of the HwAEs, combined with the superior per- formance of rhese economies in creatring and allocating human capital, provided the means bv which they attained high rates of productvity- based catching up and TFP growth. 3I6 USING RESOURCESC How did the unusually high rates of growth of exports and human capital contribute to the productivity performance of the East Asian econonies? Most explanations of the link between TFP growth and ex- ports emphasize such static factors as economies of scale and capacity utilization. While these may account for an initial surge of productivity soon after the start of an export push, they are insufficient to explain continuing high TFr growth rates. Rather, the relationship between ex- ports and productivity growth may arise from exports' role in helping economies adopt and master international best-practice technologies. High levels of labor force cognitive skills permit better firm-level adop- tion, adaptation, and mastery of technology. Thus, exports and human capital interact to provide a particularly rapid phase of productivity- based catching up. Of course, it is possible that the move to a higher production finc- tion occurred before the growth in exports-that TFr growth caused ex- port growth rather than the reverse.41 But even if expors began on the basis of productiviry change due to such domestic efforts as plant reor- ganization, the cumulative magnitude of productivity growth over nany years is most unlikely to have been a result of purely domestic ef- forts.42 It stretches credibility to suggest that the large cumulative effcts of TFP growth in Japan, Korea, and Taiwan, China, could have been achieved by the plant floor innovations proposed as important sources of productivity growth at lower TFP growth rates (Pack and Page 1993). Were that the case, it would be difficult to explain why these gains far outstrip productivity increases in the industrial economies during their own rapid-growth periods, when TFP growth was attributable largely to domestic factors. Clearly, then, an increased ability to tap world tech- nology has been an important benefit of exports.43 Exports' Role in Impedect Knowledge Markets Why do exports facilitate the move toward internationaI b&, :-practice technologies? The knowledge that pernits this shift is available only in quire imperfect markets. Often the markets do not exist-some knowl- edge is simply not sold, because its owners fear that licensing or direcc foreign investment will eventually Ieak the knowledge to future com- petitors. Even where markets exist, they are likely to be characterized by bilateral monopoly, so that a variety of difficulties confiront finns trying to purhase such knowledge (Arrow 1969). 317 4MIfA CL E In markecs with imperfect knowledge, mechanisms that help an econ- omy or firm obtain technology can confer an cxternality-a pecuniary one if knowledge is obtained at a lower cost, a real extemality if the knowledge turasfer is more efficient. For example, the transfer may be more beneficial if knowledge is obtained as a result of the self-interest of purchasers of ex- ports in the OECD economies, rather dtan through arm's-length purchase. Exporting helps to overcome some imperfections in the market for knowledge and permits its acquisition through a variety of mechanisms. These include: m Purchase of new eqiupment. The purchase of new equipment is a straightforward method of obtaining new technology insofar as it is em- bedded in equipment The two main impacts of exports in this case are in providing the competitive pressure for firms to make such purchases (if the equipment is used to produce exports), and, more generally, in earning the foreign exchange to finance the purchase of machiner)y a Direct#iriezt investment The firms that generate much ofthe world's new technology are reluctant to part with it. They perceive that the best use of tieir new knowledge with respect to developing economnies often is through exporrs of producrs to them. In East Asia, however, most of the incoming DFI has been intended for production of exports rather than as a strategy for domestic sales. An economy's recent export performance fre- quendy signals whether it is a desirable location for export-oriented DPI. Economies with rapid export growth are preferred in part because such growth often reflects good macroeconomic management Moreover, econornies with significant export growth have generally provided more infrastrucure useful to exportes than internally oriented economies44 DFI makes crucial production and marketing knowledge available to de- veloping economies. It permits them to begin manufacmring along the world's best-practice producton fiLnction by substituting foreign physical and human capital for absent local factorsfi5 As local labor learns from the presence of best-practice knowledge and equipment, knowledge rends to become diffiLsed through labor mobility and informal contacts among managers. Even when the investing firn attempts to hinder such diffu- sion, signicant knowledge ranfers are inevitable.6 * Technology licensing. licensing exsting technologies, both rights to proprietary equipment and derails about production pocesses, offers de- veloping economies substantial opportunities for improving dteir levels of best practice. During the 1950s and 1960s, Japan benefired considerably fiom licensing (see, for example, Nagaoka 1989). The net gains were 3I8 USING RESO URCC large, as the licenses were obtained at relatively low cost compared with the domestic sarch expenses avoided. However, there is some evidence, and a growing subjective sense, that arz's-length licensing is decreasing as an option for dosing technology gaps. Technology developers, the licen- sors in the OECD economies, have become wary of helping potential com- petitors, even if contracts predude exports to ocher economies for the duration of the license. Particularly in R&D-intensive sectors such as chemicals, machinery, and electronics, firms are increasingly unwilling to license technology-, they believe royalties provide an inadequate return for actions that may impair their own long-term competiuveness. This creates an irnperfect or nonexistent market for critical technology transfers. Exporting economies have an advantage in coping with this situa- tion. Licensing firms prefer cross-licensing agreements in which they obtain access to the licensee's own technology or to its manufacturing skills. Given asymmetries in knowledg. the best test of the potential parner's abilities is its performance in export markets. Moreover, the fact that a firn has exports and the requisite technical and commercial skills to produce them improves the bargaining position of the licensee Sales of licenses contain a significant element of rent (Pack 1993c), so the cost of licenses will be lower where the seller perceives the strengths of the purchaser as signaled by exports. Finally, where a mark-er does exist for technology licenses, the foreign exchange to pay for fees is more ikely to be available in economies experiencing rapid growth in exports. * Trasfir of nonpmprietary technology. The free or inecpensive trans- fer of nonproprietry knowledge is easier if an economy begins its in- dustrialization effort in labor-intensive sectors using relatively old machine designs and production technology. When exports are based on comparative advantage, precisely these conditions are met. Hong Kong, Korea, Singapore, and Taiwan, China, all began rapid industrial- ization in labor-intensive and low-technology manufacturing. The equipment was readily purchased, and the production knowledge was availale at low cost from engineering publications, trade literature, and independent consultants.47 The simpler equipment was conducive to liocal productivity-enhancing improvements, often on the part of blue collar workers (Ranis 1973). Moreover, in labor-intensive industries ic was easier to acquire and absorb information from customers withour a large stable of educated engineers. * Infibmwonfrom custorme. In the presence of imperfectly traded information, knowledge provided by purchasers of an economjs cx- 319 -ALRACLE ports can be quite important. This has been shown by Wesrphal, Rhee, and Pursell (1985) in Korea and independendy corroborated in Taiwan, China, by a number of researchers (see, for examnple, Gee 1989). Buyers want low-cost, better-quality products from major suppliers. To obtain this, they transmit tacit and occasionally proprietary knowledge from their other, often OECD-economy, suppliers. Such knoNvledge transfers are more common in simple production sectors such as dothing and fbotwear and with older technologies that are either in the public do- main or are not closely guarded. a Knowledgefrom returning naiondls. Drawn partly by the high wages made possible by exports, many residents of Korea and Taiwan, China, trained abroad, particularly in new sectors such as electronics and com- puting, have returned home to work. Many retuming nationals have received education in OECD economies and then workled for OECD- economy firms.48 Their return has provided significant transfer of besr- practice mnethods. For example, foreign-educated nationals account for al/the postgraduates employed in the industry of Taiwan, China (Pack 1993a). This source of knowledge of international best practices be- comes more important as changing factor prices dictare a shift to more capital- and technology-intensive sectors in which higher-level skills are needed to unlock knowledge that may be embodied in patents, licenses, or the use of specialized nontraded equipment. * Domestic rseanrck. In both Korea and Taiwan, China, a considerable proportion of R&D has been devoted to improving exports and reduc- ing production cosrs. There is no evidence that the gap between R&D and commercial needs frequently seen in the import-substituting econo- mies has been a problem. How Exports and Human Capital Increase Productivity Access to intemational best-practice technology and rapid fiomnarion of human capiml supplement and reinforce one anothec It is doubfl that tie HPAEs couldchave made as prductv use offoreign knowledge and imported capital without highly silled domestic engieers and workers. Convey, without foreign knowledge it is vey untlikely that total fiLctor prductivity gowth would have been as lamrt Intensve efForts by ghly skilled managers and technicians in individual plants in inward-oienred Lain American economies to improve the productiviLy of eitig capital smckwith internal innovations did not generate high- productivity growIY49 3Z0 USING RESOURCES :EFI:d3r The HPAEs' rapid export growth has often generated positive interac- tion between human capital, physical capital, and knowledge. The ex- ternalities generated by manufactured exports in the high-performing Asian economies in the form of cheaper and more effective knowledge transfers would have undoubtedly been less productive had there been fewer skilled workers to facilitate their absorption, while the HPAEs' rapid increase in education levels reflected in part rising private rowards for greater education made possible by exports and export income.50 Thus labor force skills, flexible markets for labor, low domestic disror- tions, and export incentives all interacted to promote high rates of tech- nological upgrading and productivity change. We can use the cross-economy regression framework developed in chapter 1 to look at the impact of trade and indtLstrial policies on growth (see table 6.17). We introduce two variables into the basic cross- economy regression: the index of openness (as described above, the de- gree to which domestic prices conform to international prices) and measures of export performance.51 Openness to the world economy captures not only effidencr gains induced by the need to remain glob- ally competitive but also the likely allocative benefits of having relative prices dose to international prices. Introducing these two trade-related variables substantially increases the explanatory power of the cross-economy regression. The effect of openness is positive. Economies with limited relative price distortions grew more rapidly.52 We choose as our measure of export performance two indicators-the average share of manufactured exports in total ex- ports and the share of manufacrured exports in GDP.53 Manufactured export performance is strongly correlated (at the 1 percent level) with high rates of per capita income growth. When the share of manufac- tured exports in totl exports is introduced together with the openness index, only the latter is significant. Conversely, when the share of man- ufactured exports in GDP is used, the openness index remains positive and significant. One possible interpretation of these results is that a high concentration of manufactured exporcs relative to total exports, rather than openness, contributes relatively more to productity change in a cross-economy framework, which would be consistent with our reason- ing concerning export externalities.54 Estimates of the sources of total factor productvity growd (de- scribed in appendix 6.2) confirm the results of the cross-economy re- gressions and add some insights into possible interactons between exports 321 - MIRAC LE Table 6.17 Output, Growth, and Investment (dependent variable: rate of growth of real GDP per capita, 1960-85) NrinnJerefobneations: 92 86 69 97 79 Intercept -0.4237 -0.0124 -0.2324 -0.0055 -0.3943 - (0.1650) (0.0083) (0.1550) (0.0085) (0.1748) GDP relative to U.S., 196O -0.1033 -0.0459 -0.0837* -0.0381 " -0.0892 (0.0268) (0.0108) (0.0256) (0.0123) (0.0280) Primary enrollment, 1960 0.0259 0.0210** 0.0216' 0.0221 0.023!W (0.0081) (0.0064) (0.0079) (0.0069) (0.0085) Secondary enrollmcnt, 1960 0.0317 0.0211 0.0218 0.0206 0.0156 (0.0203) (0.0134) (0.0210) (0.0148) (0.0210) Growth of population, 1960-85 0.1322 02693 0.2054 0.1856 0.1222 (0.2699) (0.2346) (02806) (02376) (0.2870) Average invesrmenrlGDr. 1960-85 0.0568 0.0659 0.0625 * 0.0444 0.0436 (0.0259) (0.0232) (0.0291) (0.0242) (0.0276) Dollar openness index 0.0042 * 0.0023 0.0040 (0.0017) (0.0016) (0.0018) Average manufactured expotslrotal 0.0002 0.0003 exports, 1960-85 (0.0001) (0.0001) Averac manufactured exportslGDP. 0.0011 0.001a oi 1965-85 (0-0003) (0.0004) Adjusted R2 0.3947 0.4912 - 0.5217. 0.3530 03938 RMSE 0.0159 0.0124 0.0132 0.0147 0.0155 Statisicailysignifieanrax the 0.01 leveL - Smotisdally significantar the 0.05 level Notw Coefficient is top number. Standard error is bottom number in parnthees. Suree. World Bank staff estimtnes- and human capital. We attempt to explain variations across economies in TFP growth rates in terms of relative income, educational attainment (as measured by the average stock of education per person), openness, and our measures of manuFactured export perforrnance (see table 6.18). The education stock variable, while positive, does not explain variatons in TiFP growth among economies. This is appropriate, since we have measured TFP growth net of human capitals contribution. Openness is consistenTly 322. USING RESOUIRtCE £ PT associated with superior TFP performance, controlling for other variables. Both indicators of export performance are also consistently and positively correlated with higlher rates of T PP growth. We also find some evidence of.- positive interaction between the share of manufactured exports in total exports and in national income and the stock of education. The coefficient of the interaction term between these two variables is positive but not significant at conventional levels, and dte export share variable becomes insignificant. When we consider the con- tribution of the variables taken together to explaining the variation in TIP growth rates, however, it is statistically positive. We conclude chat export Table 6.18 Determinants of Total Factor Productivity Growth, 1960489 (dependent variable: rate ofgrowtlh of real GDP per capita, 1960-89) Numdrmrofobservatons: 67 67 51 51 Intercept -64.9123 -71.1186 -72.0692*u -70.8604 (14.0585) (14.8657) (15.6104) (15.4608) *GDP relatve ro US., 1960 -4.8047 -5.5757 ** -2.3509 -2.1562 (1.9771) (2.0637) (2.3225) (2.3008) Educational attainmcnt, 1960 0.1471 0.0680 0.1574 0.0738 (0.0874) (0.1082) (0.1064) (0.1207) Dollar openness index O.6493 * 0.7154 IA 0.7225 0.7134 (0. 1417). (0.1508) (0.1574) (0.1558) Average manufactured cxports/mtoal exports, 1960-85 0.0314 - 0.0159 (0.0066) (0.0142) Interaction ten: Educarional attainment 1960 times 0.0032 manufactured exports/total exports, 1960-85 - (0.0026) Average manufaccured expors/G DP, 1965-85 . . J.0686 (0.0966) - .. . . m: 0.02 0.0284...................., Interaction term: Educational attainment 1960 times 0.0625 . 0.0284 manufactured cexpors(GDP, -1965-85 (0.0269) (0.0201) Adjursed R2 0.6333 0.6376 0.4507 0.4628 ** Smzisically significant at the 0.01 level. SmziricaHllyiignificantat the0.05 leveL. ; XNoeC - C dicmt is top number Standard error is bottom numbernpathes.- SoureWorld Bakstaffesdma.--- 323 - r&IAN"MIRACLE performance and education interact positivelr; higher levels of education raise the contribution ot manufacrured export concentration to TFP growth.55 This is consistenc with our hypotheses that manufactured ex- port orientation and high labor force skills interact to faclitate the ac- quisition and mastery of technology with attendant spillovers. The evidence from our cross-economy estimates is supported by a number of recent microeconomic studies that attempt to test the link be- tween exports and productivity growth. Pack and Page (1993) present ev- idence from Korea and Taiwan, China, that ar the sectoral level rapid export growth is correlated with the pattern of productivity change; ax- porting sectors have higher seoral rates of rF growth. Wei (1993) uses city data from China and finds a statistically significant relationship be- tween exporrgrowth and productivitygrowdi. Perhaps most compelling, Aw and Hwang (1993), using firm microeconomic data from Taiwan, China, find a statistica[ly significant relationship between productivity differences among manufacturing firms and export orientation. How Have Trade and Industral Policies Contibuted to Growth? The early admirers of HPAE trade policy dearly overstated the neu- trality of incentives between domestic and foreign sales and understated the variation across sectors. But the revisionists who see in the HPAEs' success evidence that highly targeted industrial policy worked overlook the fac that the calculations on which borh relied wer- static, that firms based their decisions not only on current levels of protection and factor prices but on the certain knowledge that thy would need to compete in the fuiture. Effective rates of protection may shape short-term tactics but the long-term strategies of currendy protected firms are more likely ro reflect their expectations of the future. Thus, the emphasis on exports was decisive for the entire manufac- turing sector; promotion of individual sectors was less important, since all bur the very dull knew that their turn to export would come sooner than they would have preferred. Given the widespread national under- standing established by government statements and actons in Japan, Korea, andt aiwan, China, that exportingwas thestandard by which all economic actity would be judged, even firms benefiting from higher- than-average rares of protection in the domestic markcet understood that in the near future they would be forced to compete in world markets. Sustained reduction in import protecton sent a similar message to pro- 3274 USING RESOURCES ducers in the Southeast Asian newly industrializing economies. Govern- ments were credibly committed to export competition. Exports were important because they ensured that, given the HPAEs' high human cap- ital base, productivity growth would be facilitated by the improved abil- ity to tap international knowledge. *.. What are the m2in faors that contributed to the HPAEs' superior al- location of physical and human capital to high-yielding investments and their ability to catch up technologically? Mainly, the answers lie in fim- damentally sound, market-oriented policies. Labor markets were al- lowed to work Financial markets, although subject to more selective interventions to allocate credit, generally had low distortions and lim- ited subsidies compared with other developing economies. Import sub- stitudiorn, although an early objective of public policy in all the HPAS except Hong Kong, was quickly accompanied by the promotion of ex- ports and duty-free admission of imports for exporters. The result was limited differences between international relative prices and domestic relative prices in the HPAEs. Market forces and competitive pressures guided resources into activities that were consistent with comparative advantage and, in the case of labor-intensive exports, laid the founda- don for learning international best practice and subsequent industrial upgrading. Does this mean that selective interventions played no role in East Asia's superior growth? Our conclusion is that selective interventions were neither as imporant as their advocates suggest nor as irrelevant as their critics contend. All eight economies succeeded in establishing au- tomatic access to credit fior exporters; this simple contest was an impor- tant component of their export-push strategies. Other directed-credit programs were less successful Where rhey appear to have resulted in al- location of credit to high-yielding activiries, mainly in Japan, Korea, and Taiwan, China, it was because of careful szieening and monitoring of projects and an orientation toward the private sector. Where they failed, governments limited the damage by offering more limited subsidies than the credit programs in other developing economies. Finally, we have conduded thac industrial policy, with the possible exception of Japan, did not alter industrial structure or patterns of productivity change. The most successful selective intervention in the HPAEs-the commitment to manufactured exports-was also the most general. 3Z5 x&EXkMIRACLE Flexible markets, low price distortions, high levels of human capital, and industrial and trade policies that tilted incentives in favor of exports- all combined to allow rapid growti of manufactured exports and, through it, rapid technological catching up. Appendix 6.1: Testing the Impact of Industrial Policy on Productivity Change G OVERNMENT POLICIES THAT CORRECT MARKET FAILURES should give rise to growing total facror productivity in the en- tire industrial sector, though the resulting pattern ofrFrp growth across subsectors is not easily predictable. Knowledge-based imperfec- tions are likely to yield TFP gains, primarily in the subsecaors in which the intervention occurs. Subsidization of knowledge acquisition in tex- tile production will mainly benefit other textile firmns as knowledge seeps out to them or trainecd engineers go to other textile firms. Some knowl- edge in textiles may benefit dosely related sectors such as dothing through production of ftbrics that are better suited to local conditions, that is, are not obtainable on the world marker. T-he secroral locus of gains is important for several reasons. First, em- pirical efforts to relate government interventon to productivity growth are desirable as aweak rest ofthe efficts of intervention. Productivitygains in rhe entire industrial sector or the entire economy attributable to other factors might mistakenly be ascribed to the intervention, even if it had ac- tually reduced productivity growth. Undoubtedly, the productivity gains ftom correcting various market Milures may diffse to many sectors, bur most should have a major impact in the initial sector of intervention. Empirically, many of the intersectoml interactions will be measured as intrasectoral at the two-digit level. Improved production engineering in spinning may improve TFP in weaving. Both, however, are induded in the Ewo-digit isic textiles branch. Equal diffusion to all sectors seems implau- sible. While there are individual cxdmples in economic history (Rosenberg 1976), there is no evidence that this is the general prototype. Moreover, recent research on the pattem of spillovers of R&D in industrial economies demonstrates that the major beneficiaries are dosely related sectors, often sectors that would be identical within a two-digit classifica- tion Uaffe 1986). The nonsubstantiated claim that the spread of external- 326 USING RESOURRCES ities is both large and diffiuse provides an all too easy refuge for those who argue for the benefits of selective intervention and then offer as evidence effects that are plausibly attributable to the efforts of market agencs. How We Compared the Relative Importance of Industrial Subsectors To compare the tansformation of the sectoral structure of manufac- turing in Asia with the transformation projected from international norms, we utilized the following equation: log( VA/GDP) = b0 + b1log(GDrPIOP)- + b2(PoP) where VA, is value added originating in sector i in economy j, GDP is gross domestic product, and vcl; is total population (Chenery 1960). The equation attempts to capnure dte influence of demand elasticities and dte evolution of supply. POP factors in scale economies, although in a world of free trade this may not matter.56 The equation is purely de- scriptive and has no normnative content Indeed, a given pattern might indicate that all economies follow some initial leader and undertake policies to emulate its sectoral evolution. Despite its flaws, the equation provides benchmarks to determine whether the relative importance of industrial subsectors in the HPAEs differed significandy from the weight- ings observed in other economies. For purposes of comparison, we utilized equations estimated by Syrquin and Chenery (1989) to derive the predicted sectoral shares of value added in GDP. Table 6.15 presents the results for eight branches of manufacmring, showing the ratio of actual, VI, to predicted, vp, shares of gross domestic product for various sectors. Thus a value of 1 repre- sents confbrmance t the international norm, values less than 1 indicate that the sector is smaller than predicted, and values greater than I indi- cate that the sector is larger than predicted. In general, the results conform to our expectations of these econo- mies. Three economies known as manufacturing powerhouses have larger overall manufacturing sectors than international norms based on the international norms for economies with similar incomes would pre- dict Hong Kong (1.26), Korea (1.26), and Singapore (1.38). But there are also several surprises; these are the focus of our narrative. Given the rough nature of the cross-economy analysis, we have not attempted to test the statistical significance of deviations fom the norm. 327 43 -yASrAN MIRACLE Efects of Industrial Poricy on Aggregate Productivity Growth Aggregate TFR in any period can be decomposed by weighting each secror's level of total fitcror productivity, A,0 by the sector's share in value added, v.r The growth of TFP will then depend on changes in A,, and changes in v2.. Algebraically, this relation can be written as: (6.1) ALogA = `,v,tlogA4,- via-1 logAUr.l Equation 6.1 gives the growth in A due to the increase in productivity of existing sectors, logAi.> log Ai r- or the growti in th.e value-added share of these sectors, vi,> z-, 1 whose productivity is growing.57 As will be seen belowv in table A6. 1, small differences in die rate of growth of sectoral TFP, A*i,, lead to large cumulative differences in Ai, As the text of chapter 6 points out that on the basis of international comparisons two sectors were generally overrepresented in rhe industrial sectors of the HPAEs: metal products and machinery, and textiles and ap parel. Only the former sector was promorecd in HPAEs that used indus- trial policy. We define two types of sectors, those fiavored by government policy, f and those subject to benign neglect, h. The implicit model of those who believe that industrial poLicy has had an important effect is that changes in the deployment of reso' rces were nor based on conven- tional competitive advantage but that the government created competi- tiveness in new sectors. Sectoral growth was stimulated in capital- or technology-intensive sectors rather than unskilled labor-intensive sec- tors that were more "natural" fur economies such as Korea or Taiwan, China, in the 1960s. Moreover, it is implicitly assumed but not empha- Table A6.1 Effect of Sectoral Composition on Manufacturing-Wide Growth of TFP TFPgrowh, :TFPSrow*,, Economy actua vae-ad weighn -adIjal wegkh . Korea, Rep. oE 1966-85 6.7 6.1 Japan, 1960-79 2.3 1.9 NeArWeightod by value-added shae datwould have prvaied ifthe meal products and macduneysectorhad conformneddt redictp d by txcquatuion estimated by Chenery and Syrquin (1989).- SurarzPack (1993b). Based on Kuroda.Jorgenson, and Nishimizu (1985) fbr TFP- esimares for Japan. 328 USING RESOURCESOEUF$ sized that rFP growth was greater in the selectively promoted sectors as a result of static and dynamic scale economies.58 If industrial policy was important in determining the rate of indus- trial productivity growih in the HPAES, it should be the case that a. Sectors were begun or grew more rapidly in terms of factor com- mitment than would have occurred without intervention. In terms of edquation 6.1, Vf1increases more than woulc have been "normal" from its initial level, vf Thus, ir is necessary to demonstrate that the sectors to which resources flowed were not those that would have grown on a simple comparative advantage basis. If, however, the sectors that grew1, 1 > 'M' are characterized by low wages, or low capita intensity as measured by capital-labor ratios, this is con- sistent with growth according to comparative advantage. b. Favored sectors had higher levels of TFp, Af> A,,,l. Even if (a) can be demonstrated, it is necessary to show that the government cor- rectly forecast fitture growth of TFP in sectors or that it created dif- ferential rates of TR) growth resulting in * > A*t c. There was a quantitatively important introduction of new, mod- ern sectors, v,, which exhibited higher levels of TFP than existing ones. In terms of (a), the sectoral structure of output is radically changed by the growth of modern sectors that are introduced at high levels of TFP, and this raises average manufacturing-wide productivity. Because the HPAEs were such successfhl exporters, predicting the impaca of selectve interventions on the sectoral pattem of production depends to a considerable extent on one's preferred theory of international trade- Ricardian, Hecksher-Ohlin-Samuelson, neotechnology, and so on. To demonstrate the impact of selectve intervention, it is necessary to posit a counterfictual model. This is obviously exceptionally difficult, and to be rigorous requires a general equilibrium model that in turn has difficulties of its own, such as con-crdy specifying sectorl production fimctons and learning patterns. Again, the "natural" evolution of an economy also de- pends on one's preferred theory of international trade- Nevertheless, a weaker test can be applied. If the evolution of the secaorat stucmre of in- dustry in the HPAEs can be attributed to comparative advantage, it is diffi- cult to support the view ex post of a strong impact of industrial policy. Implicit here is that the "naturatl evolution among industrial sectors can be predicted by the factor endowments theory of comparative ad- 329 ;M-IRIARACLE vantage, or Hecksher-Ohlin-Samuelson (Hos). Although large numbers of addirional explanations of the patterns of crade have been put forward in the last two decades, many of them are actempts ro explain intrasector trade and specialization arnong products that broadly require the same Factor proportions. HOS theory yields predictions about the factor con- tent of exports and imports, not directly about Ehe structure of rotal pro- duction. However, given the extraordinary growth of exports in the Four Tigers and rheir size in the GDP, the changing strucaure of production in these economies was decisively affcted by trade patterns. This is less true of the larger economies, Indonesia, Japan, Malaysia, and Thailand. Predicting the Pattern of Sectoral Growth The newer literature setting out the merits ofintervention has not rested whether the sectoral evolution of successfil Asian economies has differed fiom what would have occurred had there been neutral policies. Neverthe- less, showing that many instrumenms were selectively employed is not the same as proving dat they had a significant quantitative impact on the sec- toral structure of production. Even 'where effective rates of protection dif- fer among sectors, the impact on sectoral structure is not easily predictablc. In general equilibrium, the movement of resources has no monotonic rela- tion with ERPs (Black and Taylor 1974; Bhagwati and Srinivasan 1979). For each economy, four simple regressions are run of the form: (6.2) ei= f(x,) where ei is the change in the current price share of value added in sec- ror i, relative to value added in all man uficturing. The independenr variables, xP, are the wage per worker at the beginning, zv5, or end of dte period, we, or value added per worker at the beginning or end of the pe- riod, vZab or va. La competitive labor markers, the wage per worker should measure the skill intensity of workers in each sector. Value added per worker should reflect both skill and capiral intensity and is a proxy for technological complexity. Although value added per worker and the wage rate are likely to be correlared, both are used, as there is a possibil- ity of divergence. Governments also may have employed one or the other as indicators in targeting sectors. Wage per worker may be trans- lated as "good jobs,' while value added per worker is a natural measure of "high productivity." 330 USING RESOURCES 'E!jl Consider the movement from inward orientation to more neutral in- centives. The static Hecksher-Ohlin model predicts that for low-income economies sectors that exhibit low physical and human capital-labor ra- tios by international standards are the natural candidates for growth. Hence, the prediction is that dvu/zdwb and dvl.Idvab c 0. While many newer theories of trade, including neotechnology and those emphasiz- ing increasing returns subordinate differenccs in factor endowments. they are more likely to describe intra-industry patterns of tradc among industrial economies than the evolution of the sectoral structure of pro- duction of developing economies whose production structure is likely to reflect differences in factor proportions. If intervention has been of great importance in promoting the expan- sion of capital- or knowledge-intensive sectors and, eventually, affecting trade patterns, the data should show dvjldwb> 0 and/or that do1.Jdva6> 0. An alternate test of the impact ofselective promotion is to testwhether the sectors that expanded exhibited high capital intensity or wages at the end of the period. A result fvoring the interventionist view would be ehi-iIdw or dzIlIdva > 0; the sectors that grew most exhibited a high. capital or high-wage intensity at the end of the period. We have performed these tests on each of the Four Tigers plus Japan, wsing two-digit lSIC classifications of industrial sectors. Table A6.2 shows the sign of the regression coefficient and its significance for each period for each economy. The years chosen reflect a distillation of rhe existing literature for each economy on significant intervals during which intervention was practiced. It is not possible in all cases to utilize the ideal interval. In Japan, for example, data from before 1953 vary considerably in definition from those after 1953. These are clearly very simple tests of the determinants of structural changes in production but capturc the spirit of the opposing views. The statistical results must be employed carefully. Even if it is slhown that there was a significanr negative correlation between 9i and w1, support- ing the HOS view, some of the individual sectors that grew most may still have been high-wage sectors. Thus, we will examine some individual sectors as well as report the statistical results. Korea. For most periods, the signs of the wage and value-added vari- ables in Korea are negative. For the period 1973-80, the one character- ized by promotion of the heavy and chemical industries, the negative coefficients of w and va for the beginning of the period are significant. The same pattern occurs in 1980-88 as well as the entire 1973-88 33' *;EA?rM1RACLE Table A6.2 Signs of Regressions Explaining Change in Value-Added Share of Sectors Economy Period Wu WE VB VE IHeng Kong 1973-80 - + - + 1980-88 + + + + 1973-88 - + - + Japan 1953-63 1963-73 + 4- + + 1973-80 - + - + 1980-89 + + + + Korea, Rep. of 1968-73 - + 1973-80 - - _ : 1980-S8 - - 1973-88 - --- - Singiporre 1969-73 - - - + 1980-89 + + + -; 1973-89 + -+- + + Taiwan, China 1966-86 - - - + and - signs re sign of cDdenciLent - Signficncatthe0.10IeveL ' Significnurar tie 0.05 klveL Significant at tie 0.01 leveL NbOwtWSmZ a Wage per emPIoyeeac the b%eing (end) of he period. VB = Value added per workr at the bginning (end) of the period. Souxer Pack (1993 bl) period. This implies that at the two-digit level the relative size of more labor-intensive seccors increased during each period On the face of it this result is surprising. Quite apart from policies of selective promotion, this was a period in which both labor skills and the capital stock were grow- ing. One would have expected, according to the Rybcyznski theorem, that the labor-intensive sectors would have declined in importance. Part of the explanation for the relative growth of labor-intensive sec- tors despite the promotion of ostensibly capital-intensive suzbsectors is that the major branches that grew most rapidly in these years, and in- deed over the entire period 1968-88, were machinery, electrical ma- chinery, and metals-all sectors exhibiting somewhat less-than-average labor productivity at both the beginning and end of the period- Con- versely, some of the declining sectors induding iron and steel, industrial 332 USING RESOURCES:E'FPtC , chemicals, and paper and paper producrs, were sectors with greater- than-average u and va1, A more complete explanation of the contnu- ing growth of labor-intensive sectors is that they experienced more rapid growth in TFr, and the cost-reducing effect of productivity change stim- ulated expansion of output by more than the output growth induced in capital-intensive sectors by the decreasing relative cost of capital and skills due to accumulation and subsidies. In sum, the regressions for Korea indicate that: (a) the more labor- intensive sectors maintained their relative position or grew, (b) many of the promoted sectors themselves were not capital intensive at the tWo- digit level; and (c) subsidies to the cost of capital may nor have been suf- ficient to overcome the ongoing differentials in TFP growth among sectors. Moving away from the anecdotes about individual chaebol, the quantitative importance of government intervention to alter the struc- ture of production is not confinned at the sectoral level. Japan. The regression results shown for Japan in table A6.2 have no significant coefficients. At least a simple version of HOS does not work, probably nor surprisingly as Japan by the earliest year considerea in the table, 1953, was sufficiendy advax±ced so that intra-industry rather than interindustry trade would have become an important determinant of the sectoral producton structure. Scale economies or a neotechnology explanation of production and trade patterns are more useful at this stage of development. Going back to 1945, which the data do not per- mit, would not change this picture much insofir as this earlier period was one of reconstuction, in some cases of firly advanced sectors that had brken quite important before and during World War IL A simple HOS story would be interesting to test for the period from the Meiji restoration onwards. Once an economy is in a world of technology- and scale-economy-based development, the requisite conditions exist to allow almost any assertion to be made about the determinants of the sec- toral evolution of production. Hong Kong and Singapore. The cwo island economies reveal diffcrent patterns of development. In Hong Kong there is no relationship be- tween our explanatory variables and patterns of production. In this sense it is similar to Japan but derived fiom different causes. The slow growth of capital stock per worker in Hong Kon& partly reflecting large immigration, miliated in favor of a slower change in industrial struc- ture. There is simply very litde variance to be explned in the value- added shares in Hong Kong. 333 RAC L E In Singapore for the period 1980-89, outpur grew more rapidly in those sectors that at the beginning and end of the period exhibited grearer value added per worker, supporting the view that the Singa.- porean authorities intervened to encourage increasingly capital-intensive development. Given the rapid growth in the capital-labor ratio in Singa- pore, this result also conforms to the Rybcynski prcdictions. Taiwan, China. Like the results for Korea, the signs of the wage and value-added variables for Taiwan, China, are negative, although insig- nificant. Again, this implies that the relative size of more labor-intensive sectors increased during each period, albeit weakly, at the two-digit level. Growvth was market conforming; government intervention did litde to alter the structure of production at the sectoral level. The Evoluton of Sectors: Market Conformning or Not? The preceding eflorr to cdifferentiate besween a comparative-advantage- based evolution of industrial structure versus one resulting fiom inter- vention is notable mainly for a number of negative results. In particular for Korea, during, 1973-80, the years of greatest intervention, the most rapid growth in sectoral shares of value-added occurred in lower wage or lower value-added per worker sectors. Despite government intentions, the manufacturing sector may have evolved roughly in accord with neo- classical expectations. It is also possible that the government undertook measures such as the provision of social overhead capital and technical services such as the Korean Institute for Science and Technology, which facilitated the 'neodassical' tansformation. Singapore, in contrast to Korea, confirms the casual empiricism of many observers, as well as nurnerous accounts of rhe goals of govern- menr policy that industrial growth emphasized capital- and knowledge- intensive sectors. The sectors that grew most exhibited higher wages and labor productivity than the laggards. There is no evidence of a factor- proportions-based expansion. In contrast, the evolution of sectors in Japan and Hong Kong does not permit a simple view of its determi- nants. This is nor surprising for Japan given that HOS is designed to pre- dict the determinants of trade patterns rather than production, and trade constitutes a relatvely small share ofJapanese GDP. In tie postwar period, certainly after 1953 when our data begin, Japanese trade par- terns are more likely to have been dominated by technological develop- ment and scale economies. The failure of Hong Kong to conform to 334 USING RESOURCfSts$r. neoclassical comparative advantage is at first glance surprising, given hie enormous importance of its trade. There was, however, relatively litde variation in the dependent variable, the change in the sectoral shares of value added. Patterns of Sectoral Producfivit Growth This section considers two issues, namely, the pattern of sectoral pro- ductivity growth and the impact of the changing composition of pro- duction on the manufacmuring-wide rate of growth of productivity. The issue here is whether given the observed values of sectoral TFr growth, A*i, the unusual evolution of the sectoral pattern of production in some of the economies had a significant quantitative impact on overall TFP growth. Going back ro equation 6.1, the analysis focuses on whether the changing strucure of vj, given the pattern. of A*j, contributed to more rapid growth. The value of A* is taken to be exogenous. Of the economies considered here, detailed sectoral growth rates ofTFP are only available for Japan and Korea (surprisingly, sectoral TFP growth rates are nor available for Hong Kong or Singapore), but these are two of the three economies in which activist industrial policies were employed most consistendy to achieve productivity-based cardting up. Korea. Calculations of A*i have been carried out for a varicty of peri- ods by Dollar and Sokoloff (1990) and Lim (1991). What is mnost strik- ing is the high absolute values of.Ati in most sectors. From these studies a number of patterns can be identified that are broadly consistent with one another. There is no simple correlation between promoted sectors and TFP growth. Was selective promotion important in the sense that re- sources were shifted toward sectors in which TFP growth was (exoge- nously) high, increasing A* for the entire industrial sector? To answer this we first calculate the manufacturing-wide growth of A* using the observed value-added shares at the end of the period 1966 to 1985, uti- lizing the A> values calculated by Lin. We then recalculate the value- added weights assuming that the sectors in isic 38 had been at their predicted value, VP, on the basis of intemational norms as a share of sec- torwide value-added and, reassigning the residual value added to all other sectors equally. Thus, instead of accounting for 38 percent of manufacturing value added, sector 38 accounts for 14 percent in this calculation.59 335 SIRANL MIRACLE The result is shown in row 1 in table A61. The actual secrorwide growth rate, A* was 6.7 percent, the recalculated one 6.1 during a period in which the growth rate of manufacturing value added was 17.5 per- cent per annum. Even if the "excess" growth of the MPM sector may have been attributable to selective intervention, its rare ofm'FP growth was not sufficiently above that of other sectors to make a large contribution to overall sectoral TFE growth. The major reason for Korca's manufacturing success lay in high individual values of A;. for most secrors in most periods. Japan. As in the case of Korea, a considerable literature argues that Japan's selective industrial policy was critical for its industrial success. Nevertheless, many Japanese analysts are skeptical of its effects, as are many analysts from other economies.60 The existing literamure indicates that many of the critical acts of the Japanese government with rcspectro industrial policy occurred in the ten years following World War II. In- deed, a verv good case can be made that many of the major policies oc- curred in the period from 1868 to 1920, but these are more difficult to analyze with systematic quantitative evidence rather than anecdotally.61 Chemicals and the metalworking machinery complex have unusually good performance. These industries are those thar observers usually point to as having received significant governmenr support, induding efforts to stimulate productivity growth. A calculation, similar to that for Korea, to determine the impact on A*for the entire manufacturing sector of the unusual size of sector 38 can be done for Japan. Table A6.1 shows rhe actual growth of A* for 1960-79 and the estimated growth, had the sectors constituting 38 been at the international norm, 24 percent of manufacturing value added rather than 41 percent in 1979. The value of A' would have de- dined from 2.3 to 1.9, a relatively small decrease given the growth rate of manufacturing value added of 8.7 percent per annum. If selective promotion is interpreted to mean that only a few sectors rather than all are aided, Japan's growth rate would have been only slighdy slower with- out an industrial policy. It is possible that industrial policy significantly increased the values of Ai in sectors other than those promoted due to one or more externalities. While this may be the case, it is likely that most externalities occur within individual sectors or in closely related ones. Given the many branches constituting the metal products and ma- chinery sector, the externalities shoulld havie been revealed in its own TFP value. 336 USING RESOURCES. -FFYCZR Appendix 6.2: Tests of the Relationship between TFP Change and Trade Policies T HE ECONOMY-SPECIFIC TFP GROWTH RATES CALCULATED IN chapter 1 allow us to make an alternative test of the relationship between trade and industrial policy and growth. We specify a regression of the following form: (6.3) TFPG = f(RGDfI PRIM, SEC, OPEN, rXPTh) where rFPG is the estimated rate of total Ecror productivity growth with physical capital, human capital, and labor as inputs; RGDP is GDP relative to the United States in 1960; PRIM and SEC are the primary and secondary school enrollment rates; OE'EN is the index of relarive price conforrmity; and EXPT is our measure of manufactured export performance. Table A6.3 summarizes our attempts to tcst rhc relationship between these variables and rates of TFP change derived from our production fiunction estimates. The regressions control for other sources of variation in TFP by introducing a -variable that has been shown in previous work to be significandy correlated widt TFP growdt-relative income level- Table A6S3 Regrssion of TFP Growth, 1960-89 (&tpenkt variable: totalfactorproductivity growth, 19610-89) Na&refdsa . - 67 67 St 51 69 53 5i - inrn *419 *.634 f 50A684 *- 4&5615- -I .376- 4.584 *-5L119I -53.0356 (034n7 (035712 23.W095) (135162 (0.437) (0A3041 (24.A871. (t4.5 rwktivm US. I960 .L7120- 49925 - *5F311J * .*t3538 .33W .2.40r .5.0403 .4.4383 .:- * - (0.)1483 (0.9791 (189294) (2067) (ld0219) OM095) (2.3467) (. 2599 wwPu7 iu. lnnlem I,%0 2. 3S53 2-2-1 2.0941- 2-6731 2A.1- I.6WS 1.6382 : (0.573C (0.56501 ( 0.6722) (44757) (0.6646) (0.65939 (5.5121) (L25) SalmlSycndlmr9G. -9) 2399 1.1991 LOW60 390430 2.604 * 3.069* 3.443' (2.0633) (1.045 (2.55)55 (O.59 11.1417) (1.1406) (1.5790) (I.6325) Dallwqrnamkrla t4943 - 047!50 015543 L52Z (0.13Z) (0-23712 (LI441) (0.150 . AmW mliauraape&.n 004 025D0 0.0255 E0L=75 %0641 OD354 : a4u6 .356 *' apomn. I9GWS - (0o1073 p(0593 (a.00W (0.0073) (.U0246) (0D33) (MG02M5) (e.ecs HiAtt -.at7G 0 55 -0 aAIeSs7 : 4817. - r ,- (04261) (0Q4531) M0 54), (0.50 M)uud IV 0R5579 O J CO50 .604 0UA Q4457 CAW67 - e70 03673 KNG,'tr. Mba. Sma_rd effr its N -->- - o.dtr - : : - - . . - - . ff :m n-mrn - : - . : - -- - 337 Fr~~X MI.RACLE as a control (De Long and Summers 1991; Mankiw, Romer, and Weil 1992; Pack and Page 1993). The coefficient on relative income (RGDP60) is generally interpreted as summarizing the productivity (TFP) gains realized as a consequence of moving from lower to higher techno- logical levels. This catching up is the component of TFP, change made possible by the diflerence between average practices and best practices. As we nored in discussing the cross-economy regressions, ir rime-series analyses of aggregate coral Factor productivity growth, estimates of TFP consist of two components-intraseccoral TFP growth and the impact of factor reallocation among sectors. Thus, the shift variable RGDP60 may capture the reallocation effecc of structural change on TFP in addition to productivity-based catching up. The use of school enrollment rates is an attempt to test for externali- ties in human capical formation. In our growth-accounting fiurnework, the contriburion of human capital to output should be reflected in the elastcity of output with respect to human capital. We have argued that primary education in particular can have growth-enhancing externali- ties through the better ability to use and master technology. These ex- ternalities may not be captured by rhe human capital variable and would then be included in the estimates of TFP change. We attempt a similar rest for secondary education. Openness to the world economy is another imporcant can.dlidate for explaining rapid TFP growth. The theoretical case for this view rests not only on allocative efficiency bur also onl extemalities assoiated with trading activities and on "X-efficiencyf gains from creating a more com- petitive environment for domestic industry. Several previous studies have found such effects, induding Harrison (1991), Dollar (1991), and Thomas and Wang (1992). We test one definiton of openness, Dollar's (1990) measure of the correspondence between domestic and interna- tional relatrive prices. Table A6.3 presents our rests of the relationship berween exports and productivity change. We choose as our measure of export performnce two indicators-the average share of manufactured exports in toWtl e- ports and the share of manufactured exports in GDP.62 When the share of manufa-tured exports to cotal exports is introduced into the basic specification, it is positive and highly significant (at the 0.01 level). The share of manufictured exports to GDP is significant at the 0.05 level in the basic spedfication; how ever, when the degree of openness or the HPAE dummy is introduced, manufacrured exporcs drop out. 338 USING RESOURC Table A6A Average Annual Rates of Growth of the Labor Force by Sector Gnnupasdperkvl Atcunm Indury Serikes Thoe Industrial economie? 1880-1990 0.4 2.1 2.1 1.2 DDcvroping economicsb 1960-70 1.1 3.8 3.9 2.0 a. Developing economics include the counties of Arica with the exception of Soudh Africr Latin Ameica; Asia with die exception of China, the Denocmaic Republic of KoDrea, Japn; Mongoli:. and Viet Nan; and Oceania with the exception of Australia and New Zealand. b. Industrial economies indude Australia, Cnada. Japan. Newv Zeland. South Afuic, the United Scrcs, and countries in XYVesrrmn Europe. - iwrc Squire (1981). Table ASS Financial Liberalizaion in the East Asian Economies IN Dl.-prsn .fflaw-dd a~~~~J m - )Isrssien -In IVA4ime-wgnla7 Jfrg fly 19* 6kTn sop innrvmewidn Tudiduifr r&d hum 6W awi 9IS-Tbeerrcrk 39Q, DatEd mknatu /ort..r he irans. fer -i5nenuis dtpma .iln - undo th eoniumNMmforfrwva.x. Yeq ib cadmftgeablkxd liauibsnbe fRn HlU5et to Hls 1in litA dimison Jibe achuop wlln6 n&kr inarusul bsdznmg pam. 19P3I:1iedfeecI odue murafltk t andscowinnisldu& A-dairt$EI byvkita&-s Aldwien-cking kn.utim we aesnu -r pin Hanl og a y,'% eiW-Balimc omuim nu cirye me- bawwnjtunkt ?tlautwrr ' rihadjrnnnat - o inte St fini ecano. bawith die sin rdIasEqking,ts G et nrmaBen Us tanRg Uqunc d ITden imoatnspipsad qBiz.Wl .tobmU mnndma%ima ams nid by do m. neei,i-g bush sad dwpntncdadmifillsi A ,;Mstnie pidhoe hT, dti-nirnma AiedmdnpduWed inyharlw. iumadiaitsdwaWnim i-flOAbbidIwbbnS annl. ISMNWIh'NZd rwniy kshn*qjuio.n ILkand Uno" ikerdflq'ek.it amaiS otnoMpUifMttI Ad f& ksduapnd se timenateAm hek tug Kn TuingC.o i(DtC4md hEn.r. Qinkiua tsaty epdm -Mndab,odbsnaid(M.AI-fIAB tqir Ddgt-TudSC sii hSI-UniWatSde umuck t It ini ~~~~nsaa h- (¶3W0 snunreyA bynttnn kn- adanhis d iltrHn tngLn rels125elmns panytq.o dmmcniewairAn gstdi ' sibicckExbnGDe,nnilmicce u- - 6..m DTLCsd iumda ar mn W k !tmoba..It I.L . bkh . IZRninddasSihcF'un 'au-rmk dsq,. Eri.b dewluk Inde= io nm eymm.totx is kni- d , ,a nsw&ayrndeumciwa vith dumt. .syaniydsngd* FunnUmbw 195* Etsd fdsetl9dwSuikm adl Fume swjydryc-nnpa atn Usn RadA Reuian LiCe Oft C:Gti atm - ltea CRn Onindelesun. iehuuansiiYwEmk aa ben Theen,ttiian ints Lnb.aid Dlr At danethire Anie num. Cod eFda Edung cmetd puidestpeindn w-binnlnge1yforbamla enom (B 1a lmd -fitovr 4apter ut 6ih-k-m 25d5f-imdtrnim. r in pin i h &g. ;nrk"du IMtlnsnemutermsicdsmanitnl IX6-9& Fad trsrininen pe of :CntwmmI baokcahmn .3hiselnkid s anchahe bends iuqm. depCSfbUIadruttnlub dinm . thituNialneutinia tim b innS dthw ,i ntto 1tu- enag tcru --: luta~w Lrmnamo(iutehab. uw ledsnso sakvi - bcnriitss&s.-. ' awstdasd -nqtsin'nionvi_en- 'wedaspheol aiu. Cmavitccilnp S aujxMEn;MCmmi 'w u SW:Smtuib aetbl"i a] 0 line in na. inplnst nalFuiwtnats-- ahh&C u ,, -baksiiar.-i. capiaIcstwmrlnsfiS. ' n t a eneaominmhemsy ainmn tpedenn dhbsufl4hiWntoteoI- dw p- l-d ia wm &LSwpAOt MJ*b lksdiJ*l 6anchintnnas. la besL - f: : . ; qykriogiDI mi" , ' rkaalidlifirdismacmor - ruw ualrB&)inua&cnL 1Y-21ailnz_ r c lmorntd1rmllysw nadce1Jhun stm puivute huskid wid fir4ig AMS,EstdghAdNWwr-duc mimmi-T INnamDU finnkfj,rlnsnntkst5ankmn (HI) mki& --- trdnefidlkiusndtelkisteel 6iIss b Fd6ER.PwdioInwili. O )etanassp aqu dRtel puMladan - - m' idadwiegnn nmppoadi ctadseopintrtCD4) pplic ,dbmk&mxli muse hd pwrdfBtk.- '- - ~ - , Lnalnbe mi- -hn --dzctc ,l,ssi w-i eye Laureleng N.o f nedmw - :: - R - - - (^~~~~~~~ ~~ ~~~~~dcwys tna..sTsdahgaingaina plscigsghe, p im. :Xd _ quoky. Pa- SwitinamdlEueGCmrm nw.rWa' - - : . : - -- - : :-.- inaI Eri*ntd Ninmss lil - - nde h sh &eupa nmh339 339> iASJ* MIRACLE Table A6.5 (confinued) Fi_nes1 tflnwu&gku .f?e.emw ruin JImeqIad ,fiecu M." IVRf= -fnanq mya tse andns _y Jq_ Lading maawarmneguled bn lleal- flMtAlhyecd cmey.C ~n . JUb-& EarnWerd bfld 6lr elek - The MinyfFlhlna witi iwn bue is Idewldein droddidiol minks.. wIdch by IPRI wee ulpalin an all trannd Tdkyq DrLlr eA eauu mpmnimry I manimu sep. Oqee Dql dr qrml fnsinguith ldmik bcAl. lmae. U werwik pwns fr. the akae Lwa% weprr&ulyco-IaL Ifl&selandpwLeaand(pualk I97Puealing petin hebd SecuidmitedadEhnd e LnTreeleandlki lPf-lSC:drLl Iw pgalt. I"h ly bn6 pwnal. nLapuend eqeriancC ietCFi) muSs ELASp J . _J.mlrghti6d .J.lr -finlrualnefetreadensnl lilea. ceargnadrMh,lmwL mpl6. eunainenda ituahuks.. mpAmdeycaie lotion ofin hsr rm.m Feed bawur IJ uu TniT hu wesneJ by Ereiga. flfaRnictiuon.sinurinral iItStnMU.ddl pnd udea m son munibehinmrbskalkudIl b mr- hakbd eiclLd i n.UhUL Mea dia ndwepirle iaJ hmaClg7talwdmailunlmeuhaa lWtSinI9 NW iedlum boan.. /misarkar(pulickmmdAutiuL inada hig CWaet maeberamtnls (I 91 btw6n(Seg)wem endvn aul I 9eat*Gw nubeaeeet a 191: EatbiMl u hellWd dt Dq* In Ml 'nglea nEcesnlldCa (bean. nrlca and BAten t u .n Coonamtiden aeda.ceqnr prcedes eF a'el amarke ra(nS). 1 e- Inf: Batik l esgi te Iti ro ng ISW&T mmuyiuco nwaed tN cw lon& I...caaaar.L queamalAa4d dqeie rgetrorminga sudiarlsnlf iea incs fi.rm Feuip atitia Srn gpnErd Itnh beanit anenimpat herniae law t9t Rea - nemyskLst lhew maian heniphntheTLy.Snk iftiUenUIStrd. ncmlad. Ilntve rum e ExdaKJ.wafuned wme m I ag u lanird r ed - s thn tint JrpaI nraipal. -tan. Iwkueai p4mt (CC) mkask mmneiL Sesehreem . trdin thte Us Sequlds Edia nse ered. Dan usMes rf nd intne arn almond tv psiidpau inh m Ca,ndial hajneaaMrt AtktSmckindfis ma wading Ms-Tkofiac un.mj rfTey ru Minr 5ket maut. haairnl,n r 6at .Kern 4p.f 1* ll nareuanladlaetpm hrnitin- t% t4iodd Inla csneed Ilsotlrfeaciual eapelisn t InEudAaIo dqa.it inasunacule .cs. yalaminhinld lrtam nem Rorm lJuCaeGnrt toh ep- i ie td l. J9n1f lv lRni6i ll tt mt s - ttacauac rmcatlain 1973. cheat Jilt csneni pq-e takre ide pebieic-s Iinulcail e-in grcyu- IPRAWlderadeiteecinap and IWI-AS:P ?i.eri eua- mnani cn- cualdieetI.Setupcstmeaed mt mnd Madudeald cheOQee Of d'bg eedulapd ne.ro rncdv- nwn,61 Udissad ailowdaeyof ifnhtet-Sr lieiwg &eraidwa Eeaneiicn (ONSE ID enmdu Rn il.l Jdaapreliratth inse - rewdmmicaeidfrnephink. i;1.. wauphiias asally. Cpa e -bell l. Ieefarkinu eagn s up OTCtae amkte i rits UlflnS a bn Y IWSd Xlirlilnab 19-* M infimumed mem&nn but fwaegunla (enenere r`fieigs lt M*bUn ieU d ahea my an- cpened e nsi ta pe- h won ecago incnhAwL Isa btd.ia iama Bn IS is I te . PIa po4u I bgewen uhefhbne i lee na mee.tr nae niuluak hding 64 in 199li. Deei. p incndtudi- banldlagnrld nm.nkn-aae la shiauia uWhi.Ll3abnUnseA n enr CD. if cngrefAmuiei6c% ukhleaditccank% JW )mPAwdl kgwidaehID cace? mhlhiscynrf inae(MOljwapeniassal iinpne -r msigid.sunfreinrdepode tr. nbuineband erpsdamenprc- lamemat lnaancnmpudics ate pn&i6l g;dd r ma pe.tionnsftWBk tnt Mm bekake = .ukflnti- tna.2 Noa atw bankas ma any &W scaaelckshanSipuaeauadmaqqwd -~ ~~ b he.deeetw (NR loaig n qchncAhelFlsmeded JNIPouaiaac licthcc hey.mkannrioait a rnd p.*hmfia Dil insmial en (NOR) imnm eiq niem ewi s Inen FIksaJ9c. fmlap a ipe atcl jeeitk ar W inpdoeiii - dcpdtmc a ihmsm.L &i*.h4n P o 1W 6r roNaliq_iIFi ii fahnv in rej r=;o&m f6rtLirg' MON li mlnrt A-m dkpoir NllFIswrb-UL r-oel vnXd wu: end vmleuneetaStcdad maini. IA9 rlhid iFa_b alloerd M eAs - m gd t .a CD, :r.'o-Sf.th, 340 . :. - . . - 340 _:,_ USING RESOURCESSO aM.opta 1971: hmnaei aianhrfind&pdewicf Mi.hudantf194 hm o aku ~iapig fl%-KmJus"s Li cIS&E if L Rid hJAiog&t-eonpickalyddsdn nmm zho flmyw,g. hiajlty 6cua haiLbmab Howemi. dndc w NW p nab&hd &hf. Idga Jtaincfbhulu prcaci, of ffkb loUtes aam.csa11icMFumnl bmil kw Inured aainsO 50, &Ldqwia p-cer. wma- dqiWmnc anvocrUe9i maumpnamu and aural of Inkik&p.sitmuinr am pr uawiru they haw ha. napmdto insni rdubinmnbedq buil (End. ~~~~~~~bn~nchh w6iiu ne&lernn rmaota nmnduat jMs--aoVmei.fUanaic 6i.& mnd 0TC nhh.rCmbsponlosasd&4 1374-KNew Anaidelimuamsn ngmito inhclupo t..49 prenm 139W8 A h kaC.1'dohia Wiath gasiddnmen ogle 57&MA wnianuadu Laek baing .aac ouhAseleiA CD% mnd hr&r aua0cmdrm Ewa n mu' KManddhcfamdanX An budges Zudiddspli sit drpca alibe,tindaweL aaqumcninamn oa sriau dfitpakiLm rordineindng Ktndialaw Flame. h1Bwm.l melube. ma 19~~~~~IN-%O,SshSab,rdan cff-,bca 1398 mALina%ar Finadi lasituhm.Ar-~cwjw plfthklAngSram tifan man roe biWngto1UL CLCruStbaku 6, poad rtheakingAadZ5l'Manl FonmanceGimpma p wd n f Act .1M. SoL:i..tm,.lacu lc&g guidolina -)flCrmmrulbmk-alminhunced Bar iuqMaCCaKrhia. andactig awtdiinmsu.darda&afiaaCd,hainnd it LadIng 1atehrodcednmanlhe. 1:&ibdla c ue,pWcctc .aitardind rnamisffi.aeW upcidg an Linac 199-. (snt at liamfln car- 'n..- -taupary. phmuLSac cadlgeuzie aiiua Sk~~ar ii7kCamplcc~fi-Nrad libaclin- IWI 9Aanmoderelmcnminoal IAf ai. DAr Miadot-&lai 15M.k EorkWhh MqntuAki& iyaVSaypm uam. Bankouitn7aa J&odeurmia. ofbanWigumwawh bklbal p-x J"ifSupwpurS&Exhap or Alonig wmAt canal bau'ngAocticm. mope-bno u- kg fitcennuccd*mow dqaodtand I= aintuudrlpmz 9laUhlmI dwSIngar pveI. maicsncfainea dut baki mow mimut Wit. IcadhguotaJimam Rial afto e Ads=O Di. AdLs I6IxMA a1 a!,1niidloocuiy E aFeat BabigAauet oidha. apiiada aiuhnerma,tin.d- etehaFo wnand watplevalldratma- nEabat boideg maskts. Hi limema MqencdcwlS1VpmrCG&1dEn Bu ala r1M 10 nms .liqddkrruaz..h*j ba- rlod by 197K in shyumgarn to.a ADwiplbikano dbze.pupcdatnimnmsd eulamts adautraiomtea.o pm.perJ lomauid lnid ISZ211lutternilaprnn h Sfiogper.l 2dIcrjta />sindl3aflhlsc so il pwpawapi.Mcnd - . ~~~~~~~~~~~17r0 asa hraiaglirnbna frran ro sliad6 INtn SitdiEuaqrlam )ni cceaDTlc±sdbi -- tSlgt had..r bankiang _iona lnnard lacun i and aa1fi Ml Stock a c..mnnpecl t^ii0ilbding aWeCD* - dhr~~~~~.M aJeimonqefi and -gi vm&g-;ia:-r- tnO -w banlh Font oeadat,i oaptiax flflSWtFocf dooic buk par- lanti in apitban orfbeeW hniari c L mir"uiomefluainidey*riwai inlocalboaokomps.Poretbasik madlbomgarvhow, nuk(hd tacnve tehdi,19 poite sula, re... aim. 1950( Lirbclsd C at. Ckit AttaInhit1 ban mw IQiEahlsbehrx T.iwa o19t D. SsxhiS ad Exitnu b Crwnuiona 194 lAI,Jna amsaimds.adini- fmnddradcpitbd Suit.tud.a ucpbcnm.pnvreaepital ni.arktpeaaks. snladugawa mnd:auccl. d a e ow t mccoy IWi tfmbl&d aedqmiIa rpnce copses . lylSCosa&lrf hanktulnuadM permttdartoiWt id.scutlq6a1aw mlet11CP. CDaamnW.4Atwit IN*-JflA.."dclSriabTrmanatdn I.A. t napinlene maewi*inanmnl .ppmnL Madaadnsmcuic- awknntctnwinal. inauacsmm. acpdwMaaenrmandb.tn auborahm Ithan ainp. bank-adaalnimnd inapt. Rrmbrd aiom fiurfoaig bade.m alm)ined 1956 Craed .W.&bovebadaig Raiwd banbinglausnD puohlbk tcrdiemn Iiewma delnt an Piam cwrugahndimun long-mm lendin hwodam. N Cepnui c.rtahn kdaiuAwld maas aada hidinpia - naaapmnicaanal InacknEacs of ~~~~~mn&,uno ltmc l napdcnacuaad miumntimarrff drpltionsadlarcog.it nof oiunp ioadligrolumbrod. mW * - iSa meien anl pIMatler4l oatior t Jfmgn 73&11J Jwifletnnrymecti6pkbocWitnca CdllhOGtoMldkrhgaenIf1 113qsskd.p ITTmanrortoa ltpleamatedo ate. ,nau.rin. bmI ficoInuSpnnz hid - ew .dn startd buioando in 197tkdua. gto f/m"an purcm fiadand a ansby~ iatyiuedbyuatybw lNtPcekd 1562 habonglwaIn - pw"-Ph. FanceCsNp, - kb;-itpebid. rofihltDam bat11strada EnltaqccftluiIp trdnStds ticrmuAdpmdu as. B da ncc.dbag lachain I*oaow aJbn r.vlmrrial pad- nTh.alc RappmhscMakanr lflnlSboib tlni rh ucccmopatuy Zn6 - *lirinn reTolwet uclikehagnharm.aidndunl. caabl Lsa Rtrctcioa;fr c ti.dindopMMimdaanigf I.Ab awidr uatrumap II"urewse E uubcd d.c trer;Anbc titmmucaL .aatnr -aciallraktimcrpoitas Ca. maktow - heiJMA&iur.jpnba.ndioemc iH mintd-i * g~~~~maetmaa.ryraL :flhauzJMvs baJutk machemmanh.wiOgdiwtcndhac Oay.r Pa * Imhbclitttallepa& nemlbnga ~~~~~~~~~~~InhatWdcCS.aI Ret. cdt cmibriuinafocf mersan ie nd a acsaaa hi- -n aaadalbaukaledimcdepm -o munsdrdlpih wmaunaal mi5i - - ~ff .bat anepm.Abormiued dnfitlcpipntadtc .na.Xcmain .h&dbAlh,v&o.5oa.dii hl.z q-yydm.rllaoap sIrkn WD million bOIL am padh clicdu caon InF ISthSaundsandSabaa 0am.d cmasal wnit - ,9,d.i4 ,,em ledwiard -- :~ . - -*,-*- . : . ; r- . J^*LdC> /Is*hrbeglarff@6f:-_IT -n /fftSBpS:r34'Omiin 5jE eMRSCINMRALE Notes 1. See Greenwald and Stiglimz (1988) on the effecr of 7. The costs do not derive solely from the misalloca- coordination fiailures in labor markets. dion of labor implied by gaps in marginal product among workcrs with thc same human capital endowments. The 2. Altematively, conscant real wages could be due to a re- costs of such static misallocations rarely exceed 1-2 per- duaion in surplus labor (Lewis 1954; Ranis and Fei 1961). cent of GNP. Dynamic costs can run much higher when, for example. employment in enterprises is excessive (im- 3. A population will double in 87 years when it is growv- plying a wedge between the wage and the marginal prod- ing at an annual rate of 0.8 percenn when it is growing at uct of labor) and the subsidies required to keep firms 3.2 percent a population doubles in only 22.5 years. See afloat divert scarce savings from productive investments. Squire (1981) for a historical comparison of population Sec Gelb, Knight, and Sabor (1991). and labor force growth in developing countries in the cur- renly high-income countries in the nineteenth century. 8. A labor market is segmented when some workers reccive higher wages than others with the sarme level of 4. For the period 1989-2000 population is esrimated human capital simply by virtue of their sector of to grow at an annual rate of 0.8 percent in Hong Kong, employment. 03 in Japan, 0.9 in Korea, 1.2 in Singapore, and 1.4 in Thailand. Indonesia and Malaysia are forecast to grow at 9. Clearly, a high share of the public sector in wage higher rates of 1.6 and 2.3 percent, respectively. The share employment growth is not definitive evidence of a rela- of the population less than fourteen yeaw old was 31 per- tively high share of surplus labor in the "modem sector.' cent, circa 1900, in the currently high-income countries. The public sector could be finding productive jobs for all In 1989 it was 30 percent in East Asia and 36 percent in its employees. Latin America; in 1986 it was 46 percent in Sub-Saharan Africa. 10. In more than one-third of the economies, private sctor employrnent actually decreased, implying that the 5. Increases in the rechnological complexity of non- public sector accounted for all the observed increase in agricultural production and in the size of establishments total wage employment. bring greater division of labor and, hence, increases in skill differentiation at the same time that greater availabil- 1 1. While there is no doubt about the role of govern- ity of educational opportunities increases the supply of meat in suppressing unions in Korea and Singapore, in cducated workers. Taiwan, China, there is some question whether govern- ment action, for example, prohibiting the Chinese Feder- 6. Labor market efficiency is assessed by comparing ation of Labor from engaging in collective action, was marginal produczivities in losing and receiving activities. decisive. Market forces, and in particular the size distribu- Evidence of changes in labor allocation suggests that indi- tion of firms, also help explain whyso fe%w workers in Tai- viduals choose amongalremative occupations, industries, wan, China, are union members. See Fields (1992) and and geogaphic areas as though they had a fine regard for chapter 4. marginal costs and benefits. Ihough behavior consistenrt with rational decisionmaking on the part of individuals is 12. See Fields (1992), who nores that until 1972 unions a necessary condition, ir is not sufficient to ensure optimal were free to bargain over wages, which were largely deter- allocation of labor. It is also necessary for workers to be mined by market forces, and Freeman (1992), who, refr- price-takers and to be free to enter any industry or occu- ring to the post-1972 labor market poliqc regime, has pation for which they are qualified. Restrictions on free- asserted, "Singaporc is possibly the most interventionist dom of cntry arise from such sources as nonopsony government in the capitalist world." It is irteresting to note power, discimination, government interventions, and that the same instrument-the National Wages Council, a workers' associations. deliberation council discussed in chapter 4-was used dur- 342 USING RESOURCBS$T: ing the period of wage reprcssion, during the accelerated 19. Other evidence on the extent of subsidies can be growth of real wages in the 1980s, and now to maintain real obtained from the reduction in financial burdens to en- wage growth in line wid productivity increases. terprises that borrow from subsidized lending pro- grams. In Japan, the cost reductions have been large in 13. Lookingbeyond thegroup ofEastAsian economies sca transport (more than 20 percent); 5-10 percent in on which we have focused, 1abor repression is norgenemlly electrical power, transporc machinery, and the coal associated with economic growth. Freman (1992) notes mining industrry but very modest in other industries. that 'most dictatorships that supprcss unions are incom- Ogura and Yoshino (1988) estimate the financial bene- petent in economic affaiirs-vide Burma in East Asia," fits arising from the reduction in interest rates of though he recognizes that this generalization is not an ad- Japan's directed-credit programs was lcss than 2 percenc equate dismissal of the argument that in some economic of total invcstments in manufacturing between 1961 circumstances, such as those prevailing in the "miracle" and 1973. In Korea, some investment projects and ex- economies, repression might be good fir development. porters received substantial interest rate subsidies dur- ing the HaI drive of the 1970s, but these large subsidies 14. See Ranis (1993) for documentation for Korea and have since declined. Taiwan, China. 20. King and Levine (1993) also show that high cen- 15. Gelb, Knight, and Sabot (1991) show how in re- rral bank shares in rotal credir are bad foir growth. Hence, sponse to unemployment the creation of public sector the macroeconomic restraint disLcussed in chapter 3 per- jobs with zero marginal product an over litde more than mitred rapid expansion of private crediL a decade reduce a moderately growing economy to eco- nomic stagnation while having litde or no impact on the 21. When exports were not used as an explicit alloca- number unemployed. ion criterion, Japan often used inremational prices as a stndard of good performance. Its industrial rationaliza- 16. The governments of Cote d'lvoire, Egypt, Mali, dion programs used as criteria cost reductions that were and Sri Lanka all have at various times explicidy acted as based on the costs of inTernaional competitors. This was "employer of last resort," particularly for university grad- enforced through the tireat of international competition uates. Guarantees of jobs to workers who cannor find as imports were gradually liberalized (for example, ma- "Suitable" employment elsewhere genrally resulted in the chine tools and steel). mushrooming of public sector payrolls. See Gelb, Knight, and Sabot (1991). 22. The merits of well-managed and fbcused directed- credit programs have long been advocated bv Japanese 17. When Indonesia attempted to direct credit to in- govemment officials. For example, a study by the Over- dustryin the late 1970s, ir was immediately channeled al- seas Economic Cooperation Fund has argued that gov- most wholly to two public enterprises. Indonesia ernment involvement in directing crcdit is warranted withdrew this directed credit scheme in 1986. when there is a significant discrepancy between private and social benefits, when the investmenr risk of particular 18. Malaysia, for excanpie, lent substantial fiactions of projects is too high, and when information problems dis- credit to public enterprises between 1975 and 1989; the courage lending to small and medium-size firms (OECF fderal govemment lent direcdy to public financial and 1991). Use of policy-based lending rather than other nonfinancial corporations. Its lending reached the equiv- forms of industrial assistance (for example, lower taxes, alent of 22 percent of banking assets in 1975, dedining to grants, and so on) is premised on the argument that the 16 percent in 1980 and 9.8 percent in 1989. These loans main constrainc facing new or expanding enterprises is carried real interest rates, which were mildly negative to their access to extemal finance at reasonable terms and slighdy positive, compared with the substantial negative conditions. Directed-credit programs involving small real races of many other economies (Caprio, Atiyas, and subsidies overcome this consuaint, but Japanese officials Hansen forthcoming). have stressed that to avoid the misuse of funds and abuse 343 LERAGLE of credit programs, strong emphasis must be placed on menr, saying thar if a firm really were to be profitable in the maintenance of macroeconomic stability to minimize the long run, it would pay for the firm to encounter losses distortions in incentives and on effrive monitoring to today. But this counteragument runs into problems: it is ensure the timely repayment of loans. based on the premise that capital markets are perfect. 23. These studies are preliminary outputs of a World 31. Export prices can be set in the international (un- Bank research program on the Effeccivencss of Credit protected) markers at long-run marginal cost. Is is possi- Policies in East Asia, which is intended to provide empir- ble to show, under certain simplifying asswnpdons (see ical tests of the effectiveness of directed-credit programs. Spence (1981), dhat optimal production of the firm en- tails setting the marginal value of output cqual to the 24. Among the best known of these arc Balassa and long-run marginal cost. others (1971), Little, Scitovsk,y and Scotr (1970), and Krueger and Bhagwai (1973). WorldDevtdopmentReport 32. In the early and mid-1970s, nominal tariff rtes 1991 provides a good summarv of the arguments for the and effectve protectior. began ro dedine (Iroh and Kiy- linkage between neutral incentives and growth. ono 1988; Komiya 1990). Although there is disagreement over the extent of the reductions in effective protection, 25. For an interesting attempt to test one of these there is agreement that effective rates of protection were propositions-extenmalities accruing to export-in a substantial before the early 1970s. gcneral equilibrium framework, see de Melo and Robin- son (1992). 33. While effective rates of prorecrion or subsidy cal- culations would provide a better measure of the quantita- 26. Among the earliest formal discussion of market tive impact of protection, the figures shown furnish a failures in developmenr are Scdovsky (1953) and Chenery notional indicator, consistent over time, of the extent of (1959). A thorough survey of many of the issues is given reduction of nominal tariff rates and noncariff barricrs. by Corden (1974). More recent discussions include Pack and Westphal (1986) and Itoh and others (1988). World 34. It would be more desirable to compare efnective Bank (1992c) provides a recent survey of the issues and an rates of protection, since it is value-added prices that mat- application to the evaluation of World Bank industrial ter for resource allocation, but the most recent systematic development projects. cross-economy studies based on comparable definitions udilize evidence for only one year in the 1960s and early 27. Economic growth will eventuaHly preclude the 1970s There are few recent estimates to provide compara- need for policies to capmre the producrivity gains from ble figures. economies of either scope or scale. 35. Singapore has been exduded since its zero duty 28. For a review of many of these, see Stewart and rates will bias the estimate downwart. Ghani (1992). 36. Pack (1 988) expressed doubt abour the robustness 29. In general, real externalities provide a valid argu- of the evidence for Korea and Taiwan, China. When che ment for trade policy interventions only if they allow paper was writren in 1985, the TP estmares for both goods to be produced at less than the imported c.£f economies were based on data that began in the mid- to (costs, insurance, and freight) price. This is nor, however, late 1960s and enided in the late 1970s, a sufficiendyshort sufficient to justifjr intervention. A socially successful in- period so that increasing capacity utilization or one-rime tervention depends on whether the present discounted benefits from scale economies could have accounted for a value (Pnv) of fumtre producer surplus exceeds the PDV of substantial share of unexplained growth. One study of the cost of subsidies. Korea (Kim and Kwon 1977) was indeed convincing in demonstrating that correc measures of capacity utiliza- 30. More recent discussions have discounted this argu- tion eliminated the residual for much of the period in 344 USING RESOURCES C question. Additional observations are now available, and in Japan, Korca, or Taiwan, China, but it has been very improved estimares of the capital stock make it seem importanr in other HPAEs. Even in Taiwan, China, DFI ac- likely that the measured TFP accurately reflects unex- counted for 15 percent of employment growth in plained growth. 1975-84 and demonstrated the viability of new exports. Moreover, much of the investment was in relatively small 37. On these issues, a particularly good discussion is firms in which employees vvere exposed to new equip- given by Itoh and others (1991), part 1. ment and the relevant production cnginecring. Since labor turnover in these foreign-invesred firms was sub- 38. For details, see the various chapters in Komiya, stantial, the total imnpact of DFI on effeaively introducing Okuno-Fujiwara, and Suzumura (1988). new technology wvas quite high (Pack 1993b). 39. Itoh and Kiyono (1988). These estimares are based 46. For efforts to assess the empirical importance of on nominal tariff rates, not implicit price comparisons. spillovers from DFI, see Blomnstrom (1989) and Haddad Given the presence of quantitative restrictions, they are and Harrison (forthcoming). likely to be minimums. 47. See Rhee and Westphal (1977) fbr evidence on the 40. For extensive discussions, see World Bank use of older technology to achieve exports of textiles in (1987a). Korea during the 1960s and 1970s and Ranis (1979) for Taiwan, China. 41. In Korea and Taiwan, China, the initial spurt in exports was in labor-intensive products as predicted by 48. On Taiwan, China, see, for exanple, Wall Stret the static theory of comparative advantage. These, how- Jounmal June 1, 1990, and Liu (1987). On Korea, see ever, are also the sectors dtar Dollar and Sokoloff find ex- Westphal, Rhee, and Pursell (1981). hibited greater TFP growth between 1963 and 1979 in Korca. Wang (1990) finds a similar pattem for TaiNvan, 49. Katz, cd, (1987) report the results of these studies. China, for 1966-86. Exports may have resulted from this The firms investigated made intense efForts to improvc initial spurt in productivity. But analysis of Taiwan, productivity despite the protection afforded by the im- China's, early export growth by Scott (1979) and the port-substitution regime, an exmple of the too often for- analysis of the Korean trade incentive system byWestphal gotten point that protecton has both income and (1978) suggest that factor proportions, supplemented by substitution efects on the effort expended to reduce coscs. a small pro-export bias, plausibly account for the earliest In the frms considered, the substituion effec dearly out- export growth. weighed the income effect. Nevcrtheless, calculations based on most of the studies suggest very low rates of 42. A recent study oFTFP growth in Chile, for example, growth of capital productivity, an exception being the re- finds that following the 1970s' liberalization, accelerated sults for one steel plant analyzed by Dahlnian and Fon- TFP change due to organizational change and product seca (1987). rTP estimates are nor easily extracted from quality upgrading lasted at most six years (Perry and Her- the data reported fir most of the srudies in Karz, ed. rara, forthcoming). 50. The impormnce of interactions may explain the 43. Undoubtedly, domestic sldlls were important in al- phenomenon noted by Behrman (1987) that between lowing the technology obtained from abroad to be uti- 1960 and 1981, the averagelevelofschoolingin all devel- lized productively. oping economnies increased by two-thirds, yet in most economies this had little effect in stimulating growth. Ed- 44. On the role of infi-astructure in Taiwan, China, see ucarion has a smaller effectwhere complementazychanges Ranis (1979). do not occur, thus, high education may be necessary bur is hardly sufficient. 45. DFI was not a major source of invesrment growdt 345 lR A CIRACLE 51. Previous studies using this approach are Dc Long port variable are not significandy different from zero (tamen and Summers ( 991), Mankiw, Romer, and Weil (1992), together, the two amr significantly diffrent from zero at the and Pack and Page (1993). 0.01 level). Where rhere is a high degree ofmulricoilinearigy berween the variables, the coefficients on Lhe interaction 52. Several previous studies have fbund such effcas, term and the export variable, despire being separately in- including Harrison (1991), Dollar (1992), and Thomas significant, should still be ueaed as besrpoinresimares. and Wang (1992). We use as the openness variable Dol- lar's measure of the correspondence between domestic 56. The discussion of scale economies has recently and international relative prices (1990). This measure been revived by Murphy, Shleifer, and Vishny (1989). uses Summers and Heston's international comparisons of price levels to develop an index of 'outward orientation" 57. An alternative to cquation 6.1 is: AO = iVj A., for ninety-five developing economies (1988). which assumes that sectoral shares are constant. 53. Alternative specifications using the rate of grotvth 58. Another interpretation would be that there was an of tomal exports were also employed with similar but less initial disequilibrium in marginal productiviEies amnong robust rcsults. Dara on the rate of growth of manufactured secors and that indusrrial policy accelerated the move- exports is not available for a large sample of economies. ment of factors ro higher marginal productivity secrors. Writh respea to labor, the issue is moor in the Four Tigers 54. This conclusion is reinforced by the arguments as they ypically had rdatively comperitive labor markets that low distortions relative tO international prices can arise from having a significant share of the economy oper- 59. Put differendy, it has been shown earlier that aring at export prices; in that case, the share of manufac- Korea's vPl4 ratio for mPp was 2.76. If this had been 1, rured exports in GDP may be more strongly collinear with the actual share of value added would have been 13 rather the openness index, than 36 percent; thar is, the share would have been con- stant ar the 1968 level. If factors had been allocated to all 55. We ran two joint E-tescs on the regressions induding the remaining sctors equally, the sectorwide averge the interactve term. The Ftests were both consistent with a value for A* would have been 0.061 rather than 0.067. high degree of multicollinearity between the interaction term and the export term. The Ftest rejects the null hy- 60. See Itoh and others (1991), Komiya and others pothesis that, taken joindy, the coefficienrson thethreevari- (1988), Saxonhouse (1983), and Trezise (1983). ables (education, manufacwured exporrsftotal exports, and the interactive term) are nor significantly different from zero 61. For a very useful account of many of the measures (taken together, the three are significantly different from in the earlier period, see Lockwood (1954). zero at the 0.01 level). Likewise, the Ftest rejects the hy- pothesis that, taken jointly, the interactive term and the ex- 62. See note 53 for an explanation. 346 q!TER 7 Policies and Pragmatism in a Changing World HIS BOOK HAS SHOWN THAT THERE IS NO SINGLE East Asian model. Rather, die eight high-performing Asian economies have used different and changing sets of policies to achieve rapid growth with equiy. In this chapter, we evaluate six key policy fundamen- als used by the HPAEs. We then examine institutions to promote growth and finally assess three policy interventions. We end by considering how the changing global economy is likely to con- strain fiiture use of the HPAEs' most successful intervention: the export- push surategy. Foundatons of Rapid Growth- Getting the Fundamentals Right T HE STRESS ON POUCY FUNDAMENTALS IS A HALLMARK OF ALL eight HPAEs. These widely shared, market-friendly policies are the foundation of East Asia's economic success. In chapter 2 we identified several broad policy areas as fundamental, and in subsequent chapters we described their relevance to the strategic finctions of accu- mulation, allocation, and productivity change. Here we briefly review some key aspects of these policies. Although implementing these fun- damentals successfilly is often difficult, the benefits are substantial, as demonstrated by their crucial role in the rapid growth of the HPAEs. Moreover, despite changes in the global economic environment, these policies remain viable options for developing economies today. 347 #PACI RACLE Ensuring Low Inflaion and Competitive Exchange Rates Macroeconomic stability and low inflation were necessary precondi- tions for rapid growth in all eight HPAEs. The key element was the man- agement of the fiscal deficit. Fiscal policies havc been generally conservative in avoiding inflationary financing of budget deficits, and some HPAEs, such as Singapore and Taiwan, China, have run sustained budget surpluses. While Indonesia, the Republic of Korea, and Malaysia acquired considerable foreign debt, their debt to export ratios were lower than other developing economies, so creditors never lost faith and they never faced a fuil-blown debt crisis requiring rescheduling. More- over, budget rules and legislation requiring char sovereign debt be used for development expenditures helped to ensure that borrowing funded investnent rather than consumption. The successful record of low in- flation and sustained growth of the HPAEs reinforces the view that there is no substitute for fiscal discipline. The HPAEs' success in export-driven growth was predicared on com- petitive real exchange rates. Prudent fiscal and monetary policies coupled with flexible exchange rate management were used to keep exchange rate movements aligned with dchanging structures of trade protection and in- flation differentials with trading parmers. The HPAEs avoided strategies of macrocconomic stabilization that stressed the role of the exchange rtc in breaking inflationary expectations. Rather, exchange rate adjustments were supported by expenditure-reducing fiscal measures. While Indone- sia, Korea, and Thailand have had periods of real appreciation, they did not try to maintain a fixed exchange rate regime in the face of inflation rates much greater than those of their trading partners. The HPAES responded to external shocks more quickly and effectively than other developing economies. East Asian policymakers have been more adept at managing fiscal contractions-cutting expenditures and increasing revenues in a growth-oriented framework-to attain viable deficit targets. In contrast, governments elsewhere committed to funding large entidement prograrins-income transfers, untrgeted subsidies, - cessive public sector employment, and loss-making public enterprises- often lacked the flexibility to deal efficiently with budget contractions. Rather than cutting spending on entidements, theY have often resorted to cutting public investment. In each HPAE, a technocratic elite insulated to a degree fiom excessive political pressure supervised macroeconomic management. The insula- 348 POLICIES CI tion mechanisms ranged from legislation, such as balanced budget laws in Indonesia, Singapore, and Thailand, to custom and practice in Japan and Korea. All protected essentially conservative macroeconomic poli- cies by limiting the scope for politicians and interest groups to derail those policies. Building Human Capital We have shown that the broad base of human capital was critically im- portant to rapid growth in the HPAES. Because the HrAEs attained univer- sal primary education early, literacy was high and cognitive skill levels were substantially above those in other developing economies. Firms therefore had an easier time upgrading the skills of their workers and mastering new technology. In addtion, rapid human capital accunularion reduced in- come inequality by increasing the relative abundance of educated worker, thereby lowering the scarcty rents associated with cognitive skils. These benefits were particularly evident in the countryside. Basic educaton in the HPAES is highly oriented toward the acquisition of general academic skills, while post-secondary education tends to be oriented toward vocatonal skills. Some HPAEs have also been unusually large-scale importers of educational services, particularly in vocationally and technologically sophisticated discplines. In addition, enterprise- level training, much of which was subsidized, accelerated technological acquisition. The benefits of dosing the gap between girls and boys in primary and secondary school enrollments have been subsmntial, even in economies in which female labor participation rates are low. Better educated modt- ers raise betrer educated children, complementing public investments in basic education. The rapid increase in the educational attainment of women contributed to Ehe early decline in fertiiy, which in turn re- sulted in a less rapidly growing school-age population. As a result, mDre resources were available per child, in homes and in the dassroom. Creaing Effective and Secure Financial Systems Financial sector policies in tie HPAEs were designed to ficilitate two crucial functions. First, they encouraged financial savings. Second, they channeled savings into activities with high social returns. There was great variety in the policies and institutions used to accomplish these 349 _RALLE ends, but several flmdamental approaches were important in all eight economies. These included generally positive real interest rates on de- posits and the creation of secure bank-based financial systems. Positive real interest rates on deposits were the most important f&ctor in encouraging financial savings. All of the HPAES had periods, usuaily early in their development, when rcal intcrest rates on deposits were negative. But in general real interest rates were more stable and more consisrently positive in East Asia than in other developing regions, where high inflation has often meant that volatile and highly negative real inrerest rates prevailed. HPAE governments also encouraged savings by fostering scable and se- cure banking systems through a combination of protection and regula- tion. Protection of established banks induded government limits on the creation of new domestic banks, the enuy of foreign banks, and the op- eration of competing non-bank financial institutions. Regulation in- cluded enforcement of prudent behavior, such as limniting speculative lending. Although the HPAEs had no explicir deposit insurance, they promoted confidence in the banking system by devising rescue pro- grams as needed, sometimes, as in Japan and Korea, pressuring stronger banks to absorb weaker banks. In addition, many of the HPAEs encour- aged savings by creating postal savings plans to lower transaction costs of savings for small savers. Balancing the advantages of protection against die need for compet- tion has been a major challenge for HPAE bank regulators. East Asia's banking sector structures, as well as some measures of bank efficiency (profitability and borrowing-lending margins), suggest that banks in the HPAEs have been at least as efficient as those in other developing regions. Even so, innovation and efficiency undoubtedly suffered from limits on competiton. Pardy in response, HPAEs gradually have opened their banking systems to increased domestic and fbreign competition. Gov- ernments have also promoted bond and equicy markets to increase the depth of mamring financial sectors. Several HPAES supplemented commercial banls-and met the need for long-term capitl-by creating specialized dvelopment banks. These banks fiuthered the growth of the financial sector by establishing procedures for project financing and monitoring that commercial banks then copied. In contrast to developmenr banks in many other develop- ing economies, which tend to be subject to extensive govemment med- dling, H1AE development banks controlled project selection and 350 P O L I C I ES AN Dn:rIcv monitoring independently, albeit within broad government guidelines. As a result, their bad debts were usually associated with economic down- turns rather than die willful default common in economies where gov- ernment officials are closely involved in day-to-day management. Limiting Price Distortions Despite the HIAEs' many market interventions, reiative price distor- tions were limited and indeed smaller than in most other developing cconomies. Flexible labor markets and capital markets with positive real interest rates meant that wages and interest rates reflected more closely true scarcities of labor and capital. Thus resources were generally drawn into labor-intensive production in early stages of development, shifting to capital- and knowledge-intensive activities as physical and human capital deepened. A combination of competitive real exchange rates and moderate pro- tection levels meant that domestic prices of traded goods were closer to international prices in the HPAEs than in other developing regions. The large weight of exports in total output, duty-free imports for exporters, and extensive domestic competition meant that all the HPAEs had higher proportions of their manu&acturing sectors operating at or near interna- tionl prices than the import-substituting economies of atin America, South Asia, or Sub-Saharan Africa. Absorbing Foreign Technology While none of the HPAEs except Hong Kong and Singapore was en- tirely open to international trade, al were open to foreign technology. Japan, Korea, and to a lesser degree Taiwan, China, relied heavily on li- censing, imports of machinery, and reverse engineering during their rapid growth periods. Hong Kong, Singapore, and the Southeast Asian HPAEs, in addition, welcomed direct foreign investment that came bun- dled with technical, managerial, and sometimes labor force skills. Li- censing has become more and more problematical because owners of technology are reluctant to share it, even for a fee. However, the success of the Southeast Asian HPAEs with other forms of technological acquisi- tion suggests this need not be a serious barrier to growth. The alternative path of self-reliance has litde to recommend it. Some developing economies, such as Argentina and India, have restricted cap- 35' 0 A: -X1RACLE ital goods imports to promore the growch of domestic machine-building industries. In these cases, however, domestic machinery generally falls short of international standards. Such machinery becomes yet another high-cost import substitute that inhibits exports. Limiting the Bias against Agriculure Agriculture has played a more central role in the H1IAEs than in many other low- and middle-income economies. Wide adoption of Green Revolution technology, high investment in rural infrastructure, and lim- ited direct and indirect taxation of agriculture meant that rural incomes and productivity rose more rapidly in East Asia than in other regions. As in other economies, agricultural sectors in the HPAES were a source of capital and labor for the manuficturing sector. But in East Asia these resouroes were generally pulled into manufacturing by rising wages and re- turns, rather than squeezed out of agriculture by high taxes and stagnant or declining relative incomes. As a result, rural-urban income differentials were smaller in the HPAEs than in most other developing economics. Creating Institutions to Promote Growth rT-%HE SUCCESS OF THE EAST ASIAN ECONOMIES STEMS PARTLY Sam the policies they have adopted and partly from the insti- .1trutional mechanisms they created to implement thenL AUl of the HPAEs created secure institutional environments for private invest- ment that led to very high levels of private sector-led growth. HPAE civil services range orom the highly meritocratic and insulated bureau- cracies of Japan, Korea, Singapore, and Taiwan, China, to the less effective and less insulated public administrations of Indonesia and Thailand. Nevertheless, each of these economies has a core of techno- cratic managers. In Indonesia and Thailand, their scope is limited to management of the macroeconomy. In the other HPAEs, competent civil services administer a much wider range of policy instruments. How did these economies create a reputable bureaucrac? First, pay mattered. The salaries of bureaucrats (except in Singapore) usually fel short of those for equivalent positions in the private sector, but they were sufficiently high to attracr and retain good economic maragers. 352 POLICIES` N Second, in the HPAEs with high-quality bureaucracies, rules and proce- dures governing public sector employment were institutionalized and insulated fromn political interventions. In particular, recruitment and promotion were merit based. Third, public employment was accorded high status. These factors improved the quality of the bureaucracy, dis- couraged corruption, and created an esprit de corps among civil servants that helped insulate the bureaucracy from political pressures. Each of the HPAE governments created institutions to improve com- munication with the private sector. Formal deliberation councils esrab- lished in five of the economies-Hong Kong, Japan, Korea, Malaysia, and Singapore-included government officials, journalists, labor repre- sentatives, and academics. The economic and political benefits of these councils, and of the more informal mechanisms in the other HPAEs, are impossible to measure systematically. It is likely, however, that econom- ically they improved coordination among firms and improved the flow of information between businesses and government Politically they helped establish a commitment to shared growth and reduced rent- seeking. Information sharing made it more difficult for firms t curry spe- cial favors fiom the government and for government officials to grant special concessions. Thus the deliberation councils helped check oppor- tunistic behavior. The deliberation councils in Japan, Korea, and Singa- pore also performed an important monitoring function, assessing performance at the industry or economy level and administering contests. Consultative mechanisms in Indonesia and Thailand were less for- matl, and their role in facilitating business-government communication was much more limited. Malaysia's experiment with deliberation coun- cils is quite recent but appears to hold some promise in the area of macroeconomic management. Inteirvening in Markets M OST OF THE HPAEs, ESPECIALLY THOSE IN NORTHEAST Asia, intervened in markets in an effort to hasten growth. All interventions carry with them costs, either in the form of direct fiscal costs of subsidies or forgone revenues, or in the forrn of implicit taxation of households and firms-for example, through the structure of protection or interest rates controls. One of the main char- 353 MIRACLE acteristics of interventions in the HPAEs is that in general they have been carried out within well-defined bounds limiting the implicit or explicit costs. Thus, price distortions were present but not excessive, interest rate controls generally had as benchmarks international interest rates and were binding at positive real levels. Explicit subsidies were also kept within bounds. Given the overriding importance that each of the HPAEs ascribed to macroeconomic stability, interventions that threatened to undermine thar policy fimdamenmally, were modified or abandoned-for cxuample, the heavy and chemical industries drive in Korea or the heavy industrialization push in Malaysia. These limits to intervention stand in sharp contrast ro many other developing econ- omies, where interventions have not been consistent with macro- economic discipline. Whether these interventions contributed to the rapid growth made possible by good findamentals or detracted from it is the most difficult question we have tried to answer. It is much easier to show that the HPAEs limited the costs and duration of inappropriately chosen interventions-itself an impressive achievement-than to demonstrate condusively that those interventions that were maintained over a long period accelerated growth. In this book we have evaluated three sets of policy interventions: the promotion of specific industries or industrial subsectors, directed credit, and the export-push strategy. We conclude that promotion of specific industries generally did not work and there- fore holds little promise for other developing economies. Directed credit has worked in certain situations but carries high risk. The export-push strategy has been by far the most successful of the three sets of policy intervention and holds the most promise for other devel- oping economies. Promoting Specific Industries Generally Did Not Wark Industrial policy narrowly defined-that is, attempts to achieve more rapid productivity growth by altering industrial structure-was gener- ally not successful. In Japan, Korea, Singapore, and Taiwan, China, pro- motion of specific industries had little apparent impact. Industial growdt tended to be market-confbrnning, and productivity change was not significantdy higher in promoted sectors. Although governrents in these four economies were undoubtedly trying to alter industrial structure to achieve more rapid productivity growth, with the exception of Singa- 354 POLICIES AE 1.A pore their industrial structures evolved largely in a manner consistent with market foirces and factor-intensiy-based comparative advantage. It is nort altogether surprising thar industrial policy in Japan, Korea, and Taiwan, China, tended to be market conforming. These economies, although selectively promoting capital- and knowledge-intensive indus- tries, still aimed to create profitable, internationally competitive firms. Taiwan, China, placed great emphasis on price comperitiveness in eval- uating the performance of its large public investments in intermediate and capital goods. Moreover, in Japan especially, but also in Korea, in- stitutional mechanisms that stressed cooperative relations and exchange of information betveen firms and government introduced a largc amount of market information into the process of industrial policy fbr- mulacion. And the yardstick used to evaluate industrial policies' success-mainly export performance-provided a market test of the suc- cess or failure of the policy instruments chosen. Thus the seeming success of industrial policy in these three economies probably rests nor on picking winners-that is, on the adroit selection of which industrial subsector to promote-but on the setting of export targets for pro- moted industries and the use of export performance to assess policies. Our tests may have failed to captIe one important aspect of indus- trial policy in Japan and Korea: information exchange in a competitive framework-in short, contests-may have lessened investment coordi- nation problems and resulted in earlier and better investments in what ultimately became market-conforming sectors. But among the HPAEs, onlyJapan and Korea had the instiutional capacity to organize and ref- eree the kind of contests used in formulating and implementing indus- trial policies. Despite their potential benefits, collaborative arrangements, espe- cially when combined with access to scarce resources such as foreign ex- change and cradit, have a high risk of capture by the participants. Capture was avoided-and resource allocation may have been improved-in Japan and Korea by the combination of repeated rela- tionships between business and government and the use of exports, not domestic GDP growth, as the yardstick by which the success of the in- dustrial strategy was measured. In other HPAES, such elaborate contests were not used-for example, in Hong Kong, Singapore, or Taiwan, China-or were unsuccessfil, as in the cases of the heavy industriafiza- tion drive in Malaysia or recent attempts at technological leapfrogging in Indonesia. 355 .X4 N-MIRACLE Repressing Interest Rates and Directing Credit Worked Somefimes All HPAEs controlled interesr races on deposits and loans during much oF their rapid growth. Low real interest rates on deposits and loans in- creased corporare equity by transferring resources from savers (mostly households) to borrowers (mosdy firms). At the same rime, by limiting the spread between deposit rates and loan rates, govemments prevented banks from earning excessive rents. On the downside, such controls un- doubtedly discouraged banks fiom taking on riskier, nontraditional dients. Interest rate repression was key to the HPAEs' directed-credit pro- grams. Low interest rates on loans resulted in excess demand for credit, which permitted governments to take an active role in credit rationing. This in turn gave governments strong leverage with which to influence the beihavior of firms. In this section we discuss why directed credit was relatively successfid in hie HPAES but often had large negative conse- quences elsewhere. WVe then consider whether such polidtes could be ef- fectively adopted in the increasingly open capital markets of most developing economies. Interest rate subsidies were smaller in the HPAEs than in other devel- oping economies. In several HPAES, government-directed credit served primarily to signal commercial banks and private investors which sectors enjoyed govemment support Banks and private investors then selected projects within those sectors and offered them additional funding at commercial rates. Moreover, credit was frequently allocated according to export performance. Our evidence leads us to condude that credit programs directed at exports yielded high social returns and, in the cases of Japan and Korea, other directed-credit programs also may have in- creased investment and generated imporcant spiliovers. Even within East Asia, however, when interest rate subsidies were large, or export performance criteria were not imposed, directed credit often went to in- efficient ventures and those with low social returns. The financial sector policies used by the HPAEs in the 1960s through the early 1980s took place in a world economy in which it was possible to close domestic capital narkets. Since the early 1970s, developing economies have become ever more integrated with the global capital markeL While this has reduced government? already limited abiity to restrict capital flight and speculative capital inflows, it has also given them access to less expensive capital and offered them better possibilities 356 POLICIES AND:P to hedge external shocks. At the same time, increased integration has constrained policymakers, who are finding it more and more difficult to rnaintain interest rates at variance with international markets. The developing economies' growing integration with global financial markets can be seeni in the evolurion of their stock markets. Until the mid-1980s, foreign investors' access to domestic securities in developing economies was often restricted. Since then, fourteen of the twenty largest developing-economy stock markets have become substantially more open to foreign investment. The influence of international inter- est rates on domestic rates reflecrs a similar trend toward global integra- tion. Indonesia, for example, began to open its capital accounts in 1971 and freed domestic interest rates in 1981. Since then, interest rates in In- donesia have followed international rates. Although capital outflow restrictions are generally ineffective, they do hinder the movement of some institutional capital, for example, do- mestic pension fimds. They may also hinder the flight of household and individual savings in instances when the transaction costs of sending capital abroad are greater than the incentive for doing so. Because these barriers do marer, developing economies continue to be less open fi- nancially than they are conmmercially. This can be seen in the ratio of gross capital flows to CDP compared with the ratio of exports-plus- imports to C;DP. For the aggregate of all developing economies, the former is only 13 percent, while the latter is 45 percent. Nonetheless, the limited impact of capital restrictions in deterring capital flight has persuaded a growing number of developing-economy governments to open to global financial markets. In some cases, for ex- ample, Chile and the Philippines, the removal of capital outflow regula- tions has had the paradoxical effect of stimulating a net inflow by assuring investors they can expatriate their money at will. Even where govermnents have not liberalized restrictions, the globalization of the economy, particularly increases in trade and advances in transport and telecommunications, is rapidly eroding the isolation of closed econormies. Under these conditions, the choice for developing-economy policy- makers is not between financial integration with the global market and isolation. Rather, it is between two-way integration that permits out- flows and thus facilitates inflows, on the one hand; and de facto one-way integration that consists primarily of unauthorized capital flight, on the other. The HPAEs have recognized this dcanging global environment; all 357 '&1N IR A CLE are moving-albeit at different speeds-to liberalize their financial sec- tors and integrate them more closely into international financial markets. Financial sector interventions-specifically repression of interest rates and contest-based direcEion of credit-may have contributed to rapid growth in such economies as Japan, Korea, and Taiwan., China. Bur eadi of these economies had competent and insulated bureaucracies and banks to select and monitor projects, and each applied export perfor- mance as the main yardstick for credit allocation. In addition, these gov- ernments were able during the early stages of rapid growth to close their financial markets to the outside world. Few developing economies today have the institutional resources to consistendy impose performance- based criteria for credit allocation. Moreover, the capital markers of all economies are becoming more and more integraced& Thus, successful re- pression of interest rates and contest-based credit allocation may only be possible for the very few developing economies with strong institutions and then only to a very limited degree. Export Push: A Successful Mix of Fundameitals and Interventions The export-push strareg;y-the use of findamentals and interven- tions to encourage rapid export growth-was the HPAEs' most broad based and successful application of selective interventions. Furthermore, of the many interventions tried in East Asia, those associated with their export: push hold the most promise for other developing economies. While the changing trading environment will limit the use of highly tar- geted policies-similar to those of Japan and Korea during their early export growth-the broad pro-xport stance characceristic of all eight HPAEs remains viable. Trade policies in all the HPAES (except Hong Kong) passed through an import-substitution phase with high and variable protection of domestic import substitutes. In all cases, however, policies that strongly favored the production of import substitutes to the detriment of exports were aban- doned, and HPAE governments adopted strategic pro-export policies that established a free trade regime for exporters and offered a range of other incentives for exporrs. This export-push approach provided a mechanism by which industry moved rapidly toward international best practice, de- spire highly imperfct world markets for technology. Export-push strategies were implenented in three very different ways in the HPAES. Hong Kong and Singapore established free trade regimes, 358 POLICIES AN linking their domestic prices to international prices; the export push was an outcome of the very limited size of the domestic market coupled with neutral incentives between producing for the domestic or international market. Both economics made export credit available, although they did not subsidize it, and Singapore focused iEs efforts on attracting foreign investment in exporting firms. In Japan, Korea, and Taiwan, China, incentives were essentially neutral on azverge between import substitutes and exports. But within the traded goods sector, export incentives coexisted with substantial remaining pro- tection of dte domestic market. Export incentives, moreover, were not neutral among industries or firms. There was an effort in Japan, Korea, Singapore, and Taiwan China, to promote specific exporting industries. Protection was combined with either compulsion or strong incentives to expor- In Korea, firm-specific export targets were employed; in Japan and Taiwan, China, access to subsidized export credit and undervaluarion of the currency acted as an offser to the protection of the local market. In the HPAEs that intervened selectively to promote exports, contesrs based on performance in global markets played the allocative role that is normnally ascribed to neutral exposure of both import-substituting and exporting industries to international competition. But these contest- based incentive strucnures required high governmenr institutional capa- bility. One of the keys to success of the export push in some of the HPAES, especiallyJapan and Korea, was the govermnent's ability to combine co- operation with competirion. They were able to do this first, because their civil services and public institutions were largely staffed by competent and honest civil servants, and second, because rirms and bureaucrats knew that there Nvas a single yardstick for performance: exports. Export targets provided a consistent yardstick to measure the success of market interventions. When protected sectors interfered with the exports of other sectors, exporters could successfilly seek redress. For example, domestic content rules imposed on foreign direct investors in Taiwan, China were suspended when they interfered with exports. The emphasis on exporc competitiveness gave businesses and bureaucrats a transparent and objective system to gauge the desirability of specific ac- tions. Interventions could not be made arbitrarily, as these could be ap- pealed at a higher level of government if they interfered with exports. The more recent export push-efforts of the Southeast Asian newly in- dustrializing economies (NIEs) have relied less on highly specific incen- tives and more on gradual reductions in import protection, coupled 359 S'IWN' M[RACLE with institutional support of exporters and a duty-free regime for inputs into exports. Recent strategies to attract direct foreign investment in In- donesia, Malaysia, and Thailand have also been explicitly export ori- ented. Direct foreign investment in many other developing economies is roo often directed at serving a protected domestic market and hedged by restrictions on exports. lhe Export-Push Strateg in a Changing Trading Erwiroment The early HPAE export drives-those of Hong Kong, Japan, Korea, Singapore, aiLd Taiwan, China-took place amid the cxpanding world economy of the 1950s to the mid-1970s. Economic growth in the in- dustrial economies was rapid, barriers to trade were declining, and global efforts to spur free trade under the General Agreement on Tariffs and Trade (GArr) were the order of the day. Times have changed. Eco- nomic growth has slackened in the mature economies thar traditionally provided the main markers for developing-economy exports. Regional trade groupings are becoming more and more prominent. Industrial economies have begun to retaliate agnst allegedly unflair export prac- tices by imposing quotas or countervailing duties. In such a trading en- vironment, can today's developing economies successfi&lly emulate the export-push strategies of the early HPAEs? In the following section we break this question into two parts. First, will there be adequate, open markets for developing-economy manufac- tured exports? Second, will the need for market access constrain the de- veloping cconomies' use of export-push strategies? We conclude that markets will be adequate and open and thar an export push, therefore, remains possible. However, global realities will limit developing economies' abiliry to adopt rhe more interventionist instrumcnts of ex- port promotion. The experience of both the northern HPAEs, with their mixed export-push regimes in which protection of the domestic marker coexisted widt incentives or requirements to export, and the Southeast Asian NIEs, with their gradual trade liberalizations coupled with im- provemencs in export insitutions and availability of export credit, sug- gests that economies thar are in rhe process of trade liberalization would benefit from providirg specific incentives to manufactured exports. Modest subsidies to exports could be linked, for example, to the bias against exports in the domestic economy and bound by strict time lim- its linked to the pace of trade policy reform. 36o POLICIES ANDP Moreover, if there are important externalities to exporting, a firther tilt toward exports can be justified on welfire-economic grounds, even if the trade regime is neutral between import substitutes and exports. This might argue for functional export promotion measures such as guaranteeing access to export credit, providing tax incentives, or im- proving market access for small and medium exporters. But, these mea- sures will necd to follow internationally acceptable rules to assure market access in importing economies, and must be of limited dura- tion. Fiscal discipline, moreover, wiUl require that the costs of promo- tional programs be kept in check. Why There Will Be Adequate Markets The success of the southern-rier HPAEs' export drives in the late 1970s and 1980s, along with the continued rapid growth of exports from the northern HPAEs, indicates that export drives can succeed evenr when industrial-economy markets are not expanding rapidly. In con- trast with the success of earlier export drives, which depended primarily on increased consumption in the industrial economies, the later drives have depended on consumers i industrial economies buying goods made in Asia instead of goods made elsewhere. If industrial-economy markets remain open, developing-economy exports can continue to ex- pand in this manner. Contrary to concerns in some high-income economies about a flood of exports from the developing world, these exports remain tiny relative to the size of the major, industrial economies. In 1980, developing-economy exports of manufactured goods accounted for just 2.4 percent of total manufactured goods consumption in the Eu- ropean Community (Ec), North America, and Japan. By 1988, despite the success of Asian and other exporters, this had grown to just 3.1 percenL Even if exports from all developing economies had grown as fast as those from Korea during this period, their total share would srill have amounted to only 3.7 percent of total consumption by 1988 (World Bank 1992b). The scope for export expansion remains substantial. Indeed, indcid- ual developing economies, particularly smaller economies currently contemplating an export-led expansion, could safely assume that de- mand for their products is infinitely elastic This is even true for Hong Kong's exports, which in 1990 accounted for a hefty 2.4 percent of 36i RASW&NMIRACLE world merchandise trade (Riedel 1988). Emerging exporters are un- likely to approach this level for many years to come. In 1989, for exam- ple, Sub-Saharan Africa's exporcs of manufactured goods amounted to just 0.43 percent of world exports (United Nations 1992). Even if industrial economy markets gradually become less accessiblc, changes in the global trading environment are creating new opportuni- ties for South-South trade. As recently as the mid-1980s, import- substitution strategies in most developing economies forced developing-economy exporters to rely heavily on industrial-economy mark-ets. Since then, formerly closed economies in Asia, Sub-Saharan Africa, the Middle East, and Latin America have responded to debt problems, slow growth, and the need to raise fresh capital by becom- ing more and more outward oriented-and therefore more open to imports. They have assumed greater obligations under the GA1T and through numerous regional trading arrangements.' Under these cir- cumstances, the potential for Soudt-South trade has increased considerably. But while global markets are growing, they are also becoming more and more competitve. Each economy seeking to emulate the East Asian axport-push model will encounter difficulties because many other economies are trying to do the same. For example, Collins and Rodrik (1991) estimate that Eastern Europe and the republics of the former Soviet Union will increase their share of world trade from 10 to 23 percent, relying initially on agricultural conmnodities and labor- intensive goods that compete direcdy with the products of other devel- oping economies. Despite this heightened competition, several trends suggest that ex- panding global mnarkets will remain open to developing-economy man- ufactured exports. Onc of these is the continuing effort to open markets under the auspices of the GAIT (see box 7.1). The other is the growing role of regional trading blocs, which present a major opportu- nity for increased global trade (see box 7.2). Who provides the mnarket if all economies push exports? The dramatic growth of trade within Asia makes clear that export-oriented economies create markets for im- port. To compete internationally, exporcers must meet global stan- dards of quality and price, for which they need access to technology, capital, and intermediate inputs at world prices. As a resulr, production in Asia has become more and more international, with capital goods coming from Japan, labor-intensive assembly being perfbrmed in low- 362. POLICIES AN: D. Pb~~ ~~~~- er' he;| G s7fFiTr Box 7.1 Increased Access under the GAiT. .'THE GA7rT HAS BEEN INSTRUMENTAL IN FACITATING FREER trade. Since the end of World War II, tariff rates around the world havc plummeted. Among the major induistrial economics, weighred- -average tanff lcevls have been lowered to blow 5 percent. The cur- rent Uruguay Round of trade negotiations is addressing - .~~~~~~~o . s is ad .. ci.r. .V.i . *Increasing market access, inudingmore-open trade in agricul- ru.-- s texties, ad na l rsourcs and e remo or reuc- -. tion ofrmainin tariff and nonrariffbarriers a Extending the GATT to new areas, principilly tirade in services, standards, and the enforcement of intellectual property rights, and trade-related investment requirements imposed by govern- ; -rnents .: - - Strengthening CAT rules, especially those covering antidump- ing subsidization, product standards, import licensing, "safe- guards" (temporary import protection)3 ard dispute selemenr In short, the Uruguay Round is covering a- broai4 ambitio agenda. Even at this late dare iris impossible to know how successfill the round will be. Success could offer inip&rmnt-ropportuiities fbr :developing-economy exports, particularly in sucb arcas as agriculture aland tropical products. Progress in eliminating the Muiltifibre Arrangement would also bring major benefits; particularly to new en--. -trans into textiles and apparel-alth ough here some economies :- -might lose eisting,rents. -Ar the same time, however, a successful Uruguay Round miy erode prefrntial reaent thusE fr accorded ..som developing economies. Developing-economy governments may. 'be: obligedt to ughihen-their7:nforcement of intellectal propeAiy - rights, while easing rade-related investment. reqwremens They' :could also -fi;e. touglie rules of oigm :and anricircumvention mnea- -.sures-concerning dumping alhgarions. On the whole, hwe 6the ?WUruguay >Round' holds considerable pornise forte, develon economies.; - --Conversy,developing CononMies -wouldlose imporntopporni- -nies if the round fails. Thcsc would be pardq. lary seriousmin agn- ..:culture because the global natuie of gn&ilrure markets,makes _ *regiosounsineffectve Devopi economis udao be- - . adversely ffecd byie increased use of unilateral and bilateral re- .-stictions in te' event oh GArile. I '. U 363 ASAM- IRACLE Box:7.2 Regional.Trading Blocs ' - ''ROGRESSTOWARD GLOBALWINTEGRATONCOULD capital flows. Stronger regional integraton need not .be reversed iffdworldconory splits into regional be' assdciared with higher exteral barriers. Indeed, trading blocs, each centered on a major currency and regional blocs could have positive global effects, pro- dlosed ro' outsiders (Thuirow .1992). Some .observeis -vided the blocs remain open to trade from outside. worry that the emerging regional trading blocs will 'Deeper integrnwion within Europe, for example, will imposc-subtle and nor-so-subtle protectionist' mea- &cilitate trade with the rest of thc world. A common esurs'. Inboth teEurpean Comrnunity; (EC) and -set ofsrandards nakes it easier forallwho wish to se[l the :iNorth Ainerican'.Free Trade. Ara (NAFMA), for in Europenor just insiders. Tough rules to prevent ex'ample, fins' are already lobbying Forules olisnit -governments from subsidizing 'domestic firms .aid all the marker 'access of extra-regionial competiors. 'competitors, not only those located in the European Typically these rules require that products entering Community. one memrber economy contain ai inimuma regional - Eierging regional arrangements may provide an ' content to quaJr for entry to the markets of oiher . incentive -for some economies to. open, more rapidly. , members. (In--the,United Stares-Canada Free Trade than they would havcr otherwise. For example, the ,Area, a 50 percent regional value-added content re- formation of the Ecwas an imortant inipulsc for the . quirement was required for eligibility for regional Kennedy Round (L.wrence 1993); As Mexico access. .In the NAFTA,tihere a`re special rules. for auto- moved into the' NAFTA negotiations, it simukmne- mobiles and tetiles.`The textile provision is particu- . ously made overture toward the Pacific and towaid .lady protectionist-it requires triple uansformation, Cential and South America. Mexico is 'now seeking such that'finished:foihing.products have'to becuiit :to join the OECD, is negotiating to form a free trade ..-"..'.and sewn flom &i bric spun rom Noth American. area wthVenezuelaandCie,andhas. signed agree- - fibers. This could resiilt in considerable trade diver-. :mcnts for freer trade with seven1 Central American sion.) Regiona:l blocs'may also,-tighten fbrcemen economes. . -. '.ofrules against -Ddeiiumping. as: another means of,' oDevelopingeconuisib' differently ,-prorecion. .by th e n of rgn b -Eco tha ,- .; -.But con.erns about regional .protectionist -enden- currendy enjoy prefrrentiaJacc to'the,U.S. and EC des are easily overstaie. Unlikc-their predessors,m..arkets under schem such as thc gene ed S th: en5wblcs go beyondliberaizationoftradeto m- -.ten-of prefer th Cabbe Ina ; dudetlibealizdion of investment Lwrince 1993).- and the Lom-nConvention are likly to 'see te vale Thi- in-rai necessitates the libealinidonf naional w -of, thee prege erod& 1f- hoever, regional .production standardsthat would otherwise raise-the; arrangements further multilaibalitidn,.thy' .'.-cc'.c of=.r'*onlly nmtegt prodcton Moreover, wl . ot new opportunities r.traedaif bl-s" -in each'insrance where-regional conomic gups-are succeed ,in stimilatig. e w of teiembcrs, em. ''erging, hie' movation -has b'e'a at.re LrOato* .hose outside.the regions will.n increased expor further.access and internationl :opportuities. - - wage economies such as Indonesia and Thailand, and more sophisti- cated operations such as design, marketing, and finance being provided from Hong Kong, Singapore, and Taiwan, China. Thus despite its rep- utation for protection, Asia has generated particularly rapid increases in rrade. 364 POLICIES AN RG liT How the Need for Market Access Will Consbtain intenweuions Developing economics seeking to increase trade and attract invest- ment will experience greater pressures to open their economies and bring their practices more closely in l ne with those in industrial economies. Those that attempt to use more interventionist versions of che export-push model will risk retaliation from industrial-economy markets or punishment under the GATT. For some policy options, such as financial repression, increasing integration with global markets will limit the scope for intervention. Over the years, developing economics have daimed and received ex- emptions from GATT rules (Whalley 1990). By invoking Article XII, to safeguard their balance of payments, or Artide XVIII, which allowvs pro- motion of infant industries, developing economies have been able to es- cape GAIT disciplines more or less indefinitely. Thus developing economies have been free to implement trade policies of their choice at bome while benefiting from the openness of industrial economies. But as GATT shifts from a narrow focus on tariffs to a broad approach that seeks agreements on rules, developing economies will find it more and more difficult to hang onto exemptions. Under Uruguay Round pro- posals, rules governing intellectual property, subsidies, countervailing duties, and trade-related investment would eventually be the same for industrial and developing economies. Although developing economies would have up to eight years to bring their practices into line, the trade obligations of developing economies clearly are increasing. Changing rules about dumping illustrate the case. Industrial economies are in- creasingly applying antidumping and countervailing duty measures to developing economy exports (Lawrence 1991). These measures will make it more difficult to follow the example of Japan and Korea, which protected domestic industry in order to subsidize exports. Developing economies that seek access to the emerging regional trading blocs may find that it comes tied to increased obligations that could inhibit adoption of the interventionist policies. For example, in- terventionist versions of the export-push model could not be imple- mented within the EC with its stringent rules on stare aid and its provisions allowing for the fiee movement of capital and for the com- munity to implement competition policies. Moreover, no economy ex- posed to European competition could operate an effective infant industry strategy. 365 MASVAN IRACLE Does this imply the Asian export-push model is a historic relic? Clearly not. Indeed, an essential feature of the East Asian export push has beeat the ability to adapt to the global environment. The world mar- ket continues to present immense opportunities for those who export. New opportunities emerge as economies everywhere libemlize and the NIEs shift to the production of more advanced products. Moreover, Asian exporters have prospered during the past two decades despite a de- cline in global growth and a rise in protectionist actions. Exports can be promoted by a var.iet of means that are consistent with developing economies' emerging obligations for market aco- rs and limited subsidies. Improved institutions-mak-ing customs services, duty drawbacks, and fiee trade status for exportcrs and their suppliers work well-are all consistent with agreements on intemational tradc and arc effective mechanisms for export promotion. Access to export credit, while more controversial, was an ePffctive instrument in all of the -PAEs and remains feasible for other developing economies. . * C We began this book with the objective of understanding East Asia's success. We found that the diversity of experience, the variety of institu- tions, and the great variation in policies among the high-performing Asian economies means that there is no East Asian model of rapid growth with equity. Rather, each ofthe eight economies we studied used various combinations of polides at different times to perform the func- dons needed for rapid growth: rapid accumulation, efficient allocation, and high rates of productivity improvement. Most of the policies that the HPAES used reflected sound economic fundamentals: they enhanced the working of markets, helped prices communicate information about relative economic costs, and fostered competitive discipline. Some economies-notably Japan, Korea, arid Taiwan, China-went beyond fundamentals and intervened in markes with industrial, trade, and financial sector policies. On balance, some of these interventions con- trbuted to their extraondinary growth, but this was only possible because of highly unusual historical and institutional circumstances. The govern- ments in these economies were more successfi than most in combining the benefits of economic cooperation with the benefits of vigorous com- petidion. They did this by creating contests with exports as the yardsick of success, with subsidized credit and other government-controlled favors as prizes and with government officials as competent and usually honest ref- 366 POLICIES AND PRAGMAS erees In other economies, induding some of the HPAES, similar interven- tionist policies have uiled, largely because governments offered incentives without a clear link to economic perlormance and strong institutional support. They offered prizes but lacked the rules and referees. What can other developing economies learn from the Easr Asian mir- ade? While there is no recipe for success, there are some positive lessons: keep the macroeconomy stable; focus on early education; do not neglect agriculture; use banks to build a sound financial system; be open to for- eign ideas and technology; and let relative prices reflect economic scarci- ties. And there are some negative ones: promoting specific industries or attempting to leap srages of technological development will generally fAil; strongly negative real interest rates and large subsidies to borrowers debilitate the financial system; and directing credit without adequate monitoring and selection of borrowers distorts allocation. Finally, we found that a successful export push, whether it results from an open economy and strong econornic fundamentls, or from a combination of strong findamentals and prudently chosen interventions, offers- high economic gains. Of all the interventions we surveyed, those to promote exports were the most readily compatible with a wide diversity of eco- noamic circumstances. East Asia's own responses to changing domestic and international cir- cumstances put these lessons in perspective. T7h HPAEs are themselves in- volved in a continuing process of reform, adapting poliy inscruments and institutions to achieve the objectives of continued growth with equity. In many cases these reforrns involve reducing& modifring, or abandoning policy instruments thar were judged to have succeeded in the past. Korea's financial sector reform, Indonesia's trade reforms, Thai- land's promotion of breign investment, and Malaysia's privatization pro- grams are cases in point. The outcome of these initiatives will provide further valuable lessons on how successful policy instruments shift over time, as the relative roles of markets, the public administration, and the private sector change in response to economic and social development The experience of the HPAEs broadens our understanding of the range of policies that contribute to rapid growth. it also teaches us that will- *ingness to experiment and to adapt policies to changing crcumstances is a key elemenit in economic success. What we have not discovered fully is why the governments of these economies have been more willing and better able than others to experiment and adap . answers go beyond eco- nomics to indude the study of institutions and the related fields of pol- 367 1rtCA'*T3 M 1 R A CL E icics, history, and culture. Taking such large realities into account com- plicates rather than simplifies the cask of development. The challenge for policynakers becomes twofold: to uise policies compatible with ax- istillg institutions and to avoid policies tha. exceed the economy7s insti- tutional capacity, while at the same time improving institutions in order to make a wider range of policy options available. Note 1. In the past, regional armngements were motivated Venezuela, with Ecuador and Peru slated to join); the by the desire to pursue import-subsrirution policies on a Central American Common Marker (Costa Rica, El Sal- larger scale and were therefore doomed to failure; the vador, Guatemala, Honduras, Nicaragua, and Panama); morc recent initiatives are motivated by a desire to con- Caricom; and a Free trade area between Chile and Mexico soildate more outward-oriented strategies. Subregional that aims at reducing tariffs to zero by 1998. While arrangements are flourishing. In Latn America, for exam- progress coward formal arrngements witrin Asia have ple, these include the North American Free Trade Area; been less apparent (although the free trade area within the Mercosur (Argentina, Brazil, Paraguay, and Uruguay), Association of Southeast Asian Nations is an exceprion), which aims at a com mon market for goods and services by incra-Asian trade has increased rapidly. 1994; the Andean Pact (Bolivia, Colombia, and 368 S Bibliographic Note THIS REPORT HAS DIRAWN ON A WIDE RANGE OF MA- terials, including speifically commissioned backgrund studies, outputs of related World Bank-sponsored re- search, economic and sector work carried out by Bank operational staff; and the errensive academic literature _L on the East Asian economies. Citations in the text and footnotes refer to specific sources used. The principal sources and specific contributions to individual chapters arc discussed below. Lists of back- ground and other commissioned papers and refrences for works cited are at the end of this note. Many members of the World Bank Staff assisted in the design of the research and made helpfid comments on one or more drafts of the report. They indude: Yoshiake Abe, Ramgopal Agarwala, Harold Alderman, Paul Armington, Ahmad Ahsan, Mark Baird, Gerard Caprio, Ajay Chhihber, Yoon-Je Cho, Jaime de Melo, Shanta Devarajan, Ron Dun- can, Wllliam Easterly, Gunnar Eskeland, Delfin Go, Jeffery Hammer, James Hansen, Gregory Ingram, Paul Isenman, EstelleJames, Emmanuel Jimenez, Christne Jones, Koji Kashiwaya, Homi Kharas, Miguel Kiguel, Elizabeth King, Michael Kline, Brian Levy, Samuel Lieberman, Will Martin, Deepak Mazumdar, Saha Meyanathan, Vikram Nehru, Arvind Panagariya, Martin RIavallion, D. C. Rao, Anandarup Ray, Setsuya Sato, Andrew Sheng, Lyn Squire, Vinaya Svaroop, Shekhar Shah, Jee-Peng Tan, Jacques Van der Gaag, Dominique van de Walle, Dimirri Vrtms, Andrew Warner, David Wheeler, Shahid Yusuf, and Heng-Fu Zou. Those outside the World Bank who contributed with comments and material indude: Alice Amsden, New School for Social Research; Masahiko Aoki, Sranfrd University; Jere Behrman, University of Pennsylvania; Robert assen, Oxford University; Gay Fields, Coomell University, Richard 369 ,JG A1':ipKNOTE Freeman, Harvad University; Richard Hanushek, University of Rochcster; Shig e.lruikawa, Ayoama Gakuin Universitr, Hirohisa Koharna, Univer- siy of Shizuoka; Kazuo Koike, Hosei University, Lawrence J. Lau, Stanford University; David Lindauer, Vassar College; Kazushi Ohkawa, International Development Center ofJapan; Kenichi Ohno, University ofTsukuba; Moi- ses Naim, Carnegie Endowment fbr Intemational Peace; Peter Petri, Bran- deis University; Gustav Ranis, Yale University; Changyong Rhee, University of Rochester, Helmut Reisen, OECD; Hilton Root, Sranford University; 11 Sakong, Institute for International Economics, Seoul; Herman van der Tak; Shujiro Uraca, Waseda University; Rovert Wade, Institute for Advanced Srudy, Princeton; and Toni Yanagihara, Hosei University Chapter 1 The data for this chapter are taken fom. the data base compiled for a research project, How Do National Policies Affect Long-Term Growth? directed byWiliamr Easterly and Ross Levine. It is an expanded version of the data contained in Summers and Heston (1991). A summary of the results of the resch projcct are contained in a symposium issue of the Journal of Monetary Economics (forthcoming). Additional data are drawn from the World Bank Economic and Social Data Base (BESD) and from the World Development Report data base. The TF estimates were made possible bv access to preliminary estimates of constant price capi- tal stock series developed by the Banks International Economics De- partment, under the direction of Vikram Nehru (Nehru 1993). Stanley Fischer generously provided his country-specific estimates of TFP growth, which underlay the results reported in Fischer (1993). Maurice Schiff (1993) provided calculations of agricultural incen- tives in East and South Asia, based on the methods developed in Sciff and Valdes (1992). The discussion of export performance is drawn pri- marily from Bhattacharya and Page (1992). Analysis of the demo- graphic transition is drawn primarily from Birdsall and Sabot (1993). Ed Campos (1993) provided most of the background work on income distribution. The discussion of private investment draws heavily on Madarassay and Pfeffermann (1992). Box 1.1 draws on Birdsall and Sabot (1993). Box 1.2 draws on Pack (1993) and Pack and Page (1993). Substantial research on the economic tmnsition in China has been conducted by the Macroeconomics and Transitional Economies Divi- sion of the World Bank's Policy Research Department. This research ef- fort which is led by 1. J. Sirwh is summarized in Chen, Jeffrrson, and 370 BIBLiO' Singh (1992). The Asian Miracle Research Project has also funded re- search by professor Shigeru Ishikawa and his associates on the rapid growth of China's coastal provinces. Preliminary outputs of that research will be available in the fall of 1993. Chapter 2 This chapter draws primarily on Pack (1993), Page and Petri (1993), and Stiglitz (1993). The description of contests also draws on Leipziger and Kim (1993), Uy (1993), Okuno-Fujiwara (1993), and Ito (1993). WorldDevelopment Report 1991 provides the basis for the discussion of the market-friendly approach. Box 2.1 draws on World DevelopmentRe- port 1991. Box 2.2 draws on a paper prepared by Nancy Birdsall and John Page for the Colloquium on Lessons of East Asia, February 1993. Box 2.3 draws on material provided by Jaime de Meld. Helpful discus- sions with Vimod Thomas, Robert Wade, and partcipants in the Devel- opment Economics Study Group, Tokyo, and the Trade and Development Seminar at Stanford University contributed to the chapter. Chapter 3 This chapter draws on Bhartarcharya and Pangestu (1993); Chris- A.. Dar. Sa±nvwdla, andlu 'vicnyanona i993; Corcien (I993J; Dahlman and Sananikone (1993); Leipziger and Kim (1993); and Salleh, Yeah, and Meyanadhan (1993). Many of the data on macro- economic perfomance are drawn from a research project, How Do Na- tional Policies Affect Long-Term Growth? directed by William Easterly and Ross Levine. The concept of "export push" is due to Coln Bradford (1990). Villiam Easterly, Lant Pritchett, and Michael Walton provided major inputs. Box 3.1 draws on Corden (1993). William Easterly pre- pared box 3.2. Box 3.3 is based on Ira C. Magaziner and Mark Patinkin, The Silent War (New Yorlc Random House, 1988), by permission. Box 3.4 was prepared by Varuni Dayaratna. Chapter4 Chiah Siow Yue, Oh-Hyun Chang, Chaianan Samudavanija, Hadi Soesastro, and Bruce Tolentino provided information and analysis for their countries. Hilton Root and Mhary Shirley made imporrant contri- butions. Discussion of the role uv small and medium-size enterprises hac drawn heavily on the results of a research project, Support Systems for Small and Medium Enterprises in East Asia, directed by Brian Levy. 37' R&fl(c OTE Chapter 5 Material on human capital accumulation is drawn primarily from Birdsall and Sabot (1993). Robert Gassen provided the material on vocational training. Data and analysis on savings and investment are drawn primarily fiom Uy (1993), Stiglitz (1993a, 1993b), and Sriglitz and Murdock (1993). YoonJe Cho, Dimirri Vitras, and Andrew Sheng made important contributions to the analysis of financial repression, based on the results of a research project, Effectiveness of Credit Poli- cies in East Asia. The section on tax policy draws on Shah (forchcom- ing) and on preliminary results ofa research projecr, Tax Policy and Tax Administration in East Asia. Discussion of the relative price of invest- ment goods is based on data drawn from Pfeffermann and Madarassay (1992). Lant Pritchett made important contributions to the chapter. Box 5.1 was writen by Myriarn Quispe. Box 5.2 was prepared by Leora Friedberg. Andrew Sheng provided the material on which box 5.3 is based. Box 5A was writren by Jennifer Keller. Chazer 6 The discussion of labor market performance is based primarily on BirdsalU and Sabot (1993) and draws on contributions by Felds (1993), Freeman (1993), and Ranis (1993). The discussion of the capitl mar- ker draws primariJy on Uy (1993), SEiglitz (1993a, 1 993b), and on con- tributions by Andrew Sheng, Yoon-Je Cho, and Dimitri Vittas. The sections on industrial policy and exports arc based on Pack (1993) and Pack and Page (1993). Data on axport performance are drawn from the BESD and World Development Report 1991 data bases. Discussion at the 1993 Canegie-Rochester Conference on Public Policy were helpful in improving the sections on export extemralities and growth. Box 6.1 was written by Brian Levy. Box 6.2 was prepared by Leora Friedbeg. Box 6.3 draws on a recently cornpleted internal World Bank report on Indonesias industrial seaor. Box 6.4 was prepared by Leora Friedberg. Box 6.5 was prepared by Danny Leipziger. Chapter 7 Discussions of the international trading environment and boxes 7.1 and 7.2 are drawn from Lawrence (1993). Discussion of the interna- tional capital market is largely based on Global Economic Prospects and the Developing Countries (World Bank 1993a). Vinod Thomas made a number of helpful comments. 372. BIBLI9CAi Background Papers Samuudavanija, Chaianan. "High Speed Growth and High Pertfrmance in a Technocratic Polity: The Thai Case." Behrman, Jere R., and Ryan Schneider. "An International Shah, Shekhar, and Subodh C. Marhur. 1992. "Sources Perspective on Schooling Invesrment in the Last Quar- of Labor Markecr Flexibility in East Asia: A Review of ter Century in Some Fast-Growing Eastern and Selected Issues." Southeastern Countries." Stiglitz, Joseph E. "Sorne Lessonis from theAsian Miracle." Birdsall, Nancy, and Richard H. Sabot. "Virtuous Cirdes: Uy, Marilou. "The Role of Capital Markets and the Role Human Capital Growth and Equity in East Asia." of Government" Campos, Ed. "The Instimtional Foundations of High- Speed Growch in the High-Performing Asian Econ- Studies of the High-Peforming Asian Economies omies: Part I, Insulation Mechanisms and Public Sectr-Private Sector Relations." Thesepapers will bepublished shortly a Wartld Bank Coazn- . "The Institutional Foundations of High-Speed try Studfies u rsh;lje series dteThe Lessons of Easr Asia. Growth in dte High-PerforningAsian Economies Part Bhatracharya, Amar, and Man Pangesrn. 'Indonesia: De- II, The Bureauacy as an Instruent of Development." velopment Transformation since 1965 and the Role of Chia, Siow YuC. 'The Role of Institutions in Singaporc's Public Policy." Economic Success." Chau, Leung Chuen. "Hong Kongj A Unique Case of Corden, W. Max. "Seven Asian Miracle Economics: Development." Overview of Macroeconomic Policies.- Christensen, Scott, David Dollar, Ammar Siamwvlla, and Dekle, Robert. "Raising Saving Rates: Lessons from the Pakorn Vichyanond. "Thailand. The Insirtuional and Japanese Experiencc." Political Underpinnings of Growth." Fields, Gary S. "Changing Labor Market Conditions and Dahlmnan, Carl, and Ousa Sananik-one. 'Taiwan, China: Economic Developmenr in Hong Kong, Korea, Singa- Policies and Institutions for Rapid Growth." pore and Taiwan.' Kim, Kiwan, and Danny Leipziger. "Korea: A Case of Freeman, Richard. "Does Suppression of Labor Con- Government-Led Development." tribute to Economic Success? Labor Relations and Leipziger, Danny, and Vinod Thomas. "Country Lessons Markers in East Asia.' from East Asia." Kim, Chang-Shik "Government-Business Relationship Petri, Peter. "Government Policy and East Asia's Success." in Korea: Selective Case Studies." Salleh, Ismnael, Kim Leng Yeah, and Saha Dhevan "Trade Policy and Industrial Transformation in Meyanathan. "Malaysia: Growth, Equity, and Struc- Korea." rural Transformation." Lawrence, Robert Z. "The Global Environment for the Soon, Teck-Won, and C. Suan Tan. "Singapore: Public East Asian Model." Policy and Economic Development? Murdock, K., and Joseph E. Sdiglim "The Effect of Fi- Thomas, Vinod, and Yan Wang. Government PoLicies nancial Repression in an Economy with Posirive Real and Productivity Growth: Is East Asia an Exception?" Rates." Pack, Howard. "Induscrial and Trade Policies in the Studies of Public Policy in Japan High-Performing Asian Economies." Pack, Howard, and John M. Page, Jr. "Accumulaton, Er- These papers were presented at a Seminar on Public Poliy ports and Growth in the High-Performing Asian duringJapanf?RapidGrowth. Tolyo,Jzanry 1,993. Economies.' Horiuchi, Akiyoshi. "Government Control uJ Financial Page, John M.,Jr., and PeterA Pctd. Productivity Change Mechanism in Postwar Japan: A Tentative Report." and Strategic Growth Policy in the Asian Miracle." Itoh, Motoshige. "Industrial Policy and Corporate Ranis, Gustav. "Labor Markets, Human Capital and De- Growth in the Automobile Indusaryr Japan's Postwar velopment Performance in East Asia-" Experience." 373 IflCNTE Koike, Kazuo. "Human Resource Development Systems Asian Wall Street jolrnaL 1993. 'Matsushima Builds Ex- in Japan.' port Base in Malaysia." May 18. Kuroda, Masahiro, Takanobu Nakajima, and Kanji Yosh- Aw, Bee-Yan, and Amy R Hwang. 1993. "Productiviry ioka. "Industrial Policy and Economic Growth in and the Export Marker: A Firm Level Analysis." Penn- Japan 1960-1985: Preliminary Report." sylvania State University, Stare College, Pa. Kuroda, Masahiro, and Shoko Negishi. "Comparison of Aw, Bec-Yan, and Hong W. Tan. 1993. "Training, Tech- the Srructunml Changes in Asian Countries Outline." nological Capaciry, and Firm Level Productivity." Maki, Atsushi. "The Relation benveen Household Saving Working Paper. World Bank, Private Secror Develop- Rate and Economic Growth: Consumer Stimulus to ment Department, Washington, D.C. Purchase Durable Goods." Balassa, Bela 1982. Development Strteies in Semi- Ohno, Kenichi. 'Dynamism ofJapanese Manufacturing. Industrializing Economies. BalLiliore, Md.: Johns Evidence from the Postwar Period." Hopkins University Press. Okuno-Fujiwara, Masahiro. 'Government Business Rela- - 1991. Eco nomic Polcies in the Pacifx Area Developing uionship in Japan: A Comparative Institutional Analysis.' Countrie.t New York. New York University Press. Shimomura, Yasutami. "Industrial Policy in the Rapid Balass, Bela, and Associates. 1971. The Structure of Pro- Growth Periodc Evidence from Early Writings." tection in Developing Countries. Baluimore, Md.: Johns Teranishi, Jurn. "Emergence and Establishment of the Fi- Hopkins University Press. nancial Svsrem in Postwar japan: Government Inter- Balassa, Bela, and Marcus Noland. 1988. Japan and the ven;ion, Indirect Financing and the Corporate World Fconomy. Washington, D.C.: Institute for In- Monitorinp System." ternational Economics. Ueda, Kazuo. "The Industrial Bank of Japan and Public TheBanker. 1990 Uuly): 155-85. Financial Instirutions." Barro, Robert J. 1989. 'A Cross-Country Study of Urata, Shujiro. "Technological Progress by Small and Growth, Savings, and Government? NBER Working Medium Enterprises in Japan.' Paper 285 5, Cambridge, Mass. Yoshikawa, Hiroshi. 'Demand-Led Growth and Income _. 1990. "Govenmment Spending in a Simple Model of Distribution: The Posnvariapanese Experience." Endogenous Growth." Journal oflolitical Economy 98 Yoshino, Naoyuki. 'The Low Interest Policy and Eco- (5, pt2): SI103-25. nomic Growth ofjapan. . 1991. 'Economic Growth in a Cross Section of Countries." Quarterly Journal of Economics 105(2): References 407-43. Barro, Robercj., and Jong-Wha Leer 1993. 'Incecmaonal Comparisons of Educational Artainment." Paper pre- Aheuwal, DcherOJud. 1985. dssa . sented at a conference on How Do National Policies New Delhi: Oxford University Press. Affec Long-Run Growth? World Bank, Washingron, Aigner, D., and P. Schmidr, eds. 1980. "Specification and D.C. February. Estimation of Frontier Production, Profir, and Cost Behrman, Jere P. 1987. 'Scholing in Developing Coun- Functions," a special issue of Journal of Eoometrics rries: Which Countries Are the Under- and Over- 13:1-138. Achievers and What Is the Schooling Impact?" Amsden, Alice H. 1989. Asia's Next Giant- South Korea Economia ofEducaSn Review 6(2): 111-28. and Late tdusriaLzaeon. New Yorlc Oxford Univer- Behrman, Jere RP, and Ryan Schneider. 1991. 'How Do sity Press. Pakistani Schooling Invesunents Compare wirh Thosc Ando, Albert, and Franco Modigliani. 1963. 'The Life of Other Developing Countries?" Williams College, Cycle Hypothesis of Savings: Aggregate Implications Williamstown, Mass. and Tesrts."American Economic Review 53(1). . I992. "An International Perspective on Schooling Arrow, Kenneth J. 1969. "Classificatory Notes on the Pro- Investment in the Last Quarter Century in Some duction and Transmission of Technological Knowl- Fast-Growing Eastem and Southeastern Countries." edge." American Economic Retiew 59(2): 29-35. Background paper for The ELatAsian Miracle. World 374 BIBLIO-GR Bank, Policy Research Department, Washington, . 1993b. "Virtuous Circles: Human Capital Growth D.C. and Equity in East Asia." Background paper for The Bell, Clive, Peter Hazell, and Roger Slade. 1982. Projecr EastAsian Miracle. World Bank, Policy Research Dc- Evaluation in Regional Perspective. Baltimore, Md.: partment, Washington, D.C. Johns Hopkins University Press. Black, Stephen L., and Lance Taylor. 1974. "Practical Berry, R. Albert, and William Cline. 1979. Agrarian General Equilibrium Estimation of Resource Pulls StntutreandProdiuctiviy in Developing Counzries Bal- undcr Trade Liberalization." joUrnal of International timore, Md.: Johns Hopkins University Press. Economics 4(1): 37-58. Berry, R Albert, and Brian Levy. 1993. "Indonesia's Small Blomstrom, Magnus. 1989. Foreign investments and and Mvledium Industrial Exporters and Their Support Spillovers London: Roudedge Systems." Report prepared for the World Bank, Policy Blumenthal, 1. 1970. "Savings in Postwar Japan." East Research Departmcnr, Washington, D.C. Asian Research Center, Harvard University, Cam- Berry, R. Albert, and Richard H. Sabor. 1981. "Lubor bridge, Mass. Market Perfbrmance in Developing Countries: A Siur- Boadwvay, Robin, Dale Chua, and Frank Flatters. Forth- vey." In Paul Streeten and RJolly, eds., RecentIsues in coming. "Investment Incentives and the Corporate World Development Oxford: Pergamon Press. Tax Svstcm in Malaysia." In Anwar Shah, ed., Fiscal Bhagwari. Jagdish N. 1978. Anatomy and Consequencas of Incentivesfrr investment in Developing Countries. New Exchange Control Regimes. Cambri"-e, Mass.: York: Oxford University Press. Ballinger Press, for the National Bureau of Economic Boneili, R. 1992. "Growth and Productivirv in Brazilian Rescarch. Industries: Impacts of Trade Orintation." Journal of Bhagwati,Jagdish N., andT. N. Srinivasan. 1979. 'Trade Development&Economics39(1): 85-111. Policy and Development.' In Rudiger Dombusch and Borrus, Michael, Laura D'Andrea Tyson, and John Zvs- Jacob A. Frankel, eds., International Economic Poliey. man. 1986. "Cr- ir Policy and International Comnpe- Baltimore, Md.: Johns Hopkins Universiry Press. tition." In PawI I rugman, ed., Srategic Trade Policy Bblacachatya, Arnar, and John lvi. Page, jr. 1992. "Ad- and the New International Economics. Cambridge, justnnent, investment; and Growth in High-Perform- Mass.: MIT Press. ingAsian Economies." Paperpresented araconference Boskin, MichaelJ., and LaurenceJ. Lau. 1992. "Inrerna- on the Nev Development Economics in Global Con- tional and Intertemporal Comparison of Productive text, American University in Cairo, May 28-30. Efficiency: An Application of the Meta-Producton Bharacharya, Amar, and Mari Pangestu. 1993. Indonesist Function Approach to the Group of Five Countries." Development Transfirmation since 1965 and the Roae of Economi Studies Qucarterly 43(4): 298-312. Public Policy. World Bank Country Study, Washing- Bradford, Colin L, Jr. 1986. "East Asian Models: Myths ton, D.C. and Lessons." In John P. Lewis and Valeriana Kallab, Biggs, Tyler, and Brian Levy. 1988. "Strategic Inrerven- eds., Development Strates Reconsidered U.S. Third don and the Political Economy of Industrial Policy in World Policy Perspectives 5. Washington, D.C.: Developing Countries." In Dwight Perkins aind Overseas Development Council. Michael Romer, eds., Refirming Economic Systems in . 990. Policy Interventions and Markets: Develop- Developing Countries. Cambridge, Mass.: Harvard Jn- ment Strategy Typologies and Policy Options." In Gary sticute for International Development. Gereffi and Donald L. Wyman, eds., Manufacturing Birdsall, Nancy, Dav.d Ross, and Richard H. Saborz Miracles: Paths ofindurialization in Latin America and 1993. "Underinvestment in Education: How Much EastAsia. Princeton, N.J.: Princeton University Press. Growth Has Pakistan Forgone?" Paper presented at Brimble, PeterJohn. 1993. "Industrial Development and the Ninth Annual General Meetng of the Pakistan Productivity Change in Thailand." Ph.D. diss., Johns C..ociety of Development Economist, Islamnabad, Pak- Hopkins University, Baltimore, Md. ismt, January. Bruton, Henry J., G. Abeysekera, N. Sanderatne, and Z Birdsall, Nancy, and Richard H. Sabor, eds. 1993a. Op- A. Yusof. 1993. The Political Economy of PoverJ Eqa- portunity Forgvnc Education, Growth, and lnequaiq uiy, and Growth: Sri Lanka and Malaysia New York in Brazil World Bank, Washington, D.C Oxford University Press. 375 OG4R~KflflC-NOTE Calomiris, Charles W., and Charles P. Himmelberg. frInvestmentin DevelopingCgAnntries. NewYork Ox- 1993. "Directed Credit Programs for Agriculture and ford Univcrsity Press. Industry: Arguments fiom Theory and Fact." Papar Chia, Siow Yue. 1992. 'The Role of Institutions in Sing- presented at the Annual Bank Conference on Devel- apore's Economic Success." Background paper for The opment Economics, World Bank, Washington, D.C., East Asian Miracle. World Bank, Policy Research De- May3-4. May 3-4. ~~~~~~~~~parmacnt, Washington, I).C. Canning, David, and Marianne Fay. 1993. 'The Effect of PautnCnr, W i onenD.C. . = . . _ ,, _ ~~~~~Chiu, Paul C. H. 1992. "Money and Fimnancal Markets: Infratucture Networks on Economic Growth." Co- The Domestic Perspective." In Taiwanfrom Develop- lumbia University, New York, January. Caprio, Gerard, Jr., Izak Anyas, and James Hanson, s ing to Mature Economy. Colorado: Westview. Forthcoming. Financial Reform. Theory and F.erl- . Cho, Yoon Je, and Thomas Hellmann. 1993. 'Govcrn- Fnot. Camrnidge: Canbridge URrivrrsity Press. menr Intervention in Credit Market: An Alternative C ic, Caridge:, Camidg e, Uni Press. SInterpreton of Japancse and Korean Experiences Carroll, Crspe David and Lawrence H. Sp.ca from the Ncw Institutional Economics Perspective." - imers. 193. "savings and Growth: A rcgicerpreca- Paper prepared for a World Bank research project on dion." Paper presented a1t the Carnegie-Rochester Th Efetvns*fCei-Plce nEs sa Public Policy Conference, Bradley Poliqy Research World Bank, Financial SrecitPoriDevelmEt Dcpa.r- Cen_ r Api 23--24. Worlci Bankl, Financial Sector Dvcvdopmenc Depart- Center, April 23-2. mentc, Washington, D.C. tChan, Steve, and Cal Clark. 1992. FA&xiii Fior m-siWahigon DC (Than, Stuev an CaizCark 1992. Fesgt Cho, Yoon Je, and Deena Khatkare. 1939. Lessons ofFi- andFotun,i Thwa pment Navigating heo nancaal Liberalization in Asia: A Comparative Study. tween Scylla and Chanybdis. London: Roudledge. World Bank Discussion Paper 50. WashingEon, D.C. Chant, John, and Man Pangestu. Forthcoming. An As- Cho, YoonJe, andJoon-Kyang Im. 1993. 'Cedit Poli- sessment of Financial Reform in IndonesLiz 1983-90." - YnJ IndJoonKyg Kim.1 "CWedt Poli- s t- ' T T I A v ^ ttF a~~ces and Industralization of Korea. World Bank, In Ge_rd Caprio,Jr., IzakAti, andjames Hanson, Finanal Sector Development Departnent, Washing- eds., Financia Rqornm: Theory and Eaperince. Cam- t, Fcbra. britG: Canibrc;ge 1-jniversity l ress. bridg Cdamrd U erst Pres. Q Chrisrensen, Scott R., David Dollar, Ammar Siamwala, Che, Ewad K. 979 Hpergrow:h in Asian Ec_n, and Pakor Vichyanond. 1993. sInstitutional and Polit- omier: A Comparative Sturvey of hong k'on~g JpaP, ical Bases of Grwth-hiindztng Policies in Thailand Korea, Singapore and Taiwan. London: Macmillan. Lessons of East Asia World Bank Counnty Study. Chen, Kang, Gary H.JefFerson, and InderjirSingh. 1992. Washington, D.C. "Lesson from China's Economic Reformn." Joumna of Carke, Gerge. 1992. "More Evidence on Income Dis- ComparativeEconomica 16:201-25. rribution and Growth." Policy Research Working Chenery, Hollis B. 1959. The Interdependence of In- Paper 1064. World Bank, Policy Research Depart- vestment Decisions." In Moses Abramovit ecL, The ment, Washington. D.C. Allocation of Economic Resource. Sianford, Calif.: Sran- Cohen, S. I. 1985. 'A Cost-Benefit Analysis of Industrial ford University Press. Training." Economics of Edication Review 4(4): . 1960. "Patterns of Industrial Growth." American 327-39. Economic Review 5 0(4): 62454. Collins, Susan M., and Won-Am Park. 1939. "External -. 1986. "Growth and TransirM2aton.' In Holois B. Debt and Macroeconomic Performance in South Chenery, Sherman Robinson, and Moshe Syrquin, Kora" In Jeffey D. Sachs and Susan M. Collins, eds., eds., Industrialzation and Growth: A Comparative Developing Country Debt and Economic Peifrmance, Study. New York Oxford University Press. vol. 3. Chicago: University of Chicago Press Cheng, Hang-Sheng, ed. 1986. Finncial Policy and Re- Collins, Susan M., and Dani Rodrik. 1991. Easter Eu- fonn in Pacific Basin Countries. Lexington, Mass.: Lex- rope and the Soviet Union in the World Economy. Wash- ington Books. ington, D.C.: Insitute for International Economics. Chia, Ngee Choon, and John Whalley. Forthcoming. Corbo, Vittorio, and Klaus Schmidt-Hebbel. 1991. "Paterns in Invesrtment Tax Incentives among Devel- "Public Policy and Savings in Developing Countries." oping Countries.' InAmwar Shah, ed., Fiscal Incntive Jownl ofDevelopmenitEconomics36(1): 89-115. 376 BIBLAjt Corden, W. Max. 1971. The Themry of Protection. Lon- NBER Conference on Aging, Cancel Bay, U.S. Virgin don: Oxford University Press. Islands, May 7-10. -. 1974. Trade Poicy and Economic Welure. London: Dekle, Robert. 1988. "Do the Japanese Elderly Reduce Oxford University Press. Their Total Wealth?" Reischauer Institute of Japanese -. 1993. "Seven Asian Miracle Economies: Overview Studies, Harvard University, Cambridge, Mass. of Macroeconomic Policies." Background paper for __. 1993. "Raising Saving Races: Lessons from the The Est Asian Miracle. World Bank, Policy Research Japanese Experience." Background paper for The East Department, Washington, D.C. Asian Miracle. World Bank, Policy Research Depart- Dahlman, Carl J. 1990. "Technology Strategy in the ment, Washington, D.C Economy of Taiwan: Exploiting Foreign Linkages and Dollar, David. 1990. "Ounvard Orientation and Growth: Investing in Local Capability." World Bank, Private An Empirical Study Using a Price-Based Measure of Sector Development Department, Washington, D.C. Openness." World Bank, EastAsia and Pacific Region, Dahlman, Carl J., and Peter Brimble. 1990. 'Technology Country Departmncet I, Washington, D.C. Strategy and Policy for Industrial Competitiveness: A _. 1992. "Outward-Oriented Developing Economies Case Study in Thailand.' Industry Series Paper 24. Really Do Grow More Rapidly: Evidence from 95 World Bank, Industry and Energy Department, LDCs, 1976-1985.' Economic Development and Cul- Washington, D.C turd Change 40(3): 523-44. Dahlman, Cad J., and Valdarcs Fonsea. 1987. "From Dollar, David, and Kenneth Sokoloff. 1990. "Patterns of Technological Dependence ro Technological Devel- Producivity Groth in Soudl Koreai MaencEcturnng opmenr: The Case of the Usiminas Steel Plant in Instnes, 39-7. BraziL" In Jorge Kamz, and others, eds., Technoloy cr33(2): 309-27. Generation in Latzn American ManuracinUg frig1nd. Dombusch, Rudiger, and Stanley Fischer. 1993. "Moder- tries. London: Macmillan. ate Inflation." World Bank Economic Review 7(1): Dahuman, Carl J., and Ousa Sananikonc. 1993. E:conomic 1-44. Poalces and Instiutions in the Rapid Growth of Taiwan, Easterly, WVlliam. 1993. "Explaining Miracles: Growth China. Leson of Eas Asia. World Bank Country RegressionsMeettheGangofFour." Paperpresented Study. Washington, D.C ar the NBER's Fourth Annual East Asian Seminar on Dasgupta, Partha, Amarty Sen, and Stephen Marglin. Economics, San Francisco, June 17-19. 1972. Guidelines fir Prjecet Evaiuajon. New Yorlc Easterly, William, Michael Krerner, Lant Pritcherr, and United Nations Industrial Development Orgization. Lawrence H. Summers. 1993. "Good Policy or Good Luck? Country Growth Performance and Temporary De Long, J. Bradford, and Lawrence H. Summers. 1991. Shocks." Paper presented at a conference on How Do "Equipment Investment and Economic GrowtdL National Policies Aficcr Long-Run Growth? World Quareer4journal ofEconomics 10C6(2): 445-502. Bank, Washington, D.C., February. 1993. 'How Robust Is the Growth-Machinery Easterly, fiiam, and Sergio Rebelo. 1993. "Fiscal Policy Nexus?' Paper presented at a conference on How Do and Economic Growth: An Empirical Investigation." Natonal Policies Affect Long-Run Growh? World Paper presented at a conference on How Do National Bank, Washington, D.C., February. Policies Affect Long-Run Growth? World Bank, de Melo, Jaime, Carl Hamilton, and L Alan Winters. 1990. Washington, D.C, February. YVolunwiyExportRestminm :ACaseSrudyFocusingon Easterly, William, and Klaus Schmidt-HebbeL Forth- Eflecrs in Exporting Countries." Paper 464, Institute for coming. "The Macroeconomics of Public Sector International Economic Studies, Stockholm. Deficits A Synthesis." In Wlliam Easterly, Carlos Ro- de Melo, Jaime, and Sherman Robinson. 1992. "Produc- drigucz, and Klaus Schmidt-Hebbel, eds., PublicSector tiviry and Externalities: Models of Export-Led Defiit and Macroeconomic Perjirmance. New York. Growth." Jounl ofJnternwtional Trade and Economic Oxford University Press. Development 1: 4169. Edwards, Chris. 1990. "Prorection and Policy in the Dearon, Angus, and Christina Paxson. 1992 'Savings, Malaysian Manufacturing Sector." Policy Assessmnent of Growth, and Aging in Taiwan?' Paper presented at an the Maysian Indaried Policy Stmu (Mw) and the In- 377 LOGR ATPMI C--N O T E dustrial Master Plan (tw1), Voieme 3. Vienna: United Freeman, Richard. 1992. "Does Suppression of Labor Nations Industrial Development Organization. Contribure to Economic Success? Labor Relations Edwards, Sebastian. 1988. Exchange Rate Misalignment in and Markets in East Asia.' Background paper for The Developing Countries. Baltimore, Md.: Johns Hopkins EastAsian Miracle. World Bank, Policy Research Dc- Universiy Press. partmenE, Washington, D.C. -. 1992. "Trade Oricntation, Distortions and Growth Freeman, Richard, and Martin L Weitzman. 1987. in Developing Countries." Journal of Development 'Bonuses and Employment in Japan." Journal of Economics 39(1): 31-57. Japanese andf!nternanoonalEeonomies1(2): 168-94. Elias, VictorJ. 1990. "The Role oFTotal Factor Produc- Frischtak, Claudio R., Bita Hadjimichael, and Ulrich tivity on Economic Growth.' Background paper for Zachau. 1989. Competition Polcesfor Ina¼strializing the Worlk Development Report 1991. World Bank, Countries. World Bank Policy and Research Seriec Office of the Vice President, Development Econom- Paper 7. Washington, D.C. ics, Washington, D.C. Fry, Maxwell J. 1988. Money. Interest, and Banking in Erzan, Refik, Kiroaki Kuwahara, Serafino Marchese, and Economic Deveiopment. Baltimore, Md.: Johns Hop- Rene Vossenar. 1989. "The Profile of Protection in kins University Press. Developing Countries." UNCZAD Review 1:29-73. Fukasaku, Kiichiro. 1992. Economic Regionalization Fanelli, Jose Maria, Roberto Frenkel, and Lance Taylor. and Inta-Industry Trade: Pacific-Asian Perspecaives." 1992. "The World Development Report 1991: A Technical Paper 53, OECD Development Centre, Padis. Critical Assessment." Prepared as part of an UNCTAD- Gelb, AMan H. 1989. "Financial Policies, Growth and Ef- executed project to assist tizc Intergovernmental ficiency." PRE Working Paper 202. World Bank, Group of Twnet-Four. Washington, ''.C. Far Eastern Economic Revitew. Various issues. Gelb, Alan H., Ju..n Knight, and Richard H. Sabo: 199 1. Farrell, M. J. 1957. "The Measurement of Productive Ef- 'Public Sector Emp!oyment, Rent Seeking and Eco- ficicncy." Journal of tht Royal Stadstical Society, Series nomic Growth." Economic Journal 101(408): 1186-99. A 120 (part 3): 253-8 1. Gerschenkron, Alaxander. 1962. Economic Backwardness Fecher, M., and S. Perelman. 1992. "Industrial Effi- in Historical Prspecive. Cambridge, Mass.: Harvard ciency in OECD Countries." In R. Caves, ed., Indus- University Press. tialEEfficiernyinSilNaionss Cambridge, Mass.: Mrr Ghose. T. K 1987. "The Banking System of Hong Press. Kong." Burterworth (Asia) Ltd. Federazion of Bankers Association of Japan. 1991. Giovannini, Alberto. 1983a. "Essays on Exchange Rates 'Japanese Banks." and Asset Markets." Ph.D. diss., Massachusetts Insti- John Pci, Shirley W. Y. Kuo, and Gustav Ranis. 1981. rute of Technology (Mrr), Cambridge, Mass. The Taiwan Success Story: Rapid Growth with Im- -. 1983b. "The Interest Elasticity of Savings in Devel- proved Diributon in theRepblic ofChina, 1952-79. oping Countries: The Existing Evidence." World De- Boulder, Colo.: Westview. velopment 11(7): 601-07. Fields, Gary S. 1992. "Changing Labor Market Condi- Gordon, David L. 1983. DevelopmentFinanc Companies, dions and Economic Development in Hong Kong, State and Privately Owned-A Review. World Bark Staff Korea, Singapore, and Taiwan, China" Background Working Papers 578. "X'orld Bank, Washington, D.C. paper for The EastAsian Miracle. World Bank, Policy Goto, Akir, and Ryuhei Wakasugi. 1988. "Technology Research Department, Washington, D.C. Policy." In Ryutaro Komiya, Masahiro Okuno, Ko- Fischer, Stanley. 1993. "Macroeconomic Factors in taro Suzumura, eds., IndustrPolicy oflapan. New Growth." Paper presented at a conference on How York- Academic Press, Inc. Do National Policies Affect Long-Run Growth? Granger, Clive W. J. 1969. 'Investigating Causal Rela- World Bank, Washington, D.C., February. dions by Econometric Models and Cross-Spectral Foundation for Advanced Informarion Research. 1991. Methods." Economerica 37: 428-38. "The Development of Financial and Capiti Markets Greenwald, B., and Joseph E Stiglirz. 1986. "Externali- in the Asia-Pacific Region.' Tokyo, Japan. September. ties in Economies with Imperfect Information and In- 378 BI BLOO,G.X complete Markers." Quarter4' journal of Econonic Firms' Capital Structure and Investment Decisions: 101(2): 229-64. Evidence from a Panel of Indoncsia Establishments, _. 1938. 'Pareto Inefficiency of Market Economies: 1981-88." Paper presented at a conference on the Im- Scarch and Efficiency Wage Models." American Eco- paca of Financial Reform, World Bank, Washington, nomic Review 78(2): 351-55. D.C., April 2-3. Gregory, Peter. and Deepak LaL. 1977. "Demand, Sup- Harrison, Ann E. 1991. "Openness and Growth." PRE ply, and Structure: A Cross-Sectional Analysis of Man- Working Paper 809. World Bank, Washington, D.C. uficuring Wages.' World Bank, Employment and Hendrie, Anne. 1986. 'Banking in the Far East: Struc- Rural Dcvelopment Division Working Paper, Wash- tures and So:,rces of Finance." Financial Times Busi- ington, D.C. ness Information, London. Grossman, Gcne, and Elhanan Helpman. 1991. Innova- Hill, Anne M., and Elizabeth King, cds. 1991. "Womcn's on vand Growth in the Globe:! Fconomy. Cambridge, Education in Developing Countries: Barriers. Bene- Mass.: mT Press. fits, and Policy." PHREE Background Paper Series. Gupta, Kanhaya L. 1984a. Finance daml Economic Growth World Bank, Washington [D.C. in Developing Counties. London: Croom Hclm. Hon, Wong Seng. 1992. "Exploiting Information Tech- -. 1984b. "Financial Liberalizarion and Economic noAogy.A Case Study of Singapore." World Develop- Growth: Some Simulation Results." Jouirnal of Eco- ment20(12): 1817-28. nomicDevelvPment9(2):25-34. Hong, Won-Tal. 1989. "Korean Economy at Cross- H-adad, Mona, and Ann E. Harrison. 1991. "Are There roads." Sasang Quarterly 1 (Summer). Dynamic Externalities fiom Direct Foreign Invest- Horioka, Charles Yuji. 1990. 'Why Is Japan's Household ment? Evidence From Morocco." Indus-ry Series Savings Rate So High? A Litcrature Survey."Jounalof Paper No. 48. World Bank, Industry and Energy De- Japanese andInternaional Economics4: 49-92. parument, Washington, D.C. . 1991. The Determinants of Japatn's Savings Rate: Forthcoming. "Are There Positive Spillovers from The Impact of the Age Srructure of the Population Direct Foreign Investment? Evidence from Panel Dara and Other Factors." The Eronomic Strdies Quarterly for Morocco." Journal ofDerelopment Economics. 2(3): 237 53. Haggard, Stephan. 1990. Pathways from the Perzphery. Horiuchi, Aiivoshi, and Qing-Yuan Sui. 1992. 'The In- Ithaca, N.Y.: Cornell University Press. fluence of the Japan Development Bank Loans on 14alvorsen, Robert Forthcoming. 'Fiscal Incentives for Coporare Investment Behavior." Investment in Thailand." In Anwar Shah, ed., Fiscal Hou, Chi-Ming, and Gee San. 1990. "National Systems IncendvesJfrInvestment in Developing Countries. New Supporting Technical Advance in Indusry. The Case Yortc Oxford University Press. of Taiwan." Taipei, Taiwan, China. Hamilton, Carl B. 1989. "The Political Economy of Hughes, Helen. 1991 'East Asian Export Success" Re- Transient 'New' Protecrionism." Weirschafi'cs search School of Pacific Studies, Australian National Archive, Review of WorldEconomics 125: 522-46. University. ed. 1990. Textles Trade and the Developing Coun- Hulten, Charles Jr. 1989. "The Embodiment Hypothesis ins: Eliminating the Maid-Fiber Arrangement in the Revisited." University of Marvland, College Park, Md. 1990s Washington, D.C World Bank Huseinga, Mark, and ChrisEos Kostopolous. 1992. "Op- Hanson, james, and Roberto Rocha. 1.986. 'High Inter- tions for Reform of the Business Regulation System: est Rates, Spreads, and the Costs of Intermediation: Nepal." Country Report 8, Center for Instirutional Two Studies." Induscry and Finance Series 8. World Reform and the Informal Sector, College Park, Md. Bank, Washington, D.C. Hwang,Y. Dolly. 1991. The Rise ofa New WorldEconomic Hanushek, Eric A.., and Richard H. Sabo. 1991. "Notes Power. Postwar Thi wan. New Yorlc Greenwood. on Changes in Educational Performances." A paper Imai, Ken'ichi. 1988. "Industial Policy and Technologi- prepared for PEER Meetngs. Boston, Mass. cal Innovation." In R. Komiya, M. Okuno, and K. Harris, John, Fabio Schiantarelli, and Miranda G. Sire- Suzumura. 1988. IndustriidPolicy offapam New Yorkl gar. 1992. 'Thc Effort of Financial Liberalization on Academic Press. 379 India. Various issues. Reserve Bank ofindia Bdlletin. JDB/JERI (Japan Development Bank and Japan Economic Indones;ia. Various issues. Indonesian Financia Statistics. Researrh Institute). 1993. "Policy-Based Finance: The Jakarta: Bank of Indonesia. Expericnce of Post.ar Japan." Tokyo. Ingramn, James C 1971. Economic C&ange in Thailna Japan. Various years. &otnomicStaisticsAnnual Tokyo: 1850-1970. Scmnford, Calif.: Stanford Universiry Press. Bank ofjapan. Inoki, Takenori. 1993. "Retrcment Behavior ofJapancsc -- Neitio val Income Account ofjapan. Tolgro: Japanese Bureaucracies and rhe Role of Public Corporations: Economic Planning Agency. An Economic Interpretation of Human Resources' Jitsuchon, Somchai. 1991. "Rctrospecus and Prospects of Mobility from Central Government .o Public Corpo- Thailand's Economic Development." Working Paper ration." World Bank, Economic Development Insci- 2, Japanese Economic Planning Agency, Tokyo. tute Working Paper, Washington, D.C. Johnson, Chalmers. 1982. MITI and theJapanrse Miracle. IMF (Inernational Monctary Fund). 1993. World Eco- Stanford, Calif.: Sranford University Press. nomic Oulok. Washi.gton, D.C, May. Jones; Leroy, and Fadil Azim Abbas. 1992. "Malaysia: International Road Federation. Various years. WorldRoad Background, Malaysian Airlines, Kelang Container Statistics Washingronj D.C. Terminal, Sports Toro." Paper prcsented at a World rsu(Indonesian Economists' Association). 1990. The Un. Bank Conference on The Welfare Consequences of demandng ofEconomic Democracy. a. Selhing Public Enterprises, Washington, D.C., June 11-12. Ishikawa, T. 1987. 'Savings: Structural Determninants of Jorgenson, Dale W., Masahiro Kuroda, and Mieko Household Saving auid the Financial and Tax Sys- Nishimizu. 1987. 'Japan-U.S. Industry-Level Pro- temns? In Koichi Hamada, Malcato Kuroda, and ductivityCorrparisans 1960-70."joumalofJap Akiyoshi Horiuchi, ed., Macroeconomic Analysis of.the aIo . Japanese tmx. Amtrdam: North Holland. adlteniodEooisI() Jto, Takatoshi1992Thejapanese RncAnoerdamNorthrHolland.Katseli, Louka T. 1992. "Foreign Direct Investment and Io Takatoshi. 1992. The Japanese Economy. Cambridge, Trade Interlinkages in the 1990s: Experience and Prospects of Developing Countries." Background _. 1993. "Industial Policy for Development Japanese paper for United Nations Symposium on Globaliza- Experience and Replicabiliry." John F. Kennedy tion and Developing Countries, The Hague. School of Government, Harvard University, Cam- (Reprinted as Ccnter for Economic Policy Research bridge, Mass. Discussion Paper 687, London, July 1992.) Itoh, Motoshige, and Kazuharu Kiyono. 1988. "Foreign Katz, Jorge, ed. 1987. Techno Generation in LauinAnmr- Trade and Direct Investment.' In Ryuraro Komiya, ican Manufacturng Induarie. London: Macrillan. Masahiro Okuno, Kotaro Suzumura, eds., Inditfil Kaufman, George G. 1992. Banking Structures in Major Policy offapan New York: Academic Press. Count7is. Boston: Kluwer Academic Publishers. Itoh, Moroshige, Kazuharu Kiyono, Masahiro Okuno- Kawaura, Alihiko. 1991. "Identification of the Size and Fugiwar, and Kataro Suzumura 1988. 'Industrial Pol- Direction of the Post-War Credic Allocation in icy as a Corrective to Marker Failures." In Ryutaro Japan. World Bank, Financial Sector Development Komiya, Mashiro Okuno, and Koraro Suzumura, eds., Department, Washington, D.C. IndustiaPoficy offapan. New York: Academic Press. Kim, C. 1990. 'Labor Market Development." In Jene K 1 991. Economic Analysis of ndustrial Poiy. New Kwon, cd., Korean Economic Development New York York: Academic Press. Greenwood. Itoh, Motoshige, and S. Urata. 1993. "Small and Kim, Chang-Shilc 1993a. "Government-Business Rela- Medirun-Size Enterprise Support Policy of Tapan. tionship in Korea: Selectve Case Studies.' Back- World Bank Policy Research Department, Washing- ground paper for The EasrAsian MiracLc World Bank. ton, D.C. Policy Research Department, Washington, D.C. Jaff, Ad-am B. 1986. 'Technological Opportunity and . 1993b. "Trde Policy and Industrial Tranfrma- Spillovers of R&D: Evidence from Firms' Patens, dion in Korea" Background paper for The EastAsian Profits and Market Value." American EconomicReview liiracle. World Bank, Policy Research Department, 76(5): 984-1001. Washington, D.C. 380 Kim, Ji Hong. 1990. "Korean Industrial Policy in the Konishi, Y. 1989. A Quantitative Anadis of Eduational 1970s: The Heavy and Chemical Industry Drive." Policies in Postarjapan Research Insticute for Eco- Working Paper 9015. Korean Development Institute, nomics and Business Administration, Kobe Univer- Seoul, Korea. sicy, Kobe, Japan. Kim, Jong-l1, and Laurence J. Lwu. 1993. "The Sources of Korca, Rcpublic of. Various years. Economic Statistics Economic Growth of the East Asian Newly Industrial- Yearbook. Seoul: Bank of Korea. ized Countries." Paper presented at a Conference on _. Various issues. The FinancialSystem in Korea. Seoul: the Economic Development of the Republic of China Bank of Korea. and the Pacific Rim in 1990 and Beyond, Stanford -. 1978. Major Statistics ofthe KoreanEconomy. Seoul: University, Stanford, CaliE, April. Economic Planning Board. Kirn, K, and J. K Park 1985. Sources of Econonic Krueger. Anne 0. 1978. "Alternative Trade Strategies Growth in Korea. Seoul: Korea Development Institute. and Employment in LDcs." American Sconomir Re- Kim, Kihwan, and Danny Leipziger. 1993. Korea: A CGst view. 68(2): 270-74. of Government Led Development Lessons of East Asia. -. 1984. "Comparative Advantage and Development World Bank Country Study, Washington, D.C. Policy Twenty Ycars Later." In Moshe Syrquin, Lance Kim, L., and J. Nugent. 1993. 'Korean SMEs and Their Taylor, and Larry E. Westphal, eds., Economic Struc- Support Mechanissr: An Empirical Analysis of the tureandPerformance. New York: Academic Press. Role of Govermrent and Ore:er Non-Profir Organiza- _. 1993. "East Asia: Lessons for Growth Theory.' dions.? Worid Bank, Policy Research Departrment, Paper prcsented at the Fourth Annual East Asian 'Washington, D.C. Seminar on Economics, National Bureau of Eco- Kim, S. 1987. In Service Trainingas an Insmnrentfbr the nomic Research, San Francisco, Calif., June 17-19. DevelopmentofHwman Resources in Korea. Paris: OECD Krueger, Anne O., and Jagdish Bhagwati. 1973. 'Ex- Development Center. change Control, Liberalization Attempts, and Eco- 1988. Japan Civil Service Sys Structre, nomic Development." American Economic Review sonne andPolitict New York: Greenwood. 63(2): 419-27. ICm, Y. C., and J. K Kwon. 1977. "The Utilization of Krgrnan, Paul R 1986. Introduction: New Thinking Capital and the Growth of Output in a Developing about Trade Policy." In Paul R. Krugman, ed., Strate- Economy: The Case of South Korean Manufactur- gc Trad Polcy and the New Intenational Economics. ing.?Journd ofDevelopment Economics 9: 265-78. Cambridge, Mass.: MIT Press. King, Elizabeth, Katherine Anderson, and Yan Wang. Kumar, Nagesh 1985. Cost of Technology Imports: 1993. "Feedback Effects on the Labor Market on The Indian Experience.? Economic and Political Schooling Choice in Malaysia? Paper presented at the Weekly 20: Ml 03-114. April 1992 Meetings of the Population Association of Kunio, Y. 1988. The Rise of Ersaz Capitalism in South America, Cincinnati, Ohio. EsAsia. New York. Oxford University Press. King, Robert, and Ross Levine. 1993. "Financz, Entrpre- Kuo, S. 1983. The Taiwan Economy in Transition. Boul- neurship and GrowdLh Paper presented at a conference der, Colo.: Westvzew. on How Do National Policies Affect Long-Run Kuo, Wan-Yong. 1976. 'Income Distribution by Size in Growh? World Bank, Washington, D.C., February. Taiwan Area. Changes and Causes." Indmut of Free Knight, John, and Richard H. Sabor. 1991. Education;, Cbina 45. Produciviy, and Inequality. New York.c Oxford Uni- Kuznets, Simon. 1955. "Economic Growth and Income versity Press. Inequality. American EconomicReview 45(1): 1-28. Komiya, Ryutaro. 1990. The Japanese Econoyn: Trade, _. 1959 Six Lectures on Economic Growth. Glencoe, Induatry, and Government. Tokyo: University of Ill.: Free Press. Tokyo Press . 1966. Modern Economic Growth. New Haven, Komiya, Ryuraro. Masahiro Okuno, Kotaro Suzumura, Conn.: Yale University Press. eds. 1988. IndusialPolicy ofjapan. New York Acad- Lal, Sanjaya, and G. Wignaraja. 1992. 'Foreign Involve- emic Press. ment by European Firms and Garment Exports by 381 > NTE Developing Countries." Development Studies Work- HIID Conference on the Malaysian Economy, Kuala ing Paper 54, Queen Elizabeth House, OxfordlCen- Lumpur. Institute of Strategic and International tro Studi Luca d'Agliano, Torino, Imlia. Studies. June 1-3. Lanyi, Anthony, and Rusdu Saracoglu. 1983. "Interest Lim, Linda Y. C., and Pang Eng Fong. 1991. Foregn Di- Rate Policies in Developing Countries." Occasional ret:Investment andIndutrializtioninMa(aria.Sing- Paper 22. International Monetary Fund, Washing- apor, 7Taiwn and Thaikladn OECD, Paris. ton, D.C. rim, Youngil. 1991. 'Upgrading industrial Base and Lawrence, Robert Z 1991. "Scenarios for the World Trade: The Role of Government Policy in Korea." Trading System and Thcir Implications for Develop- United Nations Industrial Development Organiza- ing Countries." Tcchnical Paper 47, OECD Develop- tion, New York. meent Center, Paris. (Paper prepared for the Rcsearch Lindaucr. David L., and Barbara Nunbcrg, eds. 1993. Program on Globalization and Regionalization.) "Rehabilitating Govcrnment: Pay and Employment 1993. "The Global Environmenr for the East Asian Refrmn in Sub-Saharan Africa." World Bank, Policy Model.' John F. Kennedy School of Government, Research Department, Washington, D.C. Harvard Univcrsiy, Cambridgc, Mass. LiEte, Ian M. D. 1979. UThe Experience and Causes of Lee, Chinaboon. 1985. "Financing Technical Education Rapid Labor-Intensive Development in Korea, Tai- in LDCs Economic Implicatons from a Survey of wan, Hong Kong, and Singapore, and the Possibilities Training Modes in the Republic of Kora? Education of Emulation." Working Papers, Asian Employment and Trainiing Series 6. World Bank, Education and Programme, ILO-ARTEP, Bangkolk Training Deparmnent, Washington, D.C. _ 1982. Economic Development New York: Basic Lees, Francis A., James Botts, and Rubens Penha Cvsne. Books. 1990. Banking, and Financial Deepen in HraZ6;lw Little, Ian M. D., Richard N. Cooper, W. Max Corden, New Yoraki St. Martin'sPras Depening in and Sarath Rajapathirana. Forthcoming. Boom, Cri4s, New Yo& Sr. Martin's Press.Adyjustmen Th BbMatroeconondc EEpeneneofDevelop- Leipzrger, Danny. 19 87. Ko-reicManaging the Industrial AduenThMarcomiEprece/Dvl- Lzpzlgcr, Danny. 1987. Korca Mnnugmg tJyc SnivZ mg Counties. New York. Oxford University Press. Transition. Washington, D.C.: World Bank. Lltle, Ian M. D., Tibor Scitovsky, and Maurice FG. Leipziger, Danny, David Dollar, Anthony F. Shorrocks, Scott. 1970. Induniy and Trade in Some Developing and Su-Yong Song. 1992. The Distribution ofIncome Countnies. London: Oxford University Press. -and WeaLrh in Korea- Washington, D.C.: World T Bandh Liu, Alan P. L. 1987. Phoenix and the Lamne Lion Mod- Bakentiozn in Taiwan and iMainland China, 1950-80. Levy, Brian. 1993. "Obstacles to Developing Indigenous Stanford, Calif: Hoover Institution Press. Small and Medium-Size Enterprises: An Empirical Liu, Ligi. 1991. "Entry-Eir: Learning and Productivity Assessment.? World Bank Economic Review 7(1): Change-Evidence from Chile" Policy Research 6585. Working Paper 769. World Bank, Country Econom- Lewis, W. Arthur. 1954. 'Economic Development with ics Department, Washingron, D.C. Unlimited Supplies of Labor." Manchester School of Lockheed, Marlaine, and Adriaan Verspoor. 1991. Im- Economics and Social Studies 22(2): 139-91. proving Primary Education in Developing Countries. Li, KI T. 1989. The Evolution of Policy Behind Taiwun's New York: Oxford University Press. Development Success New Haven, Conn.: Yale Uni- Lockwvood, William W. The Economic Development of versiry Press. japaw Growth and Structural Change, 18681938. Liddle, William. 1991. "The Relative Autonomy of the Princeton: Princeton University Press. Third World Politician: Soeharto and Indonesian Long, Millard. and Dimirri Vita 1991. "Changing the Economnic Developmenr in Comparative Perspec- Rules of the Game." Policy Research Working Paper ive." Interational Studies Quarterly 35: 403-27. 803. World Bank, Policy Research Department, Lim, Chee Peng 1992. "Managing Structural Change Washington, D.C. and Indusrrialization-Heayy Industrialization: The Maisom, Abdullah. 1992. 'Pattems of Total Factor Pro- Malaysian Experience" Paper presented at the isis- ducrMty Growth in Malaysian Manufaucring Indus- 382 tries, 1973-89." Paper presented at ISLS-HIID Workshop velopment Policy andAnaij sis andAppication. Boul- on the Malaysian Economy, Kuala Lumpur, September. der, Colo.: Westview. Maki, Atsushi. 1993. "The Relation between Houschold Murdock, K., andJoseph E. Stiglirz. 1993. "The Effecr of Saving Rare and Economic Growth: Consumers' Financial Repression in an Economy with Positive Stimulus ro Purchase Durable Goods." Paper prepared Real Races." Background paper for The EmFt Asian for a World Bank Asian Miracle Project Workshop, Mirac& World Bank, Policy Research Department, Tokyo, Feb. 1-2. Washington, D.C. Malaysia. 1990. Quarterly Buledn. Kuala Lumpur: Bank Murphy, Kevin, Andre Shleifer, and Robert Vishny. Negara. March-June. 1989. "Industrialization and the Big Push." Jotrnal of Mankiw, N. Grego.y, David Romer, and David Wed. 1992. Political Economy 97(5): 1003-26. 'A Contribution to the Empirics of Economic Growth." Mutoh, Hiromichi. 1988. "The Automotve Industry." QuaterlyjJornal ofEconomis 107(2): 407-38. In Ryuraro Komiya, Masahiro Okuno, and Koraro Mayer, C 1987. 'The Assessment: Financial Systems and SuzLunura, eds., Industrial Policy ofJapan. New Yorlc Corporate Investment. Oxfrd Review of Economic AcademicPress. Policy 3(4): i-xvi. Myrelka, Lynn. 1978. "Licensing and Technology Devel- Mazumdar, Deepak. Forthcoming. "Labor Markets and opment in the Andean Group." World Development Adjustment in Open Asian Economies The Republic 6(4): 447-60. ofKoreaandMalaysia." WorldBankEconomicReziew. Nagaoka, Sadao. 1989. 'Overview ofJapanese Industrial McKinnon, Ronald I. 1973. Money and Capital in Bc- Technology Developmenrt" Industry Series Paper 6. nomic Development Washington, D.C.: Brookings World Bank, Industry and Energy Department, Insttution. Washington, D.C. Mellor, John, ed. 1993. "Agriculntre on the Road to In- Nalebuff, B. 1983. "Information, Competition and Mar- dustrialization." International Food Policy Research kers." American Economic Review 72(2): 278-84. Institute, Washington, D.C. Nalebuff, B., and Joseph E StigLirz. 1983. "Prizes and In- MiddkUton, J., A Ziderman, and K van Adams. Forth- centives Towards a General Theory of Compensation coming. Skillsfor ProdictivicV. Baltimore, Md.: Johns and Competition." Bel Journal of Economics 14(1): Hopkins University Press. 21-43. Modigliani, Franco. 1970. "The Life Cycle Hypothesis of Nam, Sang-Woo. Undated. "Korea's Financial Reform Savings and Intercountry Difirences in the Savings since the Early 1980s." Korea Development Institute, Ratio." In W. A. Elis, Maurice FG. ScoEt, and J. N. Seoul. Wolfe, eds., fnduccion Growth and Trait Essays in Neal, Craig R 1990. "Macrofinancial Indicators for 117 Honor ofSir Roy Harrod Oxford. Clarendon Press. Developing and Inldustrial Countries." PR£ Working Mody, Ashoka, R. Suri, and J. Sanders. 1992. "Keeping Paper 58. World Bank, Office of the Vice President, Pace with Change: Organizational and Technological Development Economics, Washington, D.C. Inperatives." WorldDeveopment2o(l2): 1797-816. Nehru, Vikram, and Ashok Dhareshwar. 1993. 'A New Monetry Authority of Singapore. 1991. "Savings-Invesc- Database on Physical Capital Stock. Sources, Method- ment Balances in Singapore: Determinanrs and ology, and Results." World Bank, Internatonal Eco- Medium-Term Outlook" Singapore: Economics nomics Department, Washingtn, D.C Department. Nelson, Richard R.. 1973. "Recenc Exercises in Growth Morris, Felipe, Mark Dorfmnan, Jose Pedro Ortiz, and Accounting: New Undesmtnding or Dead End?" Maria Claudia Franco. 1990. laan America 5 Banking American Economic Review 63(3): 462-8. Systems in the 1980s: A Cross-Country Comparort Nishimizu, Mieko, and John M. Page, Jr. 1982. "Total World Bank Discussion Paper 81. Washington, D.C. Factor Productivity Growth, Technological Growth, Mukai, Yurio. 1963. "Development of POStal Savings and Its and Technical Efficiency Change: Dimensions of Pro- Factora In Japanesr. IGnywI(eiza no. 83. December. ducriviy Change in Yugoslavia, 1965-78.' Economic Munasinghe, Mohan. 1987. Rural Electnfication for De- Journal92(368): 920-36. 383 p_E" . . 1987. "Economic Policies and Productvity Change . 1993c. "Technology Gaps between Industrial and in IndusErr. An International Comparison." Paper fior Developing Countries: Are There Dividends for Late- a World Bank research project an Productivity comers?" In Proceedings of the World Batnk Annual Change in Infant Industry. World Bank, Policy Re- Bank Conference on Developinent Economics, 1992. search Department, Washington, D.C-. Washington, D.C.: World Bank. -. 1991. "Trade Policy, Market Orientation, and Pro- Pack, Howard, and John M. Page, Jr. 1993. 'Accumula- ductivity Change in Industry.' In Jaime de Melo and tion, Exports, and Growrh in the High-Perfirming Andres Sapir, eds., Trade Theory and Economic Refonn: Asian Economies." Background paper for The East North, South andEast-Essays in Honor ofBe/ Eatsa. Asian Mirade. World Bank, Policy Research Depart- Oxford: Blackwell. ment, Washington, D.C. Noguchi, Y. 1985. "Nenkin, zeisci to kokumin no Pack, Howard, and Lary E. Wcstphal. 1986. "Industrial chochiku kodo [Pensions, the tax system, and the sav- Stratcgy and Technological Change: Theory vs. Real- kg behavior of the people]." In Kenkyu Hakoko Gaiyo ity."fournalofDevelopmentEconomics22(l): 87-128. [-Sumwy of Research Findings]. Kokuci Nin'i Seimei Packer, Frank. 1992. 'The Role of Long-Tarm Credit Hoken no Shorai Tenbo ni k..isuru Chosa Kenkyu- Banks within the Japanese Main Bank System." Co- kai [Reserch project on the Future Prospects of Gov- lumbia University, New York. einment-Managed Voluntary Life Insurancel, Tokyo. Page, John M., Jr. 1990. "The Pursuit of Industrial North, Doulas C., and R Thomas. 1973. The Rise ofthe Growth: Policy Initiatives and Economic Conse- Western Worn!c A Nev Economic History. Cambridge, quences." In Maurice FG. Scott and Deepak Lal, edts, UX.K: Cambridge University Press. Public Poticy and Economic Development Essays in Ogura, Sciritsu, and Naoyuki Yoshino. 1988. 'The Tax Honor oflan Litte Oxford: Clarendon Pres. System and Fiscal Investment and Loan Program." In Page, John M., Jr., and Peter A. Peti. 1993. "Productiv- Ryutaro Komiya, Masahiro Okuno, and Kotaro Suzu- ity Change and Strategic Growth Poliqc in the Asian mum, edmi P ndustria Policy ofJapan. New York. Aca- Miracle" Background paper for The EastAsian Mina- dlemnc Press. c& World Bank, Policy Research Department, Wash- Okimoto, Daniel. 1989. Between miT! and the Market ington, D.C. Japanese Indstrial Policy fir High Tecbnolg. Stan- Panagariya, Arvind. 1993. Unravelinj the Mysteries of ford, Calif.: Stanford University Press. China's Foreign Trade Regime." World Economy Ono. 1992. Jissenrteki Sangyo Seiskurn. In Japanese. 16(1): 51-. Tsusho Sangyo Chosakai. Pempel, T. J. 1984. "The Japanese Bureaucracy and Eco- Pack, Howard. 1971. Structurl Change and Economic nomic Development Can the Model bc Copied?' Policy in isreeL New Haven, Conn.: Yale University Economic Development Institute, World Bank, Press. Wsigo,DC -. 1988. "Industrialization and Trade." In Hollis B. Washington, D.C Chenery and T. N. Srinivasan, eds., Handbook of Perry, Guillemo, and Ana Maria Herrera. 1993. "Finan- Development Economics, vol. 1. Amsterdam: North zas Publicas, Esmabilizacion y Reforma Estructural en Holland. America Latina." Working Paper Series 165. Inter- 19Ha2. n a P C American Development Bank, Washington, D.C. _. 1992. 'Leamning and Producdvily Change in D:evel- Oping Countries." In Gerald K Helliner, ed., Trade Perm, Peter A 1988. "Korea's Export Niche: Origins and Policy Industializaton and Development. Oxford: Prospects." WorldDevelopment16(1):47-64. Clarendon Press. Pfeffermann, Guy P., and Andrea Madarassy. 1992. _. 1993a. "Exports and Externalities: The Source of Trends in Pivate Investment in Developing Countries. Taiwan's Growth in Taiwan." University of Pennsyl- 1992 Edition. IFC Discussion Paper 14. Washington, vania, Philadelphia, Pa. D.C.: World Bank and IFC. _. 1993b. Indtstrial and Trade Polcies in the High- Psacharopoulos, Georgc 1993. "Returns to Invesunent Performing Asian Economies." Background paper for in Education: A Global Update." Policy Research The ast Asia Miracl. World Bank, Policy Research Working Paper 1067. World Bank, Larin America and Department, Washington, D.C the Caribbean Region, Washington, D.C 384 BIBLO'4 Psacharopoulos, George. Jee-Peng Tan, and Emmanuel Rhee, Yung Whee, and Lary F. Westphal. 1977. "A Mi- Jimenez. 1986. Financing Educaton in Developing croeconomic Investigation of Choice of Technique." Coanmier An Exploration ofPoliqy Option. Washing- JowofDevelpmenrtEonomia4: 205-38. ton, D.C.: World Bank. Riedel, Jamnes. 1988. "Demand for LDc Exporcs of Manu- Ranis, Gusrav. 1973. "Industrial Sector Labor Absorp- fictures: Estimates from Hong Kong.' Economiclour- uon." Economic Development and Cultutra Chlange nal98(389): 138-48. 21(3): 387-408. Rob, Rafael. 1990. "The Evolution of Markcts nider De- . 979. 'Industrial Development." In W. Galcnson, mand Uncertainty." University of Pennsylvania, cd., Economic Growti and Sructural Change in Ta)- Philadelphia, Pa. wan. Ithaca, N.Y.: Cornell University Press. Romer, Paul M. 1993. "Two Srategies for Economic De- , ed. 1992. Taiwdan From Developing to Mature Econ- vclopmen r Using Ideas and Producing Ideas." In Pro- omy. Boulder, Cola.: WstviCw. ceedings of the Word Bank Annual Bank Conference on _. 1993. "labor Markets, Human Capital, and Dcvd- DevehopmmtEnomŽs 1992 Washington D.C. opmenr Performance in East Asia." Background paper for The East Asian Miracle. World Bank, Policy Re- Rosenberg, Nathan. 1976. Perspectives on ecohnoloV search Department, Washington, D.C. Cambridge, U.K.: Cambridge University Press. Ranis, Gustav, and John Fei. 1961. "A Theory of Eco- Sabor, Richard H. 1979. Eonomic Development and nomic Development." American Economic Reviewt Urban Migratin. London: Oxford University Press. 51(4): 533-65. Sakakibra, E, R. Feldman, and Y. Harada. 1982. The Ranis, Gustav, and Chi Schive. 1985. "Direct Foreign In- Japanese Financial Syttem in Comparatve Perspectiva vestrment in Taiwan's Development." In Walter Washington, D.C.: Joinr Economic Committee, Con- Galenson, ed., Foreign Trad and Investmen Economic gress of the Unitcd States. Development in &ie Newly Industriizng Asian Couen- Sakong, 1. 1993. Korea in de World Eonomy. Washing- tiim Madison, Wis.: University of-Wisconsin Press. ton, D.C.: Insritute for International Economics. Ranis, Gustav, and Frances Stewart. 1987. "Rural Link- Salleh, IsmaeL 1991. The Privatisaion of Public Enter- ages in the Philippines and Taiwan.' In Frances Stew- prises: A Case Study ofMalaysia." In Geer Gouri, ed, am, ed., Macro-polics for Appropriate Technolog in Pi vatisa n and Public Enterprise. New DclhL Oxford Developing Countries. Boulder, Colo.: Westview. Universiy Press and IBH Publishing Company, Rebelo, Sergio. 1991. "Long Run Policy Analysis and Sallc, Ismacl, Kim Leng Yeah, and Saha Dhevan Long Run Growth!. journal of Poitdcal Ecotnomy Manatan 1993. Growth, and Srcr rat 9(3) 50>21 ~ ~ ~ ~ ~ ~ . Meyanathan. 1993. Growab, Eqity, andf SDxrd 99(3): 500-21. Trsmj$rmazion in Malaysi The Role of the Pubic cc- Reid, Gary. 1992 "Civil Service Reform in Iarin Amer- tor. Lessons of East Asi World Bank County Study, ica: Lessons of Experience." Occasional Paper 6. Washington, D.C. World Bank, Latin Ameica Technical Department, Samudavania Chaian. 1992. "High Speed Growth Washington, D . and High PerD.Ciance in a Technoc.ric Poliy: The Remolona, E., and M. Lamberre. 1986. 'Financial Re- ai .n P acegin a TocreastAic o irT f[ormsaand the Balance of Payments Crisis: The Case of Thai Cae." Background paperofor Tec stAsarnMir- the Philippines, 1980-83." Working Paper 86-02. acs. World Bank, Policy Research Department, Philippine Institute of Development Studies, Manila. Washington, D.C Revel, Jack. 1987. "Mergers and the Role of Iarge San, G- 1990. "Enrerprise Training in Taiwan: Results Banks? Institute of European Finance, University fromdieVocationalTrainingNeedsSurvey? Econam- Colege of Norch Wales, Bangor, Wales. ic ofEducation Review 9(4): 411-8. Rhee, Ytmg Whee. 1989a. "Managing Entry into Intcrna- Saxonhouse, Gary. 1983. "What Is Al This about Indus- tional Markets: Lessons from the East Asian Experi- trial Targeting in Japan?" World Economvy 6(3): ence? Industry Series Working Paper 11. World Bank, 253-73. Industry and Energy Departmne Washington, D.C. Scdiff, Maurice. 1993. "Agriculture Pricing Policy and 1989b. "Trade Finance in Developing Economies." Growth in Asia.' World Bank, Policy Research De- Policy Research Series 5. World Bank, Washington, D.C. partment, Washington, D.C 385 Schiff, Maurice, and Alberto Vald&. 1992. Tle Political Singapore- Various years. YearbookofStatsticsandanEco- Eonomy of Agriulural Pricing Policy Volume 4: A nomic Survey of Singapore. Singpore Department of Synthesis of the Eamomics in Developing Countries. B al- Statistics. rimore, Md: Johns Hopkins University Pr. . 1983. Economic and Social Statistics 1960-1982. Schultz, T. Paul. 1988. "Education Invesrment and Re- Singapore: Department of Staiics. turns." In HoUlis B. Chenesy and T. N. Srinivasan, Siqgh. Ait. 1992. "'CIose' vs. 'SErategic' Integration with eds., Handhook ofDevelopment Economis voL 1. Arn- the World Economy and the 'Marker-Friendly Ap- sterdam: North Holland. proach to Developtrnent' vs. an 'Industrial Policy': A Scirovsky, Tibor. 1954. "Two Conceprs of Extemal Critique of the WorldDevelopment Report 1991 and an Economics." Journal of Politcal Economy 62(2): Altemative Policy Perspective." Faculty of Economics, 143-5 1. University of Canibridgec Cambridge, England. Scott, Maurice FG. 1979. "Trade." In W. Gal enson, ed., Singh, Ajit, and Javed Hamid. 1992. 'Corporate Finan- Economic Growth and Stral Canga in Taiwan cial Structures in Developing Countries." Technical Ithaca, N.Y.: Cornell University Press. Paper l International Finance Corporation, Washing- Scully, Michael T., and George J. Viksninsa 1987. Finazc- ton, D.C. ing EastAsias Succesr Comparatve Financal Develop Solow, Robe M 1956. "A Conterbution to he Theory menctinEght Asian Counris. London: Maanilan. ofEconomic6Growth." QuareeryJornalofconomics Shah, Anwar, ed. Forthcoming. Fiscal Incentives for In- 701)65-9.T vetcmentt in Devloig Coun Ne York Oxford -- 1957. "Tehnca Chang and die Agreae Pro- Uesnivrersity Press ducton Function.' Reizew of Economics and Sttstics Univerilty Prem ~~~~~~~~39(3): 3 12-20. Shah, Shekhar, and Subodh C. Madiur. 1992. Source Soon, Tek-Wongand Suan Tan. 1992. Public Policy of Labor bMarket Flexdbility in East Asia: A Review of andt Ecnomic Development in Singapore.' Manu- Selected Issues." Background paper for The EastAsin script prepared for a Wodd Bank Conference on the Mirrd, World Bank, Policy Research Department, Role of Goverment and East Asian Success, Wash- Washington, D.C. ingon, a.C. Shapiro, Helen, and Lance Taylor. 1990. "The State and Spence, M. 1981. 'Te Learning Curve and Competi- Industrial Strategy." World Development 18(6): on BeffJourd ofEconomics 12(l): 49-70. 861-78. Shea, J.D.Fothoin.DeeopenadstuSquire, Lvn. 1981. Employment Policy in Developing Shea, J. D. Forrhcoming. 'Developmcnt and Strruaural CCounS: A Survey of lssues and Evidence. New York- Change of the Financial System." In Patrick Hugh and Oxford University Pres Yung Chul Park, cds., Financia Development Japan Srinivasan, T. N. 1992. "Datbas for Development Korea, and Taiwan. New York: Oxford University Analysis: An Overview.' Paper presented at a Confer- Press. ence on Data Base of Dcvelopment Analysis, Yale Sheng, Andrew. 1989. "Bank Restructuring in Malaysia, University, New Haven, Conn., May. 1985-88." PRnE Worlkng Paper 54. World Bank, Fi- Stem, Joseph, Ji-Hong Kim, Dwighr Perkins, and Jung- nancial Sector Development Department, Washing- Ho Yoo. 1992. "Industrialization and dhe Stare The ton, D.C. Korean Heavy and Chemical Industry Drive." _. 1990. "The Art of Bank Resuucturing2 Isues and Research reporr Harvard Institute for International Techniques." World Bank, Financial Sector Develop- Developmenr, Cambridge, Mass., and Korea Develop- ment Department, Washington, D.C. ment Insitute, Seoul. Shouda, Yasutoyo. 1982. 'Effective Rates ofProteciion in Stevenson H., and J. Stigler. 1992. The Learning Gap. Japan." Nikd7on ieza Kmkyu 11 (March). New York. Summir. Shoven, John. 1988. Government Poicy Towards Industr Stewart, Francis, and Ejaz Ghani 1991- 'How Signifi- m the United Stares and Japan. Cambridge, U.K: cant Are Externaities for Development? WorldDevel- Cambridge University Press. opment l9(6): 569-94. 386 Stigicz Joseph E. 1993a. "The Role of the State in Fi- 1. 988. Satistical Yearbook of the Republic of Chinz nancial Markets." Paper prepared for the Annual Bank Taipei. Conference on Development Economics, Worldc 1992a.Ststical Yearbook of the Republic of Cbina. Bank, Washington, D.C. Taipei. . 1993b. "Some Lessons from the Asian Mirade." 1992b. StaSist of Sll and Medium Braines in Background paper fbr The East Asian Miracle. World T1iwab, R&asissic of mina. Taipei: Minisumy of Eco- Bank, Policy Rescarch Department, Washington, D.C. nomic Afneirs. Summers, Lawrence H. 1985. "Issues in National Savings Tan, JeePeng. and Akin Mingar. 1992. Edcation in Policy." NBER Working Paper 1710, Cambridge, Asia: A Comparative Study of Cost and Financing. Mass. Washington, D.C.: World Bank. Summers, Lawrence H., and V. P. Summers. 1989. Tankingoki. 198. "Alu m Rfni s "When Financial Markets Work Too Well: A Cau- Tanaka,Naok. 1988. AlumnumReFiningIndustry In rious Case for a^ Secuwriics Tmnsaceian Tax." In Regv- Ryuraro Komiya, Masahiro Okuno, and Koraro Suzu- uious Cse for a Securities T ransaction Ta." In Ret- mur, ed. Indrutnl Pvoliy of/Japan. New Yorkc Aca- &no'y Reforrn of Stock and Futures Markeet, a special dmim pes. issue of the Joumal of Financial Services Research. mc rss. Boston: Kiuwer Academic Publishers. Tanzi, Vito, and Parthsarathi Shome. 1992. "The Role of Sumnmers,Robert, andAan Heston. 1988. "A New Set of Taxation in the Development of East Asian Econ- International Comparisons of Real Product and omies." In Takaroshi Ito and Anne 0. Krueger, eds., Prices: Estimates for 130 Countries." Review oflncome hser Poliicagl Economy of Tax RPfrm Chicago: Uni- and Wealth 34(1): 1-25. versiry of Chicgo Press. . 1991. 'The Penn World Tables (Mark 5): An Ex- Teranishi, Juro. 1986. "Economic Growth and Regula- panded SeE of Inenational Comparisons, 1950-88.' tion of Financial Markcts: Japanese Experience dur- Quane4yJournal ofEconomics 105(2). ing the Postwar High Growth Period." Hitotsubashi Sun, C., and M. Liang. 1982. "Savings in Taiwan, Journal ofEcnomics 27: 149-65. 1953-1980." In Kwoh-Ting Li and Tzong-Shian Yu, Thailand. Various issues. Quarterly Bulletin. Bangkok: eds., Experences and Lessons of Economic Development Bank of TliailancL in Taiwan. Taipei: Institute of Economics, Academia Thomas, Vinod, and Yan Wang. 1993. "Government Sinica. Policies and Productivity Growti: Is East Asia an Ex- Sundavarej, Tipsuda, and Tniiranvorakul, Prasarn. 'Experi- ception?" Background paper for TheEastAsian Mira- ences of Fmancial Distrss in Thailand." PRE Working cle. World Bank, Policy Research Department, Paper 283. World Bank, Office ofthe Vice President for Washington, D.C. Development Economics, Washington, D.C Thurow, Lester. 1992. Head to Heatd The Coming Eco- Sussangkarn, Chalongphob. 1988. "Production Struc- nomic Battle Amongjapan, Europe, an/Amria.New tures, Labor Markets and Human Capital Investments: York: Morrow. Issues of Balance for Thailand." Thailand Develop- Trela, Irene, and John Whale,y. Forthcoming. "Taxes, incur Research Institute, Ofijprinr #7, Bangkok Outward Orientation and Growth Performance in Suzuki, Yoshio. 1987. Thejapanese Financial System. Ox- the Republic of Korea." In AnNvar Shah, ed., FiscalIn- ford: Clarendon Press. centives for Investment in Developing Countries. New Syrquin, Moshe, and Hollis B. Chenery. 1989. Paterns of York: Oxford Universiry Press. Developmenr, 1,950 to 1983. World Bank Discussion Trezise, P. H. 1983. "Industrial Policy Is Not the Major Paper 41. Washington, D.C. Reason for Japan's Success." Brookings Review 1(3): Taiwan, China. Various issues. Financial Statistics 13-8. Monrh4. Taipei: Cenial Bank of China. Tullock, Gordon. 1988a. "Why Did the Industrial Rev- Various years. Statistcal Yearbook of the Republic of olurion Occur in England?" In C. Rowley, R. Tolli- =Chinz Taipei: Directorate General of Budger, Ac- son, and G. Tullock, eds., The Political Econont of counting, and Statistics. Rent-Seeking. Boston: Kluwer Academic Publishers. 387 : _ .1988b. "Rents and Rent-Seeking." In C. Rowley, R. The Wall StreetJournaL 1990. "Taiwan, Long Notd for Tollison, and Gordon Tullock, eds., The Political Cheap Imitations, Becomes an Innovator." June 1. Economy of Rent-Seeking. Boston: Kiuwer Academic Wang, Fang-Yi. 1990. "Reconsidering the 'East Asian Publishers. Model of Development': The Link between Exports Turnham, David. 1993. Employment andDevelopmenetin and Productivity Enhancement, Evidence from Tai- New Review ofEvidence. Paris: Development Center, wan, 1950-87." Ph.D diss. University of California at OECD. Los Angeles. Ueda, Kazuo. 1992. Insitutiona and Regrulatory Frame- Warr, Peter G., and Bhanupong Nadhiprabha. 1993. worksfor the Main Bank System. Tokyo: University of "Macfocconomic Policies, Crisis, and Growth in die Tokyo. Long Run: Thailand." Background paper prepared for UNESCO. Various years. Statistical Yearbook. Paris. Litde, Ian M. D., Richard N. Coopcr, W. Max Cor- den, and Sarath Rajapathirana, cds., Boom, Crisis, Ad- United Nations. Various years, a. Demographic Yearbook. justnent- The Macroeconomic Expience of Developing New York. Countries. New York: Oxford University Press. _. Various years, b. Eneg Statistics Yearbook. New Wei, Shang-Jin. 1993. 'Open Door Policy and China's York. Rapid Growth: Evidence from City-Level Data." Pre- _. 1976. World Energy Supplies 195o-74. New York. pared for the Fourth Annual East Asian Seminar on _- S1992. World Economic Sutnvy: Current Trends and Economics, National Bureau of Economic Research, Policies in the World Economy. New Yorkl Cambridge, Mass. uNop (United Nations Development Programme). 1991. Westphal, Larry E. 1978. "The Republic of Korea's Expe- Human Development Reporr. New York: Oxford Uni- rience with Exporr Led Development." World Devel- versity Press. opmnent6: 347-82. Van Adams, A. 1989. "A Study of the Economic Envi- Westphal, Larry E., Yung Whee Rhee, and Garry Pursell. ronnent's Influence on Human Capital Formation in 1988. Korean Industrial Competenre: Wher It Came Korea and Egypt." George Washington University, Fmron World Bank Staff Working Paper 469. Wash- Washingon, D.C ington, D.C. Viner, Aron. 1988. Inside Japanese Finaial Me. Whalley, John. 1990. "Non-Discriminatory Discrimina- Homewood, Ilk: DowaJoneserFin. don: Special and Diffierential Treatrnent under the Homewood, IL: Dow Jons--Invin.CArr for Developing Countries." Economic Journal Vitts, Dimitri. 1991. 'Measuring Commercial Bank Ef- 100(403):1318-28. ficicncy: Use and Misuse of Bank Operating Ratios." Wiboonchutikula Paitoon. 1987. "otal Factor Produc- Policy Research Working Paper 806. World Bank, tivity Growth of Manufacmring Industries in Thai- Country Economics Department, Washington, D.C. land." In Productivity Changes and International -_ .1992. 'Poliy Issues in Financial Regulation." Competitiveness of ThaiIndustries. Bangkok Thailand World Bank, Country Economics Department, Development Research Institute. Washington, D.C. April. Winkler, Donald R. 1990. Higher Edruation in Latin Vogel, Ezra F., and David Lindauer. 1991. "Toward a So- Amerca: Issues of Efficiency and Equity. World Bank cial Compact for Korean Labor." In David Lindauer, Discussion Paper 77. Washington, D.C. Jong-Gie Kim, Joung-Woo Lee, Hy-Sop Lim, Jac- Wolf, Charles. 1988. Markets or Governments: Choosing Young Son, and Ezra F. Vogel, eds., Korec The Strains between Impect Altenzetives. Cambridge, Mass: Mrr of Economic Growt. Cambridge, Mass.: Harvard In- Press. stiture for International Development. World Bank. 1983. "Special Economic Reporc Policies Wade, Robert. 1989. 'What Can Economies Learn from fior the Financial Secaor." Europe, Middle East, and East Asian Success?" Annab oft heAmerican Academy of North Africa Region, Washingron, D.C Political Science 505: 68-79. . 1987a. Korec Managing the Imndrial Transition 2 . 1990. Goening the Market Economic Tho?y and volt Washington, D.C. the Role vfthe Government in EastAsian Industriai- _. 1987b. WorldDevelopmentReport 1987 New York.: tiomz Princeton, NJ.: Princeton University Press. Oxford University Press. 388 _. 1989a. 'Indonesia: Strategy for Growth and Struc- Trade Policy and the New International Econonmia. tural Change." Counny Department V, Washington, Cambridge, Mass.: MIT Press. D.C _ 1985. The Politial Economy of fapan. Stanford, 1989b. Malaysia: Matching Risk and Rewards in a Calif.: Stanford University Press. MixedEconomy. Washington, D.C. Yamawaki, Hideki. 1988. 'The Steel Industry." In Ryu- _. 1989c. World Development Report 1989. New York: taro Komiya, Masahiro Okuno, and Komaro Suzu- Oxford University Press. mura, eds., Indusrial Poliy of Japan. New York: 1. l990a. Indonesic SterategyJfr a Sustained Reduction Academic Press. in Povety. Washington, D.C. Yarnazawa, Ippie. 1988. "The Textile Industry." In Ryu- . 1990b. World Development Report 1990. NewvYork: taro Komiya, Masahiro Okuno, and Kotaro SuZU- Oxford Universiy Press. murm, eds., Industrial Polic of Japan. New York: 1. l99a. Colombia: Industrial Compettion and PJefor- Academic Press. -mance. Washington, D.C. Yan, Lin See. 1991. 'Savings Investment Gap, Financing -. 199lb. World DevelopmentReport The Chalenge of Needs and Capital Market DevelopmenL" Bank Ne- Development. New York: Oxford University Press. gara, Malaysia. September. -. 1992a. "China: Reform and the Role of the Plan in Yaron, Jacob. 1992. Successfid Rural Finaee Insitutions. the 1990s.' East Asia and the Pacific Regional Office, World Bank Discussion Paper 150. Washington, D.C. Washington, D.C Yonezawa, Yoshie. 1988. YThe Shipbuilding Industry." -. 1992b. Global Economic Prospects and the Developing In Ryuraro Komiya, Masahiro Okuno, and Kotaro Counties. Washington, D.C. Suzumur eds., Induserial Poliy ofJapa;L New York: . 1992c. World Bank Stpportfor Industridizaton in Academic Press. Korea, Indi, and Indonesia. Washington, D.C. Yoshino, Naoyuki. 1993. "The Low Interest Rate Policy -. 1992d. World Development Report 1992. New York: and Economic Growth ofJapan." Department ofEco- Oxford University Press. nomics, Kelo University, Tokyo, Japan. 1 993a. GlobalEconomicProspectsandtheDeveloping Young, Alwyn. 1992. "A Tale of Two Cities: Factor Ac- Counties. Washington, D.C cumulation and Technical Change in Hong Kong and . 1993b. World Developmnt Report 1993: Investing Singapore." Paper presented at a World Bank seminar. pmenth Rrport 1993: rdzwfng inii PFrbebruary. Health. Ncw Yorkm Oxford University Press. Yusof, Z A., A. A. Hussin, I. Alowi, S. Singh, and L C. W porld EconomicForum. 199eneWorldvCompetitiveness Sing. Forthcoming& The Impact of Financial Reform Repor- Geneva. in Malaysia" In Gerard Caprio Jr., Izak Ariyas, and Yaamamura, Kozo. 1986. "Caveat Enmptor The Industrial James Hlanson, eds., Financial Refrnrn Theory and Ex- PoliCy of Japan." In Paul Krugman, ed., Strategic perience. Cambridge: Cambridge University Press. The complete backlist of publications from the World Bank is shown in the annual hIder ofPublications, which contains an alphabcical tide list and indexes of subjects, authors, and countries and regions. The latest edition is available fiee of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, or fiom Publications, The World Bank, 66 avenue d'Idna, 75116 Paris, France, 389 J~~~~~ < ~~~~ ~ ~~~ ~ ~~~~~C C,,, I] $ ~~~~~~~~~ClHINA6 r ~ ~ ~ ~ ~ N r - 1) C't ,.. - 6r- j, >~~~~~~~~~~~~~~~L , _-. - r j2kmmx~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ %L:0 be fI/ t . - - - D D ;, ;; - ..~~~~- *. SfDN I Ais,~ ~ ~ ~~~~~~- 1