.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4!~11 44 *tT *I L,' 1. O'i;

EAi .

("'t M # k A C l., ,,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~WI,-rit,;t4

'' I!WSbi7S'XtWs6Ni SHS'WlIF .8e,.l.e iXS4

NE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4

F ~~~~~~~~~~~~~~~~~~~~~~~~~~~1

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

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

TheEast AsianMiracl eEconomic Growthand Public Policy

Published for the World BankCxCORD UNIVERSrrYPRS

Oxford Univeirm Pressw

OXFORD NEV YORK TORONTODELI BOMBaY CALC;UJrA MADRAS KARACHikUAA LUMPUR SINGAPORE HONG KONG TOKYONAIROBI DAR ES SALAAM CAPE TOWNMELBOURNE AUCKLAND

and asidad compania in

BERLIN IBADAN

0 2993 7lmriarmnd Ba&nkfor ReomtruawnwdDevaop mewl OPJE U'2LD HAAh1818 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 UnirsiyPreazA(1rgth raerd Noparjrh5pub&icatum 6be reprw l .woredin a rrtrudqsfemor Drwmm&mc4 in any farm or by any meas, deamn, merhank4 pumwpy4rewonmg& ore.z se,i w etvepdorpermrnuonfxf0uirdUriviy Prm

Mamftiazdn she Unisd s /AmeriaFon paning&Spemner 1993

Ciogun ph s-At the wp. a ricefud inJava, Indoniaig; c resy ofMatricrAiser. Atr 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 Misbok 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 thWorld Bank its affiiner or isr Bad ormember conries onwMtming the egl as1 0f ayewonM, mq rroj cirt or or /t&e anthrifia xref, or rcwemning rthedeimiwdonqfirs boitames or in natiolaie

rfCwFs Caragtoing-Publadon Dar

The EasAswaw mirade w :omwgrvww5 =d; bicpAlp. cn.-(MordBankpdiy r=* reponw)

Incdaw r 3uietpieal ner,IrN 0-19-5209934IL. AsEa;rnee-Ecwmic pli 2- tE a-Econidc cendo-,

3. Asi Sawi&s-Ecenamicpole 4. Ask CitiomSL. WorldBank f. Se,ieHC4625.A275 1.993338.95-c20 9S33046&

ismv 1G20-0851

Torprepaper anom the Ameria Natnax SxdrdfrPawzw*w wfP1pfr1Jiwd MreLw ialr Z39.48-1984

Foreword

EBATES ON THE APPROPRIATE ROLE OF PUBLIC POUCY IN

economic development have occupied policymakers andscholars since the study of developing economnies bcgan in

earnest at the dosc of World War IL. The success of many of theeconomies in EastAsia in achieving rapid and equitable growth, often inthe context ofacivist public policies, raises complex questions about therelationship between government, the private sector, and the markernSeningly, the rapidly growing economies in East Asia used many of thesame poficy instruments as othcr developing economies, but withgreater success. Understanding which policies contribuced co their rapidgrowth, 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 inEast Asia.

This volume is the summary ch. appearsas the first in a series of Policy Rcsearch Reports, which arc intended tobring to a broad audience the results of research on development policyissues 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 andcIealy 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; andthey 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 methodsused 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 theseeconomies Were also better able than most to allocate physical and

buman resources to highly productive investrents and to acquire andmaster txchnology. In this sense there is nothing 'miraculous" about theEast 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 ofpolicies, 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 becountry-specific, if they are to be effectivec But there arc also some com-mon threads among the high-performing East Asian economies. Theauthors conclude that rapid growth in each cconomy was primarily dueto 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 otherrcsearch that has sressed the essential need for dcveloping economics toger the policy fundamentals igIL The research also firther supports thedesirability of a two-track approach to development policy emphasizingmacroeconomic stability on one hand and investments in people on theocher. The importance of good macroeconomnic management andbroadly based educational systems for Easr Asia's rapid growth is abun-dandly demonstrated.

The reporr also breaks some new ground. It condudes that in someeconomics, nainly those in Northeast Asia, some selective interventionscontibuted o grwth. and it advances our undersmading of the condi-tions required for interventions to succeed. The authors argue tharwhere selective intervcntions succceded they did so because of th o es-sential prcrequisites. First, they addressed problems in the functioningof 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-performancecriteria 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 theirsuccess 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 policiesdid not succeed in mrany other economies, the report also leaves unan-swered many important questions. The market-oricnted asj ctxs of EastAsia!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 EastAsian 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 inthese areas will continue beyond this report.

The support of the Government ofjapan for the research program onthe high-performing Asian economics is gratfilly acknowledged. Thereport is a product of the staff of the World Bank, and the judgmentsmade herein do not necessarily reflect the view of its Board of Directorsor the governments they rcprescnt.

Lewis T. PrestonPresidentThe World Bank

Augur 1993

vii

Contents

7Te Rear Team xi

Ack3wwkdgmnae xiii

Defmiufons xv

Overview-r The Making of a MiracleThe Essence of the Mirale: Rapid Growth with Equity 8Policies for Rapid Growth in a Changing World Economy 23Note 26

1 Growth, Eqrity, and Economic Change 27Rapid and Sustained Economic Growth 28Declining Income Inequality and Reduced Poverty 29Dynamic Agricultural Seaors 32Rapid Growth of Exports 37Rapid Demographic Transitions 38High Investment and Savings Rares- 40Creatng Human Capital 43Rapid Productivity Growth 46Appendix 1.1: Accounting for Growth 60Appendix 1.2: What Do Tess of Cogniive Skills Show?- 70Notes 76

2 Public Policy and Growth 79PolicyExplanations 81The Functional Growth Framework 87Notes 103

3 Macroeconomic Stabiity and Export Growth 105Pragmatic Orthodoxy in Macroeconomic Management 106Creatng an Export PuLsh 123Appendix 3.1: Economric and Political Timedines 148Notes 156

ix

4 An Insitutional Basis for Shared Growth 157Achicving Legitimacy through Sharcd Growth 158Insulating the Economic Technocracy 167Wooing Big Business 181Notes 189

5 Strategies for Rapid Accumulation 191Explaining EastAsia's High Human Capital Forrnation 192Explaining East Asia's High Savings Rates 203Explaining East Asia's High Investment Rates 221Appendix 5.1: Granger Causaliy Tests forSavings

Rates and Growth Rates 242Appendix 5.2: Technical Note on the Relationship beween

Interest Rates and Growth 245Notes 255

6 Using Resources Efficienty: Relying onMadrets and Exports 259Explaining EastAsiais Efficient Resource Use 259Using the Marken Labor Markets in East Asia 261Assisting the Markct Financial Markets and Allocation 273Using the Internarional Market Trmde and Industrial Policy 292How Manufactured Exports Increased Productiviy 316Appendix 6.1: Testing dhe Impact of Indusaial

Policy on Productivicy Change 326Appendix 6.2: Tests of the Relationship berween TFP

Change and Trade Policies 337Notes 342

7 Polices and Pragmatism in a Changing World 347Foundations of Rapid Growdh-Getting dhe

Fundamentals Right 347Creating Institutions to Promote Growdi 352Intervening in Markets 353Note 368

Bibligraphic Note 369

The Research Team

T HIS POLICYRESEARCH REPORTWAS PREPARED BYATEAM LED

by John Pagc and comprising Nancy Birdsall, Ed Campos, WMax 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 ofthe seven devcloping high-performing Asian economies werc under-taken under the direction of Danny Leipziger. The tcam was assisted byMaria Ltiisa Cicogniani, Varuni Dayaratna, Leora Friedberg, Jay Gon-zalez, Jennifer Keller, May Khamis, Sonia Plaza, Myriam Quispe, CarolStrunk, and Ayako Yasuda. The work was initiated by Lawrence H.Summers and was carried our under the general direction of NancyBirisall.

The editorial-production team for the report was led by AlfredImlhof The support staff was headed by Sushrma Rjan and includedMilagros Divino, Jan-Marie Hopkins, Anna Marie Maranon, and LindaOehler Polly Means gave graphical assistance. Bruce Ross-Larson andMeta dc Coquercaumont provided additional editorial support TheMap was designed by jeffrey N. Lecksel. Mikla Iwasaki coordinatcdwork in Japan.

XI

Acknowledgments

REPARATION OF THIS REPORT DREW ON SLVERAL OTHER RE-

laced research projects. Studies of the Japanese main bank sysrcnand 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 onsupport systems forsmall and medium-size enterprises in Asia, andJohnPage coordinated projects on lapanese perspectives on public policy dur-ing the rapid growth period, and tax policy and tax administration inEast Asia.

Many individuals inside and outside the World Bank provided valu-able contributions and comments. (Specific acknowledgmcnt of theirefifrts is made in the Bibliographic Note at dce end of the book.) Par-ticular thanks are due ro the following senior officials of the economiesstudied who reviewed and comment"d on an earlier drafc Seung-ChulAhn, Isao Kubota, Shijuro Ogata, J. B. Sumarlin, Teh Kok Peng,Chilao Tsukuda, Tan Sri Dato Lin See Yan, and Cesar Virata. VinodThomas comnmented on several drafts and coordinated the comments ofhis 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. JohannesLinn, 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 theWorld 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 nationalF 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 ofeach income group may change from one Bank publication to the nextOnce the dassification is fixed for any publication, all the hisrorical dataprescnted are basod on the same economy grouping. The income-basedeconomy groupings used in this sEudy are defined as folows

* Low-income economies are thosc with a GNP pcr capit of $635 orless in 1991.

* Middle-inme ecoomzies are those with a GNP per capita of morerhan $635 but less than $7,911 in 1991. A fijrther division, atGNPper 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 $7,911or more in 1991.

* Word comprises all economies, including economies with sparsedata and those with fewer than 1 million people.

Low-income and middle-income economies axe sormetimes referredto as developing economies. The use of the term is convenient; ir is notintended to imply that all economies in the group are experiencing sim-ilar development or that other economies have reached a preferred orfinal stage of development. Classification by income does not necessar-ily reflect development status.

'V

-Anabtical Groups

F OR ANALYIICAL PJRPOSES, THIS STUDY GROUPS ECONOMIESinto 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 tothe duration of their succcssfiul record of economic growth:

. TheF : vur igers usually identified as Hong Kong, thc Republicof Korea, Singapore, and Taiwan, China, have been growingrapidly for decades and have joined or approached the ranks ofhigh-income economies.

* The newly industrialidZWg economies (NIEs) are Indonesia,Malaysia, and Thailand. They have joined the group of HPAEsmore recently, within the last two decades.

* EastAia comprises all the low- and middle-income economies ofEast and Southeast Asia and the Pacific, east of and includingChina and Thailand.

* South Asizr comprises Bangladesh, Bhutan, India, Myanmar,Nepal, Pakistan, and Sri Lanka.

* S&zb-&harn Af5ca comprises all econoniies south of the Saharaincluding South Africa but ecduding Mauritius, Reunion, andSeychelles, 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 theforrner Yugoslavia, and all the economies of North Africa and theMiddle East, and Afghanisan.

m Latin America and the GCaibcan cDmprises all American andCaribbean economies south of the United States.

The regional grouping of economies in occasional parts of the text ortables may differ from that used in the main tt of this study, describedabove. 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 BankpubLications because of continuous updating as better data becomeavailable, 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 indexDFI Direct foreign investmentGDP Gross domestic productGNP Gross national productrlPAEs High-performing Asian economnies1MF International Monetary FundISIC Intemational Standard Industrial ClassificationmI Ministry of Trade and Industry JapanNIs Newly industrializing economiesOECD 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, UnitedKingdom, and United States)

PPP Purchasing power parityR&D Research and devclopmentTFP Total factor producdivityUNDP United Nations Development ProgrammeUNESCo United Nations Educational, Scientific, and Cultural

Organization

-n.

Overview: TheMaking of a Miracle

AST ASIA HAS A REMARKABLE RECORD OF HIGH AND

sustained economic growdh. From 1965 to 1990 thetwenty-three economliies of East Asia grew faster than allother regions of the world (figure 1). Most of thisachievement is attributable to seemingly miraculousgrowth 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 theoriginsof their successful growth are necessarily arbitnary processes.' Botswana,Egypt, Gabon, and Lesotho in Sub-Saharan Africa have also beenamong the world's top growth performers in the past two decades, ashave such diverse economies as Brazil, Cyprus, Greece, and Portugal (seefigure 2). Why focus on eight economies in East Asia? In part the choicereflects 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 numberof 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 significantdevelopment 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 comparisonis problematic. We touch on China's recent development in chapters 1 and 3. Te economic 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 6GNP 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 eighteconomies do share some economic characteristics that set them apartfrom most other developing economies.

Since 1960, the HPAEs have grown more rhan twice as fist as the restofEast Asia, roughly three times as fast as Latin America and South Asia,and five times faster than Sub-Sahann Africa. They also significandyoutperformed the industrial economies and the oil-rich MiddleEast-North Africa region. Between 1960 and 1985, real income percapita increased more than four times in Japan and the Four Tigers andmore than doubled in the Southeast Asian NEs (see figure 2). If growthwere randomly distmbuted, there is roughly one chance in ten thousandthat success would have been so regionally concentrated.

The HPAEs have also been unusually successfiul at sharing the fiuits ofgrowth. Figure 3 shows the relationship between the growth of gross do-mestic product (GPD) per capita between 1965 and 1990 and changes inthe Gini coefficient, a statistical measure ofthe inequality ofincome dis-tribution. The HPAEs enjoyed much higher per capica income growth atthe same time that income distribution improved by as much or morethan in other developing economic% with the exceptions of Korea andTaiwan, 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, ChinaIndonesia

Hong Kong . -

SingaporeRep. of Korea

JapanMatta

LescothoEgypt

CYPMUS . oGabon GreeceBrazilSyria

Portga :wA

MalaysiaYugoislavia

China __ _ - - - ~ -

Thailand __ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

NorwayItay

Spain ~-.-.-.---Israel

Jordan ___

Germnany.. .

MexicoPakistanColombia .Middle 20:

United Statbes United Kingdom ----- ________- _

Philippines ----- ,.-----.------------

Venezuiela .-. ..

Keniya . -.

PeauArgentina

Uganda___India __ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _

Benin__ _ _ _ _ _ _ _ _

SenegalHaiti

UberlaSudan

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 45Percentage 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 thibdA 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 fimrde 1960s to the 190s; a negative number indicates thar income becume lss concentraredL The decade avcge is used because data are availablefor 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, Japanand 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 56years in 1960 to 71 years in 1990. (In other low- and middle-income

economies, life expectancy also rose considerably, fiom 36 and 49 to 62and 66 years, respectively.) In the HPAEs, the proportion of people livingin absolute poverty, lacking such basic necessities as dean water, food,and shelter, dropped-for example, from 58 percent in 1960 to 17percent in 1990 in Indonesia, and from 37 percent to less than 5 percent

4

in Malaysia during the same period. Absolute poverty also declined inother 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. Ahost of other social and economic indicators, from education to appli-ance ownership, have also improved rapidly in the HPAEs and now are atlevels that sometimes surpass those in industrial economies.

'What caused EastAsia's success? In large measure the HPAEs achievedhigh growth by getting the basics right. Private domesric investmentand rapidly growing human capital were the principal engines ofgrowth. 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 otherparns of the developing world. And some of these economies also got ahead strt because they had a better-educated labor force and a more ef-fecrive system of public administration. In this sense there is littde thatis Kxmiraculousr about the HPAEs' superior record of growth; it is larglydue to superior accumulation of physical and human capital.

Fundamentally sound development policy was a major ingredient inachieving rapid growth. Macroeconomic management was unusuallygood and macroeconomic perfornance unusually stable, providing The

essential firnework for private investment. Policies to increase theintegrity 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 rapidincreases in labor force skills. Agriculural policies stressed productivityand did not tax the rural economy excessively. All the HPAEs kept pricedistortions within reasonable bounds and were open to foreign ideasand technology.

But these fimdamental policies do not tell the entire story. In most ofthese 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 selectedindustries, keeping deposit rates low and maintaining ceilings on bor-

.rowing rates to increase profits and retained earnings, protectingdomestic import substitutes, subsidizing declining industries, establish-big and financially supportng government banks, making publicinvestrments 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 industrieswere promoted, while others were not.

At least some of thesc interventions violate the dicturn of establishingfor the private scaor a level playing field, a neutral incentives regime. Yetthese strategies of selective promotion were closely associated with highrates of private investmenc and, in the fastest-growing economies, highrates 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, weknow already that their interventions did not significantly inhibitgrowth. But it is very difficult to establish statistical links betweengrowth and a specific inr.w..tion and even more difficult to establishcausalityr Because we cannot know wrhat would have happened in theabsence of a specific polii); it is difficult to rest whether interventionsincreased growth rates. Other economies attempted similar interven-tions without success, and on averge they used them more pervasivelythan in the HPAEs. Because the HPAEs differed from less successfiileconomies both in their doser adherence to policy findamentals and inthe 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 interventionsthat contributed to growth from those that were either growth-neutralor 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 andmore equal growth than otherwise would have occurred. However, theprerequisites for success were so rigorotus that policynakers seeking tofollow similar paths in other developing economies have often met withfailure. What were these prerequisites? First, governments in NortheastAsia developed institutional mechanisms which allowed them to esablish dear performance criteria for selective interventions and to monitorperformance- Intervention has taken place in an unusually disciplinedand performance-based manner (Ainsden 1989). Second, the costs ofinterventions, 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 onthe ability ofthe Minisrry of Intcrnational Trade and Indusry to carryout subsidy policies, and in Indonesia and Thailand balanced budgetlaws 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 ofprudent macroeconomic management. Price distortions arising from se-lective interventions were also less extreme than in many developingeconomies.

In the newly industrializing economies of Southeast Asia, govem-ment interventions played a much less prominent and frequently lessconstructive role in economic success, while adherence to policy finda-mentals remained importamt These economies' capacity to administerand implement specific interventions may have been less than in North-east Asia. Their rapid growth, moreover, has occurred in a very differentinternational economic environment from the one that Japan, Korea,and Taiwan, China, encountered during their most rapid growth. Thusthe problem is not only to try to understand which specific policies mayhave contributed to growth, but also to understand the institutional andeconomic circumstances that made them viable. Indeed, the experienceof the Southeast Asian economies, whose initial conditions parallelthose 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 growthwith equity and uscs economic models to attempt to account for thisgrowth. Chapter 2 reviews policy explanations for East Asia's economicsuccess 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 ofpolicies that led to two impormnt characteristics of the HPAEs' economicperformance: macroeconomic stability and rapid growth of manufac-tured exports. Chapter 4 disLcusses the role of institutions. Chaprer 5looks at the role of public poliqc in the HPAEs' unusually rapid accurnu-lation of physical and human capital, while chapter 6 analyzes the meansused to achieve efficient allocation of resources and productivity growth.Chapter 7, in condusion, assesses the success of East Asian polices andtheir applicabiliy in a changing world economy. The remainder of this

7

-' -.-- V IICACLE

overview parallels the organization of the book, highlighting the centralarguments 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 sharedcharacteristics cause them to be grouped together and set apart

from other developing economies? First1 as we noted above, they hadrapid, 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 inthat 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, andunusually 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 highrates ofprivate investment, combined with high and rising endowmentsof 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 quiteunusuaL In fact, productivity growth in the FHAEs exceeds that of mostother 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 rapidlyaccumulate human and physical capital and to allocate those resources tohigh-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 entirelyaccount for the high-performing economies' success, as the presence ofunsuccessful economies in the same region attests. Among the variety ofpolicy explanations, two broad views have emerged (see dcapter 2).

Adherents of the neoclassical view stress the HPAEs' success in getting thebasics right. They argue that the successful Asian economies have beenbetter than others at providing a stable macroeconomic environment anda 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 distortionarypolicies 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 capitalin the HPAEs' success.

Adherents of the renisionist view have successfilly shown that EasrAsia does not wholly conform.to the neodassical model. Industrial pol-icy and interventions in financial markets are not easily reconciledwithin the neoclassical fiamework. Some policies in some economies arcmuch more in accord with models of state-led development. Moreover,while the neoclassical model would explain growth with a standard setof 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 theneoclassical 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 overalleconomy. In East Asia, the revisionists argue, governments remediedthis by delibemtely "getting the prices wrong"-altering the incentivestruccure-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 itsproponents have not claimed to establish that interventions per se accel-crated growth. Moreover, as we shall show, some important govemmentinterventions in East Asia, such as Korea's promotion of chemicals andheavy 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. Thusneither 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 rapidgrowth in developing countries has been associated with effcctive butcarefully limited govcrnment activism. In the 'market-friendly" strategyit articulates, the appropriate role of government is to ensure adequateinvestments 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, governmentsare likely to do more harm than good, unless interventions arc markerfiendly. On the basis of an cxhaustive review of the experience of devel-oping economies during the last thirty years, it condudes that attemptsto guide resource allocation with nonnarket mechanisms have generallyfailed to improve economic performance.

The market-friendly approach captures important aspects of EastAsia's success. These economies are stable macroeconomically, have highshares of international trade in GDP, invest heavily in people, and havestrong 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 thesenortheastern 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, eachof the HPAEs rnaintained macroeconomic stability and accomplishedthree fiunctions of growth: accumulation, efficient allocation, and rapidtechnological 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 arethose 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 thatpush nontraditional exports. We try to understand how governmentpolicies, borh fundamental and interventionist. may have contributedto 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 filuresin the working of markets. Otherwise, markets would perform theallocation function more efficiently. We idenrify a dass of economicproblems, 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 coordinationproblems-responses that emphasized cooperative behavior among pn-vate firms and clear performance-based standards of success.

Competitive discipline is crucial to efficient investment. Mosteconomies employ only market-based competition. We argue that someHPABs 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 simplenonmarket allocation rules, such as access to rationed credit for ex-porters, to very complex coordination of private investment in thegovernment-business deliberation councils of Japan and Korea. The keyfeature of each contest, however, is that the government distributesrewards-often access to caedit or foreign exchange-on the basis ofperformance, which the government and competing firms monitor. Tosucceed, selective interventions must be disciplined by competition viaeither markets or contests.

Economic contests, like all others, require competent and impartialreferees-that is, strong institutions. Thus, a high-quality civil servicethat has the capacity to monitor performance and is insulated from po-litical interference is essential to contest-based competition. Of course, ahigh-quality civil service also augments a government's ability to designand implement non-contest-based policies.

Our framework is an effort to order and interpret inlformation. Weare 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. Pragmaticflexibility in the pursuir of such objectives-the capacity and willingnessto 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 inwhich the HPAEs achieved tvJ important objectives: macroeconomicstability and rapid export growth (sec chapter 3).

Achieving Macroeconomic Stabilty and Export Growth

More than most developing economies, the HPAES were characterizedby responsible macroeconomic management. In parricular, they gener-ally limited fiscal defcits to levels that could be prudently financedwithout increasing inflationary pressures and responded quickly whenfiscal pressures were perceived to building up. During the past thirtyyears, annual inflation averaged approximately 9 percent in dteseeconomies, compared with 18 percent in other low- and middle-incomeeconomies. Because inflation was both moderate and prediaable, realinterest rates were far more stable than in other low- and middle-incomeeconomies. Macroeconomic stability encouraged long-term planningand private investment and, through its impact on real interest rates ai, 4the real value of financial assets, helped to increase financial savings. TheHPAEs also adjusted their macroeconomic policies to terms of tradeshocks more quickly and effectively than other low- and middle-incomeeconomies. 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 periodsof exchange rate protection. Adjustments to exchange rates in otherHPAEs-validated bv policies that reduced expenditures-kept themcompetitive, despite differential inflation with trading partners.

In addition to macroeconomic policies, the HPAEs used a variety ofapproaches to promoting exports. All (except Hong Kong) began with aperiod of import substitution, and a strong bias against exporEs. Buteach moved to establish a pro-export regime more quicldy than otherdeveloping economies. First Japan, in te 1950s and early 1960s, andthen 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 withmoderate but highly variable protection of the domestic marker. A widevariety 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 creditand supportinginstitutions. In these economies exportdevelopment hasrelied less on highly selective interventions and more on broadly basedmarket incentives and direct foreign investment.

Building the Instutional Basis for Growth

Some economists and political scientists have argued chat the EastAsian miracle is due to the hiigh quality and authoritarian nature of theregions institutions. They describe East Asian political regimes as "devel-opmental statee' in which powerful technocratic bureaucracies, shieldedfrom political pressure, devise and implement well-honed interventions.We believe developmental stare models overlook the central role ofgovernment-private sector cooperation. Whle leaders of the HPAEs havetended to be either authoritarian or paternalisdti they have also beenwilling to grant a voice and genuine authority to a technocratic elite andkey leaders of the private sector. Unlike authoriruian leaders in manyother 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 thesupport of the sociey at large, East Asian leaders established the princi-ple of shared growth, promising that as the economy expanded allgroups would benefit. But sharing growth raised complex coordinationproblems. 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 thecooperation of the middle class and the poor, the leaders had to showthem that they would indeed benefit from future growth.

Explicit mechanisms were used to demonstrate the intent that allwould have a share of future wealth. Korea and Taiwan, China, carriedout comprehensive land reform programs; Indonesia used rice and fer-filizer price policies to raise rural incomes; Malaysia introduced explictwealth-sharing programs to improve the lot of edtic Malays relative to

'3

- IRfLMIRACLE

the better-off ethnic Chinesc; Hong Kong and Singapore undertookmassive 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, theseprograms demonstrated that the goverment intended for all to sharcthe 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 torecruit 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 ineach economy, economic technocrats helped leaders devise a credibleeconomic strategy.

Leaders in the HPAs also built a business-friendly environment. Amajor element of that enviromnenr was a legal and regulatory strucmrethat was generaly hospitable to private investment. Beyond this theHPAiEs have with varying degrees of success enhanced communicationbetween business and government. Japan, Korea, Malaysia, and Singa-pore have established forums, which we call deliberation councils, inwhich private sector groups are invited to help shape and implemnent thegovernment 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 rulesdear to all participants.

Using Deliberation Council. In Japan and Korea technocrats used delib-eration councils to establish contests among firms. Because the prvatesector participated in drafting the rules, and because the process wastransparent to all paricipants, private sector groups became more will-ing participants in the leadership?s development efforts. One by-productof these contests was a tendency to reduce the private resources devotedto 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, andbetween management and labor. The councils thus supplemented themarket'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 Japanand Korea to a more indicatve and consensus-building role, along finc-tional as opposed to industry-speific lines. In Malaysia the councils appearto 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 EastAsian governments have changed with changing circumstances.

Accumulaing Human and Physical Capital

Drawing on the strength of their institutions, East Asian economiesused a combination of fundamental and interventionist policies toachieve rapid accumulation of human and physical resources. Funda-mentals included such traditional government obligations as providingadequate 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 interms of human capital and have since widened their lead over other de-veloping economies. In the 1960s, levels of human capital were alreadyhigher in the HPAEs than in other low- and middle-income economies.Governments built on this base by focusing education spending on thelower grades; first by providing universal primary education, later by in-creasing the availability of secondary educations Rapid demographictransitions facilirated these efforts by slowing the growth in the numberof school-age children and in some cases causing an absolute decline.Dedining ferdlity and rapid economic growth meant that, even wheneducation 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 HPAEsimported educational services on a large scale, particularly in vocation-ally and technologically sophisticated discplines. The result of thesepolicies has been a broad, technically inclined human capital b±ve well1--stited to rapid economic development.

.IHPAE education policies also contributed to more equitable incomedistributions. To be sure, initial conditions helped to set up a virtuouscircle: initial low inequality in income and education led to educational

I5

.:~~~~~~~~~~~~~~;n

11 :t4iSjA-N MIRACLE

expansion, which reinforced low inequality. In addition, by focusingspending on primary and secondary education, and leaving demand fortertiary education to be largely met by' a self-financed private system,governments served large segments of the population that otherwisewould 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 instiutionsFigu 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:ederolRepublic of Gcmany bebore reunificaion, welfare c -r. On the other, the benefits are apparent in the rapid growth ofGreece, Iceland, Iredand. Imly, Luxem- these ecovomies. Savings, forced or not, generaed high payoffs based onbour, dte Netherlands, Norny, Portugal. consistendy high rates of return to investments. In conwnt to otherSpain, Sweden, Switxrland, and theUnited 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 aPakistan, Pem, Sri Lnka, Turkey, better job than most developing econom>es at creating -infastructure

Urugua. Ycnzuels the fone upEraUrgan y Venezuela. the YugosIiaI that was complementary to private investment. Seco.d, they created anand 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 termedfinancial repression.

Japan, Korea, Malaysia, Thailand, and Taiwan, China, had extendedperiods of mild financial repression. To be sure, increasing interest ratesfrom negative to zero or mildly positive real rates and avoiding fluctua-tions (by avoiding unstable inflation) encourages financial savings. Butbecause savings are nor very responsive to marginal changes in positivereal interest rates, HPAE governments were able to mildly repress interestrates on deposits with a minimal impact on savings and to pass the lowerrates to final borrowers Because savers were mostly households and bor-rowers were mostly firms, this resulted in a transfer of income fromhouseholds to firns and in a change in the form in which savings wereheld, from debt to corporate equity.

Holding down interest rares on loans increases excess demand forcredit, which in turn leads to rationing of credit by the government it-self or by private sector banks working with goverrunent guidance. Thisheightens 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 inJapan, 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 andgovernment-ided allocation; mirroeconomic evidence from Japansupports the view thar access to government credit increased investment(see chapter 6).

Generally, financial repression is associated with low economicgrowth, 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 comparativstudies (see chapter 5). While we cannot establish condusively that mildrepression of interest rates at positive real levels enhanced growth innortheast 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 publicand private banking institutions in Hong Kong, Japan, Korea, Malaysia,Singapore, Thailand, and Taiwan, China, involved the banking sector inthe management of troubled enterprises, increasing the likelihood ofcreditor 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, couldhaVe adversely afFected the allocation of resources. Similarly, industrialtargeting 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 allocationand productiviy change fall into findamental and interventionist cat-egories. Labor marker policies tended to rely on fundamentals, usingthe market and reinforcing irs fiexibility. In capital markets, govern-ments intervened systematically, both to control interest rates and todirect credit, but acted within a framework of careful monitoring andgenerally low subsidies to borrowers. Trade policies have indcuded sub-standal protection of local manufacurs, but less than in most otherdeveloping countries; in addition, HPAE governments offiet some dis-advantages of protection by actively supporting exports. Finally, whileinterventions 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 efficientallocation, increased acquisition of foreign rechnology, and more rapidproductivity 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 lessvulnerable 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 effectivelyboosting the demand for workers. As a result, employment levels havernsen first, followed by market- and productivity-driven increases inwage levels. Because wages or at least wage rate increases have beendownwardly flexible in response to changes in die demand for labor, ad-justmenr to macrocconomic shocks has generally been quicker and lesspainfiul in East Asia than in other developing regions. Rapid adjust-ments helped to sustain growth, which in turn made more rapid realwage growth possible.

IHigh productiviry and income growth in agriculture helped to keepEastAsian urban wages close to the supply price of labor. In contrast tomany other developing- economies, where the gap between urban andrural incomes has been large and growing, in the HPAEs the incomes ofurban and rural workers with similar skill levels have risen roughly at thesame pace; moreover, the overall gap between urban and rural incomesis smaller in the HPAEs than in other developing economies.

In Sub-Saharan Africa, Latin America, and South Asia, where wagesin the urban formal sector are often pushed up by legislated minimumwages and other nonmarket forces, urban wage earners often have in-comes twice their counterparm' in informal sectors. In contrast, the gapbetween the formal and informal sectors in East Asia is only about 20percent. Smaller income gaps contribute to overall social stability, thusenhancing the environment for growth.

Caprtal Markets and fllocatio. Most HPAEs influenced credit allocationin three ways: (i) by enforcing regulations to improve private banks' pro-ject selection; (ii) by crearing financial institutions, especially long-termcredit (development) banks, and (Li) by directing credit to specific sec-tors and firms through public and private banks. All three approachescan be justified in theory, and each has worked in some IIPAES. Yet eachinvolves progressively more government intervention in credit marketsand so carries a higher risk.

Goverrunent relationships with banks in the HPAEs have variedwidely. 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 publiccommerial 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 enterpriseswere trgeted sectors in Indonesia, Malaysia, and Thailand. Taiwan,China, has recently targeted technological development. Japan andKorea have used credit as a tool of industrial polic34 organizing conteststhrough deliberative councils to promote at various times the shipbuild-ing, chemical, and automobile induscries.

The implicit subsidy of direcred-credit programs in the HPAEs wasgenerally 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 frompreferential credit was large during the 1970s, resulting in a large gap be-tween bank and curb market interest rates. This gap has dedinedsharply in recent yea, as Korea has shifted away fiom heavy credit sub-sidies to selected sectors. In Japan imnplicit subsidies were small, and thedirection 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 achievepolicy 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 applicantsthat fell within government sectoral trgets. In Korea, the govemmentindividually monitored the large conglomerates using market-orientedcriteria such as exports and profirabiliy. In some cases, major enterprisesthat filed to meet these tests were driven into bankruptcy. Recenr as-sessments of the directed-credit programs in Japan and Korea providemicroeconomic 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 tohave improved credic allocation, especially during the early stages ofrapid 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 largelyunsuccessful. Even in the norther-tier economies, the increasing level offinancial sector development and their increasing openness to internationalcapiml 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 foreigntechnology through a variety of mchaisms. All welcomed technologytransfers in the forn of licenses, capital goods imports, and foreigntraining. 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 aggressivelyacquiring foreign knowledge through licenses and other means.

In contrast, odter low- and middle-income economies such as Indiaand 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 exchangefor trips to acquire knowledge, been restictive of DFI, and have at-tempted prematurely to build up their machine-producing sectors, thusforgoing 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-knowninstances includeJapan's heavy industry promotion policies of the 1950sand the subsequent imitation of these policies in Korea. These policiesinduded import prorection as well as subsidies for capital and other im-ported input& Malaysia, Singapore, Taiwan, China, and even HongKong have also established programs-typically with more moderateincentives-to accelerate development of advanced industries. Despitethese acions we find very litde evidence that industial polices have af-fected either the secroral structure of industry or rates of productivitychange. Indeed, industrial structures in Japan, Korea, and Taiwan,China, have evolved during the past thirty years as we would expect venfactor-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 informationand 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 notsucceeded. This has been the case with the ambitious industrial policyprograms 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 theHPAEs has been a significant source of rapid productiviy growth: the ac-tive promotion of manufactured exports. Although all HPAEs cxceptHong Kong passed through an import-substitution phase, with highand variable protecrion of domestic import substitutes, these periodsended earlier than in other economies, typically because of a compellingneed for foreign exchange. In contrast to many other economies, whichtried to preserve foreign exchange with sticter import controls, theHPAEs set out to earn additional foreign exchange by increasing exports.Hong Kong and Singapore adopted trade regimes that were dose to freetrade; Japan, Korea, and Taiwan, China, adopted mixed regimes thatwere largey free for export industries. In the 1980s, Indonesia,

Malaysia, and Thailand have adopted a wide variety of export incentiveswhile 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 weredoser 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 equalbetween exports and imports, this was the result of countervailing sub-siclies rather than trade neutrality the promotion of exports coexistedwith protecion of the domestic market. In the Southeast Asian HPAEs,

conversely, govermments used gradual but continuous liberaization ofthe trade regime, supplemented by institutional support for exporters,to achieve the export push. In both cases governments were crediblycommitted to the export-push strategy; producers, even those in theproteced 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, allrelied 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 andfrequently adjusted. Thus, many of East Asia's "industrial upgrading'programs of the lare 1970s and early 1980s were substantially modifiedor abandoned when they failed to produce satisfactory results. Using theexport rule meant that even programs of selective industrial promotionwere indirectly export promoting.

Manufacmred export growth also provided a powerfil mechanismfor 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 theeconomy that are not reflected in market transactions. These information-reated extemalities are an imporrant source of rapid productiviygrowth. Both cross-economy evidence and more detailed studies at theindustry 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 liberalizationwith 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 aChanging 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 havebeen 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 opennessto direct fbreign investment and licensing were crucial to rapid produc-tivity growth. Public investment complemented private investment andincreased its orientation to exporcs. Education policies stressed universalprimary 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 benefitedfrom careful policy interventions. All interventions carry costs, either inthe direct fiscal costs of subsidies or forgone revenues, or the implicittaxation of households and firms, for example, through tariffs or inter-est rate controls. Unlike many other governments that attempted suchinterventions, HPAE governments generally held costs within well-defined limits. Thus, price distortions were mild, interest rare controlsused 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 condusivelythat 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 carrieshigh risk. Export-push strategies have been by far the most successfil

combination of fundamentals and policy interventions and hold themost 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 theirmost sustained and forceful repression of interest rates-during a periodwhen 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 theirrestrictions on capital outflows were more effective than similar restric-tions in many other economies. In today's increasingly global economicenvironment, 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 capitalflight is sharply narrowed. However, in some exceptional instances, verymild financial repression of short duration to increase corporate equityremains a viable optiOII. This has been the case in Malaysia, which haswide open financial markets but nonetheless succeeded with very mildfinancial repression for more than a year.

The export-push strategy appears to hold great promise for otherdeveloping economies. But the conditions of market access under theGeneral Agreement on Tariffs and Trade (GArr), and odlier tradingarrangements, will hamper developing economies' use of policies viewedas unFair in major industrial-economy markets. Subsidies to exports anddirected-credit programs linked to exports are not generally consistentwith 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 inmost developing economies. Fortunately, many powerful instruments ofexport promotion are nor only within the institutional capacity of manydeveloping economies but remain viable in today's economic environ-ment. Creating a free trade environment for exporters, providing financeand support services for small and medium-size exporters, improvingtrade-related aspects of the civil service, aggressively courting export-oriented direct foreign investmnent, and focusing infrastructure on areasthat encourage exports are all attainable goals that are unlikely to provokeopposition from trading partners. Indeed, some or all of these have beenpart of the export push in Indonesia, Malaysia, and Thailand. Thesethree economies, the most recent participants in the "economic mirade,'may show the way for the next generation of developing economies tofollow export-push strategies.

The phenomenal success of the HPAEs is already inspiring attempts atimitation. 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 policiesprecludes drawing any simple lcssons or making any simple recommen-dations, exccpt perhaps that pragmatic adherence to the fundamentals iscentral to success. These market-oriented aspects of East Asia's experi-ence can be recommended w;ith few reservations. More institutionallydemanding strategies have often filed in other settings and they clearlyare 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-SaharanAfrica and Latin .Amnerica, and elsewhere in Asia where such institu-tional conditions are lacking, activist government involvement in theeconomy has usually gone awry. So the fact that interventions were anelement of some East Asian economies' success does not mean that theyshould be attempted everywhere, nor should it be taken as an excuse topostpone needed market-oriented refonn.

The success of the HPAEs broadens our understanding of the range ofpolicies that are consistent with rapid development. It also teaches usthat 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 andinterventionist policies to East Asiis remark-able growth, and the crucialrole that institutions have played in their evolution and application. Aswe shall see, maaking a miracle is no simple matter

Note1. 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 modemseem to be an inappropriate subject for study. However economic growrh, it may provide some useful insightsmany 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, andEconomic 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 ofthe postvar period. Moreover, the eight differ in thedegree of government intervention in the economy

and the manner in which their leaders have shaped and implementedpoticies. Korean policymakers, for example, have intervened heavily inindustrial, labor, and credit markets, while policymalers in Hong Konghave been consistently hands-off.

Despite many differences, however, the eight economies have muchin common. In a number of ways, their postwar experience distin-guishes them as a group. Their most obvious common characteristic istheir high average rate of economic growth. During the same period,income inequality has declined, sometimes dramatically. These twooutcomes-rapid growth and reduced inequality-are the definingcharacteristics of what has come to be known as rhe Easr Asian eco-nomic miracle.

The eight economies share six other characteristics chat set themapart. 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 savingsa* 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 equitablegrowth. Some are sources of growth, some are outcomes of growtl, andsome are unique features of HPAE growth, but most fall in more than oneof 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 from1960 to 1990. They averaged 5.5 percent annual per capita real

income growth, outperforming every economy in Latin American andSub-Saharan Africa (except diamond-rich Borswana). Another EastAsian economy, China, has grown 5.8 percent a year since 1965 andcould stake a claim to join the ranks of the HPAEs.1

Figure 1.1 shows the relationship between income level relative to theUnited States in 1960 and per capita income growth for 119 econormiesduring 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-economyaverage.2 More disturbingly, in thirteen developing economies, percapita 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, theRepublic of Korea, Singaporc, and Taiwan, China, were particularlynotable.

Other developing economies have grown ist fr several years, partic-ularly befbre the 1980s, but few others have sustained high growth ratesfor 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. The11 that achieved rapid growth during both periods are in the northeastcorner. Of these, five are East Asiau success stories: Hong Kong, Japan,Korea, Singapore, and Taiwan, China Japan's shift from exatremcly rapidgrowth in the 1960s to rates more typical of high-income economies indie 1970s is apparent. The other three IIPAEs-Indonesia, Malaysia, andThailand-all show accelerating growth, with higher growth rates in thesecond 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 III0 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 ratesbetween 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 andcontemporary 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. Fortyeconomies are ranked by the ratio of the income share of the richest fifthof 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 PerisbenceGDP growth rate (percent), 1970-8S

- Sin.s .oret Taiwan, China6$|- _.!._ ._ _>_______ : * @ ~~~Singapore**6%- d

*Mai*a.siaRep. of Korea

4% -w*,*f

. t *Thland Japa

0%~~~~~

0 I

. | * U High-Income economies4%- * *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 4.0percent) and low relative inequality (ratio of the income share of the topquintile to that of the bottom quintile less than 10). There are 7 high-growth, low-inequality economies. All of them are in East Asia; onlyMalaysia, 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 ofequality and growth using Gini coefficients illustrate (see figure 3 in theOverview). The developing HPAEs dearly outperform other middle-income economies in that they have both lower levels of inequality andhigher levels ofgrow&h Moreover, as figures Al .7-A1.9 at the end of thechapter show, improvements in income distribution gmerally coincidedwith periods of rapid growth.

30

GROWTH. EQUITY. AND ECONOX-rO

Figure 1.3 Income inequalit and Growth of GCP, 1965!89GOP 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 - ;.

.Spaino France

* BelgiumUnited Kingdom AuS ilia i * Colombia

e Sri LankaSwitzerland' * iine Mec,

Ia Pakistan ' 0Kenya

. India * Venezuela1- ._ lMalaw_ _ _ .A.t--

Ne * ChIIe Argenttna'j

Bangladesh *PBolivia * *eruge!tn

- Cote d'lvoErr

* Maurita nia

*. Ghana Ia 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 Koreahave shown increasing inequality in recent years; however, most of thisis 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 comparesthe declines in poverry, defined as the inability to attain a minimal scan-dard of living (World Bank 1990b), in some HPAs and other selectedeconomies (the period varies depending on data available). ncreases in lifeexptcy 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 rapidlythan other developing economies.4 But the decline in the relative im-portance of agricUlture in the HPAEs is not because agriculture has lackeddynamism. Across developing regions, agriculture's share of output andemployment has declined most and fistest where agicultral outputand productvity have grown the rnost (see table 1.3). From 1965 to1988, growth in both agricultnual output and agriculmtal productvitywas higher in East Asia than in other regions. Many factors contributedto the success of agriculture in these economies. Land reform (notably inKorea and Taiwan, China), agriculural extension services, reasonablygood fastructure (especially in the former Japanese colonies), andheavy investments in rural areas (notably in Indonesia) all helped.

East Asian governments have actively supported agriculatral researchand 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, andthe 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 ofpoprdauionbelow the poverty line N mber ofpoor (millions)

F.rst Last Fis - Last PercentA&vnenq Year year yar aAnge 'ar year chang

HPAEs"Indonesia 197242 58 17 -41 67.9 30.0 -56Malaysia 1973487 37 14 --23 4.1 22 -46Singapore- 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 -25C6reds'Iire 1985-86-1 30 31 1 -3.1 3.3 .6.4I-ndia 1972-83 54 43 -9 311.4 315.0 1Morocco- 1970-84. -43 34 -9 . 6.6 7.4 :12..-R; .-. > Pikismn0; . 1962-84 . ; 54 23 -31 26.5 21.3. -19Sri-Laik 1963-82 37 27 -10 3.9 . 4.1 5

:NoeThiis able uses economy-specificpovertylines Official orcommonlyused povertyrlines have been used whenav2iblec In otherC3SC5 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 ruraland urban regions is limited, and it is difficult to make good compar-isons among economies, but available data suggest that the HPAEs haveallocated a lager share of their public investment to rural areas than didother low- and middle-income economies. Of critical importance in thisrespect has been the build-up of infrastructure-roads, bridges, trans-portation, electicity, water, and sanitation. Table 1.4 shows that therehas been a more even balance between rural and urban public invest-ment in sanitation and water &cilities in Indonesia, Korea, and Thailandthan in other developing economies. The data on rural electrificationalso 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 78Indonesia 46 59-K(orea, Rep. of. 53 772Malaysia 58 71Singapor . 65 74Th- Nailand: - -: . ;- . -: - - 52 .68

: OtherAsW: .0 .62-Chin. 43 -69In&dia .47 58.

-Sub-SaharnAlea . 43 52Lad 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:LSountWorld 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 ruralelectrification. Endonesia has not done as well, but even there die relatvedisparity 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&eof1965-88l. A t,i - d Atcrds:,,-;,: 196 198 -irc ,--: .;-:-;:e kiborfirce rru8,,,A A'd'-:.-.:'..' "

t SouhAsa ,3 2. ' <2., 6 1.l9U .::

:' 4 asrArn 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 directand indirect taxation on agriculture in East Asia. During the past threedecades, dozens of governments in other regions, eager to promote m-dustrial growth, have funneled surpluses from agriculture to industrythrough 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 domesticindustries thar manufacture agricultural inputs and the goods purchasedby 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 EsKorea, Rep. of ;54 101Thailand 1- -26 102Indoncsia -168 113

OtcrAia - - : - 64.6 38.6Latin America 58 4665SubS - ibw bn Africa 43 30

.- Prcatage of rrurl adnd uran pop ukonser&d by ecaicity

co~namy Ur Rura

' PAErIndonesia,-1984 39 10:M.a:ilays.;1;983 - : -:0: g 5- 85 55Thailand, 1984 . 78 :40

-OrhlerAsik:

-Bangradash, 1981 :20 2:India, 1981 <25 15

-S -Lakna,j982 -.. 35 8- LeninAme-ic

*~ArcndinaJ 1982- ->95 : 5Bolivia, 1981: 72 . 9.. Bra;Ail, 1981 N :-.i .......... :0 0- - : >95 19.

;C6tc -dvoir, 1981 93. : 20i 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 60Level of piotection (percent)

Rep. of Korea

alys ... .i -............I i I

Thaland . --

PsidalsiU =96 i___ ____ I ___ ~~~~~N1970s

I - 3---- 197ksPhEippines . . I k

.~ ~ ~ ~~~ ~ ~ ~ ~ _ l_ __ _ _ _ _ _ _

1 ~ ~~~~~~~~~~ I Sri lanka IE~

-2 0 2 4 6 8 10Agrutural growth mte (percent)

Sour= Schiff(1 993).

36

GROWTH. EQUITY. AND EC_OO

agricultural sector receives positive protection. Thailand's taxation of theagricultural sector was similar to South Asian levels in the 1960s and1970s 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, transrswere often voluntary-financial savings, for example-and were nearlyalways 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 steadilyIlk 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 to13 percent in 1980 and 18 percent in 1990. Manufactured exports haveprovided most of this growth. From 1965 to 1990, Japan emerged as theworld's biggest exporter of manufactured goods, increasing its share ofthe 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 ofmanufactured exporrs grew nearly fbur tmes faster than Japan's (seetable 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 surgein manufactured exports.

Some analysts have, with hindsight, attributed these achievements tounique 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 pessimistsand did not anticipate that Korea's exports would grow four times as fistas 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 orientedthroughout 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 -7Fourflgcr- 1.5 3.8 6.7 6.0 13.3 33.9SoutheastAsian NEs - 1.5 - 2.2 2.4 6.2 7.8 12A

HPAE subtoal- 7.9 -13.1 18.2 12.2 21.1 56.3All developing economies .242 28.7 198 100.0 100.0 100.0World 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 61.5SoutieasAsian 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 onlysightly 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 inthe postwar period; in ali eccept Sub-Saharan Africa, birth rates havealso fllen significantly (see table 1.7). However, compared with otherdeveloping regions, the transirion to low fertility began sooner in EastAsia (in the 1960s in the north and in the early I 970s in the south-andof 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- 1.988HPA&

'Hong Kong. 1.50 1.52 1.78 2.82Indonesia- . - 0.25 0.46 : -0.38 -0.42-lorea, Rep. of: - 0.32 0.63 :0.66 0.66Japan 0.19 '0.25 0.23 0.11Malysia 0.89 .1.00 0.85 -1.09Singapore 2.12 3.70 -2.77 3.47-Taiwan, China -.0.53 0.95 ... 0.82 0.90Th1ailand 0.28 049 0.44. 0.35

SubR -SSI,arnAfr-k - 0.24 : 0.30 0.27 0.455out6Asia -- ; ; -: ; -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 AccountsSMtaddcs: Analysis ofMainAggegates, 1988-89 (Unietd Nations). -

During the period 1965-80, all the developing regions of the worldexperienced 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 muchamong regions. There was, however, substantial regional variation in theextent 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 around2-10 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 EastAsian economies, in some cases quite sharply. For example, in Korea itfell from 2.6 percent a year in 1960-70 to 1. I percent in 1980-90; inHong Kong, from 2.5 to 1.4 percent; and in Thailand, from 3.1 to 1.8percent. In Latin America, fertility dcdines were also sufficient to reducepopulation growth rates, though generally not to the levels observed inEast 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 ofsharp declines in death rates and modest declines in birth rates resultedin an acceleration of population growth-for example, in Ghana from2.3 percent a year in 1960-70 to 3.4 percent in. 1980-90, and in Kenyafrom 3.2 to 3.8 percent.

39

Table 1.7 The Demographic Transition(percent)

Change in -ange in Aventgeannulcr-de crude grovtt ofbirth . eadl popadationrante, rate,

Eonomy - 19650 1965-80 160-70 1980-90EaswAsia

* Hong Kong --52 -54 2.5 1.4Indoncsia -40 -55 2.1 1.8

. Korea, Rep. of - -54. - 2.6 1.1Malaysia -25 -58 2.8 2.6Singapore 45 -16 2.3 2.2Thailand -46 -30 3.1 1-.8Latin Amen'caBazil - -31 -36 -2.8- 2.2Mexico -40 -55 :3.3 2.0Peru -33 -50 2.9 2.3Venezueda -31 -38 3.8 2.7-Sub-Saaran Afri-.Ghana -6 :-2 23 -3A-Kenya -13: -50 -3.2 3.8SierraLeone -2 --29- -1.7 2.4Tanzarta -2 -22 2.7 5.1

SouthAj~aBanglad -27 --33 2.5 2.3India -33 , -45 2.3 - 2.1Nepal -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 asone of the driving forces of economic growth. In a dosed economy, sav-ings is the only source of invesrmenr, and the two, by definition, mustbe 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 increasedmarkedly 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 LatinAmerica 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 figure1.6). 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 2.2Relative 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 isremarkable is their high share of private invesrment. Figure 1.7 com-pares average private and public investment shares in GDP for 1970-89in the developing HPAEs and other middle-income economies. Privateinvestment 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 about18 percent between 1975 and 1984. Private investment contracaedsharply between 1984 to 1986, reflecting the global recession, then re-covered by 1988.6 In contrast, private investment in other low- andmiddle-income economies has remained relatively stable at about 11percent of GDP.

.427

GROWTH. EQUITY, AND

The story is different with respect to public investment. In the 1 970soverall levels of public investment did not differ markedly between theHPAEs and other developing economies; during the decade publicinvestment rates in all economies rose from about 7 to 10 percent (seefigure 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. Inthe HPAEs, conversely, public investment shares actually rose between1979 and 1982 and then remained at a level nearly 4 percentage pointsabove their 1970s average. Only after 1986 did they begin to dedinetoward 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 trainingduring the past three decades has been dramatic The quantity of ed-

ucation children received increased at the same time that the quality ofschooling, and of training in the home, markedly improved. Today, thecognitive skill levels of secondary school graduates in some East Asianeconormies 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 nationalincome for more than 90 developing economies for the years 1965 and1987.7 Enrollment rates are higher ar higher levels of per capita income.But the HPAE's enrollment rates have tended to be higher than predictedfor their level of income. At the prmary level, this was most obvious in1965, when Hong Kong, Korea, and Singapore had already achieveduniversal 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 systemswere evident at the secondary level. Indonesia had a secondary enroll-nment race of 46 percent, well above other economies with roughly thesame level of income, and Korea had moved from 35 to 88 percent,

43

jA C L EitACLE

Fgure 17 Public and Private InvestnentPrivate Investment/GOP (percent)

25

.5

-D- HPAEs (excluding Japan) Other LMlEs

0

1969 1974 1979 1984 1989

Public Investmfent/GDP (percent)

20 - ------

5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I10

-U111- FPAEs (excluding Japan) -4- Other LMIEs

I ' I } t i ~~~~i ; ! i . ; 1 : i I I i i , I0 -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 wasthe 28 percent secondary enrollment rate well below the income-predicted 36 percent and the 54 percent mean for middle-incomeeconomies.9 In recent years Thailands weak educational performance hasbeen 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 rates140' - -

:120 .. ......... . 120- - n -- . lindonesia (1118VBrazil (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_ [7.018 [10.8]

Indopes15 72)1 -; I60 ~ pesI5ji2y -- '[8 d _ _ _ _ _ _ _Banigtadesh (59)

*Bangladeh(4.9) ,Pakistan (52)40-its . 1965 - -.[O79]1 9 74-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 rcsiduaLsSourre: 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 todose 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 perpupil at the primary level rose by 355 percent in Korea- In Mexico andKenya, expenditures rose by 64 and 38 percent, respectively, during thesame 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 incomegrowth and in the number of children entering schools, both of whichfavored the East Asian economies. A somewhat better measure of schoolquality 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 performbetter 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 rates100 - . -. ,U , ,~- - -, _- ; , . ,

. ¢ | I , ', . . Rep. of Korea (88)

so - -- -- i

60 - -.. Malaysia (59)

Singapore (45) Indonesia 1 _-

40 BraA 9*-[18. .119.5] >I40~~~~~~~~ - - 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 Asianeconomics had advantges in the 1970s and 1980s. Using a simpleindex taking into account the mother's education and the number ofchildren at home, we estimate that the learing environment in Koreaduring twenty years was enhanced 114 percent more than in Brazil and147 percent more than in Pakistan (Birdsall and Sabot 1993b).

Rapid Prdu Growth

HE FINTAL OUTSITANDING CHARACTERISTIC OF THE HPAFs IStheir rapid productivity growth. That labor productivity in theHPAEs increased rapidly is self-evident per capita GDP grwth

depends primarily on rising outpur per worker Some of this increase inoutput 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-,FAAIF.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.1o y a=,;-.,us J---ygra uaUj;'..'ib;gdsj gender' Th d il� iffii:ntii f6�cyiii-,.,in ca-d6i. 1�.o-- �n";�' evc �-rnenr iv ary�an sc6ndzrj'k:�&

-th -1--'-'6d---'c�, ower� ataun.��-..,,.�-,,,�to-aiaineas-incomes�maeast--- --bf thi Aia� tir).Ji-I 9 65i at c pd-n-.----�

yztdj,.,1mpr6ve, I the' 'c[ p

rnaryp_ c ccorioriuq�,-,g�nderV V

'1987'thi,&ur,,;,.t-',"Jt)r,outst e,--.tlii --hi" The- 'j!aI[70Wid-,- e:gcn-':-'.,-',;;,'e. omm`- e' --HPAB_ WM �S�wl"cr-,,E=ticxpected.,,-By�-pmg jan!eqonoa,Lies-f6i.W.W&6ca.ot er eV 0 aM

Ir�; ;�cconomjqs,'� �splteZ�wr'u''iat, norms-,,ttl2t. put, greater--, achieved universai p6mary cationC , :r

p-at' a-'t-1e'%'rib7SLLbi6iLWqq01L- ucation-ofs6diaahi, tisci; -Ov- iiiiirii'n'aungthe''gendir th'C gender-gaps, continik46"dvt�:� I Mmii ai -d6cationl., dau M' um mtes

10 , i'. E. -'. r -�;- " -, - -. � - , -- -. 4.-, -"s. I. .......... Slg

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- ooni ive4rgu -64c�4,,th qarrip e,,f G�61 64iS31S =&T.1fij,'-1991 :F6v7lecon6nues, GNP pq.qLpibi '$500'-%

'I. A -in aTsmoolv0'ff -om Miienr,� etwccwiboys.AD"Mak-Yea 6f 'Edud�fialfl ges-2 6�fita- i� pio,in, fs-,--=,a, as, 'n�o-r4n

p

In!; 1965, 1-ir '(6c, 6 Thi-.4 cq4 gcn..d&,gaps-,-avtqe-k' d�inAmi Iiid 'd'-"-16iill: cc aryschool-ppincreastrig.second

e;M as�;mu .2- CIj 'an

C�5 7 0, v-166'.1-970i cl -,q80s!2,S1t-.1Srhl --

ey opmg. -wo&

-Ccon6mli SUCCCSS �Ilh laL%ngjcntsto PV .98

V4.1 g�n !cr:ga,p-wJ- `6ii6,

er 1) Ja ii I4�-. -j� kcapim-41 -6�

A-siadeconornies.5nour,stuZ.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 beaccounted 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 unusuallyrapid growth in physical and human capital. The remaining third of ti-sgrowth 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 thereforepartly explains why these economies have been catching up writh the in-dustrial economies, while most other developing economies have not. Inbox 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 Regressionsof 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 across-economy regression in which we estimate the relationship betweenthe rate of rcal per capita income growth and the share of investment inGDP, and two measures of educational attainment, controlling for dterate of growth of the economically active population, and the relatvegap between per capita income and U.S. per capita income in 1960 at1980 U.S. dollar prices.'2 Our measures of educational attainment areprimary and secondary school enrollment rates in 1960, the area ofhuman capiral accumulation in which the HPAEs showed a substantialdifference from other low- and middle-income economies.

The results of applying our statistical analysis to 113 economies aresummarized in table 1.8. The estmates indicate that fcror accumulationmatters. Investment in both physical capital and schooling contribute sig-nificantly to economic growth.'3 An increase of 10 percentage points ofGDP in the rate of investment-about the difference in private investmentrates 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 enrollmentrare would raise per capita income growth by 0.3 percent

Beyond accumulation of physical and human capitl, initial incomealso has a significant relationship with per capita income growth.

48

GROWTH, EQUITY, AND�E'CO.'i

�M�77 i 7,x ,Zrl%�evjc,;i-.Ii_-.Ip"��-��."- _2�_�_� A

�A

)NOMISTS, ROAD, mmmenr uman an1-4

�accumuw- er:onj'am Murpur

VC.'. �.,Jf.

-w-V hton;t,&I rexten--o 3K

-Flj e 3Vlcwlls"_b"lV4, 1. _T'. y aIC um2n ppi ajt� it

r=enini

p nonluesi- irk*, -o'wth' anA ysi&pfil- &72ACt1.06

".w -,,-,,.amo4ccon6n3jcs_ cr cromki&mrz-k iE66��-Er; esame, 6i&bffinva= �ccononiic,_-or-4� gan,=qon;,-jnnovaunonuc-.,gr -0 o ogyarc o arm y.14666iassem- 'lik" r tLonMR P 31iiAiii '446ccum non-, cKami)t M_-llFAE:f a ii'ibilbimer vict-, ruo have

"d zin 4 "W o .. ........CoDrifth lpor�pcmicnVor economc gm.

M, Om ccum07Ww"Ah -,-tccon-._t-,,, oi�l ficior-Z.- Q

_9M Y ---- 2ssum reci=cs 0 �,p LW2�'CX (SOIW49- 5.7)N

-nu e-mcoifi 0 f -T, P, P'. J:i_7

coiiiii.econoun vll,z, Om r onspi cOj_ Cj;V;.

a�upajmong de,�-,'2ni� t

V 19P-J;_�2 tcrcmeat&jj06, 'gro

er, 0 �gt

6i�basji-,tVoutpu poon 0

.4;_fItaL

A

econo nV4

OD

'p tSd'

49

L i-MIRACLE

S j ': . $ ':~~~~~~- --. r~ l'--'rr' ~i. r:*rt ',, ,.. :.;,..:,,

?%e 7 aWe to acqwre technologyhniore~ =Žie4iy; daii nical efnaen4 chan ge- Bo6h tho y -and empi icprgres&more rapidlj than induiial -- economics. evidence: leadus to, nclude chat tol&cr pruc-

K>Thisiis pnid4crrybased catcbiiig: up.. vy growthl thdeindusirialecmomiioicilrerU- radbe- i" a ; < . mov dcj-- 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-icomei--- 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 betteron 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 in1960, 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 effectof 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 crnhsest.Also indiude Tun".isiand South.AMica.S6an Woild Bank snff csdnnrcs.

Armned with thes e statistical results, we can examine two facrs aboutthe role of accumulation in growth in the HPAES. First, our estimates ofthe parameters of the cross-economy growth equation enable us to do asimple "accounting for growth" in the HPAEs. Table 1.9 shows the pro-dicred contribution to growth of investment, human capital, populationg'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 ispredicted 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 17-191car 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 10S 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 theHPAES 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 schoolenrollment in 1960 makes a particularly strong contribution to itsgrowth (41 percent), more than investment, while the laggards in sec-ondary enrolhment rates, Indonesia, Malaysia, and Thailand, have thesmallest proportion of their predicted growth attributable to secondaryenrollments (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 initialincome, the HIPAEs have a significantly higher rate of growth than allother 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 andSub-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 EastAsian high performers.

WhIiat is most striking, however, is how lirde we are able to accountfor differences in growth rates between the HPAEs and other economies

on the basis of conventional economic variables.14 We are able, in theend, to prcdka only about 17 percent of the actual difference in growthrates between the HPAEs and Latin America. We do somewhat better be-tween Sub-Saharan Africa and the HPAEs, predicting about 36 percent ofthe difference. Controlling for their superior rates of accumulation, theHPAEs stil outperform while Sub-Saran Africa or Latin America -derperform the statistical relationship between accumulation andgrowth, leaving much of tie regional difference in per capita incomegrowth unexplained (even though a large fiacton of HPAE success is ex-

piained).1 5 They have been apparendy more successful in allocating theresources that they have accumulated to high-productivity activities andin adopting and mastering catch up rechnologies.

Evidence from Estimates of TFP Gnxvb

iTP change captures both of these important aspects of productivitygrowti. Te is estmated in a neoclassical fiamework by subtractingfrom output growNt the portion of growth due to capital accumulation,to human capital accumulation, and labor force growth. In keeping withthis 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). -013CM! (11i2 .- L9 .< . 15i200 ; - 4 (I

3,0'.. ~~~~4.IM 2.32

using a produc-ion function esdimated with cross-economy data. (Thetechniques are discussed in appendix Li..) Figure 1.10 plots the result-ingr Tr growth rstimates for 196049 against relative income levels in1960. There is great va-riance in the rates of n.' growth among low-in-come economies. A number of develo'ping economics show higher ratesof 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 outsharply, witha high absolure levels of TF. Five HPAEs-Hong Kong,Japan. Korea, ThAand, and Taiwan, China-are in the top decile. Theother three HimaEs-Indonesia, Malaysia, and Singapore-aze closer toTF' growth rates in high-income economies (about 1.5 percet), but arestill 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 relativelycompact distribudton ofTFP hrates around 1.5 percent a year, andthe tendency for IT? growth to dedine with rising income among thehigh-income coonomies (see figure 1.10). In low- and middle-incomeeconozPrwst, howevr, changes in *rTFP must rcflect more than technicalprogress, othierwise we would never find negative TF growh ratev. Wehave 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 changeand 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 1.2Ratio 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 capitcalto 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. Thusour estimates of TEP growth answer the following question: on the basisof the average efficiency with which physical and human capital are usedin the world economy, does accumulation over- or underpredict incomegrowth? The answer is that for most low- andt mididle-incomeeconomies it overpredicts, while for the HPAEs it underpredicts.'

Under very restrictive assumnptions (see appendix 1. 1), we are able tooffer 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 theelasticities of output, which should be used to calculate TIP change, arethose that apply to high-income economies only. These more alloca-tively efficient economies have elasticities of output with respect to bothphysical and human capital that are higher than those for the wholecross-economy sample. We subtmct the average rate of TFP change forthe high-incomne economies, which we have associated with movementsin international best practice, from TFP change estimated using thosehigher elasticities to get an estimate of technical efficiency change. Usingthis metlhod, Hong Kong (2.0 percent); Japan (1.0 percent); Taiwan,China (0.8 percent); and Thailand (0.1 percent) are the only HPAEscatching 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 ofIndonesia, Malaysia, and Thailand were falling behind internationalbest 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 seznmssomewhat more remarkable. Using the same method, we have estmatedrhe average rate of technical efficiency change for Latin America (-1.4percent) and Sub-Saharan Africa (-3.5 percent). Against these bench-marks, all the HPAES except Singapore stand up well in their ability tokeep pace with the world's shifting technological frontier. 7

Although the absolute magnitude of TFP in HPAEs is higher, so isgrowth. Is the proportion of growth due to TFP high? On the basis of ananalysis ofsources of growth forasample of economies, Chenery (1986)devc-cped a typology of the contribution of TFP growih to total outputgrowth by income. A typical low- and middle-income economy had arelatively small contribution of TFP growth to total output growth, be-tween 10 and 20 percent. High-income economies, conversely, derivedabout 30-50 percent of dteir total output growdt from TFP growdt.

Figure 1.11 largely confirms this pattern for low-, midcle-, andhigh-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 muchhigher. The HPAEs fill into two distinct groups: investment-driven andproductivity-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 toGmwth 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~~~~~~4r _ ~ N .4 . 4.__ '

.4 -2 4 2 4 8TOtal bector hipLrt growth (percent per year)

Note: Dashcd line:s tcprcscnc: toul 2vcngF GDP growch raus, 19650 89. Soli rays reprewnsiEhe 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 recentgrowth (see box 1.3) shows some of these characterisEics.

The EIPAEs are unusual arrong developing cconlomies because of therelatively 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 beenhigher than in other cleveloping economics and is importat to the EastAsian 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 -wcm<tj~ ra - -;Ghina's ' WrtL JWiflg'staiildsgo r. comaierdmds. thia.za,utOr19

e.orublihgnedpopuation oFalleigh-sbonomf - l sour; & W:nimal2 'peror inp the mC iesebq dir als-

..recorde.av-nna GNP- gewdoflocDeta;ose gnll;6roucy 'g non ar>a.a-.

gpal e -s gntao

%AAuciMAdannbaawaL.I 5> 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

fro<i;vasmessof,CltinE.lIiaks 1979 tonenili7 1 u.4;>pcrct}en t,;saviing ta armcsinvsted inXrw nsip-ndAf&0MF1982wjt 198~Fo bier-retreii.chmin il aeti~se.CzitharS.- 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 ipour-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 achievinghligh rates of growth and redluctions in inequality- These were accompa-mied by-.agricultural transbrmnaions, rapid fertility dedinci, andmaufactured. export: growth; and accounted for by rapid ac cumulation ofphysical'and human capital and by productivity growth. What policieslie behind this success? In the next chapter, we examine alternative cx-.

59

CLE

planations of the relationship between public policies and economicsuccess and particularly the ways that public policy can lead to improvedeconomic outcomes. In subsequent chapters, we explore the relevantpolicies 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 TOestimate the relationship between accumulation and growth usingboth 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 ofother 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 usingHeston-Summers' measures from 1960 to 1985; INV is the averageshare of investmenr in GDP for the period 1960-85; ED is a measure ofeducational ataiunment LEG is the rate of growth of the economicallyactive 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 in1960.'; Mankiw, Romer, and Weil (1992) have recently demonstratedthat 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 ofthe regressions is good, and the coefficients of the variables are ofthe ex-pected sign and are significant at conventional levels (0.05 level).'9 Wehave also estimated the basic relationship fir two subperiods, 1960-70and 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 .196-55Nwnberofobsennrtianr 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 paraheeSOUIrcZWorld Bank staffestiniares.

two periods, but in most respects t-hc' parameter estimates have similarvalues and significance. The major exception is the coefficient estimateon investment, which becomes insignificant in the period 1970-85.

Table AM. also reporcs on two e%frcs to examine the impact ofchanging the basic specification. DeLong and Sununers (1991, 1993)have argued that equipment investment, rather than 'total investment, isa superior explanatory variable for per capita income growth- When theinvestment 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. Oursample size is much reduced, however, and we condude that the reduc-tion ini residual variance is not sufficient to justifyr a change in our basicspecification'. 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 alternativespecification.

Deviations fiom the regression are estimares of labor productivityclange that cannot be attributed to accumulation-investment in phys-ical or hunan capiral or to the component of TFP change associated withrelative income levels. The patterns of productivity change resulting fromthe basic specification of the cross-economy growth regressions can beseen in figure Al.1. The figure plots the partial scatter of the relationshipbetween growth of output per capita and investment, controlling forhuman 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 regressionline, 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, 1960-85Y.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 20Orthaganal 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 dummyvariable 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 ameasure of human capita endowments, and L is a measure of labor servicesin 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 thecontribution ofaccumulation to growth in output per worker. TEP changecan 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 returnsto scale, s,K and SE are equal to the income shares of factors. Thus mostempirical 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 oursample, we instead estimate s5 and sE direcdy using a simple, cross-economy production fnmction. The data for this analysis indude a newconstant price capital stock data set (Nehru and Dhareshwar 1993).Measures of human capital are incorporated in the specification usingBarro and Lee's (1993) measure of educational atainment. We regressannual log output growth on log capital growth, log human capitalgrowth, and log labor growth during the 1960 to 1990 period,constraining their coefficients to sum to unity (that is, specilig theproduction function to be Cobb-Douglas). We also indude cconomy-

63

- RLIRACLE

specific dummy variables to estimate individual rates of TFP change foreach 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 lowelasticity of output with respect to capital from the wholc cross-economy sample. This is not altogether surprising. As we pointed our inchapter 1, there is a subset of developing economies that has positive netinvestment and human capital growth rates but negative output growthrates. In effect, the marginal product of both physical and human capi-tal was negative in these economies, reducing the elasticity of output ofboth physical and human capital in the production function based onthe whole sample. We have reestimated the production function on thebasis of only high-income-economy input-output relationships. These re-sults and the estimated TFP growth rares are displayed in table A1.2. Theelasticity of output with respect to capital in the high-income-economyproduction 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 1.49High-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 2.4113Indoncsia 1.2543 -0.7953Japan 3.4776 1.4274Kore,.Rep.of 3.1021 0.2355

M2;ysia - 1.0755 -1.3369sing2pore1 1.1911 -3.0112Taiwa,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 largesample of economies to test their robustness. Using a data base derivedfrom 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 thosepresented above are summarized in figure A1.3, which plots the scatterof Elias's estimate of average TIP growth during 1950-87 against ourestimates. Fischer (1993) has also presented a set ofTFP growth rates fora large sample of economies. A similar plot of his results against ours isshown 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

is0- 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 4TFP growth, 1960-99 (percent)

Snuru:v Elias (1 990a); World Bank data

Both samples show sim-ilarly great variance in rates of TrP growthamong low-income economies, while the high-income economies arerather closely distributed around a mean ThFP growth rare of about 1.5percent a year. There is also great consistency between the relative rank-ing of TEL? growth rates for the HPAES. The same productivity~-driveneconomies-Japan, Korea, Thailand, and Taiwan, China (Hong Kongis a bsent from Fischees TF' estimrnaes)-are in the upper quintile of bothdistributions, with the except:ion of Thaiand, which under the Fischerestimate is still in the upper half of the distribution but is not in theupper half under the Elias esrimara-

Technical Prmgress and Technical Efficiency Change

As we pointed out in the conduding section of chapter 1, it is difficultto disentangle the relative contributions of allocative mistakes, technicalprogres, 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 4TiFP 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(.) satisfiesthe 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 beachieved without the introduction of new techniques. Firms can movealong 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 asdefined by Solow (1956). Observed performance in a sample of

economies or firms reveals that few are at best practice.2 1 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 timeyields:

(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. Thisis 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 duringperiod t. While i; is always non-negative, u *t) can be either positiveor negative.

We define the rate of total flacor productivity change as the variationFigure 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 thechange 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 suchas A to D, combining accumulation with a movement toward bestpractice.

T!-,s reinterpretation of TFP diange is useful in understanding thesources 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 haveTFP growth rates exceeding the rate of technological progress if technical

-efficiency change is positive, icit) > 0. It is also possible for TFP changeto be negative, if technical efficiency change is negative and greater inabsolute 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 productionfunction aiLd derive estimates of both the rare of technological changeand the rate oftechnical efficiency changc3 4 This approach is notsuitedto the cross-economy data available, and we do not employ it. An alter-

native is to measure TFP change direcdy by growth accounting. Sincemeasured TFP change consists of both technological progress and tech-nical efficiency change, we impose the strong assumption that techno- Table A1.3 Technical Efficienqlogical change, the movement of best praaice, is constant and does not Change Estimates for the HPAEs

vary across economies.2 5 Under this assumption all of the variance in I enbikid

rates of TFP change derives from variance in the rate of technical effi- effdciencyciency 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 -0.2044CM 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 :17estimated 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 Testsof 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 5Newf wkC lh 100 -.5 ; -65 18 1N-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 Korearanked first in cach of the five categories of problems. The performancegap between Korean students and the others, all of whom were fromhigh-income economies, was greater for higher-order skills.

The International Association for the Evaluarion of EducationalAchievement also administered mathematics tests to thirteen-year-oldsin 1980-82 and included children from Hong Kong, Japan, and Thai-land among the children from twenty economics in the sample. FigureA1.6 summarizes the results for algebra, which are very similar to thosefor arithmetic, geometry, and measuremenL Japanese students rankledfirst, students fiom Hong Kong were in the top half of the distribution,and students fiom Thailand -were near the bottom, at about the samelevel 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 100Mom score

Souwe Hauvsheh and Sabor (1991).

7'

- M 1 RACLE

These results reinforce the conclusion that in some East Asianeconomies 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 EastAsia 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 coefficint0.7 ;-

* F ' * Gabon

-16 .--- - ~~~~ i Peru L-eio'Brazil

*Colonmbia

j Venezuei au Chile Malaysia Singapore

0.5 ,- -~- U~~~~~~~ -~~~s~Phiil;Dlr - ~ EZamba EHong KangU 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 0.1GDP 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 * SingaporeI E~~~~~~~~~~~Horg Kong

OA t -- i- . . .: , ,_,,_hdiaE *ArgenU;tinda Indong0.4j - ------ 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 0.08GDP 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 reO l-~----------EPeru-- * Philippines - - Ehilane - M Hong KongEThailand

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 0.08CDP 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 1987120.61 I I 1965 987

di*Pahistan

6 ~~~~~~~~~~-~~~~~~I ~~~~~[19.21E Indonesia >- I

* [-5, I .-

E maliand i1nd %ralaysna Ii [-10.t6] [5.1] N '-

jRlof Korea A-ndneit 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 resiEduaSowre: 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 77E a

IndonesaKo3g.913 ~ ~ ~ ~ ~ ~ ~ ~ 29 PoakKontan

E ~~~~'Thailand -J~~~~~~~~~~~~~ 18.2]I [-.2

Sn apore__ _ _ _ _ _ _ _

* Av~~~~~~~~~~~~26 N Maayi0 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. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~3.4Scarce 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 individualment 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 investmentthree reasons: (a) China's recent very rapid growth has dedined continuously during the 1980s from a peakof2Obeen 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 theFour Tigers and than thar of Malaysia and Thailand 7. Behrman and Schneider (1992). The regressionsamong the southeast Asian NIEs; and (c)-and most control for a polynnomial in average per capira income inimportant-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 abovegrowch 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 rateally 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 andbourg, Netherlands, Norway, New Zealand, Sweden, Stigler (1992), among others, who report results oftests inSwitzerland, 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 theeconomies, with growth rates exceeding those in indus- cognitive skills levels of East Asian chidren to othertrial economies, is defined as 'unconditional conver- students.gence' in the new literature on growth dteory (see box1.2) and would be expected to occur only under very re- 11. See appendix 1.1 for a formal definition of TPPstnricve 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 offerences in variables contributing to growth. cross-economy regressions that have been used in recent

years. These studies have also introduced other variables3 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, Dollarthe uniqueness of persistent growdt among the HPAEs. (1992); endogenous investrment, Barro (1990). We have

- not included these variables because they are not central to4. 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 withneoclassical 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 inother 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. Wetwo subperiods, 1960-70 and 1970-85. The fit of the re- hlave not included these variables because they arc norgrsion is mnrkedly better in the earlier period, but in central to our theme.most respects the parameter estimates have similar valuesand significance. The major exception is the coefficient 19. The magnitude of the coefficient on investment isestimate on investment/GDP, which becomes insignificant lower than that reported in Dollar (1992) for his sanmpie ofin the period 1970-85. 114 economies (0.113). The magnitude of the coefficient

for education is abour that given in Dollar's results, and the14. 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 asmaller 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 whcheadtcnilicfcecyTrcisbnoapopulation censuses, fails to perforn better than the p- which he cllled technical inefficiency. There is by now amaty 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 aother independently derived estimats for a large sample Soloof countries (Elias 1990; Fischer 1993). The pattem Of wtype neoclassical mode. Proponents of endogepr.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 oncontrasting Asia with other devcloping economies, te measurement of best-practice or frontier productonreaches similar condusions, both with respect to the esci- fiunctions and their relationship to traditioral estimates ofmated magnitudes of TFP change and the relative contri- the production funcdion is contained in Aigner andbution of TFP change to output growth. Thomas and Schmidt (1980).Wang (1993) and S. Edwards (1992) also reach broadlysimilar results concerning the pattern of productivity 24. This is the approach adopted by Nishinmizu andchange 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 fiomtion 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 dustrialeconomiicsgenerallyyieldacompactdistributionofhousing 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 firsttherefore undervalued the rate of growth of output and approximation ofthe rate of technological change.hence the rate of rvp change. Similar derailed criticsmscould be made for other economies, both HPAE and non- 26. The rests were administered to a random sample ofHPAE, aLw our TFP results are rherefore best regarded (as in- 2,000 students in each county. The test questions weredeed are the :ross-economy regression results) as indicatve drawn from the National Assessment of Educationalof 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? Someobservers attribute success to geographyand culture rather than economic or otherpolicies. But geography and culture cannot

explain it all; East Asia indudes Myanmar and the Philippines as well asthe Republic of Korea and Malaysia. Nor is there a simple policy story.Of the successful East Asian economies, Hong Kong and Thailand havehad relatively little activist intervention in economuic sectors by theirgovernments, while Korea and Malaysia have had a lot.

in this chaprer, after examining the significance of geography andculture, we suggest a simple framework for understanding growth inthe 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 differentpolicies can be used, if circumstances are propitious, to address thethree 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 outputfor 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 relativegeographical proximity are the most obvious shared characreristics ofthe 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. Intraregionalcvnomic relationships date back many centuries to China's relationswith 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 yearsbefre 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 commontrading 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 tradeexpansion. Cheap ocean transport and shared historical experiences fir-ther knit together a far-flung, culturally disparate region. U.S. assistancein rebuilding Japan after World War 1i, followed by massive U.S. eco-nomic assistance and military spending in the region rbroughout thecold war, also helped to set the stage for rapid gowth. Positdve impactsincluded 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 readymarket for emerging export industries. Japanese industry received a sub-stantial boost from the provisioning of U.S. troops in Korea; just assome of Koreis biggest conglomerates got their start selling goods andservices to the U.S. nilitarv during the war in Vietnam.

Regional linkages fadilitated the adoption of imitative strategies, inboth public and private sector activity. Policy imitation-specifically ofJapans industrial strategy-was an explicit objective in Korea andMalaysia. 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 ofJapanese success undoubtedly impressed policnfakers throughout EastAsia, engendering a sense of confidence as well as providing models ofpotential 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 laterfrom 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 evenwhere foreign investment policies have been restrictive, informal creditand information networks have helped investors to move capital rela-tively freely.

In addition tO direct linkages and imitation, East Asian economiesmay have benefited from positive regional externalities. Through theirearlier trade with Japan, Western importers had become farmiliar withAsian 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 rareopportunities for other East Asian producers to enter internationalmarkets. Producers of garments, shoes, television sets, automobiles,and other products in Korea and Taiwan, China, took advantage ofthese episodes to establish lucrative market positions (Petri 1988).

Of course, regional cl'2racteristics alone cannot account for EastAsia'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 thecase. Indeed, cconomies that are part of the same matrix of geography,culture, and history as the HPAEs but followed different economicpolicies-the Democratic People's Republic of Korea and the Philip-pines are two widely divergent examples-have yet to share in the EastAsian miracle. We turn, then, to the heart of this book, an examinationof 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 ihedifferences 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 relativelysuccessful developing [economies]-Hong Kong, Malaysia, Singaporc,the Republic of Korea and Taiwan, China-have greatly benefited fromdecisions and policies that limit government's role in economic decisionmaking, 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 inJapan 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 performtieir functions."

The neodassical interpretation of the axperiencc of die HPAEs, and e;pecially of Korea and Taiwan, China, presents a coherent and powerulview 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 werequickly corrected (S. Edwards 1988). Interludes of intensive import-substituting industrialization in Korea and Taiwan, China, were briefManufacturers were thus able to concentrate on improving productivityperfbrmance rather than coping with mpidly changing relative prices ofinputs and outputs.

Although neoclassical interprettions of the success of Hong Kong,Japan, Korea, Singapore, and Taiwan, China, are ofren nuanced-andadvocates offer them with varying levels of consumer warnings-theystress 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, hasbeen limited, inputs havr- flowed to sectors roughly on the basis of staticcomparative advantage, anti inrernational competition has provided theimpetns for cost disciplne and technological upgrading. Traded inputshave 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 positivereal 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 ofrapid 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, sometimesdubbed revisionist, have systematically documented that governmentsin these three economies cxtensively and selecivelypromored individualsectors. They have convincingly shown that levels of protection and thevariation 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 encouragedheavy and chemical industries by setting targets and offering a variety offinancial incentives (Pack and Westphal 1986; Westphal, Rhee, andPursdl 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 advancedtechnology and establishing rationalization cartels to &cilitate the exit ofinefficient firms?3 Taiwan, China, used public investment in large-scale

manufacturing enterprises to ensure inputs for predominantly small andmedium-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 shiftfrom negative to positive levels (McKinnon 1973). Buit while interven-nion declined, it nonetheless contnuedL In fact, Japan, Korea, and to alesser extent Taiwan, China, did not rely solely on markets to allocatesavings. 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 acceleratecatching 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. Itcondudes 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 strategyis to ensure adequate investments in people, provision of a competivecdimate 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 economicperformance through attempts to guide resource allocations by other

than market mechanisms. Attempts to guide resource allocation ininternational 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 torapid 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 offour 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 willThe- 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 theBrazil 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.otjamnaica,howevr c to CarS inprov eVe. i mnt ocr th otimes, ofKislogw.wth. .Governments must ais - . Mr overspens, terdui senoonowinaj '- stmonetayryl 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; nTalidhiemnK L .markeslcqi .k%versely; retuctionson competition' gecL tod "keep. thefir rmacroononic:. polid&s on.' I

(fo e5pe;Indsiadcensn o nappopriate. course adteir-.bioaUednoic jerfbr~cehaihes ?P opeoi) .biddk'.b ' betarizlyo S >

, ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.,

human capital CtiaaCion, openness tO nternational trade, and an en'n-romnmenr that encourages private invrestment and comjpetition. Effectivepolicy in one diension (such as human capital f hrmation) imprves.the results from effectve policies in others (such as openness or mac oeconomic stability)t In this view, the success of many economies in EastAsia has been due tO teinforcing policy feedbacks. No single policy hasensured success; strong and effictive policies in all four critincal aEcas, andover a sustained period, have been hkey. -f -p. 'csumThe marke-ffiencdly. approach captures important aspects of EastcrAsia'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 worldeconomy, 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 ofdomestic 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 policychoices have been less interventionist.

-NMultip Paths to Gmwth

As these three views show, while it is easy to see how elements of EastAsian experience contributed to rapid growth, it is nor easv to identifyr asingle recipe followed across the region. An alternative approach is torecognize this diversity ofpolicies and to assess whether and how variousmixes of policies contributed to successfil implementation of three cen-tral functions of economic management accumulation, allocation, and---productivitygroth. 4 Thi4 s approach recognizes that policies, like tac-tics, can and should vary depending on the situation, while the centralfunctions, which are crucial to development, musr always be addressed.

One of the hallmarks ofeconomic policymaking in the HPAEs was thepragmatic -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 wereabandoned. Thus, the Korean and Malaysian governments reduced in-tensive promotion of heavy and chemical indusries when the fiscal costsand strains on the financial system thrdatened macroeconomic stability.Indonesia abandoned capital marker controls ancd attempts to controlinterest rates when they werc found to be ineffeccive. Japan shifted fromhigly selective interventions to promote exports in the 1950s and1960s 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 policyof 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-governmentconflict to highly structured mechanisms to encourage business-government cooperation. Chapter 3 discusses the pragmatic flexibilityof HPAE governments in pursuing two goals, macroeconomic stabiliryand growth of manufactured exports.

Obviously, our three functions of growth-oriented economic man-agemnent were not objectives in themselves for policymakers in therapidly growing EastAsian economies. Instead, they are simplya frame-work for analyzing how the varying mix of policies and instrumentschosen by HPAE governments contributed to or detracted fiom rapidgrowdt. At the same time, we do not want to imply that growth was dueto mere luck in the HPAEs. Though policies varied over time and place,the formulation and implementation of policies were purposeful, withan overriding objective of 'shared growth." (This tcrm is described anddiscussed in chapter 4.)

In contrast to the diverse, changing, but purposeful mix ofpolicies inthe HPAEs, ocher economies have either rigidly adhered to policies thatwere demonstrably unsuccessfil-fbr example, extended periods of im-port substitution in Latin America and Sub-Saharn Africa, policies todevelop domestic technology in India, and state-led industrialization inEasteem Europe-or have lacked a purposeful direction, as the history ofadjustment pograoms in some economies in other regions attests. In-deed, as we illustrate in the next chapter, the flexibility and pragmatismofpolicyrnaking in most of the HPAEs reflected governments' willingnessto 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 (fiudamentalsand 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

ComnpetitivePolicy choices discipline Growth functions Outcomes

Fundanmentabe Accnnmulatn Rapid andsustained growth

. Stable macroeconomy * Increasing humancapital * Rapid growth of

* High human capital *Hg,aigsexports

-~~~~~~~~~~~~~~~~~~~ Hihsa.g

: Effective and secure MdHgll Rapid demographicfinancial systems Metetransti

Um.t-ig price distortions isin Export RFoapid agriculturalOpenness 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 rnartI 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 * PmduftJvpbasedW 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 toboth growth and improved distribution ofincome.

The arrows indicate that the system has numerous self-reinforcingfeedbacks. For example, rapid growth and relatively equal income distri-butions contributed to the HPAEs' superior accumulation by increasingsavings -rates and generating largr and more efective investments inhuman capital; they also contributed to superior allocation of humancapita 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 thatthey affect the attainment of growti functions primarily through market-based mechanisms of competitive discipline. Three-macroeconomicsmbifiry, effective financial. systems, and limited price distortions-assistmarkets. 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 withvarying degrees of intensity to alter market incentives. The broadest in-

rervenrions were generous incentives for manuhactured cxports. (Theseare discussed in some detail in chapter 3.) As part of this export push, allthe 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 basisfor awarding access to foreign exchange, investment licenses, or creditIn Taiwan, China, large-scale public investments, induding in publicenterprises, 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 rationcredit; favoring some would-be bonowers over others. In addition to theimportance of these directed-credit schemes to export promotion, someT 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 chapters5 and 6.) Some HP'AE chose very narrow promotion iargets. selectivelyassisting knowledge- and capital-intensive industries and firms in hopesof upgrading the economjs overl level of industrial technology: InJapan andt Korea the governments at timnes concentrated on promotingprivate domestic investment in) heavy and c-hemical industries (Hcas). InTa-iwan, China (and to some extent in Indonesia), the public sector in-vested directly in technologically sophisticated industries. Singapore andThaiand have attempted to attract foreign investment in capita- andknowledge-intensive sectors. Malaysia initially undiertook public invest-menus in HCI firms but has recendly shifted to a strategy of promotingprivate investmnent more akin to that of japan and Korea.

Some selective interventions went beyond helping markets performbetter. Rather, theyr guided and in some cases even bypassed markets. Inthe following section of this chapter, we explain why, in light of marketfailures, some interventions make sense even within the neoclassicalframework. 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 amechanism by w-hich the production decisions of the myriad firms thatmake 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 (includingwages) depend critically on what- other firms are doing, there ar-e greatpotential benefits ofsharing information. Even in well-developed marketeconomnies, information is conveyedi in many ways other than prices.Trade journals, trade association meehings, and information newslettersare al inportant minsitutions for the trsnsmission of informiation.

Adam SmiteHs invisible hand paradigr argueso hat ach indcividual, inpussuing his or her self-inters also maidmizes the common welare;cooperation is therefore unne lssary- -In rafiuty modern economies arecharacterized 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 informationis 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, andfuture 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 undoubtedlygre'ater. When economic change is.siow, making predictions about fu-ture prices is relatively easy: a simple projection of the recent past willdo. But when economic change is rapid, predicting the future is muchmore diffi'cult. There is often need for coordination. beyond what mar-kers can provide; hence there may be a greater role for govermecnt tocreate institutions and facilitate coordiination.

Missing Infonnatio and Credit Markets. Among the markets most af-fectedi by information problemns are capita markets. Even in industrialeconomies, equity markers generally do not finance much new invest-ment, largely because of information asymmetries (Stiglitz 1993a). Ineconomies where'equity markers are weak- or absent, credit mecha-nisms become the primary vehicle for raising capital and. diversifyringand spreading risk- But credit markects-even those flee of interest ratecontrols-are often characterized by credit rationing. Credit is seldomallocated to the highest bidden. This is for an obvious reason: bidders

* are bidding promnises; those who bid the most may not be able to fulfillteir promiss Lenders are concerned nor with the promised return

but the actual return. As a result, capital is allocated by a screening andevaluation 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 addressthis problemn in the credit marker, going beyond the normal regulatoryfunctions 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 andother 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 otherforms of interventions, reduced the risks borne by investors.

A'

Scale Econmnies, Externalities, and Ciooperation. Coordination may yieldsubstantial benefits when large indivisibiliries in investment lead toeconomnies of scale. One such case arises when investments are inter-

dependenc. For example, if a steel plant and a steel-uising industry areneeded 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-usingindustry unless there is a plant. If each awaits the other, nothing hap-

pens. Market failures due to incomplete markets, such as the absenceof 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 meanthat 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 arerelated to learninge spiovers associated with developing markets,

spillovers associated wit discovering what goods can be produced, andspillovers assocated wevith the incomplete appropriability of tecnologicalknowledge. In each case, fims undertaking a new actvity face long-term,noncollateralizable ivestments that are difficult to finance. Coordinationfailures loom especialy large where "diFuse externalities exist and thereare 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 forplanning in the 1950s and 1960s. Such planning ulled in part because

it attempted to concentrate all rclevant infonadvon nm the goverumenitplanning bureau. This was sinply inmpossible; infrroation is too diffuse, too complex. Plnmning ministries were not adept at generating the

infornation needed to support investmnt dceisions; as a rexsultemnyplanned 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 upmechanisms through which business and government could exchangeinformation and coordinate investment decisions. Japan and Korea, andlater 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 isdiscussed 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 slackor 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 cooperationavoid these problems? They combined cooperative behavior-indcludingsharing of information among firms and between the private and publicsectors, 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 ; AtiZNpetition has usually been vigorous. The Japanesc government hasproceeded- on the assumption that competition among fewer, more '--

evenly matched firms is preferable to having one large firm competing s4with many smaller rivals, a principle that is well-recognized in athletic .

- -competitions (NalebuffadSii193.vContest-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 and 7somchmes 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, contestg. ! 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-toauut7laeyp"' 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 iX for cwii.c peduiioh~nL aidnducingpartipantsto cooper- !. . ffhtolerance ofz5nconfdrmss has d offandsomely0 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.e0~ ~~~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 ineA&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-scalep- 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-sported successfilly had an easier time getting fiure loans. Because suchcredit was fungible and credit generally was tight, firms sometimes dilverted 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 monitoringproblems were greater, since both export perfomnance atid the relatedprojects 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 whichcommercial banks competed to provide financial services in line withgovernment policy-for exampic, by opening rural branches-in hopesof winning highly sought-after urban branch licenses. The contest alsogave 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'ti156 7

:--- - : - . - . - 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 ulnabSInInxhlnhm.um udncmhhAnadit . -,two?"

44-. ~~~~~~~~~~~~ . t~~~~~iflncrwimm"AS1 ain ls kMhlngaahmla.

Am* rudpmiadl bmulmyrnL

* h~iwIUcDlvsnn andvbmmrad. Mise a-. h. zdlIc.Mlie. *. -~~~~~~~~~~~~~~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 complyewirsthaguidaceiabou the spreatsdubetwee depositiand lni-rngrtes were-nrepermitteo t rexpandt thesepoir brnhnewrs (Yohin o19)

Duringtherapidgrothperiod the acMtinsr ofd Ineratonlt Tradandonduisrysal (Mrr lieamnsedIt capacit Thpasio in sevral keytibindu-tderies-usueall heavyc andichemicald induestiesgI makt'hea195s ands 1960s,Dmflpimted entrydoandi mrkestrce compewtinga irmpworts. ofim in-the pro-

tected ndustrywerereuireds tos cordnaeivsmnstopeetapcsite 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 exportmarket sharcs.

Our final example is that of the coordination of investments in Koreaunder 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 economictargets fbr performance. Its economic success is still not clearly estab-lished, and many observers have judged it a failure, but it illustrates verywell 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 throughcoordinating 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 rewardsconsisted of quasi-rents due to import restrictions and restrictions onentry, 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 Koreawas the result of rwo performance-based rules. Fi-st, government estab-Iished a timetable for the attainment of international competitiveness ineach industrial sector Firms that failed to maintain the pace of cost re-cductions in line with the international norms faced both political andeconomic sanctions (Kim and Leipziger 1993). Second, because accessto credit, even at early nonexporting stages of the development of HCI

investments, was linked to the export performance of other productsproduced 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 rewardsare unrelated to effort, contests will not increase compe'tiion. There isone 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 industriallicensinwere 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 rulewas broadly shared among HPAE governments and is almost unique tothe 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 tocreate 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 domesticeconomy 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_noR 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.14o<tjf, -- rSi,'22...{

e Sz >o 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 theprocess of entering international markets, firms learn a great deal andnor 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, successin the production of intermediate goods requires producing to standardsthat are typically higher than those tat prevail within developingeconomies. The contacts made in the process of exports may also be ofvalue when the firm decides to enter related markets. It will know towhom to turn to acquire advanced technology.

While from a social perspective, success in exporting may be a betterindicator of whether a firm merits additional fuinds than success in sellingdomestically, 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 concerningexternal markets than they are concerning internal markets, and from thebanks 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 thathave attempted selective interventions. Though they made mistakes ofjudgment, they generally did not force decisions on others who were will-ing to risk their own capital (The creation of the Mitsubishi AutomobileCompany in) 965, rejectng m's gtidance ro refrain from entering theproduction of automobiles, is a case in point.) This is one ofthe strengthsof decentralized decisionmalcing it provides insurance against mistakenviews being given too much dominance.

Running the Gamen The Role of Referees. Referees are also fundamental tocontests. Someone must enforce the rules, administer the rewards, anddiscourage cheating. If one group of participants in a contest capturesit and turns it to their advantage, they will reap rewards without thenecessary effort. If some participants feel that others are excessivelyfavored, 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 essentialto repayment Banking supervisors and customs authorities could also

.-,use their coercive power to enforce the rules. In the case of banldng inJapan, the supervisory authorities of the Bank ofJapan were responsiblefor 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 politicalleadership and the economic administration monitored performance.

The. Liits and Coustaints of Cootests

What determines whether contests can be effiecively used to promorebetter allocation and fitster growth? There appear to be two importantelements the relative benefits of coordinated behavior and the institu-tional costs of implementing nonmarket, contest-bared compeiitivediscipline.

Where the benefits of coordinated behavior are not great-for exam-ple, in small, highly open economics with good entrepreneurial skillsand 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 beefficient, and social and private returns will coincide- This may explainwhy HPAEs such as Hong Kong and Singapore have not found it worth-while to run contests and establish the types of highly structuredinformation-sharing mechanisms (deliberative councils) found in Japanand Korea, despite the high quality of their bureaucracies.

Even in cases where coordinadion benefits are potentially large-innontraded goods, externalities, or interdependent investments-govermnents may not be able to enforce performance-based coordinationrules. Taiwan, China, is an example. Despite benefits fiom coordinatedbehavior that were presumably potentially as large as in Korea, and thepresence of a high-quality-civil service, contests of the types employed inJapan and Korea were not used. This may have been due to Taiwan,China's, different industrial structure. With a large number ofsmall andmedium-scale firms, in contrast to Japans and Korea's relatively smallnumber of major industrial groups, the institutional costs of attemptingto 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, canalso be interpreted as a case of high institutional costs. Civil services inthese 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 selectivetO functional incentives and fiom extensive reliance on contests duringthe rapid growth period tO market-based competiuon in the 1990s. In

-large measure this is due to perceptions among poliyrnakers that thelarge 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<ol^ nuL. o "mtsincl-u) ^c. or-moseb thnoi tsucr-asi--.:o;

~~~~~~~~~~~~~~~~~~~~~&~ig- dj#et.uiibyi nnNaarnlnen,-\nttniioegn cmPulnscV a ura.N-U-gmii ' w3 s = 3 <m . ,v-¢.v; P1 71 ¢2-. l p;., $st

%~~~~~~~~~~~~~~~~~~~-Wr -- -vS -iissS ; t->. <,<-3 V ; <gX* -%w,7 jb, -t4i s H: ^_s .I- 4, 4~zC g -; s ' f £'e 2 J 8,'. m g'4 g:(~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

m usg£aousmfluestrXn^SDriomeg-¢w li tel\97osaaI-S-lsineznotXiiaNgi yjmr&|w[1Sim;-W*t;S>-rt-4r4---R,-F,3 98,' ;;~- w-t C q i£M t- sovv

- )sD~~~~~~~ ~~~~ inreqmnemure,nnn;otdaterneationalreemisvaiSSnieEtR ;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<Xsor;cell!loLeaElDr-u=¢Es{

~ =>. > $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:$rDvdcpcyTa-ytt '-ru blcne le.d>ndu ngn l5W.^grLtg'@eS=< @X<' ncg- 4cot.2-Xti.>C{vlsgt/4't. =.E$*-.w'xjl2>(g,.<t,'.X>an >i

: i Q. ut9. Wa | = 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. Simnilarconcerns 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 hasnotbeencompletey successfulL Each of the broad views of the relationship be-

tween public policy and rapid growth-neoassical, revisionist, andmarkcet-friendlyadds important elements to our undcrstandingl butnone fuiily captures the compleity of public poiicy and rapid growth inthe Hfl'AE. We have proposed using a fuinctional approach to under-standing the relationship between policies and gowth. In this vieewpolicies 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 formedthe basis for their rapidh growth. More interventonist policies have thepotenti to contribute to growth in cases wohrec the address economiccoordination problems, but to succeed they must combine the benefitsof 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 inJapan and Korea. The historical origins of these economies' high-qualityeconomic administrations and the mcchanisms that they used to main-cain quality and limit political influence and corruption are the subjectof chapter 4. The second is the pragmatism and flexibility of govern-ments in the high-perfonning East Asian economies where contestswere 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 nextchapteL

102

PUBLIC. PO'LICyA i

Notes1. 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 arates of protection (Em's) are based on triffdata, not on es- single-minded direcrion to economic activity, promotingtimates 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 commonOn Taiwan, China, see Wadc (1990). On Japan, see Itoh set of goals.and Klyono (1988). Wade does not present or dce estimatesof 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 areERPS 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 aChina, as a modd of limited intervention, therc are no esti- dicussion of the instiutional features of the Ha drivemates 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 StabIltan ort

ACROECONOMIC STABILiTY AND RAPID Ex-

port growth were two key elements in scarningthe virmous cirdes of high rates of accumula-tion, ficent allocation, and strong p roductiv-

ity growth thiat formed the basis for East Asia'ssuiccess. Governments achieved macroeco-

nomic stability by adhering to orthodox policy prescriptions-in partic-ular, by holding budget deficits to levels that could -be prudentlyfin2nced. They achieved rapid export growth through an export push-a combination of orchodox, market-oriented mechanisms, miany linkedto Macconomic stabilitye and omplementary pro-export incentives.'

Policy initatives in these two areas shared two features that furtierillustrate the framework outlined in chapter 2: respect for certainfindamentals-for example, fiscal pridence and avoidance of exchangerate overvaluation; and quick and flexible responses to changing eco-nomic circumstances. Policies to achieve macroeconomic stability andrapid 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 packgensthdrugh trial and error.

We define macroeconomic stability to mean that inflation was keptunder control, inti eal and external debt remained manageable, andmacroeconomic crists tat emerged were resolved quiddy, usuaiywithin a year or two. Short recessions and policy adjustments to macro-economuc stress somtimces sq the private scector. But these wearetransidonalperiods to new episodes of rapid growth, quite unlike theyeam 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 frombroad, export-friendly measures such as avoiding an appreciated exchangerate, which are esmployed in all the HPAEs, to government-mn export con-tests, which have been used primarilyin Japan, the Republic of Korea, andTaiwan, China. Universal export incentives, such as tax breaks and credirguarantees for all exporters, have been the main instruments in HongKong 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 thesuccess they have achieved. The second half of this chapter reviews theevolution of export-push policies in each of the HPAEs.

Pragmatic Orthodoxy in MacroeconomicManagement

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 ratesand 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 successfulin keeping deficits within bounds and have therefore had more troublemanaging 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 toSwitzerland7s, 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 at18 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 somegovernments have run substantial deficits, none has financed a deficit ina manner that destabilized the economy. The level of the deficit that isaffordable, and hence nor destabilizing, is specific to each economy. It isgenerally larger the faster the rate of growth and the larger the pool ofprivate savings (both at home and abroad) relative to private investment.In both these dimensions, the HPAEs have perforned better than manyof their developing-economy counterparts during the past thirty years.Because of this, while some economies have had higher inflation thanothers, none has had to endure the very highi, debilitating inflation thathas troubled other developing economies.

The HPABs chose a variery of macoeconomic policy paths because ofdifferent economic conditions and preferences. All lay within thebounds of prudent stabiliy, and whenever the macroeconomy appearedto be in danger of moving out of control, swift action was taken to re-store stability. This was true even when the source of macroeconomicinstability was policies intended to promote growth in the real economy.

For example, the heavy and chemical industries drive in Korea, to whichwe shall return below, was modified when its adverse impact on the keyindicators of macoeconomic stability, inflation, and the real exchangerate became excessive.

In the following section we discuss how low deficits enabled thesuccessful 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 varieyof 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 sectorcould count on relatively constant prices and interest rates. Here weconsider 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 notdramatically better as a group than other developing economies'. Butthere are two distinctive things about the HPAEs. First, dtey almost al-ways kept the deficit within the limits that could be financed withoutmacroeconomic destabilization. Second, these limits were higher thanin 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 outprivate borrwing; large external financing of the deficit leads to debtaises 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 currencyboard 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 whyKorea was able to keep inflation, extrnal borrowing and interest rateswithin bounds. Malaysia and Thailand are more complicated. Thai-lands budget deficits were about average for devdopi economies inthe 1980s, while Malaysiais were substantally bigger than average. Bothran bigger budget deficits than such troubled economies as Argentina,Brazil, Mexico, and the Philippines. Unlike these and other economiesthat 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 Asianand 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 higherin Malaysia and Thailand than in less successfil economies, ahigher budget deficit could be financed. Because high growth in-creases the demand for financial assets, Malaysia and Thailandwere able to absorb higher levels of monetary financing without arapid rise in inflation. Moreover, their rapid GDP growth raised thelevel of sustainable domestic and external borrowin (an economy

can borrow more for a given debt to GDP ratio when GDP is risingrapidly). In contrast, many Latin American economies fell into avicious cycle of low growth and unsustainable deficis

a Second, there was fredback from bigb financial savings. Savingsrates 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-savingsfinancial policies described in chapter 5. This firther increased thedemand fir money and other domestic financial assets, makingincreased domestic financing of the deficit possible without re-sortng to inflationary financing. In Malaysia, the government'sProvident Fund mobilized domesuc savings for the goverrment'suse in noninflationary financing of the deficit

109

~~~~~~~LE - - _~~~~~rIR C L E-

* Third, there were low init debt ratios. In Thailand, the initiallevet 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 burscsof money aceation that afflict other developing economies. Figure 3.1shows money creation as a ratio to GDP in Korea, Malaysia, and Thailandand in three unstable comparators-Argentina, Mexico, and Zairc. Thecontrast is striking while money creation has been relatively constantamong the HPAEs, each of the comparators experienced two episodes ofrapid 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. Table3.2 shows the low inflation in Malaysia, Singapore, Thailand, and Tai-wan, China. Indonesia and Korea have higher inflation but sdll farbelow Latin America. Jnternational experience suggests inflation below20 percent, a level not breached by any of the FPAEs during their rapidgrowth 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 toHong .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 Indonesiaand Taiwan, China (and to a lesser extent Korea), aversion to inflation

MiK& ld^y^ l grew out-of traumatic inflationary spirals tiat accompanied economicand 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- - ZaireMArge.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 nominalO 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 ::-

;4 S 'MIRACLE;

Korea, Malaysia, and Thailand, compared with Argentina, Ghana, andMexico. As with money creation, the contrast is rcmarktble. In the EastAsian cases, low inflation and flexible financial policies kept real interestrates within a narrow range. For the comparators, the combination ofnominal 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 Asiaand Other Selected Economies

Real Interest rate (percent)

20 -~ ~ ~ I

-70 J j Rep.ofKorea 1I 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 thefour with foreign debt has faced a aisis, in the sense of having toreschedule 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-GNPratios have been quite high compared with other indebted economies(see table 3.3). As with fiscal deficits, however, fivorable feedback fromother policies enabled the HPAE debtors to sustain higher external debtto GDP han other economies. High levels of exports meant that foreignexchange was readily available to service the foreign debt. Similarly, highgrowth implied that retrs on borrowed capital were sufficient to paythe interest.

Korea's successful handling of a very high foreign debt illustratesthese 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 itequaled more than half its GNP. Yet because of its high export-GNP rauioand rapid overall growth, Korea neverlost creditwordliness. From 1986the govermment pursued an actve debr-reducton policr, drawing on

Table 3.3 Intional lndebtedness

>Akioftvnal4g-." ;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 14percent (In contrast, when Mexico faced scvere problems with its cred-itors in 1982, it had a much /owerdebt to GNP ratio than Korea in 1984but 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). Incontrast to such economies as Bolivia and Ghana, the East Asianeconomies did not cling to a given nominal exchange rare (or inade-quare rare of nominal depreciation) in the face of continuing inflationbut depreciated when necessary, someumes quite sharply. Fiscal pru-dence prevented the cxcessive demand pressures that appreciated the realexchange rate in such economies as C6te d'Ivoire and Nigeria.

The evolution of exchange rate regimes in the HPAES has been broadlysimilar. Hong Kong, Malaysia, and Singapore pegged their curencies tothe British pound during the Breon Woods period, then floated themin 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 25Ko 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 1984the baht was devalued and floated as part of an adjustment program. In-donesia fixed the rupiahto the U.S. dollar from 1971 to 1978. Becauseinflation 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 toi-the dollar but resorted to four major devaluations between 1961 and thestart 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 afixed linked) rate regime in 1983 in the face of fierce speculationagainst the currency prompted by political uncertainty. Under the man-aged floating regines that began in the early 1980s, policymakers nolongrer set rates but attempt to influence them at the margin, generallyto move in parallel ro the U.S. dollar Because the United States has beenthe region's major export market, maintaining a stable and at timesslighdy undervalued exchange rate vis-a-vis the U.S. dollar has assistedexporters.

The HPAES' success at maintaining stablc exchange rates is apparent infigure 3.3, which contrasts the remarkable stability of real exchange ratessince 1970 in Korea, Malaysia, and Thailand with rhe severe exhngerate instability in Argentina, Peru, and Sri Lanka. Argentina repeatedlyattempted to use the exchange rate as a nominal anchor against high in-flation (fr example, in 1973-74 and 1980-81). But failure to keepother macroeconomic fundamentals in line led to the collapse of the realexchange rate and sharp real devaluations. In contrast, the East Asianeconomies' pragmatic macroecononic management eambled them toavoid swings of the real exchange rate, even in the face of major externalshocks (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 andtighdty supervsing banks, governments reduced the spillover from the

"5

Fiupre 33 Exmples of Real Exchange Rate variabiryy in East Asiaand Other Selected Economies

Real exchange rate (percent)

300 1 Rep oKorea-

25D 0-

200

| SD- __ | I__ I_ I___ __I., - __

I I ; I' : I i .01970 1972 1974 1976 1978 I980 1982 1984 1986 1988

300-- -- Arg ntarM300- ------- - -- 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 exacerbatedfiscal woes. Second, flexdble labor and capital markets enabled the rcalsector to react quiddy to government initiatives, seting off new growthcycles tha eased the recessionary impaa of stabilization measures. Thusthe 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" "' 4 t.>-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,<w;-;='i--- <6_. f $X

sjeoomamr £t'F"'vhe awer"'ou '.-s - -wlWtora-- -..

ico, din flled to tak toiml ac5tidon inepos to uc shockl s i-tand *su

Somxe nIobsrvr hv a rge ta Ea'noooS>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&q8w 4 d4x,O Aursou1s4 ipA o

-o . - ;DeB errcostlsXnomy<naewdroDalln - trto': ; *mrC&jp- ntoJran XanonianadiLiebanksav-w1rkt1iaru&oPmnsm;amapcemad

^n~~~~~t .4l~~a 4

: > 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 percentOfGDP measures the extent by which absorption had to be reduced-theects--e of increased principal repayments of long-term debt, higher in-

terest payments, and reduced income from oil and gas imports. Thismassive resource shift, also associated with trade liberalization, wasbrought 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 ina boom of manufactured exports that pushed the export growth rate to25 percent a year for the period 1985-91. Since then Indonesia haschalked up average annual growth of about 7.1 percent.

Respodimgto M economic Crisis in Korea. In 1979, Korea encountereda variety of problems that threatened to undercut the 1970s' impressivegro wthu Rising oil prices battered Korea's termns of trade, the world re-cession dampened exporc demand, and high interest rates boosted debtservice costs. Korea was not unique in these troubles, of course; thesewere the same woes that led to debt cises in many economnies outside ofEast 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 Park1989).

Korea had a few structural advantages, however Unlike economiesthat 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. Moreimportant, Korea responded quicldy to its troubles with an aggressive

January 1980 stabilizaton package backed by UmF standby credits. Thegovernment ended the fixed exchangre rate regime, devalued the won by17 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 accountdeficit approached 9 percent of GDP. The strong medicine was pardy re-sponsible for the economys worsning symptoms: the devaluationspurred inflation, while tighter aggregate demand polcies excerbated

:i9g

* ~~~~~the drop in output. Even so, and despite political outraies over Kordsrising 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, thegov:edr nes proumpt and effective response to a potential crisisstresngthened the conomy, preparing it for rapid growth in the 1980s.

Mjusting n to in Sho i Thailand. Thaland only partaly adlusted tothe first oil shock and in the late 1970s engaged in a mild private andpublic spending boom Th then mcmeh second oil shock and the rise inworld interest rates. By 1980-81, thet cornlidated public sector deficitwas 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 therelatively open capital marke. The government therfci re took the alter-native path, fiscl contraction, movig gradualy but consistenty dudrgthe 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 ofGDP in 198142 to 1.6 percent in 1986-87, when adjustment was es-sentially complete. Meanwhile, steeper tax rates and tougher colectionefTorts boosted cepntal gavernent ax revenue f'rom 13 percent of GD?in 1982 to 16 percent in 1988. The amdjustment process was facilitatedby a 1984 devaluation.

Thai gradualism was possible because foreign borowing had beenmode:ra and the economy did not yet fatce a crsis. But grdualism wasnot 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 1980-81world 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 unusualepisode included government encouragement of high wages between1979 and 1981, which was intended to speed the decline of labor-intensive production but backfired by eroding competitiveness, and theapprecation of the Singapore dollar's trade-'weighted exchange rare.These were eamcerbated by a sharp decline in public investmentL due tothe simultaneous completion of several major projects. Private invest-ment, which had been declining gradually for several years, also droppedin 1985, apparently because of flling external demnand resulting fromSingaprs decining comperitiveness

Recognizing the problems, the government devised an integratedpolicy package that reversed the high wage policy to restore Singapore'scompetitivenes and stimulated domestic demand. The govermecntalso cut the emnployers compulsory contribution to the Central ProvWdent Fund from 25 percent of wages to 10 percent, reduced corporatetaxes, and introduced accelerated depredation. Income taxes were cutandt 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 reached9.5 percent, a level maintained through 1990.

How Macreconic Stabiriy Contnb-- to Gmowth

It cannot be a coincidence that all of these seven economies have hadexceptionally high growth by world standards, and all have had unusualsuccess 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 fivorableenvirorinmentfor growth.

Thaere are four main reasons why high inflation is likely to be adverseFor growth. and why rhe conservatve policaes followed in the HPAs arelikely to have been favorable for growth. Fsrl economies that are notfuly adjusted to a given rate ofinflation usually suffer from relative pricedistortionspcaused 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 relappreciatons 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 mostHPAEs, compared with other developing economies, which meant thatthese 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 basedon 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 fiscalproblem is intensified by inflation, and public savings may be reduced.Public savings have been an important component of the unusually highlevels of total savings in the HPAEs, compared with other low- andmiddle-income economies (see chapter 5).

Third, high inflation is inevitably unstble. There is uncertaintyabout 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 thehigher the inflation, the more likely are measures to reduce it. These inturn have a contractonary impact on private investment in the shortrun. Finally, high and variable inflation imposes substantial institutionalcosts in many economies. During periods of price volatility, scarce man-agerial resources in the economy are drawn into financial as opposed toreal 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; aneconomy with a slightly lower-than-average inflation rare for a longerperod 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 inflationrates and low growth. The most important case is that of India. Thereare also cases of high inflation associated with high growth. The mostimportant example is Brazil from 1968 to 1980. More recenty, Turke/sgrowth rate from 1981 to 1990 averaged 5.4 percent, while its inflationrate averaged 46 percent (ittle and others, forthcoming). Low to mod-crate inflation may be a necessary condition for growth, but it dearly isnot sufficent.

Relatively cautious fiscal and fbreign borrwing policies meant thatserious debt crises were avoided, which reduced the stopgo pattern ofcrisis and response that characterized many developing cconomies in the

In

MACROACCjj_~.

1980s. As noted above, there have been problems, notably in Korea andMalaysia, but they were dealt with swiftly. The Indonesian and, evenmore, Thai reactions were more gradual but also effective. Avoidingcrises and the need for rescheduling meant that creditworthiness wasmaintained, and it was easier to borrow in the shorc term and to avoidvery deep cuts, especially in invesmenL Sudden reductions in aggregatedemand and in investment compelled by debt crises were major causesof the sharp declines in growth rates in many of the heavily indebtedeconomies of Latin America and elsewhere.

Although macroeconomic stability and prompt responses to macro-economic shocks were not the whole story of the HPAES' success, thesefactors created a basis firom which policies intended to affect the realeconomy-the supply side-could be launched in an environment ofstable real interest and exchange rates. We now tum to the most broadlyshared 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 andby 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, sincethese economies differed in the degree and selectivity of promotion, andeach economypassed through several stages. Because of tis and becauseecport promotion has played such an important role in the East Asianmirade, we condude this chapter with a survey of the evolution ofexport-push polices in all seven developing HPAEs (these arc summa-rized in table 3.5). Appendix 3.1 givcs brief policy histories of eacheconomy in the fiorm of timelines.

Macmecononic Stability Facilitated omre Open Economies

The success of the HPAES rests pardy on what they have done andpartly 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-substinicingindustrialirzatiLon-,1953-57

J AF:K~~~~~~~~~~~Eport promotion1958-72 -

ZTilting.twr

~ incur 1967 73 tiC .;;J*-Export promotion. Fvoring import sbtSelectiv6 inter:vendion>- 2andirnPorts subst-~ Indi'srial. con-solida- sio 197180:

Omlandcommoity dioi4hii:heavy 2nd stontut7-o5 . no-boomIn,974-81 cciakuwe growth, 1973-080

* ~~Functionalabdlbcraizaon,

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 paymentsdeficits. 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 andKorea. 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 particularlyhard 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 accountproblems, as consumers and firms buy imported goods not previouslyavailable. The trade regimes of the eight H-PAEs have differed widely, buteach 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, andforeign aid or concessional loanis to offset a tempomarily deterioratingcurrent account.

The dose link between successfuil macroeconomic policies and tradeliberalization can be seen in the experiences of Indonesia, Korea, andTaiwan, 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 widespreadremnoval of quantitative import restrictions. In Korea, devaluations andtrade 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 chatmarked the start of the Suharto era. The new government abolishedcomprehensive import licensing and in 1970 united and devalued theexchange rate. After an inward-looking period in the mid.-1970s, tradeliberalization resumed in the 1980s and has since proceedecd togetherwith 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 ofimport 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 carryit 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 discussagl three belowo

The very large current surpluses that Taiwan, China, ranin 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? Exportswould have become less competitive, reduced export growth wouldprobably have had a deflationary effect, and this in rum could havereduced savings. Alternatively-and more realistically-thc poten-tial deflationary effect would have been offset by increased publicexpenditure, leading to a budget deficir.

* Korea used exchange rate protection from 1986 to 1989 when itran a current account surplus (which peaked at 8 percent of GDPin 1988). A desire to protect the export industries was certainly afictor in Korean exdcange rate policy, but tie main concern was toreduce the debt ratio and build up reserves to avoid repeating theclose 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 realappreciation of the rupiah on non-oil exports had been offset bythe rise in the quantity and value of oil and natural gas exporcs.Rather, the aim of devaluation was to encourage non-oil exportsand slow import growth. By 1982, when the balance of paymentshad sharply deteriorated, the wisdom of the 1978 devaluation wasclear.

One can see a fairly clear relationship between devaluations and ex-port growth in the 1980s. Taiwan, China's, real exchange rate relative tothe United States depreciated sharply from 1980 to 1985 (the period ofgeneral dollar appreciation), and the result can be dearly scen in an ex-port boom to the United States. The Korean real exchange rate was keptfairly stble by numerous nominal devaluations during the period of thefirst great Korean export boom, 1963-72. The effects on exports of realdevaluation from 1982 to 1988 can also be seen very dearly. TheMalaysian 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 exportgrowth were quite dramatic, even though there were also external fiatorsexplaining export growth, notably the availability of capitl fiom Japanand Taiwan, China, for developing export indlustries. From 1986 to1989 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 wasparticularly 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 currentlyin the midst of a fourth Unlike several of the larger economies in East

X Asia, which evolved from protectionist, inward-looking trade regimestoward 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 importingraw materials and intermediate goods for processing and export withadded value.

War, P oanstrclion, and Land Refonn (1956O-01. Korea had a relaivelywell-developed infrascructre at the end of World War II, but the parti-tioning of the peninsula by U.S. and Soviet forces severed economiclinks between the heavily Midustrial north and agricultural south. Thewar took a heavy toll, taking 1.5 million lives and destroying two-thirdsof 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 warmay have helped to prepare Korea for an industrial takeoff by looseninga rigid social structure, opening the way for fundamental changes inoudook. While developmenr efforts in tie 1950s induded several falsesrans, progress was made in reconstruction, including the restoration oftransportaton 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 biasedin fiavor of exports as a whole but essentially neutral with respect to dtecomposition 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 exchangeearnings to import restricted commodities, and permission to borrow inforeign currcncies. This system not only avoided hampering exporterswith restrictions on capital and intermediate inputs for their own use,but it also gave access to he firavomble exchange rates determined byscarcity 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&i<E&YZY.5. suGbsud - u - a '.' equwa.. inSu-Yen .wE rceMt) (percent)

-1958 1.79 97 3 ~ '01959.19 0-19611.962 10.7173 23`19638 7 29 .

1964 114 '~~ ~~29: 21 .5K19 . .4 -225

:.~~~~~~~~~16 113 - ; -69 ,- .J1.,--:-:5

...196., 148 44 33,1970 118 .44 3: .-. :

...

r'c onnrts;E Efct. 'i=bunos ,

xz8

MACROECON

Tariff exemptions were given to indirect as well as direct exporters, andgenerous wastage allowances on imported intermediates allowed someresale (table 3.6). These enabled exporters to avoid distortions from pro-

tection and in some cases to benefit from the protection of the domesticmarker.

Support for exports was also channeled through rhe stare-controlled

banking system. Objectves were implemented through bank loans ex-

plicitly earmarked. by the government for particular activities or indus-tries, lent passively by banks at preferential interest rates. Following

explicit government directives, banks increasingly used export perfor-mance as the criterion of creditworthiness.

Heavy and Chemical Indostries Drive 11973-791. The heavy and chemical

industries (HCI) drive was a major policy shift, away from the neutral in-

centives of the takeoffperiod to a commitment by govremment to use allits levers to steer resources into specific sectors to rapidly alter the indus-

trial structure. Special legislation singled out six strategic industries-steel, petrochemicals, nonferrous metals, shipbuilding, electronics, andmachinery-to receive support, induding tax incentives, detailed engi-neering, subsidized public services, and preferential financing. Thce gov-ernment chose the fir dutee sectors to enhance self-sufficiency inindustrial raw materials, while the latter three were meant to begroomed inro technology-intensive export industries

Unlike other governments thac have attempted to build a heavy-in-dustry sector, Korea was at least partly successful. One reason is that the

government made dear fiom the outset that these industries were ex-pected to become internationally competitive. As a result, projectstended to be forward-looking--only current technology was imported,and U.SI-trained. Korean scientists and enginaeers were recruited. (See

box 33.)However, interventions were so pervasive that botdenecks emerged,

largscale debts were incurred, and labor-intensive industries were

starved of credit. When the second oil shock hit, inflation was alreadyhigh, and the exchange rate had apprecated; capacity utiliation in the:HC: seccor was low, and exports were faltering. The government

switchled course.Functionl Incentives and Liberalization (1980-90J. Support for strategic

industries was curmiled and abmpdy reversed. The currency was de-

valued, and credit allocation policies switched, with a termination of

large-scale preferences to the HcI sector. The five-year economic plan

129

-* fr:tU.C L E' .

: ~ rE . -%. -Cj

,-cmpnc ,,-,-

P1...L ,,mdopmentBaEcL: ySps i4B . 2 -_ )ctytpscueebeRayEng,:mlcK0wav,es0vens,3nfNthe:Earler tioizhzr-nontss ays-ng.cheap loan- 5

033

apancsg-'co ramadcaelz cnt S-It?' __~~~~~~-w ,

B EW4 mt.'w"gWW $8!stft47 ha&obtide3".*>¼mt'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~- aMoi~~~~~~~~~_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.tiME- Z><dustM-iDsSoxrxSStesrtrialed poSosoulcylis in it apprs:oa o fnaLncasetor rneform. The overn-si

mfed nt1has been opngithd thersdeat theompexcesse of the 1970m ws and

oh bnd cartiue pn glarlyaact eateinbaleo fo susete mand wlesacdcu industries. ith remnants offnancsr liae anr port-oitepizaie Ko grais pay-

*Infoprointervention policies79and foruedo the r iue oest ablishg andkme

industries, sidpdetfnna h devsector. en of t-chnology and the promo-

-1309.y, ssovemmenn1. officms Wwortion ofcompctiton.

mdn*e haien 197 incgnwized th reideopei of the ex onesomy th was and

has bendciv parclarlyn andv in brailors role sunrt and oveaseveragedlly

; . ~~~~~~~~~~utrieslathedita remants ofps financies are ipart-libtepraiceio progreamis.p

-- ~ ~ ~ ~ ~~~~~m for pror ntrvention sicp9 a olcies ad for the failucure tof edbishresse

: - - ~~~~~~industrident fiupport foste dvlpetofrc.lg adtepoo

MACRO: EcoNP.

The Evolition of Export Push in Taiwan, China

Development policy in Taiwan, China, has consisted of five stages inwhich the government has implemented comprehensive but changingpoliqc packages. Throughout, low inflation and macroeconomic stabil-ity have been a foundation for growth-enhancing policy initiatives, andsince the late 1950s -export growth has also been a fundamental goal.

Land Reform and Reconstruction (1949-52). When the Taiwanese author-ities took over in 1949, one of their first initiatives was an ambitiousland reform program. The program fostered social and political stabilityand increased agricultural production. Greater agricultural output pro-videdraw materials for exports and earned foreign exchange to fund im-ports of machinery, equipment, and industrial raw marerials. These inturn helped to make possible subsequent cxport-led rapid development

Impat-Substitg Industializaoun (1953-571. During the second stage,the government attempted to develop industry as the base for economicself-sufficiency The government invested heavily in infrastructure, ex-panding transportation and power necworks built by the Japanese; U.S.aiid was an important source offiriance, funding 49 percent of public in-vestment in infiastructure. Excensive quantitative restrictions and highta-riff rates shielded domestic consumer goods from foreign competition.To take advantage of abundant labor, the government subsidized somelight industries, particularly textiles. Consumer goods industries such astextiles, apparel, wood andlearher products, and bicycles developedveryrapidly. By the end of the 1950s, industrial production had doubled.

-'However, the costs ofimporc substitution increased over time. By stim-ulating the import of capital and intermediate goods while penalizingexports, imporc substitution contributed to a growing trade deficit, fi-nanced largely by U.S. aid. As the small domestic market became satu-rated, overall growth declined, fiom 9 percent in the early 1950s to 6.5percent in the mid-1 950s.

Expot Prmotion 11958-72). Anticipating the termination of U.S. aid

2and hence a need to obtain foreign exchange, the govermnent shifted toa policy of outward orientation and export promotion. Staring in 1958it adopted a series of measures aimed at promoting exports and foreignmnvestment. Amultiple exchange rate system was replaced with a unitaryrate, and appreciation was avoided. Tariffs and import controls were-gradually reduced, especially for inputs to export. In addition, the Bankof Taiwan, China, offered low-interest loans to exporters. The govern-

131'

R~ ~ ~~TWL

ment also hired the Stanford Research Institute to idendifjy promising in-dustries for export promotion and development. On the basis of Tai-wan, Chinas, comparative advantage in low-cost labor and existingtechnical capabilities, the institure chose plastics, synthetic fibers, andelectronic components. Other indusrries subsequently promoted in-duded apparel, consumer eectronics, homc appliances, watches, anddocks. Direct foreign investment (Dn) played a catalytic role during thisperiod and replaced U.S. aid as the main source of foreign capitl. Al-though DFI was only 6 percent of gross capital formation in the 1960s,nearly 80 percent of it went into manufaacuring. More important, DFI

facilitated technology and skill transfers, leading to much improvementin quality and the diversification of industries.

The impact of these measures was dramatic. Exports, which hadgrown less than 12 percent annually berween 1953-62, grew 28 percerta year between 1963-72, rising from 123 million to almost 3 billiondollars. The transition firm import substitution to export promotionwas the most important policy change in Taiwan, China's, economic de-v-elopment. It shifted the economy from a relatively dosed to an openeconomic system and exposed it to the forces of international competi-tion and technological change.

Indusirial Consoidation and New Export Growth (1973-80). As hic 1970sprogressed, intemal and external challenges threatened the contunuationof export-led growth. The rapid manufacuring increases of the 1960sstrained transportaton, electric, and communications systems. Theisland suflrd profound extemal setbacks. More important, Taiwan,China's, light manufacuring industries faced new competition fromlower-wage producers abroad. As foreign investors rushed to the newlyopened mainland Chinese market; international confidence in Taiwan,Chin;:s, economy dedined. The 1973-74 oil crisis had dramatc leper-cussions for the Taiwanese economy. Real GNP grew only 1.2 percent in1974 and inflation climbed tO 47 percent, while exports declined in realterms by about 7 percent.

Beginning in 1973, the govenment chose a more self-relianr devel-opment strategy based on industrial consolidation and renewed export

growth. Once again, it mrned to foreign experts and commissioned theU.S. management firm ofArthur D. Litde to find solutons to the eco-.nomic crsis. Based on Taiwan, China's, economic needs and capabili-ties, the Americans recommended heavy investuents in infiasrmrure,industrial upgrading and secondary import substitution. A government

132

MACRO'ECO

plan incorporating the recommendations focused on development ofcapital-intensive, heavy, and petrochemical industries to increase pro-duction of raw materials and intermediates for the use of export indus-tries. Thle government also launched ten major public sector projects, at

a total cost of 8 bilion dollars, to rcvitaize the economy and removebottlenecks to economic growth. These included highways, milroads,airports, and construction of nuclear power plants..

High Technology and Moderunihaon 11981-I. As it entered its fourthdecade, Taiwan, China, confionted a challenging domestic and inter-nacional environmenc. The spectacular growth of the 1960s and early1970s sputtered to just below 7 percent in the late 1970sX Taiwan,ChinaTs, continued inrcgration into the world economy revealed struc-tural weaknesses, particularly the financial system's inability to matchthe increasing demands of induscialization and external trade. Exter-nally, Taiwan, Chinais, persistent trade surpluses with major tradingpartmes led to growing protectionism. In the later half of the 1980s,Taiwads exports faced an additonal loss of competitiveness due to theappreciadon of the Taiwan, China, dollar and rapidly rising wages.Manufacuing wages rose, undercumrng Taiwan, China's; advantage,and local firms moved production overseas. As in other first-generationEast Asian NiEs, Taiwan, China, manufacturers were squeezed betweenlower-wage NIEs in traditional, labor-intensive manufacturing on theone hand, and high-technology products from industrial economies onthe other.

Once again, the government moved to rescrure the economy. Afterextensive consultations with domestic and foreign advisers;, the govern-ment decided to focus on high-technology industries: information,biotechnology, electro-optics, machinery and precision instruents, andenvironmental technolo&rindustries. Theshift to ahigh-technologyecon-omy has necessiated the close coordinaton of industrial, financial, sci-ence and technology, and human resources policies. In 1984, thegovernment revised laws to provide tax incentives for manufacturerswho allocate a percentage of their revenues to research and developmenr(R&D). Incentives were given to industry to diversifyr and improve pro-duction techniques. The government encouraged the establishment ofventure-capital firms and revised university curricula to stengthen sci--nce, mathematics, engineering, and computer education. It began torecruit rechnical manpower from abroad by offering competitive salariesto forner Taiwan, China, residents living overseas.

'33

~1R'A CLE

In addition, in 1985 the government launched fourteen major infra-structure projects, including expansion of the energy, telecommnunica-tions, and transportation networks and development of water resourcesand national parks. The government's development plan for 1991-96calls for 330 billion dollars in public sector projects. To support these ef-fbrcs, the government has adopted an overall strategy of economic liber-alization and internationalization, including the lifting of foreigiexchange controls.

While it is too early to assess these policies fillly, a few observationsare possible. A growing number of small, high-technology firms pro-duce incrcasingly sophisticated and higher-value-added products. Formost firns, however, the transition to hi-tech industries has been diffi-cult. On one hand, the rapid rate of technological change and risingprotectionism in industrial economies make it increasingly difficult toobtain advanced technology. On the other, the small-scale structurc ofindustry is not conducive to the costly investments in R&D and skills-training needed to shift toward high technology. As a result, most Tai-wan, China, manufacturers are still assembling imported high-techcomponents.

Malaysia's Shift knm Resource-Based tD Manufacturd Exporus

During the first twentyyears afer independence, Malaysia continuedthe essentially free market trade and industrial policies of the colonialgovemrnment, aldtough it intervened extensively to promote mral devel-opment and provide social and physical infrastruacure. While the govern-ment protected import-competng industies, protection was generallyless strong than in other developing economies. Government was re-strained fiom heavily biasing incentives against agriculture by the eco-nomic and political importance of the plantation and mining sector In

1957, Malaysi's exports of tin anct natural rubber accounted for a thiidof GDP (Bruton and others 1993).

Import Subsitution [1950-701. The objective of Malaysia's limited im-porc substitution was the same as that in odter developing economies:reducing imports of consumer goods and increasing processing of nat-ural resources to create industial employment opportunities. The gov-ernment did not promote individual sectors, and effective rates ofprotection were very low, averaging 7 percent, compared with a range of25 percent to 92 percent in other economies at similar levels of income.3

134

MAC ROE CO

Nevertheless, between 1960 and 1980 import substitution combinedwith domestic demand expansion accounted for virtually all of Ma-laysia's growth of manuFacuring output (Salich, Yeah, and Meyanathan1993).

Combining Export Promotion and Import Substituion (197145). Malaysia's1969 ethnic conflicts triggered a reexamination of developmenc policyThe New Economic Policy (NEP) launched in 1971 had many dimen-sions intended to promote growth with equity. In trade policy, the gov-emmnent began more acive promotion of natural resource exports,particularly rubber, timber, palm oil, and petroleium, and light manu-factured exports, particularly textiles, footwear, and garmens Duringthe 1970s, export-promotion efforts were designed to complement themodest import-substitution regime inherited from the 1960s. MajorNEP export incentives induded taxable income deductions linked to ex-port performance and domestic input content, tax allowances forexport-related promotional expenses, and accelerated depreciation forfinms expormng more than 20 percent of their output. Credit policiespromored exports through guarantees and automatic rediscounting ofexporc financing at low interest rates.

Export-processing zones, fiee trade zones, and licensed manufactur-ing warehouses that permitted duty-free import of materials to beassembled or processed for export were crucial to the successfill combi-nation of import substitution and export promotion. Foreign invest-ment, particularly from Japan and elsewhere in Norcheast Asia, pouredin. By 1980, about 70 percent of manufactured exports originated in theexport-processing zones, primarily from foreign-owned firms. As theeconomy has developed, the proportion of exports from the zones de-dined to about 40 percent in 1989.

Adjustment and Liberalzation (1986). A combination of terms of tradeshocks and fiscal imbalances (described above) prompted the govcrn-ment to move in 1986 away from state-led industrialization. Promotionofprivateminvestment across a broad range of sectors was combined withmacroeconomic adjustment and continued efforts to increase manufac-tured exports.4 Tax incentives for exporters were increased and importswere liberalized. As a result, the average effective protection to induscrydeclined from 31 percent in 1979-80 (an internal World Bank report)to 17 percent in 1987 (C. Edwards 1990). At the same time, tariff re-ductions and changes in export incentives increased the variation in cf-fective protecCton across sectors.

[135

I -ALCL E

Indonesia Tries It Both Ways and Opts for Export Push

Indonesian trade policies have swung from protectionism to open-ness. The current open phase has gone fiurther than any earlier effortsand obtained the most impressive results.

Nationalism aud Guided Development (1948-66E Following independencein 1948, economic policy was shaped by a strong sense of nationalismflavored with anticolonial and anti-Chinese sentiment. DLespire severalearly attempts at liberalization, the policy regime became increasinglyinward oriented and interventionist. A pervasive and complex regime ofimport and investment licenses fostered a new group of Indonesian im-porters and traders who earned substantial rents and became a powerfullobby for trade restrictions. Following a centralization of political powerin 1958, President Sukarno expounded a populist platform of 'GuidedDemocracy and Guided Economy," and the economy entered a periodof more direct state control of production and trade. Dutch enterpriseswere nationalized, and stare enterprises took over all aspects of theeconomy-including the recendy emerged import monopolies.

Growing mismanagement resulted in chaotic economic conditions..Inlation accelerated to 1,000 percent by 1965. Exports and foreign ex-change reserves dwindled, and debt service exceeded foreign exdcangeearnings. Economic growch stagnated. By 1965 per capita income was15 percenr lower than in 1958. Widespread social unrest and violenceser the stage for a change in leadership.

OitwardOriented New Order Goverment (1967-731. The new governmentof President Suharto moved quiddy to restore macroeconomic stabilityand adopted a more ivorable stance toward domestic and foreign privateinvestment Some nationalized enterprises were returned t previous own-ers, and a new Foreign Invesament Law that provided a thirty-year guaran-ree of nonnationalization was enacted. The exchange rate was adjustedthrough large devaluations. In 1970, the governent established the mcli-anism for a market-driven unified exchange rate and abolished controls oncapital movements. Sweeping changes were introduced in the trade and in-centive regimes, ildg the abolition ofiport licensingthough litdechange was made to tariffi. As a result, non-oil exports increased at anannual rate of nearly 26 percent from 1967 to 1973, compared with an-nual increases of barely 2 percent during the previous seven years.

Oil and Commodt Boom (1974813. Indonesia benefited from the suddensurges in c-i prices in the 1970s, and the boom in primary commodity

x36

MACROECOO

prices. The surge in revenue gave the govenment an opportunity tointensifr development effort but it also posed the problem of how toprotect the competitveness of the non-oil economy from the adverseconsequences of oil-windfiil spending. While Indonesia was relativelysuccessful in moderating exchange rate appreciation, the appreciationthat did occur increased pressures to protect domestic industry fromimports.

Compared with many oil exporters, Indonesia used its enhanced re-sources well. Daring the 1970s, about 40 percent of the governmentbudget went to infrastructure for the economy. The government alsoplaced strong emphasis on improving the availability of education,health services, and family planning. As a result of these physical and so-cial investments, Indonesia was able to make extremly rapid progressduring the 1970s in reducing poverty and improving social conditions.

Economic growth averaged dose to 8 percent through the 1 970s and

early 1980s, as a result ofstrong expansion of public and private invest-ment- Non-oil activities reinained buoyant, especialy agriculture andmanufacturing. The government maintained good macrocoonomicmanagerment and, following a debt crisis with the state oil firm Pertam-ma in the miid-1970s, a conservative foreign borrowing straregy, By1980, the current account of the balance of payments was in surplus,and debt service payments were less than 13 percent of exports

Despite these achievements, the trade and investment regime becameincreasingly state dominated and inward oriented. Fluzh with cash fromoil and commodities the goveme nt invested heavily in such capital-and resource-intensive sectrs as oil refining, liquid natural gas, chemi-caL, pulp and paper, fertilizer, cemenc, and steeL As the states role in thecconomy grew, the govemment tightened regulations on foreign anddomestic private investment Meanwhile, gradual appreciation of theexchange rate eroded the competitiveness of non-oil exports, and pres-sure from domestc interests caused the trade regime to become moreprotectionist and variable.

AdMjusuentt Extenal Shocks (1982-85. The high-cost nature of the in-dustrial scaor and die end of the oil boom sparked a debate and re-assessment of industrial strategy in the early 1980s. The govermnment

initiated a broadly based adjustment program designed to maintain bal-ance of payments and fiscal stability while reducing the economy's de-pendence on oil revenue. As with previous macroeconomic imbalances,the government responded with a combination of fiscal, monetary, and

'37

.rs>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, moreinward 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 theface 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 regimefavored 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 tradeand 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 1985Tramnston 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 alsoBrinmbc (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 onwardas a resurgence in domestic demand accompanied the continuedgrowth of non-oil exports. Plywood and textiles remained the largestmanufactured export items, but other manu&ctured exports such atsshoes, 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 domesticinvestmcnt rose by a similar proportion, to $30 billion. Perhaps morcimportant, 70 percent of the foreign invesrment approved was exportoriented, compared with 38 percent in 1986. Since 1986, GDP growthhas averaged 9.3 percent annually, the highest sustained growth ratesince independence.

Indonesia continues to encounter occasional difficulties. In 1990 and1991, the economy experienced an import surge and deceleradon of ex-port growth due to increased aggregate demand prompted by a prioreasing of monetary policies. This led to a widening current accountdeficit 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 ofexport-push policies can be expected to generate momentum for firTherexport-oriented reflrm. If this is the case, future swings toward protec-tionism, should they occur, are likely to be less pronounced and lessdamaging than those of the past.

Thailand's Shift fom Resource-Based Exports to Manufacturing

Likc Indonesia and Malaysia, Thailand historically exported primaryand agricultural products derived from its rich natural resources. Earlytrade policies emphasized taxation of the resource-based rent associatedwith the production of these exportable commodities. Trade was heavilycontrolled. 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 exportproducton (Jitsuchon 1991).

Natural-Reso.rce-Based Exports 11955-70). In 1955 the exchange rate wasunified, and the state marketing monopoly on rice exports was abol-ished. Although these policy reforms encouraged natural-resource-basedanrd agricultural exports, Thailand maintained substantial import pro-

'.39

ffl ]\ ,, ,, .RWCLE

tection for specific industries. As late as the mid-1960s, however, overalllevels of effective protection to industry were modest by developing-economy standards. This was especially notable in consumer goodsmanufacruring, where nominal tariffs were in the range of 25-30 per-cent ad valorem. Tariffs on machinery and intermediate inputs were inthe range of 15-20 percent.

FavoiingImportSubsfition (1971401. In the 1970s Thailamd, followingthe import-substitution strategies favored by many other developingeconomics, raised tarifli on consumer goods to a rangc of 30-55 per-cent. Capital and intermediate goods continued to enter at low dutyrates, leading to an increase in the effective protection to value added inimport-substituting industries and to dedines in the eff-ective protectionto agriculture and traditional exports (figure 3.4). The bias against agri-cultural and export production increased.

Textiles, pharmaceuticals, and automobile assembly were particularlyfavored. 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 upstreamsuppliers ofcomponents and parts. These nontariffbarriers to competingimports of components and parts raised producton costs in the assemblysector 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 ofmacroecononic polic, the second oil shock exposed weaknesses in theThai economy that were the result of the import-substitution policies ofthe 1970s. In 1981 Thailand's trade poliqc shifted explicitly in the di-rection of export promotion. Remaining export taxes were reduced, andas described above, the exchange rate was devalued and subsequentlymoved 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 theaverage rate was reduced. The tariff reductions were alrnost immediatelyoffser, 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 ofvalue added, higher thaL other EastAsian economies such as Korea (28percent) and Malaysia (23 percent).

During these early stages of the export-push period, the govemmentBoard of Investment (BOI) played an important role in promoting cx-

140

M ACROECO?rw1. .~~~~~~~~~~~~~~~~~~~~~~~~.

port growth. The BOI, which had been established in 1960 to supervisethe application of investment incentives, had long been a tool for im-plementing the import-substimtion strategy. One of its key roles was todistribute (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 auNDP-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) andshifting 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 laborintensity. Box 3.4 describes the impact of this shift on direct tariff in-vestment. Other incentives for exporters induded tax exemptions andrebates, 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 unnecessatyregtuations to expedite export shipments.6

Automatic and concessionary credit was another major element inthe export-push strategy. The Bank of Thailand had taditionally ex-tended refinancing flaciiities (RFFs), through the commercial banks, tokey economic sectors.7 These RFFs were mostly used to rediscount cx-

* - port bills of large exporters handling traditional producs As part of theexport incenuves package, the Bank of Thailand revised its rediscountirules to focus more explicidy on small, nontraditional exporters and onother productive activities.

Many of the export-push incentives were specifically designed to off-set remaining distortions from Thailand's import-substitution era. Taxrebates, 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 exportpromotion tookplace in the context ofa domestic market that remainedmodctey protected from the rigors of foreign competiton. Thailandhas recently begun a more broadly based liberalization of imports. In-1990 tarifs on capitl goodswere reduced, and in 1991 large reductionswerc 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 - ~~~~~~~~~~nstetsm 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 .13.3Traditional 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 plethoraof 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 timeand in any particular economy, the magnitude of the incentives variedacross activities. Some governments favored nontraditional exports,some favored particular types of manufacmured exportsh Even so, thereare significant commonalities among these ponlCis. Each contributed toone or more of the four elements of a successfl export pushi accss forexporters to imports at world prices; access for exporters to lon- andshort-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 ewho do not face similarly expensive imports. To combine import pro-

tection with export promotion, these ill effctcs must be mitigated. HPAEgovernments, 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, thehigher thelevel of irmport-substitution protecton, the greatcer the level

wof adminisrative cormiptenc needed to cseate countrvailing imcentvesfor 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-orientedeconomies, 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 difficultyexpoers Fce in cracking fbreign markers but again chose various meansto encourage exporters to overcome these hurdles. Some directly subsi-dized export activity (direct income tax incentives), some subsidizedmarker penectration (through exporter associations), some subsidizedsmall and medium-size exporters to offier their difficulties in markerpenetration, and some promoted the creation of international tradingcompanies.

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. Theright strategy depends on the circumstances. It changes as the economychanges. 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 largecommitment to promoting heavy and chemical industries4 the Koreangovernment nevertheless changed course when circumscances changedand 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 requiredmajor revLsions.

In chapter 1 we showed the impact ofthe HAEs' export-push sategyon the rapid growth of manufactured exports. Another outcome of theexport push can be seen in their changing composition of exports (seefigure 3.5). With the exception of Indonesia (whose exporcs of refinedpetroleum products are induded under chemicals) all of the HPAEs dra-matically increased the share of exports in machinery and equipment atthe 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 otherdeveloping economies. This shift clearly shows the response of exportersto changing market opportunities, rising labor force skills, increasingtechnoogical capabiliy, and changing comparative costs.

What is surprising, however, is the extent to which textiles and apparel remained important exports in the face of shifting comparative ad-vantage. As late as 1991 most of the developing HPAEs had shares oftextdles in manufacured exports which cxceeded the average for otherdeveloping economies, suggesting that these sectors retained substantialcost competitiveness despite rapid capical deepening and rising realwages (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 otherdeveloping economies. They achieved macroeconomic stability by ad-hering to macroeconomic fundamentals-particularly by keeping fiscaldeficits 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 microeconomicincentives. In contrast to the many less successfiul economies that haveclung to filled policies, the HPAEs have pragmatically chosen and as-sessed policies according to their impact on macroeconomic stabilityand export growth. They have adapted policies as necessary, kept thosethat worked, and dropped those that failed or outlived their usefilness

'Why did the East Asian economies choose macroeconomic stabilityand exporrs as their policy yardsticks? And why have they been moresuccsfil than other economies in selecting, implementng, and alter-ing policies as necessary to attain these goals? We mrn to these questionsof political economy in the next chapter.

3-47

R'A CLE

Appendix 3.1 Economic and Ploflitcal rimermesIndonesia

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&,wontz.kdqaSns,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.7. Gicderiy "~"_...

148

MACRO

4 it

i i- - - Rqmn.

'S~~~~~~~~~~~~~~F

.h 1.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 GuarimeuREr;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?Lntaeflundd 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 - lctorFmna , Dndh p mn Praadnfpn)- aunAvl&mmmac=:-cdtdldnrc RelasionsAct

kbca~hsrt' Ddpsasll emda.A toa rsc apusrake f Epl'e q byp(uuimInanics .lard indusrs indusif Vzmz * A Cud adlcldorrr

ad~~~~~~~~~~~~~~ asnrtarn -~ ~ ~ ~ ~ ~ in cir

dvilnnm Z

152~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.

A 'Ip 1.1 ijr Ii I:

S.� 4

5.I

N

U

E. . .. . .

� .2

� � Liii rU Fgfl � j

iI� jS *j.* H* I

I:

* lB*

K

* i i.irr a* up . '� rI ¶ LI"U r I, '(UI

* - N - IIdII JJ'L[ illi-I�i

** Ia

C L E

Tainan, China

141gb ma aflntmt ~~~~~~~~Tbhir liur-Yac; Ema,noncaauvlapdcpnr 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

Notes1. 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 intained movement toward parity of incentives between ex- chapter 6.port and import substitutes, combined with institutionalsupporr for exporters. 6. Onc impormtn example of these efforts was the es-

tablishment of export-processing zones (Ens) and bonded2. 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 temporarydated public deficits, as well as the resr of the data in this admission of inputs for exporters in the zones. The zonessection except where otherwise indicated, are from the also offered lower-cost elecrricity, frcedom of immigrationsame source. Consolidated public deficits, though less for managers and technicians, and. common-facilitywidely available than central government deficits, are a bonded warehouses to reduce customs and storage costsmuch 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 macroeconomiccases. 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 GrotN EACH H-PAE EXCEPT JAPAN, NEW LEADERS PACED AN URGENT

need to establish. their plitical viability before the economictakeoff. The Republic of Korea was threatened by invasion fr-omthe North; Taiwan, China, fr-om China; and Thailand, fromViet Nam andt Cambodia. In Indonesia, Malaysia, Singapore,and Thailand, leaders faced formidable commnunist threats. In

* addition, leaders in Indonesia, Korea, and Taiwan, China, having takenpower, neededt to prove their ability to govern. Leaders in Malaysia andSingapore had to contend with ethnic diversity and attendant questionsof political representation. Even in Japan, where the competition wasless immediate, leaders had to earn public confidence after the debacleof 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 ofsbazredgrowth. The leaders hoped chat rapid, widely shared improvements ineconomic 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 beshared. 'W1hat distinguished the HPAEe' leadershp was thec extent towhich, they. adoptedt specific institutional mechanismns tailored to thesegoals, and that workced. In this chapter we briefly review the history ofthe adoption of an explicit commitmnent to shared growth. We then dis-

* cuss the main institutional mechanisms adopted to achieve this goal andhow 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 politicalpressures

o InsEirutions and mechanisms to share information and win thesupport 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 responseto racial riots in 1969, three dominant parties, each representing

a major ethnic group, widened their coalition to indude other partiesunder a national firont. The coalition's goal was to share wealth more eq-uitaMly without stifling growh The coalition's solution was the NewEconomic 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 nationalunnity through the two pronged strategy oF. (i) eradicating poverty byraising the income levels and increasing employrnent opportunitiesfor all Malaysians, irrespective of race and (ii) accelerating the processof restrucruring Malaysian society to correct economic imbalance, soas to reduce and eventually eliminate the identification of race witheconomic function.... The efforts to attain these objectives were, in

-turn, to be undertaken in the context of rapid structural change andexpansion of the economy so as to ensure that no particular group ex-periences any loss or feels any sense of deprivation. [Salleh, Yeah, andMeyanathan 1993]

The principle of shared growth is also explicit in Indo- .esia, wherePresident Suharto advocates "economic democracr. For example, in1990 President Suharto himself strongly endorsed research results ofthe Indonesian Economists Association stressing the importance ofsharing 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 ofthis vital issue. This will enable us to reach national consensus as westrive forward toward the full implementation of Economic Democ-

: 58 ::- -:

A N..:I NsTn'

racy" (isEa 1990). Economic democracy advocates reliance on the freemarket to stimulate growth. But it also explicitly acknowledges themarket system's inadequacies in distribution and actively seeks to re-dress them.

In Taiwan, Cbina, commitment to shared growth grew out of theTaiwanese authorities' searching analysis of their failure in mainlandChina The leaders' five-point diagnosis included three equity issues:agricultural tenants had rebelled against exploitation while the partycontinued to identify with the landlords; labor unions had run out ofcontrol; and government was beholden to vested interests (Wade 1990).These realizations, combined with pressure from the United States toundertake social and economic restructuring, led the Kuomintang toadopt 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, thesequence of policies it pursued, and the flattening of wealth resultingfiom the Korean War effectively produced egalitarian growth. Pardy byaccident and pardy by trial and error, the leadership gradually adoptedthe principle of shared growth.

In both postwar Japan and post-independence Singapore, growthwith equity was the leaders' primary concern. In Japan, the war's de-struction of landed wealth actually helped the leadership to implementits redistributive goals. In Singapore lack of a significant rural sector andthe 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 beenunusually successfud in achieving shared growrh. Growth-rate andincome-distribution performance measures show that the HPAEs signifi-cantly outperformed other low- and middle-income economies. Howdid they do this? If adopting the principles were sufficient, certainlyachievement wouid be widespread. The answer must partially lie in theinsttutions and mechanisms and in how they worked.

To win the support of non-elites, the leaders of the I4PAEs introducedmechanisms that drastically increased oppormtnitues to share the benefitsof growth. These mechanisms varied from economy to economy but in-cluded education (in all the HPAEs); land reform (in Japan, Korea, andTaiwan, China); support for small and medium-size industries (HongKong, Japan, Korea, and Taiwan, China); and government provision of

'159

such basic amenities as housing and public health services (Hong Kongand Singapore). Nearly all HPAE governments walked a delicate line re-garding labor by limiting the power of unions and intervening to checklabor 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 grantingdirect income transfes or subsidizing specfic commodities (for exam-pIe, food or fuel), HPAE leaders have favored mechanisms that increaseopportunities for upward mobiity. The frequent result is that indivlidu-als and families, provided the opportnity and convinced that effortswill 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 contributedsubstantially to opportunities for upward mobility. This mobiity in turnmirigated the feeling of non-elites that society is unjust and made themmore 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 wanIn rhe mid-I 960s, Park Chung Hee's export-promotion drive boosteddemand for educated labor, providJing jobs for the first crop of graduatesto benefit from the postwar education push.

Equitable Land Holding and Land Refonn

Theory and empirical evidence suggest that widespread ownership ofland not only improves equity but also improves land prodlccrMry(Berry and Cline 1979). All the lPA.Es with subsEantial agrarian sectorshave 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 havedominated agriculture since the colonial era. But with a relatively smalldpopulaton and ample land, it has avoided equity problems common-toother 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 simnilarcircumstances. 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. InTaiwan, China, rhe government seized land from the landlords, com-pensating them with shares in state enterprises. It then sold d-ie land tothe tillers at fitvorable credit terms and fivorable prices. The governmentthen helpcd the tillers upgrade production for domestic and export mar-kets. The program worked, economically and politically. Land reformhelped Taiwan, China, achieve one of the world's most equitable incomedistributions (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 stateenterprises, 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 byU.S. forces in 1947, distributed the land confiscated from the Japaneseat the end of Wurid War II to the tillers and put a ceiling on rents ofother land. The second, begun in 1950 and completed after the KoreanWar, was undertaken by the Korean government after a lengthy debatein the legislatum The government took over landlord properties, paidthe latter nominal compensation, and distributed the land to 900,000renants, 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 marketforces rather than government intervention. But several of theseeconomies supported SMEs with preferential credits and specific supportservices. Rapid growth of labor-intensive manufacturing in these firmsabsorbed 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 theexport sector, producing about 60 percent of the total value of exports

r16i

- g,sji~~srA 1 N 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- - -. PercentageofCue.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. 90.32Constructon -. , 28,419 26,456 93.09.Commerces: 489,864 - 475,106 .96.99Tmansporr,-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 towarddcveloping small and medium-size enterprises. Public financial insti-

rutions have allocated an average of 10 percent of lending towardSMES. 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 provenparticularly helpful during times of transition and rapid change (Itohand Urama 1993). Although there are mixed evaluations of the successof these programs, one thing is dear. The SME sector has become animportant cornerstone of Japan's economy. In 1989, SMEs accountedfor 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 of100 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, theSME sector began to grow rapidly.2 SMES' share in total manufacturingemployment 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 rosefrom 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 guareeprograms. 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 higbunmemploymenrt subsequendy, the wide availability of low-cost housingfor 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 HongKong's population from 600,000 in 1945 to 2.4 million in 1950spawned 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 thegovemment launched a Thst- track public housing program that hassince expanded to include the construction of entre new towns outsidcthe city propet By 1987, more than 40 percent of the population livedin 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-incomepublic housing and self-sufficient new towns similar to those in HongKong. Today 80 percent of the population lives in public housing, andmore 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 havebenefits 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 extentMalaysia), 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 cUrox iiGwn~~Skj&tiEboy,

*~.~FL t ~L !A? ia.tj. 1 t -,1 new.urbanj ,gos aalHrnlaaf.ok unablc os afford, ...

.,,peratlvece; isarbcle uron, lu; d.:-oLo6.i.>rohow nri r schaSuetck&^le conca4 ctconianrsbdI'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 atMrE 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-basedunions. 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 tosystematic restructuring, were nonetheless routinely suppressed at thefirst sign of radicalism, primarily because of the governments' fear ofcommunism. In Hong Kong, conversely, union radicalism has beenheld in check by a steady supply of new labor from China and by the in-fluence of Beijing, which has encouraged moderation in the unionsunder its Hong Kong umbrella organization.

Governments in several Asian NIEs, particularly in Malaysia andSingapore, have modeled their systems of labor relations after Japan,where labor radicalism was quickly suppressed after World War It andlabor-management consultations were gradually institutionalized. InJapan, these consultations have led workers to expect that adjustmentsin their wages will correspond to trends in the national economy despitethe unions' lack of political weight. 'Labor has been compensated for itsdecreased political role through wage policies tied to increases in pro-ductivity" (Johnson 1982, p. 151).

Like the government in posrwar Japan, the government in Singaporesharply lirmited union autonomy soon after independence due to fears ofpolitical instability Lacking natural resources following the separationfrom 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 suggestnew directions for the union movement.

The NTUC, which is in effect a branch of the ruling People's ActionParry (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 NationalWages Council (Nwc). Each year, the NWC reviews wage and economictrends before advising the government on wage adjustment guidelines.Although recommendations are officially nonbinding, the govemmentusualy endorses them wholesale, and business and labor use them aswage 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 ofabout 12 percent. The rAP also reorganized the NTUc itself, transformingit from a negotiating agent to a provider of goods and services. The NTUChas since established a number of worker cooperatives, ranging from dheFairprice 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 KLW. ,>NtwOACoA't bf.lrodany, byOf&r( fridhbYii ari

a I I ..... t

f,ruung_. -r eforp1crncuorVrne a arry rJo 80.ebrrelxelw:pa6N-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 hrt 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 politicallypowerful unions with less formal arrangements. In Taiwan, China, forexample, 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 generatedrapid increases in the demand for labor. Consequendy, as figure 4.1 il-lustrates, from 1952 to 1979, inflation-adjusted basic wages increasedmore than sixfold in manufacturing and more than fourfold in agricul-ture. The total increase in compensation has actually been higher, sincebasic wages are supplemented twice annually with profit-sharingbonuses.

Insulating the Economic Technocracy

IMPLEMENTING SHARED GROWTH TO SOLVE THE PROBLEM OF

political legitimacy requires sharing but it also requires growth. Wenow 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 policiesin kceeping with politicaUly formulated national goals with a minimum oflobbying for special fivors fiom politicians and interest groups. Withouti, technocrats in the high-perforningAsian economies would have beenunable to introduce and sustain rational economic policies, and someviral wealth-sharing mechanisms would have been neutralized soon aftertheir 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 0089 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, AnnualtSt ~ & '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 colonialgovernment, which has been strongly committed to free marker forces,need ior depend on the political support ofprivate interests to maintainpower 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 areheavily insulated from outside pressures.

In Japan, insulation chacacterizes not only the economic technocracy

but nearly the entire bureaucracy. Japanese bureaucrats draft laws inconsultation with the policy committees of the ruling liberal Democra-tic Party and private sector representatives; the diet (parliament) merelyratifies what the bureaucracy has prepared. Because the bureaucracy hasindependent power, it can often ignore pressure from individual actorsin the private sector. Moreover, rhe bureaucracy can achieve policy goalsthrough 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 C0:00000~~~~~~~~~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 themfrom 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 draftinglaws. Though the.legislature must approve the laws, the bureaucracystudies, analyzes, and drafts the bills. It has considerable control over theagenda and can use this to minimize political pressure.

IThejapanese bureaucracy is further protecred fiom political pressure bythe Natonal Personnel Authority, an independent body that sets the bu-reaucray's pay scales and promotdon policies, administers civil serviceexams, and makes most appointments. Japan's prime minister names onlyhis ministers and, except in a few caes, one of the two vice-nunisters ineach 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 GeneralPark Chung Hee seized power and reorganized the govemment to pro-mote rapid development. He createa a powerful Economic PlanningBoard that has had broad budgetary authoriry. Acdministraive controlover the banking system has given the bureaucracy strong leverage overbig business, which relies on government-backed financing to fundrapid expansion.4 The partnership has generaly been uneven, with thebureaucracy dominating the relationship.

Whe few of the HPAEs have such thoroughly insulated bureaucraciesas Japan and Korea, each boasts at least a small, ideologically consistenttechnocratic core that answers directly to the top leaders and thereforehas some independence from the legislature and other sources of politi-cal pressure. Indonesia and Thailand, where bureaucratic traditionsmore 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. Thailandis unique in Southeast Asia for being the only economy in that regionnot to have been colonized, in fact one of the few in the.developingworld. One reason for this remarkable alchievement was the openness ofthe 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 controlover the kingdom's many regions. Historically, then, the bureaucracy inThailand has always had a powerful influence over public administra-tion and policymaking. But it was-not until 1957, when Gcncral Sarittook over the reins of government, chat bureaucratic reforms were in-troduced to address macroeconomic imbalances. Since then, keepinginflation low and efficientdy managing debt have been among the gov-ernment's key concerns.

These same issues have been extremely important to the Indonesiangovernment since President Suharto took power in 1967. The previousGuided Dem'ocracy regime of President Sukano had let goverrnentspending run out of control, leading to annual inflation rates of 1,000percenic and other serious economic problems induding a rapidly dete-riorating infratructure. Suharto's New Order govemment ascended topower 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 infrastructured development.

To achieve the rwin objectives of low inflation and prudent debtmanagement, both governments created mechanisms tO insulate their

-: .economic redcnocracies. In Thailand, the government's Budget Bureauhas tight control of the budget drafting process and has maintained astable exchange rate and low inflation. To draft the budget, the BudgetBureau consults with the National Economic and Social DevelopmentBoard 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 withoutincreasing 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 itto a parliamentary budget committee for evaluation. But since neitherthe 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 fiorroutine 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 spentservcing the foreign public debt (the current cap is 13 percent). Parlia-mentary rules offer a fiirther unusual guarantee against runaway deficitspending 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 figure4.2). 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. Thisdemonstrates how a small technocratc core, insulated from politics, canset a positive tone for an entire economy.

In Indonesia, too, technocrats keep a tight rein on the budget. UnderSuharto, 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, hefinance 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 financeministry has established a macroeconomic environment favorable togrowth. 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 thepotential 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 oforoduction 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 tendto be more willing to leave the policy process to the technocrats.

Reozganization of labor from industrywide unions into companyunions, 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 ofcollective action. Thus, in contrasr with wokers in many other developingeconomies, 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 laborthey, too, have had less incentive to press demands on the reclnocracy.

Of course a powerfil bureaucracy insulated fiom external pressurecan also be dangerous. Indeed, the description ofjapan's bureaucracy-an independent force that controls generous incentives and issues vagueguidelines-is nor unlike that of bureaucracies in ocher economies thatengage 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 resistpoliticization.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 theirdecisions. How have the HPAEs kept insulated bureaucracies honest andresponsive 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 thecompetence 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 theHwAEs, Hong Kong, Japan, Korea, Singapore, and Taiwan, China, havebeen successfiLl in building relatively competent and honest bureaucra-aes. The Southeast Asian NiEs, Indonesia, Malaysia, and Thailand, havegradually introduced measures to upgrade theirs, with Malaysia in theforefront 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 toincrease the appeal of a public service career, thereby heightening com-petition and improving the pool of a1 plicants. The overall principles of

these mechanisms, readily applicable to any sociecy are:

-'74

ANN I NSTUT

* Recrauitment and promotion must be merit based and highlycompetuve.

* Total compensation, induding pay, perks, and prestige, must becompetitive with the private sector.

* Those who makc it to the top should be amply rewardecd

Merit-Based Recnuibnent and Pmmotion. Recruitment in Japan revolvesaround highly competitive cmil service examinations administered eachyear 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 thebureaucracy is based on a combination of seniority and a host of perfor-mnance indicators that dif&r across the ministries. Because the numberof 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 supplementedthcse filkers by,,recruitingheavilyamongacademics with advanced degrees.(Box 4A describes the turnaronund in the Korean civil service in the1960s.) This has resulted in a prolifcration of research instiutes associatedwith ministries. Since die Economic Plamning Board established theKorea 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 inthe 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 acionguidelines 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 everythingelse, 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. Relativepay in Mvialaysia and Thailand is about the same as the average for otherlow- and middle-in&me econo-mies but is brill signEficantly higher than-in the Philippines, which is widely perceived to have one of the weakestbureaucracies in SoutheastAsia. In general, the more fa-vorably the totalpublic sector compensation package compares to compensation in theprivate 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 -copnisbu W-pine aflownrs hlp to narrowr th;e ; *iffcen- Tain aconof salary plus* bonuss in thlet pvteo seaorzan sahlary plus hatl towne in

~d. .

;1 tNrnsnorecdRBErmwt 1 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 sseconn Jeapan th@e Nationalt Per,.sonnel kAu*thoit annuallyS*,-,{F.4 survesthu dso o mpanis toderin :job pay rates$ (Kimo 1988 p.) InKoea 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 ;EnnyluasHPA&; ,. 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 7,9540LMa1-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 °

~~2 tmgu ; - - 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 notavailable 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 attactcompetent individuals to the ci-vil service.

Prestige is of course partly a function of culture; thus East Asiancultures that regard dvii servants highy have an advantage in buildingbureaucracies. 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 ofapplicants can produce high status for civil servants in any society. Pro-vided that the exams are sufficiently competitive, individuals who passthem will be relat-ively rare. Thus passin the examn or winning a com-

petitive divi service promotion becomes similar to being inducted intoa hail of fame; it signals the public that an indLividual is special. 'Whilethe value of prestige cannot normally be quantified, the monetary valueaccorded to some Thai royal decorations, described in box 4.5, suggeststhat 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 seriousmistake. Security of tenure translates into lower income variability,which in turn provides incentives for public- employees to accept a lowersalary. 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 bureaucracyis a well-defined, competitive career path with a substantial prize for thosewho make itto the top. In Japan and some other E*AEs, retrement comes

early, and the rewards to a successfUl bureaucrat are substantial, extendingbeyond the pay; perks, and prestige to indude a golden parchute. The

Japanese have a special word for it- amakuckr4 which means "descetfrom Heaven"' and connotes a lucrative job in a public or private corpora-.tion or occasionally election to political office. Similar practices exis inKorea, Malaysia, Singapore, and Taiwan, China, in differing degrees.

Unlik the much-criticized revolving-door, which whisks bureaucratsin other economides fiom the pubfic sector to the private sector and backagain, 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.fi &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 gzovenmersternnoaafti 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>><is* "

M i %ehanimsthtidc cospoDBSimpeteny ciamonalso Ienace hownestyCompetiimrlon dscoraesdhoetsice.sgdistar<acsuovry o a corrupt or dis-^hoet c wull4ould diquaif nyiapplicant,. in.rsocietiEes havlue publicr

179

i;RA ACLE

service, dismissal brings disgrace on the entire family; hence familymembers have an incentive to discourage corrupt activities. This mech-anism appears to be especially cffective in Korea, where tradition hasplaced a high premium on goveniment service.

Strictly enforced dismissal policies also help reduce graft. In Singa-pore, large or small corruption is pcnalizcd by expulsion from the civilservice and withdrawal of one's pension. In addition, distissed publicemployees are unwelcome in the private sector. Under such circum-stances, the cost of engaging in comrpt acts becomes nearly prohibitive.Other economies in the region have government anticorruption squadsthat focus on enforcement and ptnishment

As Japan and Singapore in the early 1 990s illustrate, rich econo-mies have an advantage in building honest and competent bureaucra-des that further the creation of wealth. For example, economies withhigler per capita GDP can afford to pay bureaucrats better and tofund the high-qualinr education that results in a better pool of appli-cants. The challenge for developing economies is to concentrateavailabie resources where the can most effectively replicate theadvantages of the industrial economies. This is nor to say that cor-ruption and bribery do not occur in ithose HPAEs with reputable bu-reaucracies. These social evils exist in Japan, Korea, and Taiwan,China, and even in Singapore. The critical point is that they haveoccurred less frequently and in smaller magnitudes than in other low-and middle-income economies. Indeed, most of dte HPAEs scored sig-nificantly better than other newly industrializing economies in surveyrankings of the prevalence of improper practices and special interestgroup lobbying contained in the 1992 World Competitiveness Report(World Economic Forum 1992).

Developing Better Business-Govemment Relations

Whereas most other low- and middle-income economies have onlyrecently turned to the private sector to stimulate growth, the HPAEs havedone so for many years1 often decades. They have done this by provid-ing a legal and regulatory enviro=nent for private invesment, by creat-ing support systems for SmEs, and by establishing cooperative relationswith big business and in some cases labor.

Mmlh Legal and ReguLaztoy Envirounet. Compared with other developingeconomies, the HPAMEs have been more successful in creating a legal and

-80

- t;- - - ~~~~~ ~ ~~~~~~AN 4N.l

regulatory environment conducive to private sector development. Basedon selected criceria or fourteen newly industrializing economies in the1992 Worlz Competitiws Report, in which lower scores indicate a betterenvironment, the HPAES had an average overall score of 32.3 (with astandard deviation of 18.4) compared with an average of 56.9 (witha standard deviation of 13.3) for the odter economies.8

EncouragingSrall and Medinm-Size Enterpriss Small and medium-size en-texprises generally have difficulty geridng credit to finance capital improve--ments, obtain better technology, and break into markets, particularlyexport markets. Hence, a nontrivial SME sector cannot be established with-out external assistance. Studies indicate that the northern-tier HPAEs-Japan, Korea, and Taiwan, China-lhcv been able to develop relaivelyeffecdve financial supporc programs (Itoh and Urata 1993; Kim and Nu-gmt 1993; Wade 1990). largdy because of weak institutional capabiities,Indonesia has not had much success with itS support progams(Berry and

Levy 1993); however, the private banking system appears to have beenhdpf3i irn developing some sNEs. Both Malaysia and Th h V atiVsupport prograams but not much is known about their effectiveness.

Wooing Big Business

T HE HIGH-PERFORMING ASIAN ECONOMIES TEND TO HAVE

fbrmal instiutions that fciitate communication and coopera-don between die private and public sectors, whereby rent-

sharing rules can be made transparent and whereby each participant canbe assured of a share of rents. These art, in effect, an institutionalizedform of wealth sharing aimed primarily at winning the support and co-operation of business elites.

East Asia's Deliberation Councils

Japans efforts to establish such institutions are the most widely rec-ognized and have been the most thorough. Since the beginning of thepostwar period, the Japanese government and private industry have en--

gaged in serious policy deliberations though deliberation councilsoftwo

ype The first is organized along functional or thematic lines, such aspollution or finance; the second is organized according to industry, fr

: 8ii

exaple-, automobiles or chemicals. Government ministries establishthe councils, which are generally associated with. a specific buLreau

within the establishing ministryEach councl provides a forum for government official and represten-

tatives from the private sector-business, labor, consumers, academia,and the press-to discuss policy and marker trends and gencrally to ac-

change information. Consensus is encouraged. Figure 4.4 shows theprocess of formulating a policy ac the Ministry of Trade and Industry.MITI officials first invite interested parties to a series of hearings, the re-sults of which are forwarded ro a Mm research group. The researchgroup prepares a report for the approprate deliberative council, wherepolicy negotiations may take place. On the basis of feedback from the

council, the group revises the report and releases it to the public to ac-

plain the objectives of the new policy. Final approval by the diet is usu-ally routinc. Without consensus from the relevant deliberative council,however, a MITm policy stands little chance of success.9

Figre 4A Indusial Wsion Fonnuation Process

Conductsurver r| MMjuniorstaff Hearingcompile data study group - Learned individuals

- Interested parties- Overseas erployees

- Local rpresentadvesPrepare draft - oUers

> 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 andprivate sector relations were also dlose and cooperative, although somein the p'rivate sector argue that goverment was coo strong handed anddictatorial.10 Goverment and business leaders met oftn and regularlyalthough 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 channelswere monthly rexport-promoton meetings, at which the president ofKorea himself presided over discussions berween the economic ministersandictop 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-ofsion groups involving company managersi middbe-level government of-

-ficials, and experts or scholars. As in Japan, there were basicaluy twotypes: fumnctional (for example, taxation) and sectoral (usually focusedon a particular industry). In both, bureaucrats invited incerested partesand 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, irlationsbetween goverinmet and business have become more distant and themeetings 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'o1985, as the goverment responded to the racial drots-of 1969 wich theN 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 1986business-government relations have gradualy imprOved, culminating in1991 ind he New Development Policy- (ND), which has struck a moreNe cn balance between growth and distributon. For example, unerc venalac beCe rwhaddsrbnn o xml, under theNEP, 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-businessconsultative mechanisms, a determinat-ion reflected in Primec MinisterMahathir 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 publicizedbudget 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. In1983; the govenmment established the National Joint Pubfic and PrivateConsulkative Committe (NJPPCC) with the prime munister as chairmanand 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 theThai BankersAssociation. Provincial consultative comrnitees have beenestablished 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 privatesector paricipates in policymaking in many ways. For example, privatecitizens serve as directors on government statutory boards and as mem-bers ofad hoc government advisoy commirtees. In both capacities, theyreview policies and programs and make recommendations for official

considemraion. In addirion, the government regularly invites chambersof commerce, trade associations, and professional societies to subniittheir 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, simultaneouslyfiurthering the govenment's guidance of business and of labor, as wel asbusiness-labor cooperation. Having rejected more typical denocratic ifrmsas 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 tiroughlarge public entrrise that provide basic inputs to manu £cturers andthrough enterprise owned by the Khomintang the dominant political

184

AN NST.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- ----- - ---

parry.-1 Indeed, public enterprises, broadly defined, account for a largershare 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-scalesensitive investments and a means to influence the generally smaller firmsthat oomprise the private sector. Over the years, the large role of the pub-lic sector has become less obvious, as some parry and govemment firmswere sold to cooperative private interests. Coordination is also facilitatedby middle- and lower-level officials of the Industrial DevelopmentBoard, who make routine, infornal visits to thousands of small andmedium-size enterprises within their jurisdictions.

Like Taiwan, China, Indonesia lacks the formal government-businesscoordinating links of other HPABs; this may be pardy explained by theweakness of its bureaucracy. Coordination appears to be handled by in-fbrmal networks linking senior officials with major enterprises.'2 Littleis known about these networks, but they appear to have produced a highdegree of cooperation in the economy. In contrast to the deliberationcouncils, 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 thesociety 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 ofdeliberation councils has been so importanr in winning the support ofbusiness elites, we now look in greater derail at the operation of thesecouncils in two very different economies, Japan and Malaysia.

Japan's Councils: e Foreunnes The concept of deliberation councilsin Japan is most closely associated with Mm. Because of m1Im's role indeveloping and implementing the economys industrial policy, thepublic-private consultative process was most dearly manifested in thecouncils attached to it.

Government-business collaboration has long been a hallmark of theJapanese economy. The zaibauu-Japan's large corporate holdinggroups-were dosely linked with the military-bureaucratic complex ofthe 1920s and 1930s and were a key element in building the economy'sindustrial and military base. At that time, the influence of big businessmatched and sometimes exceeded that of the bureaucracy.

The U.S. occupation forces, ideniliing the zabatiu as a key elementin 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 policyformulation-but also faciliraced the resurrection of industrial holdinggroups, organized this time around banks (the so-called kid'mu system).These new groups becamc the main. players in the government-initiatedconsultation procss.

The principal instrument of the bureauicracy in managing thisprocess has been administrative guidance, the use of threats and rewardsto induce privatc parties to follow or implement measures suggested bythe bureaucracy. Administrative guidance does not depend on any lawor regulation. Rather, it employs eidter threats of adverse consequencesshould a private party refuse to conform to a 'suggestion or recominen-dation" or, alternatively, rewards dhat can be granted should the partyconcede,I MITE is well kcnown for having employed this with great skill.

Until the mid-1960s, administrative guidance was a powerfil4 tool forMm bureaucrats. Mm had firm control over the allocation of foreign ex-change and over importation through the Foreign Exchange and TradeControl Law. But intermational pressure, combined with the need tomaintin 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 ControlLaw was dismanded and the Foreign Capital Iav liberalized. Of coursethis meant that MmT was losing key components of administraive guid-ance. By the late 1970s, guidance had taken a distinctiely differentform-mergers of industies spiced with financial support from thegovernment-and coercion declined.

Ealaysia's Councls: mnulalingJapan and Korea. Malaysia's experience withpublic-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 andland 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 "MalaysiaInc.,' 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 growthwith 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 withcutting red tape. Many had litde private sector participation; those thatdid, however, produced concrece results. Since then, more deliberationcouncils have been formed, dealing with broader and more complicatedissues (the budget and trade, for instance). One of the most ambitious,the Malaysian Business Council, which was formed in 1991, resemblesKoreas highly successful Export Promotion Council.

Although it is too early to know how effective these new councils willbe, thev would be hamstrung without an efficient and rcpurable civilservice 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 ofregulations domestically and abroad, synthesize the information into anaction plan, and communicate the plan back to the private sector. Self-interested behavior is more or less constrained by the repeated nature ofthe collaboration. This in part helps establish credibility-private sectorparticipants believe cheating and reneging are less likely. Politically,these councils serve as proto-democratic institutions, providing directchannels 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 concentareon market competiton and not worry about others tring to curry spe-cial &ivors from the government. In effecr, their share of industry profitsis determined more through competition than through rent-seeking. Inthis sense, a deliberation council can be viewed as a wealth-sharingmechanism.

: . 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 oftheir peoplc to stimulate and sustain growth. Heavy-handed treatmentmight subjugate the ruled, but it would also increase the perceived riskof exprpriation, driving away capital. Lack of capital would slowgrowth, in turn driving away skilled workers and entrepreneurs. Thus,even authoritarian leaders cannot rely on coercion alone Rather, allleaders need a strategy that will induce people to participate willingly ingrowth. Whether by design or accident, the leaders of the HPAEs have tovarious degrees devised mechanisms thar effectively deliver the promiseof shared growth.

Although policies to promote growth are generally well known, theytend to be either difficulr to initiate and/or to sustain over time. Growthmakes those at the top of the economic pyramid uncertain about the fiu-ntre of their wealth. It introduces changes that are likely to open avenuesfor fellow elites and new competitors to contest control of their e-nomic fiefdoms Elite groups therefore generally find it in their interestto try and maintain the status quo unless they are assured that they willbenefit from rapid growth. Furthermore, the initial stages of growthoften require that the general population make sacrifices. Hence, even ifgrowh 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 sharethe growth dividend. They must demonstrate to non-elites thar part ofthe dividend will indeed accrue to them. Governments must createacedible 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-allhelped 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 implementeffective policies. To varing degrees, and with a variety of mechanisms,the HPAEs have performed these tasks unusually welL In the next twochapters, we examine the policies:and instruments that HrAE govern-

ments used in their effort to rialize the promise of shared growth.

; I88

Notes1. 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) theductivity. 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. The2. 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 thesurvey 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 itsboth vice ministers, he is constrained by custom to ap- rankings for each of the criteria. The overall rank score forpoint 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 their310-80, 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 theUnited Kcingdom, where the range is 50-90. 10. Though dhe channels of communication are still

operational, the democratization of the polity has led to5. 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 reestablishgood 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 betweenassumed 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 largenot -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 midsizcensure 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: "Theamong 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 large7. 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 waterare 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 RapidAccumulation

CCUMULATION OF PRODUCrIVE ASSETS IS THE

foundation of economic growth. It is therefore noEsurprising that, as shown in chapter 1, the high-performing Asian economies (HPAEs) accumulatedboth physical and human capital much more rapidlyand 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 achievethese rapid rates of accumulation and, in particular, to what extentpolicies were responsible. This chapter examines the HPAEs' superiorperformance 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 effectivelyi ActivisE policies: selective interventions underraken in some HPAEs

to increase accumulation.

Rapid economic growth and the attendant changes in economicstructure not directly related to accumulation policies were importantfactors in increasing all dtree forms of accumuladion. For example, theHPAEs' rapid sLift from high birth rates to low birth rates increased the re-sources per child potcntially available for education at home and in rhedassroorn. At the same time, the rapid deceleration of population growthincreased 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 morewisely, 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 saversagainst default, all of which encourage savings and investment. TheHPAEs 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 Favor0 ble general condidions and getting policy fun-damenrals right, some HPABs have deliberately intervened in marker tosolve specific coordlination problems related to accumulation. These in-terventions aimecd at boosting savings and investment rates are the mostcontroversial of the HPAEs accumulation policies. Generally, interven-tions have been more comrmon and more successful in the three northernHPAEs-Japan, the Republic of Korea, and Taiwan, China-al of whichhave 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 tosome extent were chosen as a group on that basis), we also knoyw, evenbefore examining their policies, that they wer not inconsistent withrapid 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 rapidaccumulation. We do thtis, noting that our findings are based on antalyticand empirical judgmecnts rather tha-n derived from statistical models.

Explaining East Asia's Hi'ghHuman Capital Fonnation

XCEPTr FOR IN THAE AND, THE QUANTITY OF BASIC EDUCATIONprovided to boys and girls of school aehas been consistentlyiEtiigierin 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 thatlevel.

What accounts for this extraordinary performance? We focus first onthree 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 twopolicy variables: the overall budgemry commitment to education andthe distribution of the education budget. We seek to show that the allo-

* cation of public resources to primary and secondary education was themajor determining factor in East Asias successful educational strategies.

Rapid Economic Growth

Rapid economic growth is the first factor that accounts fbr East Asia'sextraorlinary quanrity of education. For example, during the decade1965-75, CDP growth averaged 6.7 percent in Malaysia and 4.1 percentin Argentina. This implies that over the decade a conscant share of GDP

allocated to education would have doubled the resources available foreducation in Malaysia, while in Argentina they would have increased by

- less than half Rapid growth also creares jobs, increases real wages, andraises the rate of return on labor force skills, thereby increasing the deomand for education (Ing, Anderson, and Wang 1993). Thus, rapidgrowth operates on both the supply side of the market for educationalservices, 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 adampening effect on educational opportunities, if the relative wages ofteachers rise. Fortunately, rapid accumulation of hurnan capital facilitatessubsequent educational expansion by increasing the potential supply ofeachers, thereby reducing the premium teachers command.' The wages

of educated workers, including teachers, tend to rise at a slower rate thanaverage wages. Therefore, the impact of rising wages on the cost of edu-cation and hence on enrollments is weakened.2 For example, per pupiloperatng costs at the primary level average roughly 13 percent of percapita 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. InSub-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 growthis a deline in the rate of growth of the school-age population. During the1980s, the growth rare of the population age 6-11 years was very low inEast Asia-so low that the absolute number of schoolchildren in Korea,Singapore, and Thailand actually declined-but was phenomenally highin 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 inKenya it rose from 47 to 51 percent. Similar contrast are evident betweenother HPAES and Sub-Saharan African economies (see table 5.1).

When the school-age population is growing rapidly, as in Kenya orPakistan, rising expendirures on basic education are needed just to keepenrollment rates constant. With declining or slowly growing school-agepopulatons, however, similar increases in expenditures can go for moreschooling or better quality. Alternatively, expenditures as a proportion ofGDP 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' . ...-' ...

1 Banglaesh .X' ' ' .. '-.'43 . .

* I94 -0 . ; Brazl t, 4 35 120 - ' "1-

... :,..:a' 'C: '.47 0 ''.,

T- 47s51_3 4igrn.~~~~~~ ~~. , 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 allof 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 proportionof the population of school age.5 Consider the most striking example: ifthe share of the school-age population in 1988 were as large in Korea asin Kenya, the Korean govemment would have had to spcnd 5.6 percent

Table 5.2 Percentage of CNP Allocated to Education SavedDue to Lower Fertility Rates

t f 9 ;\; _ i X~- dne torowthm;s#eExpendtre '.on ; rhuol-aepopnltdon #tbasic 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 achiieveKorea's high enrollment rates. This suggests that lower fcrtility rates savedKorea 2.8 percent of GNP. Conversely, had Pakistan's school-age popula-tion grown at the relatively low Korean rate, the govcrnment of Pakistancould 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 morethan eighty economies, there is a strong and statistically significant neg-ative correlation between basic education enrollment rates and the levelof 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 enrollmentratcs 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 inEast Asia.

Due to greater income equality, the income of the poorest 20 percentof the population in East Asian economies is higher than in economieswith the same average income but greater inequality (see table A5.4, atthe end of the chapter). Because of budgetary constraints (and capitalmarket imperfections that preclude their borrowing to invest in educa-non), very poor households are unable to invest in their children evenwhen the returns are high: the pressing need to subsist crowds out high-return investments. This is less likely to occur in East Asia because theincomes of the poor tend to be further above subsistence level.

Measures of the income elasticity of the enrollment raio derivedfiom cross-economy education expenditure equations are 0.31 and 0.43at the primary and secondary levels (Schultz 1988). This implies that inEast Asia the positive effect on school enrollments of lower inequalitycan 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 inBrazil as in Malaysia, enrollments among poor Brazilian children wouldbe 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 HinderHuman Capital Fomation

Numerous studies show that a person's carnings are higher the highertheir human capital, as measured by their education and 'iealth status. Asshown in chapter 1, the same is true for nadons; we know that invest-ments in education contribute significandty to economic growth. Butneither 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 aneconomywide 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 resultin fnilies investing less in their children's education than is in society'sinteresL 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 implythat families thar invest in education are not the sole beneficiaries.6

The difficulty of borrowing to send children to school affects thepoor 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'sschooling than would make sense even from a strictly private point ofview.7 These market faulres in principle suggesc making loans availableand improving information about future retuns. A simpler, and morecommon, alternative is for government to reduce the direct costs ofschooling 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 governmentshould 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 andinnovations spill over to others. Similarly, education may have socialbenefits not directly compensated monetarily, for example, a reductonin 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 andin 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 primaryPri_ 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. Vocational05 ID L training 2may also havc high social payoffs, if it improves worker pro-

X 0~~~' 5 10 15- 20 25 30 3!i3 SocIal return U Pilvato ductivity not only for the trainee but also for her co-workers. More im-

*; Social retrns N Nvaf mtportantly, 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 andincome distribution.

Table 5.3 Public Expend4tnre onEducation 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 bothoegg :~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 forT 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, 3.7PakfisraaV$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<developing econom ,ies subsequent increases in primary and seconday

wa1 -ourc es£daral enrollmnent rtates account for Koreas present wide lead in enrollmentsover other middle-income economies. A comparison of lndonesia, asuccess story, and Pakistan, a laggard, is also illustative. In 1987 In-

;' .- donesia had achieved universal primary cnrollmcnt and a 48 percent0 - ~ - D g s q Ni secondary enrollment rate. By contrast, Pakistan's enrolment rates were

52 percent at the primary level and 19 percent at the secondary level.-. Q: - t> 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 34percentage points, implying that inost of the current gap is explained bythe 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 thepast thit-seven years

Prmawy and Secoudary Education. The allocation of public expenditurebetween -basic and higher education is the major public policy factorthat accounts fbr East Asias extraordinary performance with regard tothe quantiry of basic education provided. The share of public expendi-ture on education allocated to basic education has been consistentlyhigher in EastAsia than clsewherm: Korea and Venezuela provide an ex-treme example that nicely illustrs the point. Table 5A indicares thatin 1985 Venezuela allocated 43 percent of its education budget tohigher education;, by conwzst, in the same year Korea allocated only 10percenc 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 fk ~~~~~~~~~* ; 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, 0fiKoeaRepo 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 Asiahas tended to be low, averaging roughly 15 percent during the past threedecades.9 In Latin America die share has been roughly 24 percent.'0 InSouth Asia, the share is dose to the Latin American leveL This had beenthe case in Sub-Saharan. Africa as well, but in recent years the share hasdedinedrto East Asian levels.

By giving priority to expanding the primary and secondary basesof dheeducatoa pyramid, East Asian governments have stimulated the de-mand for higher education, while relying to a large extent on the privatesector to satisfyr that demand. Eni all developing regions the probability ofgoing to university is miarkedly higher for secondary school graduatesfrom high- thian from low-income famirnls. Typically, in low- and middle-income economies government subsidies of university education are norrelated 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 inpoorly qualified children fr-om low-income backgrounds being forcedinto the private sector or entirely out of the education system. Becauseof the higher conce-ntration on basic education in East Asia, publicfinds for education are more likely to benefit children of low-incomefamnilies who might otherwise have difficulty remaining in schooL

Vocational Twining. Human resources and the trainingr to upgradethem have been important to the HI'Ans successful export drives, despitethe 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 theshop floor, which are responsible for a major share of productivity in.manufacruring, demand both higher- and lower-level skills. But whilevocational taining is widlreonzda imotant such traing israrely cost-efficient when provided in the school systems. Firms prefer todo their own trairnin pardy because many skills are firm-spedific..

-200

STRATEGIES FO.R. k-A�i

-WlUs N 31SZEdMud n :p quesm--l-gg

0

. ou 00

to .,H, citi0ii"bf- propoipomp

Y,

$61-5-iPC-= -r. n.ZOM ulca'don- ud et,' 00Pff pq*"

90r w=ntpmbamr- ucauonz,�-,F��-,,"j,,�.

C� uc2uoqdiiii2b �M-ffiki6

0,-percmrp -,-�PFF2=enta o

Bft a c;-,MC:�rcmamclcri,,-,,CUOn.

arsozaLs= R -'zcu -my,'one in�

C

O j'an

jop�PrO

7 o non,�ZhQ?

1987':ax .11;2pcm gr ;",O

"ea ):PM m 's, ihamliEfic,ar-kniv P C

LiC=O caiiNwduj���

&�-Co

zm

owcvcrLti 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 foundchat returns from training were higher in industries with well-educacedworkers 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 andonly 7 percent of technicians received it. As would be expected, trainingwas positively affected by the level of technology and negatively affictedby 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 reTurof 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 goneawry According to one study, Koreis 1974 Special Law for VocationalTraining, which required firms to provide six months' training in amproved schemes, discouraged firm-level training; firms considered theperiod too long and opted to pay a fine instead (Kim 1987; van Adams1989). More narrow governmentefforts to forecast occupationdemandor provide tnained workers to fill anticipated jobs have also generally notbeen successfil.

One exception is Singapore's use of training to promote the inforna-ton technology sector. Singapore has achieved world leadership ininformation-related services thrugh a concerted program thar involvededucational insintitons (specializng in business and engineering softwaretraining), training subsidies to schools and office workers, computeriza-

don of the dvil service, and establishment of TxadeNer, an interatonalinfornation network (Hon 1992). This success illustrates the importnceof 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 dearconclusion. 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 increasesthe number of skill-intensive jobs, thereby ensuring that new skills donor erode. Training, then, appears to function like many other aspects of

development in East Asia. It contributes to rapid, susained growth butdoes 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 werealso betten They responded more appropriately to coordination filuresin the market for education. Emphasis on universal, high-quality pri-

maryeducation had importantpayoffs both for economic efficiency andfor equiy. The excess demand for secondary and tertiarv education,generated by rapid ateainment of universal primary education, was metlargey by a combination of expansion ofa public secondary system withmentocratic entrance requirements and a self-financed private system.

This stands in stark contrast to many other low- and middle-incomeeconomies, 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 themaintenance 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 invokedcultural factors to explain high savings rates and used those high savingsrates to explain high rates of investment and growth. Neither element ofsuch explanations has received much empirical support. Indeed, studiesof the income-savings relationship in a broad cross section of economiesindicate that while income and savings growth are highly correlated, in-comes often have risen befbre savings rates rather than after, suggestingthat growth drives savings rather than the other way around. Recenteconometric studies support the idea that rapid income growth boostssavings rates as households acquire resources fister than they increaseconsumption (Carroll, Weil, and Summers 1993). According to thisview, 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 relationshipfor 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 ofgrowth. Results are mixed for Hong Kong and Malaysia, and causationmight run either way. In Singapore income changes were not a signifi-cant fctor in the speccacular rise in savings rates in the 1970s and1980s, consistentwith the view that demographic fiactors and, to a lesserextent, 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 mayhave cont-ributed to rising savings rates for many of the same reasonsthat rapid demographic transitions fostered human capital formation: asbirth rates fell, the dependency ratio-that is, the ratio of non-woring-age people to working-age people-decreased. Theories of savings basedon the life cycle of increasing then decreasing income relative to con-sumption predict that societies with a high proportion of prime-ageworkers will save more than those with higher proportions of young orold 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; Horioka1990).

For the HPAES other than Japan, there is litle evidence of the rela-tionship between demographic changes and household savings. TheSingapore study mentioned above showed that most of the increasedsavings rate fiom 19 to 46 percent betwi. :n 1970 to 1989 was attribut-able to dedining dependency ratios (Monetary Authority of Singapore1991). In Taiwan, China, where household savings rates are generallyvery high, households headed by prime-age workers tend to save more,possibly reflecting in part preparation for old age. (Bur savings in olderhouseholds are also remarkably high and stable, contrary to the view ofthe 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 ageat reduced wages appears to contribute significantly to higher savingsrates (Horioka 1990).

Policy Fundamentals and High Saings Rates

In addition to benefiting from higher growth and demographic shiftsconducive to savings, East Asian governments selected, ofien throughtial 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 budgetarychoices.

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 areoften negative and unpredictable, discouragng domestic financial sav-ings. Moreover, since unanticipared high inflation erxdes the ral valueoffinancial assets, volatility of inflation ,increases the riskassociatedwitthholding, 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 inflationgave 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 rareson 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'.41935-91 ~~~~~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 inLatin America, 14 in Sub-Saharan Africa, and 6 in South Asia. Malaysiaand Singapore have maintained the most consistendy positive and stablereal interest rates on deposits, paralleling interest race trends in the

--.. United States. Japan maintained negative, but highly stable, real interestraces on deposits until 1983 and has had very low btut stable positive realmiterest 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 betweenreal interest rates and the growth of savings deposits and broad moneyaggregates (Fry 1988; Lanyi and Saracoglu 1983; Gelb 1989). This isespecialy evident for changes from highly negative to positive real ratesof interest. For example, when Taiwan, China, raised real interest rateson 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 rosefrom 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 thesame.

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 publicsector 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 arefaced with liquidity constraints and are not able to act on expectaions.Thus total savings may rise with an increase in public savings (Summers1985). Empirical evidence shows that government savings does not fulycrowd out private savings. Furdtermore, the method of raising publicsavings 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 48-65cents (Corbo and Schmidt-Hebbcl 199 1).

Most HPAEs have maintained high public savings compared withother low- and middle-incomececonomies througi. -ombination of taxpolicy and expenditure restraint. The extent to waich HPAE govern-

ments have restrained current expenditures relative to other developingand industrial economies is shown in figure 5.2. which gives averagepublic consumption as a percentage of GDP between 1970 and 1988 fora sample of 111 economies. Indonesia, Japan, Korea, Singapore, andThailand are in the bottom third of the distribution of all economies interms of the share of public consumption, with public consumptionshares below 15 percent of GDP. Malaysia and Taiwan, China, are ecx-ceptions, faling in the middle of the distribution. In Malaysia the highlevels of public consumption reflected the explicit redistribucional ob-jectives of the New Economic Policy As a group, the avcrage share ofpublic consumption in the HPAEs is below both the OECD economiesand all other regional groupings of low- and middle-income economies.

In contrast to the situation in most developing economies, in theHPAEs, 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 from22.6 to 24 percent. Malaysia's public savings increased from an average3.2 percent of GDP in 1961-80 to 10.3 percent in 1981-90, while itsprivate savings increased slightly. Th2iland's public savings was high inthe 1980s (87-15 percent of GDP), but declined as a proportion of GDPin the mid-1980s, when the privare savings rate rose dramatically (seetable 5.6).

While in some other economies (for example, Pakistan, Zimbabwe,and Venezuela) public savings was relatively high and accounted for amajor part of national savings, generally public savings in low- andmiddle-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. Bothpublic and private savings dcdined in the mid-l 980s in Argentina andChile. Brazils public savings dropped from 8.3 percent of GDP in 1980to 3.2 percent for the period 1983-85.11 Some Sub-Saharan Africaneconomies (for example, Ghana) had negative public savings m the1980s, 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 40Average public contsumptlon (percentage of GDP)

.Soe _, --orl dB2nk6 -

209

!&4iIRfACLE

Table 5.6 Public and Private Savings of HPAEsand Other Selected Developing Economies(pereL'ntage of CGP)

u -ble ; saings V

i4o6J-apa, ,*

')0 z | g 95 5-70 ' . S 0 t ̂6.2 17.2--19714=0 4.6 '201 1

6 12 150 ; 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. 2350 :: : 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 ofthese 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, tomore 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 totru:st.an insurer than they do the bank: either or both could faiL In-deed, -a firm that insured several banks would risk nearly simultaneouslosses during a general economic downturn or-a bank run. The risk ofsuch losses m-akes aprivare insurers promise to deliver.even-more sus-pec than. the banks. Thus, withour government effors to guaranteethat banks will be. able to honor their liabilities, potential depositors

.would rend-tosave less and keep a higherproportion of theirsavings innonfinancial assets..

Governments have. a second-motivation for insuring deposits.- Fman-.ial.hlures. of banks have a strong negative- excernality; when somebank fail, depositors lose confidence even in. sound. financial institu-dons, as the history of bank runs attests- Govemrnments therefore ty topreventrovert bank 6ailures-, fiequendy acting explicitly or implicidy asthe-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-noefforrro attact ithe savings -of workers, poor people,_ andLresidents-ofsnulLi 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 suggestedthat individuals acting on their own and their flimilys behalf may saveless 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 decisionson others. If savings decisions could be coordinated, each individualwould save more and aggregate savings would rise (Dasgupra, Marglin,and Sen 1972). In such circumstances social returns on savings exceedprivate 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. Allgovernments take steps to alleviate the lack of deposit insurance marketsand to promore stble financial institutions. These generally indudeprudential 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 developingeconomies. East Asian governments have artached great inmportance tomaintaining saver' confidence in financial institutions. Often theirpolicies have gone beyond prudential rgulations to include protectngbanks fiom competition to increase the financial strength of banking in-stitutions.12 Sometimes this has been done at the expense of long-nmefficiency.

When necessary, all EPAR govenments have bailed out troubled fi-nancial institutions through financial and managenent assistance ormergers 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 economiesthat are expanding as rapidly as those in East Asia Nonetheless, thecrises that have occurred, for exanpie, in the wake of the 1 980s' real es-tate speculation boom in Hong Kongi Japan, Korea, Malaysia, andThailand, 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 manydeveloping economics. For example, one study of twelve Latin Ameri-can economies found that nonperforming loans accounted for nearly 14percent 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 debtsconstituted 14 percent of tota pordolios in 1989 (an internal WorldBank report). In Sub-Saharan Africa, many banks have failed, and25-40 percent of bank portfolios have had to be written off. Whilemacroeconomic instability and low growth are partly responsible for thispoor 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 internationalcapital adequacy requirements set by the Bank for International Settle-ments (Bis) to ensure that banks do not take on inappropriate levels ofrisks. Hong Kong, Japan, and Singapore began to stengthen prudentialreguation 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'4 rements. Indonesia, the last to do so, has had some diffi-culty achieving BIS capital requirement levels because it introducedproper prudential regulations rather late.13

By conast, in Lain America only Ecuador adhered to BEIS standar4iwile Colombia was in the process of adoptng the standards (Morris andothers 1990). Moreover, some Latin American economies have notestablished minimum capital requirements for new banks. In Sub-SaharanAfrica, the net worth positions ofbanldngstyerms siarply deteriorated in the1980s, and adopting BIS standards wold entail subsantial rcapialization.

Most East Asian economies continue to rcly on dose contact betweensupervisors and banks to encourage prudence. This routine, often dailyinteracton enables regulators personally to assess the riskiness of a bankAportfblio. 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 companiesare dosely heldl and dlisdosure rules are geCnerally weak. Moreoverinm sev-erl economies, notably Indonesia, Japan, and Thaiad, banks and firmsrend to have inrerlikng ownership.

Many East Asian financial systems appear to be moving to aramixed system of enfbrce-menr, in which the highly structured exposurerules that characterize international supervision are backed by tradi-tional moral suasion in which bank regulators use their control ofbranch licensing, rediscounts, and other regulations as leverage to ensurecooperation. The Bank of Japan, for example, has used constant inter-action between city bankcs and the supervisory authorities, coupled withits 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 goodpmudential regulation, there have been financial crises in Eas Asiabrought 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 beenable to maintain confidence by acting as implicit insurer without exces-sive fiscal costs. Rapid growth and spiraling real estate prices have oftentempted 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 ofthese economies were facing diffliclties, as nonperformning loans rosedue -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 inthe 1980s, many firms had large excess capacity and poor -debt servicecapability. Among them were gvrmecnt priority projects in shipping;overseas construction, texties, and machinery. Because the firms werehighly leverged, they were financially vulnerable to changes in demandtand interest rates.

'When fnnildistres has occurred, HPAE governments have cometo the rescue_ Table A5.5 describes the origins of and responses to someof 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 insureddeposits by stepping in whenever necessary to prevent bank failures.

In some cases, especially in Japan, stronger banks were pressed to takeover banks in financial trouble. When mergers have been impractical,East Asian governments have served as lender of last resort. Only HongKong and Thailand have liquidated insolvent financial institutions. Yetiecause of rougher prudential supervision and better econornic growth,the costs of bank rescues in East Asia have been lower than in manyother economies. During the 1980s the estimated fiscal cost of bailoutsin the HPAEs was 0.5-1 percent of GDP, compared with 2.3 percent inthe 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 financialinstitutions to try to ensure thar newcomers are solvent and will remainso. But many East Asian governments have gone fiirrher, limiting com-petition and protecting existing banks both fiom foreign banks andfrom new domestic competitors. With the notable exceptions of HongKong, 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, rhebanking sectors in the HPAEs are no more concentrated than in otier de-veloping economies or OECD economies (see table A5.6, at the end ofthe chapter). This means that banking sysrems differ litde in structureamong 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 thannet 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 theyforgo the additional profits from asset growth. Net worth requirementsdo not provide goveriments this avenue of influence.

Have these restrictions on compention contributed to increased fi-nancial savings? The answer is ambiguous. Larger, more profitablebanks 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 thebanking system (provided banks do not distribute profits to share-holders). These effcts would raise saver confidence and hence increasesavings. Yet, in most economnies regulators have argued that the financialsectoes tendency toward concentration is so great that govermnmentsmust intervene to maintain competition within the banldng sector, pre-venting monopolistic practices that squeeze dlients, reduce efficiencyand innovation, and hence, lower savings.'

Therefore, whethe-r controlling competition results in &avorablc orunfaivorable 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 (Vitrs1991). But in many developening econoes financial markets are domi-natal by a few large, veq inefficient banks whose size is bad less oneconomies of scale thn on restrictions of bank licenses and othecr limitson 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 others1982). Thee is no dominant commercial bank in Japany The top fiveban are roughly equal mn size, and they accunt for about 40 percentof total commercial bank assets (see table A5.6). In comparison to largeGerman banks, Japanese banks tend to have lower lending margins(spreads), although smal German banks have lower spreads than smallJapanese banks. The Japanese banks also have lower lending marginstha equivalent bankcs in the United States but approximately equalrates 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 SriLanka, and Turkeys and selected economies in Latin America (Hansonand 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, indicaungthat controls on intermediation magins were generally effcidve. Figure5.3 shows that avage terest spread of dte iHPAEs (about 3 percent) islower than aveg spreads in other developing regions, which range from4.0 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). InKorea and Malaysia they were generally igher at about 4.5 percent (seetableA5.7, at the end ofthe chapter). Thus, while restricingcompetitiongrenerated 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 AfricLS:ournranamdonal MoncrccyFund dan.

2;7

..- A GL EL

stricdions made banks in East Asia rlaively efficient compared widt thosein other low- and middle-income economies.

Regulated spreads are not always binding, however, because bankscan circumvent them by charging fees or requiring compensating bal-ances. Banks in Japan and Korea, for example, charged compensatingbalances. There is some evidence that these practices were sufficient tocircumvent 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 ondcposit 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, andinterest rate spreads increased slighdy after the mid-1980s due to a dropin 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 encouragedsavings by ensuring stable banks, Japan, Korea, Malaysia, Singapore,and Taiwan, China, established government-run postal savings systemsto attract small savers. Postal savings systems offrecd small savers greatersecurity and lower action costs than the private sector and weretherefore particularly effectivc in attracting to the formal financial sectorthe savings of low-income and mral households

Japan established the region's first postal savin progam in 1875, withthe explicit goal of fostering savings of rual dwelers and people with lowto moderate incomes in the cities and towns. Undl then, such people wereeffectively exduded from the financial ssten, which lacked rural networksand discouraged small depositors by requiring high minimum balances orpaying very low interest races on small deposits. The Japanese governmentheavily promoted postal saving among low-income households and madethe 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. LikeJapan, both Korea and Taiwan, China, have granted tx- exempt statusto 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 businesshours than other financial institutions (Shea, fbrchconiing). In Malaysiaand Singapore, where postal savings have also accounted for a large pro-

,.. - - :.... .:

STRATEGIES FOR. RAPD

portion of domestic deposits, the governments separated managementof the savings system from the post office in the early 1970s, when theproportion of postal savings in total savings dedined, cvidendy becausepostal 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 postalsavings systems piggyback on the mail-delivery infrastrucumre, thus min-ining overhead and fixed costs. In the early years in Japan, the systemoften 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 beenmuch less than that of the private banks (Mukai 198 1).

Fored Savins Besides the above measures, some HPAEs have tried tocompel savings through several measures, including mandatory pensionschemes and restictons on consumption and borrowing for consump-tion. Because these measures consttute more active interventions inmarkers, their efficacy, even in East Asia, is open to question. Restric-tons on consumer choices, induding the basic consumption-savingsdecision, have welfare costs. Moreover, similar efforts in othereconomies have been spectacular failures. Examples indude the Wide-spread deprivation and massive waste associated with forced savings inthe now-defunct command economies.

Three economies in East Asia, Japan, Malaysia and Singapore, havewell-developed, mandatory pension plans The impact of these plans onaggregate savings depends on the degree to which they subsdtute forvol-untary savings. Evidence of the impact of pension funds in Japan andSingapore, the only two economies where the issue has been studied, isincondusive. One study of savings in Japan found pension funds had nosignificant impact on tocal savings, although they did channel householdsavings 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 ofvolunary savings Conversely, there is evidence that Singapores CenalProvident 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 savingsin 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 Japanafter World War II was driven by the need to acquire consumer dumables,and he shows the same pattern existed in Korea and Taiwan, China. Ashousehold incomes and the demand for consumcr dumbles inacased, sav-ings as a proportion of income rosc rapidly. Once the excess demand forconsumer 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 formost of East Asias superior savings performance. Maoeconomic sta-bility combined with solid prudenrial regulation of financial institutionscreated an environment in which savings were secure and retained theirvalue. Limits to competition among banks and the willinpess of gov-enments to rescue financial institutions in trouble had costs-reducedefficiency in the former case and fiscal costs in the latter-but these weremoderate and may have been compensated for by the increase in savingsengendered by public confidence in banks. Postal savings had the virtueof being both secure and accessible instruments for small savers. Gov-ernments in Japan, Korea, and Taiwan, China, as wel as Malaysia andSingapore, 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 theywouldhave voluntarily increased economic wvelfare? The answer is ambiguous.Restictions on consumer credit raised savings rates tempoarily in Japanbut at a cost in terms of consumer wdfar Forced savings through pen-sion schemes have not been a major ficaor in increasing saving rates inEassia, 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 increadsavings must exceed the privatc oppomruity cos detenined by the rateof return on investment. This would be the case, for example, if total fic-tor productivity growth were positive and a significant component ofoverall 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 beenpositive and rates of return to capital have been good. In Singapore, how-ever, evidence suggests that aggregae total fctor productivity growth isnegative and rates of return to capital are low (Young 1992). Singaporemay 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 viruouscirde 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 tonvesntmen. 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 orpartial expropriation of their investments, private investment initative

is naturally reduced. Thus secure property rights embodied in the legalframework were a key to high investment rates. In addition, formal andinformal 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 theconditions 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 economicgrowth. These investments havc bccn highly complenenary to privateinvestment, especially in export-oriented manu&cauring activities.

Table 5.7 shows some indication of infrastructural development forthe 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 ofeconomies at similar levels of per capit incomc. Similarly, from 1975 to1985 the five developing HPAEs with nontrivial mral sectors increasedtelephone service much faster than other economies with roughly thesame 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 odaivrroads 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 '205coe.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: '- 1Gn . ..528: 79.7

India L 34W ~~48.0- 209 89.5 '352, -20

muridu 0 935M'-'co ,60.0 3 21 KU W---hilippine. - - - :-- - ..12 5 -1 .7

. Si tnlt 66 1P7o-mljFL 7~~~~~~~~~~~~~~~!

=2.2

STRATEGIES FORRK R.A%W

tius, and Peru had smaller gains. Al though increases in infiastructureare clearly correlated with income growth, recent econometric evidencesuggests 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, mostHPAE 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 alleconomies, but the problems and the consequences tend to be worse indeveloping economies. For example, such developing-economy informa-don problems as poor accounting standaris and a dearth of merchantbanks and other institutions to monitor corporate performance meanthat bond and equity markets are often weak or entirely absen.' 7 Simi-larly, while market mechanisms for sharing risk are limited in alleconomies-firms in developing and industial economies often cannotbuy insurance for the most serious risks they face-the consequences aremore severe in developing economies in which lower wealth levels makehouseholds and firms more vulnerable. Furthermore, the transforma-dons inherent in development, such as the creation ofnew firms and newindustries and the absorption of new technologies, heighten uncertainty.

Polcies to Pmoute Inmeshment HPAE responses to these problems can begrouped into three broad categories FLrst, as these economies have ma-tured, policynmakers have gradually increased their atrteion to the creationand improvenent of bond and equity markets. Hownrer, while these mar-kets are playing an increasingly important roe they were not generay akey factor in mobizing investment during the HPAEs economic mkeoffi.Second, many Hwss created development banks to ease constraints inlong-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 inother 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 waseffective in increasing investment. We devote particular attention to thethird category, whiich includes several policies that involve a high degreeof intervention in markets. At the simplest level, many HPALE boostedcorporations' retained earnings by adopting low corporate taxes. So'mealso reduced the cost of investment by adjusting tax, tariff, and exchanigerate policies to hold down the relative price of capital goods. Several gov-ermecnts, particularly japan, Korea, and Taiwan, China, tried to reducerisk for investors by shouldering some of it themselves or spreading itacross 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 onsavings depo'sits and corporate borrowing. This last mechanism, whichwe 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 playeda relatively small role in Asian finance. As table 5.8 shows, bonds onaverage between 1970 and 1990 accounted for less than 1 0 percent of thenet financing of corporations for the five HPAEs for which dataexist This ishigher than the share of bond finanding-in the oECD economies, however

Ironically, pofides that contributed tco some of the HPAEs success indeveloping other financial market institutions may have been inimicalto the early development of a corporate bond market. In Japan, for ex-ample, the long-ternm credit banks govermnment-protected monopoly onthe issue of long-ternm debentures discouraged the development of bondimarkets.'8 TeTa 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 minimizingfiscal deficits, has deprived embryonic bond markets of the most widelyutilize-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 governmentdid not need the financing, to establish a risk-flee benchmark for thecorpomte bond market. Malaysia and Sina ar now consideringfollowing suint.

.22.4

STRATEGIES FOWIRAP-r"DA&

-table 5.8. Net Financing Soumes of Nonfinancial Corporationsin 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.uu 3 .I97-Z5~~~~~ 80. 5 '0i 10d

!P jan23.2.&) 1.LuSti.

v~74;- 23 2 7 8 '

T..wi

b.CaWuW.Km.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 withtheir 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 ulatoryagendes, -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<lkl o play arnincreasing role in its:.financial-fiicure~.

25'

.~~~~~~~~~~~~~~~I .

The story for equity markets is broadly similar to that for bonds: de-velopment has tended to follow rather than lead the HPABs economictakeoff. In this, the HPAES are similar to other economies. Indeed, recent

evidence shows thar equity markets do not play a dominant role in cap-ital formation in either developing or industrial economies (Stiglitz1993a). Even so, as the East Asian economies have matred, govem-

ments have moved with varying degrees of success to promote their

growth by streamlining listing procedures and strengthening safguards

against fraud. In Hong Kong, for example, the govemment pushed the1986 merger of four competing exchanges into the unified Hong Kong

Stock Exchange, and it has since gradually strengthened discosure re-

quirements and antriaud safeguards. Korea has prodded firms toward

the equity market by limiting bank lending. Since 1987, Indonesia hasliberalized restrictions on share issues and allowed foreign-owned jointventures to operate secur;'ies firms.

Even so, before the remarkable bull market of the late 1980s, Asian

stock markets played a relatively minor role in the mobilization of capital-

except in Korea and Taiwan, China, where issues of shares accounted for

more than 10 percent of net financing. Since 1985, new share issues

have increased, and the proportion of external financing from equity has

nisen steeply.Recent policies to support the development of equity and bond mar-

kets have been largely successfUl. The growth of stock marker capital-ization in the HrAEs was very rapid in the 1980s, altiough the volatilityof stock prices in response to speculative pressures and capital inflowsmay have reduced their value in signaling efficient investments.20

However, the Asian stockmarket boom is a relatively recent phlenome-non, and its timing-beginning long after most HPAEs' economic take-of 3s were well under way-suggests that it is a result rather than a cause

of East Asia's rapid growth.Devopmnt Banks. East Asian governments created a wide range of fi-:

nancial institutions to fill perceived gaps in the types of credit providedby private entities (see table A5.8, at the end of the chapter). Thcy3ad-

dressed the need for long-term credit for industry by creating develop-

ment banks. Most have also aeated specialized institutions that providecredit to agriculture and small firms.

-Industrial development banks have been substantial. long-termleders in Indonesia, Japan, Korea, and Taiwan, China, but not in the

other HPAEs. In Japan, the development banks-the public Japan De-

:226

STRATEGIES FORP RAPDCG '½

velopment Bank ODB) and the private Industrial Bank of Japan-accounted for about two-thirds of loans outsmnding for equipment in-vestment in the 1950s and about half in the early 1960s. Their share intotal lending to industry was small, however, and has dedinedL At itspeak in 1953, theJDB accounted for 18 percent ofnewfimnds lent. Sart-ing in the mid-1950s, job lending fell to 1-6 percent of new lending;the rest was accounted for by private banks (Kawaura 1991).21 The

Korean Development Bank made an average of a third of all loans andguarantees in the 1970s, and the development bank of Taiwan, China,the Bank ofCoimunications, holds about half of the assets of the bank-ing-system. Conversely, Malaysias development financial institutionsaccounted for 2.9 percent of the assets of the financial sysrem in 1980s.Thailand's industrial development bank has only 1 percent of dte assetsof the financial system. Hong Kong has no development bank

Many other developing economies have also attempted to remedythe perceived failure in long-term capital markets. Nearly all have beenunsuccessful in creating development banks. Development banks inSouth Asia, Latin America, and SubSaharan Africa are beset with lowrepayment rates; a representative sample of eighteen industrial develop-ment banks in developing economies had on average nearly 50 percentof the value of their loans in arrears (World Bank 1989c). The mostcommonly cited causes of development bank flilures have been politicalpressiure to finance bad projects and the poor incentives for and capabil-ity oFfinancial instimutions to screen and monitor projects.

Development banks have performed much better in the HPAEs, espe-cially those banks in the northern tier economies concentrating on in-dustrial finance.2 Fmancial performance has been adequate to good,and the capacity to evaluate and monitor projects has at least in Japancreated spillovers for the resr of the formal financial.systen (JDB/JERI

1993).23 Successfil development banks in East Asia have applied com-mercial criteria in selecting and monitoring projects and firms, even- bwihin the constraints set by government prioriiy activities. Develop-ment banks in Japan and Singapore, for example, must select projectsthat -will repay, since the banks are expected to repay with interest the

--funds thy obtain from the government Taiwan, China, introduced an-other mechanism when in the early 1960s it required borrowers receiv-ing fumds from the U.S. Agency for International Development to put-up mathing funds and made lending to a nonperformer a criminal of-fensei imposed on: the loan Officer.

227

Most Eaist Asian governmnents have devised means to contain wilfulpolitical interference in devlopment banks. Japan, Korea, and Taiwan,China, have appointed senior officials from ministries offinance as chair-men, so they may withstand pressure from other parts of governmentOthers have controlled the tyes of lending. Thailand simply disallowedits development bank froam lending to state enterprises. The high pro&cs-sionalism and institutional identificat-ion of sraflj which is charactristicof successful development banks, is a positive &actor as well.

Policies to lacrease the Attractiveness of Invetmnt

Besidies fostering a dimate conducive to investment, encouragingbond andi equity markets and, in some instances, establishing develop-ment banks, each of the HL'AEs responded to coordination problemsthat hinder investment with a variety of mechanisms to make privateinvestment more attractive. Like most other governments that havetried to increase private investment, the HPAEs embodied many of thesepro-investment policies in their tax codes. These codes varied acrosseconomies and over time, both in their specific intent and in the degreeof succe-ss, but have nearly always aimed broadly to encourage invest-'ment. Because tax codes are often central to the quait-y of the invest-ment climate, this section begins with a brief survey of tax policy ineach of t-he wpAEs. We conclude the section, and the chapter, with a dis-cussion of foiur highly specific mechanisms usedt in some of the HrWAEs.

to encourage investment. low relative prices for capital goods, risk slar-ing, controls on capital outflows, and mild financial re-pression.

Tax Policies, to Encorage Investment The HPAES have employed a widlevariety of tax policies to increase investmnent by raising the retainedearnings of companies. On one extremne, Koreca and Taiwani, China,used complex tax codes not only to favor investment but also to directthe pattern of their industrial development. On th: other extreme,Hong Kong has maintained a simple and largely neutral tax structurewith lo croaetax rates. Other tax policy regimes fll inx between..Singapore uses taes to promote industrialization and to attain socialobjectives. Indonesia and Malaysia have relied heavily on oil revenuesandt so until recently have paid little a-ttention to other aspects of taxpolicy. Like Korea and Taiwan, China, Thailand has tried to use taxpolicy to provide investment incentives; however, weak administrativecap.abilites mean such efforts have been distortionsxy rather thin

2oi.

STRATEGIES FOR RA A

helpful. Box 5.2 further explores these issues for Korea, Malaysia, antThailand.

Below we review briefly the investment-related aspects of the taxstructr in each of the developing HPAEs.

* In Taiwan, China, income taxes generate only a small share of rev-enue because of extensive exemptions. Income from dividends and in-terest were tax-exempt until 1981, as were capital gains until 1989.Corporate tax incentives have also been extensive. Research and devel-opment expenditures are exduded from the tax base. New industries re-ceive a five-year tax holiday, and industry generally is offered accelerateddepreciation of investments. Industries that the government deemsstrategic receive additional special tax credits (Tanzi and Shome 1992).

* Korea was the first HPAE to estblish a value-added tax (VAr). The vAT

in Korea provided revenue to support incentives for investors in indusry,

particularlyexporters Tax incentives induded credits, accelerated deprec-ation, and special exemptions. In contrast with the other HPAEs, Korea hascontinuously fine-tuned itS tax policy to promote industrial development.From 1953 to 1986, there have been nine major tax reforms.

* The most distinctive feature of Hong Kong's tax policy is its lowrate strucmre and the low level of revenue it generates. The maximum

personal income tax rate is about 15 percent, while oorporate income

taxes are 17.5 percent. Hong Kong uses low taxes to attract investmentsin contrast to the mix of high tax rates and high tax incentives offered inKorea and Taiwan, China.

* Singapore's tax regime is as interventionist as those in Korea andTaiwan, China, but focuses special incentives on foreign investors andon promoting savings. (Contributions to the Central Provident Fundare tax-exempt for both employers and employees.) Malaysia has thehighest level of taxation. Its tax-GD? ratio averaged 21.45 percent in1979-88 but fcil steadily during the period because oil revenuesdeclined without compensating increases in non-oil revenucs, and thegovernment continued to expand tax incentives to sustain its industri-ilization drive. The resulting high deficit and emergng constraints onspending have focused attention on the need for systematic tax reforms.

* Since the 1970s Indonesia has relied heavily on revenue from petro-leum andgas, and did not until the early 1980s use tax policyforspecific

incentive objectives. In 1985 in response to dwindling oil revenues, itembarked on a comprehensive reform that unified personal and corpo-

rate-income taxes, introduced value-added and urban real estate taxes,

229

2~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i ; =L

;Wni -XV

x.2'.fCTL Ei e.

r ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~- q' 4

- ; adeimntd obse n0tvsNewon a eeroesysc- . - fill in rKlucing clisroriions and increasing mx rerenuesP~~~~~~~~~~~~~~~~~~~~~ _,

hii~~~~ Thiln fzsoeo i otcmpe xEsn mn i- H2AEs. Business and incorne ~~~~tahnueabwidrgary frts

-~~pa ~. Crmpevi aind ii~,lack FTof b rai capci haecsltti oo m - pe.n.men.I conrs wihtora n Taiwan, Chinaniheri s, in

-~ ~~~~~A cetie -ae1nealwrkda itl, inceninres m hiIn-~~~~4 apert aedoemr amtl ode h oermn a e

-~~~~~~~~~~~~~~~~~~~~~~~~~~k compl iuiesnaX' .-. ..y .n -:f :.~ ~ ~ ~~~~~~~~~~~~~~~~~~11ta-t _ , . .i:~~~~~- raS: -2i :xI ac- . - - .n

.. ;.'-V ~ ~ ~ oX-, o1anes-- f.eos~av-~~t-*~~~~~~~~~~~~~~~~~~~~~~~~er~-n :aea.s o,

t., .,. - : 5. , , . : f -,. . S; o: f , ,- . . - : . : f - - - - - 0; :-f p-:

STRATEGIES FOIR-

- :&tu~t 4e~½~ 17 , *. 4. r,

- . Z>- n

!14r - ow .omorumavntaaeealorne hstrowCr,,>.mdlnon; xbat'flpC c-..s% rte-

{g ' E aji&ht w*kw4ti4<_4$SF-> 4nordi#Pndfr^½Mstoa ts)boni.tacnnve~idrssi. rte 'lmiberonibo£anM mictaves:ant

.~~~ .rDooe>Doe ite: baafpmeERm-@edlafsicrtoqsnutoC*Ddae- 'zi*:<firtl*8if =lMQl:}enn,>zem ,J 49.W4.,. 0,fl ' tl 5 f1

e ^D¢ 4$i.<v:eiopmencfI rmi4mlulceSeancQgi 3 &rntlolsescllubTncd alyEboult !5j7crceXu,!ri< -.-.f4 {S onoImc,3n&213 *^ jO?9S t.ct PgrOw d ~- sFMi S16

iln at n~ GDP~Its ~ -

gu 5brd

* eau:tvrgone:revcnuc. de2Tmcnc *u;lro:et ooteono4 N~S

-N ~~~~~~~~~~

tZib1c$varro~~~~~~adn ra-I

Did the Hr-s' investment-friendly tax olicics spur -growth? Tax poli-

des involve fiscal sacriflice, a trnsfer of income from taxpayer in general toinvestors. The assessment ofwhether these polices were successf hii dependson similar considerations to those applying to pubc or forced savings, Ifthe increased investments were absorbed effidiently, as ir appears thcy were

*in mostrof therHrAEswitha2ctvist tax polldies~ itris likelythatrthey raised ag-

gregate welfare But the administrative burden associated withi a compli-Cared tax incntitructure such as K-oreas or Singapore's has hi,h social

costs itself The success of ocher economies in East Asia, such as HongKon with very low but unifrm. tame on corporate incomes, suggests that:

the major gains accrue from leaving mrtaied earings with corporationssceand pionolvng fivestment i equipmmen, rather than from attempting totfine-mne incentive strucurecs to private investors through the tax system.

7231

Low Relative Pties for Capital Goods. Unlike mostrocher low- and middle-income economies, the H-iPAEs were able to hold down the relative pricesof invertment goods, especially equipment, during the 1980s (Bhiar-tacharya and Page 1993). Figure 5.4 shows the relative price of invest-ment of goods (ratio of the investment deflator to the GDP deflator) forforty economies for whi'ch consistent real private investment data areava-ilable. F-or the group as a whole, the price of investment goods in-creased about 15 percent faster than all otlicr goods during the 1980s, inpart because devaluations increased the domestic price ofinmported andimport-intensive investment goods.

The HPA.Es arc different in several ways. First, the relative price- of'mn-vestment goods remained lower than in other developing economiuesthroughout the global economic expansion of the 1970s and the adjust-ment period in the 1980s. Second, the relative price- of investme'ntgoods declined during the early stages of adjustment frm 1980 to1984-a period when they were rising in most other developingeconomies-and only began to rise in'1985, when the HiPAE economieswere well on the way to recovery. By 1989, relative prices between HPAESand ocher economies were again aligned. Thus in these economies de-.

Figure 5.4 Investment Deflator/GDP Defator

Index 1970 _120 -

:i~~~~~~~~~~~~~~~~~~~~ I

12±2

Now eafe LM ices fo=aia od niems te low- and middle-ncmconcs

- ta~~~~~~~~Suceay PeTrand Pand Mad93).a (1992) hwv h daieprc f net

232~~~~~~~~~1

letog od rtoo h necetdfao oteGrttso)fo- -foerecnoie fr hih onisenrrel riat iarsten dt0ar

104lbe oregopa awoe h rieo netmn od-n

vestme.t godsreaie lower than middler-ievlopin economnies .- ;: ir~~~~~~~~~~~~ource Pout-an th lbleooi xaso fe17sand Madaassa (I 92

: e.peidi2h328sScnd h eaiv rc f neimn

STRATEGIES FOR RAPI

cdining real prices of investment goods smoothed the impact of adjust-ment on investment.2 4

Rising investment goods prices affect the relationship between in-vestment effort and investment outcomes. As the relative price of invest-ment goods rises (over dime or across economies), more nominalinvestment expenditure is needed to achieve the same volume of physi-cal investment. For the HPAES and the economies of the non-HPAE sam-ple, investment efftort and outcomes coincided quite dosely during the1970s when the relative price of investment goods was essendially simi-lar. During the 1980s, the rising relative price of investment goods innon-HIPAEs was reflected in an increasing divergence between investmenteffort and outcomes. By 1989 the difference was nearly 3 percentagepoints of GDP? By contrast, in the HPAES, investment outcomes exceededinvestment effort (for both public and private investment) for the entireperiod 197089.25

Tax, marif, and exchange rate policies that kepc the relative price of in-vestment goods in the HPAEs below that for other low- and middle-incomeeconomies undoubtedly contributed to growth. Snce the same volume ofnominal investment bought more real capital goods in these economies,output was increased and returns on nominal inves=nenrs were higher.

Bounding Risksto Piate Iiwestwm Some HPAE govemrnments, primarily in

the northern tier, have attempted to increase investment by lowering theuncertainty associated with real investment, implicitly or explicitly shar-ing risks with the private sector. Risk-sharing mechanisms have come inmany forms-recession cartels in Japan, firm and bank workouts in

Korea, financial repression to recapitalize firms in Malaysia. signalingprorities and policy intentions through the direaed-aedicsystems ofJapan, Korea, and Taiwan, China, and credit guarantees to small andmedium-size enterprises in Korea and Taiwan, Chinat Those that wereeffective all exiibited a common feature firms benefiting from sharedrisk were monitored for performance. In this way governments miti-gated the problem that has plagued many public sector attempts to sharerisks with rhe private sector in other low- and middle-incomeeconomies: reducing the risk of fihilure to the private sector reduces theincentive to avoid failure. This can lead to a political dynamic in whichgains are private but losses are socialized.

An importnt feature ofjapanese industrial policywas the creation ofrecession carrels and the use of other forms of adjustment assistance indedining industries. Japanese andimonopoly law has been consistendy

233

: :: g -: C A C LX0 _EEFLE V-~~LE

lenient in allowving the organization of cartels under the administaiveguidance of Mm to ease the exit of firms from dedining industries.These cartels were especially common during the rapid growth periodwhen they were used to ease the adjustment problems of dedining in-dustries. The cartel allows firms to share losses. They may also be eligi-ble for subsidized loans to upgrade technology or restructure operadons.In ths way, bankruptcy or massive layoffs by one enterprise may be

avoided (Ito 1992).Japanese adjustment assistancc has been especially prominent in min-

ing, textiles, shipbuilding, and aluminum. In general it has succecded inavoiding the abrupt dislocations that often characcerize firm fiailures inmarket economies, but this orderly decline has been purchased at thecost of inaeased prices to consumers and lower reurums to investors infirms that would have survived. Moreover, the cartels provide an incen-tive for firms to overinvest in capacity during good times, since they willbe protected during downns and their adjustnent assistance will berelated to their existing capacity.

Performance criteria play an important role, even in declining indus-tries. Interacton among firms and between firms and MITI are a meansto prevent free riding and to monitor performance. Thus the organiza-tion of an industial adjustnent assistance program in Japan conformsto the model of contests presented in chapter 2 and suggests thateconomies without similarly well-developed institutional capacity maywish to avoid the commitments to orderly exit policies.

In some HPAEs, most frequently in Japan and Korea, troubled priorityprojects have been bailed out by the govemrnment. Often, the financialcost of the bailout was large. In Japan the govenment took over somelosses associated with financing decdining industries, such as in coal min-ing. When govenmuent-supported heavy and chemnical industry projectsin Korea experienced severe excess capacity and financial diffidulties inthe 1980s, the governmenz provided seventy-eight distressed firms withnew, subsidized loans (totaling abour 16 percent of commercial bankloans) and rescheduled outstanding ones. The govrncment also providedbanks, whose nonperforming loans rose substantially and whose profitsdedined (partly because lending rates were cut), with subsidized creditfrom the Central Bank and allowed them entry into attractive areas of fi-nancial services. In Indonesia, the govemment bought 35 percent of theequity of a large ement plantwhen it became financiallytroubled whileoperating only at 50 percent capacity (Kunio 1988).

234

STRATEGIES FOR RAPI

Dcvcdopmenr bank lending may also confrr implicit insurance. It fie-quendy signals areas of government commitment, providing an addi-tional measure of comfort to private investors and banks JDB/JERI

1993). These commitments may extnd to efforts to ease adjustmentproblems in promoted sectors. With government inducement, the In-dustrial Bank of Japan has led syndications for firms in distress andforced their restructuring& substituting for liquidation or externalcakeovers. In the 1980s Korea extended cheap development bank loansto firms that were nor meeting their debt obligations.

Has bounding risk contributed to growth? Bounding risk by the pri-vate sector was apparently successful or at least not a fitilure in Japan,Korea, and Taiwan, China, but this is almost unique compared with theexperience of other economies, developing or industrial. Risk sharinggenerates moral hazard problems, which Japan, Korea, and Taiwan,China, attempted to contain dtrough fostering insitutions to monitorpriority projects. One consequence of shared risks was shared monitor-ing by the private and public sectors. Several instiutions, induding de-velopment banks, private financial institutions, firms, and variousministries in the governmenr, often had stakes in the implementation ofthe project. The government's role as monitor of firms was important.For example, Korea's financial policies promoted high leverage, whichmade firms vulnerable to loan rol over decisions of financial insttu-tions. Since the government was able to influence which loans wererolled over, it was able to use them as a lever for eliciting compliancefrom firms (Sakong 1993). Similar monitoring was rarely oresent in theSoutheast Asian HPAEs, with the consequence that attempts to boundrisk there were not successful.

Coerals on Capital Oufflows. Restrictng capital outflows is one of themore controversial mechanisms employed by the HPAEs to increase do-mestic investment. Japan, Korea, and Taiwan, China, all employed suchrestrictions during the formative periods of their rapid growth. In each

case the rationale was straightforward: if people were prohibited fromsending capital abroad, they would save and invest at home. In additionthe capital controls made repressing interest rates on deposits possible,

since savers were denied higher-yielding assets abroad. Although allthree economies have since liberalized controls on capital flows to vary-ing degrees, the existence of restrictions at a time when these economieswere achieving high and rising savings rates challenges the premise thatfree and open financial markets are always best for growtk

235

f a .;;. .- 5Y

WVhat impact, then, do capital export restrictions have on accumulda-dion? The obvious fuilure of such restrictions in economices prone to cap-it-al flight suggests that capital is too fluiid to be retained wvhen domesticconditions are inimical to savings and investmenc. Therefore the sim-plest explanation for the seemning success of these restrictions in North-east Aiistathecnmsoffered returns adequate to retain capita

even without restrictions. Converely, for reasons we discussed in chap-ter 4, these particular governments have been unusually successful ineliciting the cooperation of economic elites, t-he very groups and indi-viduals who might otherwise shiip capital offihore.

Thus, for certain periods in these dthre economies, government restric-dions on outward capital flows may have contributed marginally to do-mesnc accumulation, if only by raising the risk and therefore the cost ofsending capital abroad. If so, such restrictons; which have the potential

disadvantage of driving otherwise legitimatc ecoonomnic activity under-ground or even provoking capital flight, would rank among the compara-tively risky interventions that are likely to prove successfujl only ineconomies that have able bureaucracies and have already established thefundamentals of growth. Moreover, as the trend toward capital accountliberalzation in Japan, Korea, and Taiwan, China, suggests, retrctions

are probably useful orl4 for a lImited dimc: as economnies grow, becomingmore sophisticated andi more dlosely linked to international markets, re-strictions on capital exports inhibit efficiency and must be dismantled.

In aniy event, the success of the more open BPAEs undermines anysuggestion that capital flow restrictions are somehow necessary for sav-ings and growti. Hong Kong, Malaysia, andl Singapore have all achieved

high and rising savings rates while permitidng and-in the case of thegional financial hubs, Hong Kong and Singapore-even encouragingfree flows of capital. Indonesia attempted to control its capital accountfrom independence until the early 1970s but due to the very open na-ture of its capital marker chose to shift to a gradual policy of libemliza-don. Finaluy, Thailand has retained format restricdions on capitalmovements, but the economy in peractice exhibits a high degee of opn-nes (an internal World Bank report). Thus, although capital accountrestrictions may have played a positivc role in the early growth of someHPA:s, conomues that have never imposed rthem or have since liberal-ized appcar to have little to gain by imposing them.

East Asia's lesson about capital account restrnctions may in fact behow to get rid of them rather thanhow to impose them. Both Japan and

1Z36

STRATEGIES FOR R

Taiwan, China, were successful in liberalizing their capital accounts verygradually throughout the 1970s and 1980s, adjusting the pace as neces-sary to maintain macroeconomic stability and minimize interest rateand exchange rate volarility. Indonesia's capital marker liberalization isdescribed in box 5.3. But while gradual deregulation may be best foreconomies in which bureaucrats can craft economic policy with rela-tively little interference, sudden deregulation may be the only deregula-tion possible in economies where governments are less insulated andthere is strong resistance to reforn. In such cases policymakers mustweigh the dangers of sudden liberalization against the risk that gradualreform may be derailed alrogether.

Moderate Repression of Inteiest Rates. The final mechanism we discuss-the repression of interest rates on corporate borrowing-is perhaps themost controversial. Financial repression is often used to describe the sit-uation when interest rates are at negative real levels. A more precise def-inition, the one we use here, is govermnent intervention to hold interestrates below market-clearing levels, Of course, in many developingeconomies with capitaI account restrictions, managed exchange rates,and undeveloped capital and money markets, it is difficult to determinea market-clearing interest rate. In the absence of such rates, HPAE gov-

ernments tended to use international interest rates, such as the London

interbank offired rate, as guides for the opporunity cost of domestic

savings. Given dosed capiral accounts or high transactions costs in mor-ing funds abroad, they have sometimes had the latitude to maintain de-

posit rates below marker-dcearing levels without provoking capital flightor significant disintermediarion into the informal markets, where higherreal rates prevaiL

Economic theory and empirical evidence agree that if a governmentholds real interest rates on deposits too low for too long people will havelittle or no incentive to accumulate financial assets, financat savings willfall, and econonic growth will be adversely affected. Fmancial repressionusually occurs when inflation runs ahead of adjustrnents in govermment-regulated nominal intcrest rates. This type of financial repression hasoften resulted in severely negative real interest rates, for example, in Ar-gentina between 1972 and 1985, in Indonesia between 1973 and 1980,and in the Philippines between 1970 and 1985. Cross-economy regres-sion analyses supporc a positive association between real interest rates

and the rate of growth of economic output; that is, large negative real in-terest rates result in lower growth (Gelb 1989).

237

AR CL

e nhgh9 L4i meore,s 4 guhuS

dn rs R k4/ flV.w% 1" .a U2

d&rgdl ~ $¶ ~ ~oi'nrArf ~4W iJ*&V

assessme T heHofNavust se fzinanchial beresoom s codnsitn n e

suggests'rle'op5 fnanr n sfr gsures to elctdco&nomic naecitlrsea. Hoevr,83acile

time or notherheldrntcrestirtes4below makt-dZaringlcv hl

nnthels HacShaeveinog used anc mia repression asgrea citen and de-nia

liiberate polic fovr time tosecurentl rsourcessfto finnc the bdgvlpentobranksanr resourcesnto selectedsaig:dcthAin economicsscos oevriaca gre-pression hasca bepenssedion, selectd peraniodsrepusually afterany extenalshockte to treAsanse mresucles?oadete iigidsra on so

time orfanotherahel finterestl raespbelow markeostclaring lEeveSmwhilnonetUyhaels achitevighgdn pigrtso aggregateintmt and find n hHancial.

banksgancoimeodpositinofordsavns diaidltheoAsianleconomiesrowes

23-8

STRArEGIE R0 RAIDXO=t~~~~~~~~~~~~~NT 37 5

. t =~~~~~~~~~~~~~~~~~~~~t

- . - - S T~~~~~~~~~~~~~~~ RA- ;7;; E SF A

X. 2

-; . . nr+ridd rd4 a4 the'interest r iid it pmasewoduc-s and servces'inuluhdprvate b s- - .i£is uMffiedylwn s has e mchiaseher mrgina- rpn 9SiytoiBhves overrnment hogusane o Cld ser,Un 12 r:Fa 1 ipnvacerafinnci

* ODW.nm_xc0xnVcbUarQto.%cn cc, .'t ;:l tt3.e ie n *r the -: s. aatet subidy

- fiom househol .to firsns inreases the oolorate!profitabljusvof invest

* - -ment.by reducUingborrowig 0 oRnted,thes cij1 j r poirsii bostp

::: ~ ~ flfiC0 ofth 'qiymru duringn to2E'eooctaefprosbpo

y.$0 d iid &MCM-.-~~~~~~~~~~~~~~~~~~~~. 2 -

- s -FandiadorpressincinthuHAEghs iffered froim latStin ro.swd o nthernasrxcs i4owe

-:cooiesinmthreea impom.nvhent dshedrefeeshiond wasmorElativdy moderate arldw sh -lir`d excepions di notcrtoet

ulbCC& ubkflt&7Q

in pen;istndy Iit In real.n interestrat non tablea5.5 ieod as. ae

- i3rprsso of euryn in a h&gbvn enon merntnenrghn

.~~- -.. ~ .+*~a.- -,.. -~ off' mareooi *1-ai an was no N. urintede conseque -of

. ... . - -.-

A.....

g ; . a --. ,J . . .

- ~ ~ ~ _ -- X - - ..-

ings is,- ' sufficiently low and t corporate sector has a highe marginal

from ousehlds o frms inceaesth crprates proitbiit1o iinvest-.trq =--

the euityoffins, nablig thm toinvet inriskier- projetsnd bo Uomore to financethese projec.' Thi ma hvepadyofe te eanes

idingfirns. anatrnatived toa tequt fiainngesrt (Sdlasiity of993a)ldsavFginancufiadpendyion ind the HopAESeseo has diardfo thathin mostgothe

penois iny thre nporthan th ways.hoFirstao,thedereenofrepressayincrwasearegav1 oeate ansmnd, wirtha ewrsore,t-lied fcdeptiontrs,tio retcsultidino peusistenlydnegtoeireal interestmes (see tablrae 5.5).tbiSecyofnd,estwe

havte showniny chapktedri3g rheprsion wasonderctakenoifa penvironmentpro

oFmnancninic rpestainintye HandEws nohes unintened rmta conmseqec ofhe

ecoomis i thee mpotan W27S.rlrt, he legee f rpresio3Wa

t - 0 _ L E . 0~~~n-t

,~~ ~ ~~~~~~~~~~~~~~~~~~~~~~ .. 1p -,w

.- ~~~~~~~~~~~~~~. : -

':~~~~~~~~9O,9 C .2

,-''' ;':~~~~~~~~~~~a e:on

-- : -lasqueezetL+Si ~ ~ xb~ unosyxyi}neetrtSadvabnod "Minte rajelnal=knsa T, 1- FQ15g"fim%1,,.,-1B g _>9-= I .;-i

.~~~~~~~~R~p . Xtn<sn oisbly neFEen^o,rmn B or4-mom eot tzSnprXdarmem; s;-

Siz.EbrincrS pl,o in,rainrasne.-dig 0_,ovn os -Wofnloin ;-bod ul er..

.:~~~~~~~a 2r- s--a et r >Et 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*<g.- ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~r - 1uuNvdeLAtsts@ oovr>navnolX -icmiemltGsaco cosraire-mti=ur.a: -

f~~~~~~~~~~~~~~~~~~~~~~ _N 2 P

i ¢ itrFt..4->><stxts,t<i/gt; ,,>,'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 > ; ; <S:t' n - *t i : - -- r., Ji I B;be>vE;uldshor-rem;tfinancrqprn=svsitrmalsmm4ia;UkI, b pem6d t -vat.FSeracser.

: . I g ; ;,= a '<l-'4' '- ^ ,, } "~~~~~~s vereLstri r

0',''~~~~~~~~~~~~~~. 'V _.I : . 9WillhOtilL9tlt'iOiEiittCrClltOnnpis-:SIi t r, Of.ii.rAdilICDO:4Wai& &1 :

E > Srsge wer fNtf<! *-5s-^st2 ¢Sg.^=^, s> .w--- ¢\*1.tPGK> -O f mof-. . S.7;1vt ; mE~~~~~~~~~~~~~~~~~-m "Proodx tssXadMeneeRnf>Bcue.i zn- S'6Stae>-enf 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 tho.- ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~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 mostlikely to be kept. Since rationing already takes place in financial systems,the additional impact of moderate financial repression can be relativelysmall. Indeed, credit rationing may actually be good for growth, if themechanism used to select borrowers generates incentives to perform. For-emmple, Japan, Korea, and Taiwan, China, allocated subsidized credit onthe basis of past performance, usually with exports as the criterion of suc-cess. Since the value of subsidized capital was high, firms perceived highmarginal 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 thisquestion by examining the relationships between income growth, realinterest rates, and inflation using two statistical methods and two datasets: a sample of nineteen economies that includes a subser with positivereal 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 interestrates. i the subset of economies with positive real interest rates, the in-wrtest rate is statistically insignificant and actually has a negative effecton 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 coefficientis statistically insignificant. In Korea, inflation and the curb market rate(our proxy for the equilibrium interest rare) significantly explain incomegrowth. In Taiwan, China, none of the variables is significandy relatedto trhe growth rate. While these tests are not condcusive, they support theview that moderate fnancial repression did not necessarily hindergrowth 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? Judgmentsare difficult, given our incomplete understanding of the costs and bene-fits of all the policies used. But on balance the common elements acrossthe 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, soundand solvent financial insctations, secure property rights and comple-mentary public investments in infrastructure, and low relative prices ofinvestment goods. Efforts to improve the institutional framework forcapital market development came later in the process and were not re-sponsible fbr takeoff. In some cases well-functioning developmentbanks were a positive but not a determining factor.

More selective interventions-forced savings, tax policies to promote(sometimes very specific) investnents, sharing risk, restricting capitaloutflow, and repressing interest rates also appear to have succeeded insome HPAEs, especiallyJapan, Korea, Singapore, and Taiwan, China Butthe potential costs of these more selectve interventions if misappliedcan be very high in terms of consumer welfare, and strong institutionalcapability is necessary. They would not have succeeded without the im-portant monitoring and disciplinary roles performed by the banks andpublic institurions of these economies. Where other East Asianeconomies have lacked this capability-in Indonesia, Malaysia, andThailand-ef%ors at selective interventions to promote rapid accumula-tion have been generally unsuccessful.

Appendix 5.1: Granger Causality Tests forSavings 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 capitai) and the gross saving rate (s). Tests were done for regression resultfir 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,+,ik-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 arejointly significant (equation 5.1) and whether the lagged growth ratesare jointly significant (equation 5.2). Up to five years of lagged rates areincluded. The results show a causal relationship from ODP growth ratesto savings rates for most East Asian economies (see table AS. 1).

The Granger definition exploits time-series relationships to identifycausality. By this definition, g causes s if the prediction of the currentvalue of s (s) is strengthened by usingpast values ofgt. Thus, gcauses s ifc2a(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, wemake two simplifjing assumptions. First, that the set A, indudes s, andg only, and not a third variable. Second, that tansformations 7Tand Texist, 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 096991'aaysi=5. 572.5

PT ~~~~~~~193'3. 314.7 -8340Ot0xnanana .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 PerCapit GOP Growth Rate and the Gmoss Savings Rate

Si ~~~~Sig

*~~~~~~~ ~ ~~~~~~~ *2 S; S

4 .m .Si

NS~~~~S3;.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 S4 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 theautoregressrve representations, s. can be regressed on its own past valuesand on past values ofgto rest whether g, causes s, by resting whether thecoefficients 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 Relationshipbetween 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 inflationin 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 inthe real instimutional rate of interest does not mean a change in financialrepression if the equilibrium rate falls equivalently because of lack of in-vestmenit opportunities. This study, thercfore, introduces better proxiesfor the degree of financial repression in the single economy regressions:the difference between the curb marker rates and the banking system'sinterest rates. (These are available fMr Korea and Taiwan, China, but: notfMr 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 experiencesa sudden increase in productivity, for example, the demnand for credit willrise, 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 studywould 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 toboth thLe complete sample and to the subsample ofeconomies with pos-itive average rates of interest. Avenage GNP growth rates were used asproxies 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 accountfor any sectoral trend, and two measures of an index of relative incomebecause standard growth models predict that poorer ecoonomies havehigher 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 twentyeconomnies the real interest ratechas astatisticallysignificant and positiveimpact 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

5 LAJIA <tLLUL -000;; 1 0.277t Subsamplcof&os o-i;

-. _ AX C5

f tri-ot >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 havehad a positive impact on growth. The coefficient on inflation remainsnegtive 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 interestrace. 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 rerainshighly 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 havethe same sign and similar manitudes as those for Korea, but the standarderrors 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 7.954Maaysia"' ..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.OzhtAidPhilPEippines000: 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 3i31983' 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- - 19.9p~. * 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~~~~~~~~~~~0i ---̂ Boiswana. =' :i -- -4- --- 19856 `2'.: 5 0 00-Cbce'Ivi-re 05 198- ;8-T 5 0 527 10 5 5Gh-.na ',-. -14-. 19s8j9' j rt

dia 1-9 '19645 6.7' 48 -9 7.21975-76 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 prudentialt* regation.

Hihinternationalg;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 ovcrby 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 dbki 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 -247 '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 etaand4 (of 47) i' naiicecompa . Bank = Iedfe ,ic

-. ,.gee' (b) Deterioration rtemsia-nieterealsoi inacial dis- li 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 tinvestorsaccounting 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 torubd -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-±rcgnkabiklion, or 26.2 percent of GDP.r ( p ' tesd interest rate sub

ltoadsupervisowo er4utpwp o eaand r2Mpaniisiderand 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 ornealy 6 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 paido- deposit money bank assets,w esi rats, coupled with off by ie deposit insurance182 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 raaasset. s .-.-ofde|itnolney ban. tienes and inadequate : epate agency under over

-- ;;:-.By-1986, centl b:.lcassis- -- piud al rigon t;meat mMnaeent. Five privatee.- :;: - -- :-.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-ir5. .-.- 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-eA 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' upQ 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 sesr ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ &rp .' ,. Tnw _.e .;g

-tuSI -. L E

* g f ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ mi

Hmlshnet ~~~193 75 77.7 .2Indaish 28-54 59 690.10lapi ' 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<N-x-~~~~~~~~~~Dhlds) - t ips-Y;an - . 74w (UGh U.f:t

X=atbr i ; g ! .9! i

Pidais ~~~~19116-U - 4 O.26hipln 3986-U 40 CLOW)

. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ a a ..C. .

CA"ac '.19156-111- 20..

M e 90 0.50

52

.~~~~~~~~~~~~~~~~&cw . .

6CM.

u~~~~~~~~~~~Ti,; .- 76

l O s-u . 5. 0040

cs~~~~~~~~~~~~~~IW..~~~~~~~~~~~~~~~~~~~~~~~~ ~ - .20. . .*~~~~ ~ ~~~ ~~~~~~ -.- l .. 3.8 67X, 4:02

2989 ' ~50 6 41 LU.Camdz, 1989 , 5~~~4 69, 431a

Pane *. . - 1359U .43 ~ - as ~ i

; ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Iy .91 r . ** .7

- 993 .r

S'ahgznbnd - ~~~1383 .. 71 -

~~4enua1abJc .. , :2

tlkHcdnMiI Sahi main&Inedbna~iuIl.,w ,.M.Ifbbrh mAh6hIfl6&lkj d k=ni htl

cineanmtflrKd ktmdinfr¶ask~.~ah.k indah d SmwcnSm t klptnDdU n(Ihhkaa& Iph.Pn ISbehbrasO tnzh,S),,3higuhie Saif wn(5 baptvv I [naMti cm(x9kzphw,JnJ wi 3,nishll

-. - - - STRAT~~~~~~~~~~~~~~~~~~EGIES FOR0 :!i ~~~~~~~~~~* ST1_s;RAT F s RA,_

-.Table A5.7 Interest Rate Spreads of Commercial Banks, Selected HPAErst S r _ps P > 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 -7.00 r ,s; a 1 ' _ L _N . r ."5

;~~~~~~~~~~~~~~~ 6

X .. @V i or:ptnsn<SiXpn:-il-cl tclHai-ners ac:

tMposrnt'ctsaresweaswond;El}i L&enm&at ~ 'calS' s4 -

:/_~~~~~~~~~2 4

e -1aitltwo i ;;

- . X .e. >nfre. .l °fl°,c.,. x,,ua,o.lpoxasedr lGo e s.v . .;-

<,>3.g,c<anddwoslE.! -tasta;6OrM-.7;Ft..f r4,4

kV'N we;ostffnsfrIhln,&-

.- ,-~~~~~~~~~~ til uo,,.ii.rn=ii

-::~~geed .te Z -r7- - 4 . .- -. :-

. E5 -IR_AL EE

Table A5.8 Outimes of Major Policy Implem on inancing Institutions

-~~~~~~~ A f it e . -. -. . . < . - . . - .0,--;0Dla -Aals

U..e ( - ;.

I - .. nnii 7.5 pantna..d

M3 manC4 (GO~~~~~mt) -- bafowa nd. Acwith bath.Bath ( -PtN009 ' - . ( 0rWSM7mIlm, - %nr . - Pksci.fwpa&M.y- M ;f0Q00Q.-- ; .4n ~tirl;.u;

LnU1S UmmlUll 23

'1135~~~~~~~~~~~~~~~~~~~~~~~~~ii -u

De - - , f - n1ef terstssX MO Fw* %

wi4 l, .*, raEd ** *t 17a= ;

Ewe uneur Das. '- - - hAS?- bik we-' 5, p=^ ' nrlmam4ea - Iaaswmi-aawha-uda h-

* t2 --b; ;- l -l,e A, t .-; r -{lwace h . - n . n - , -- - e s . -

(US$3 LEADe 38) crs) mmdmmBane C ,x l , , ,

tss9,~lnsgT 05, lat , H - 351 ,Wtau .- - ii inpesmu, - Mas.s-~

r6'w -S> '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 mnouvktPhllpphim (inw) . UISI WUDNW Caa,g--st aayTciamr

salmawn 4 youn fixwmldtus

Hai, 3 Ealcmwmaai ma

25dluftolankga.safaw,

* *. Ml~~~~~~~~~~~~~naCr) -. - . .~~~~~~~~~~~~~~~~~~~~~~agmriaa ftfiwm. Fwre iofhlsi5p).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 d 9 =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 basesmixs dhc DdJchi Kangac Dank. lad. Malayt, teat] banksa.iaa.Dn fllnAadRps 33Twhfpa.Ca ai u.uo Iqi BthtapnbahndoknetJapa Cntlnmmod Kot.spaeqtaa(nip am TDank 52ldr. w mi lee N.re hl~C"lbakum T a

Novtes1. 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 theeffect 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 enrollmentswith 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 ageteachers are roughly 5 times per capita GNP. However, be- cohort in East Asia than in Latin Amcrica (see bdow incause pupil-rcacher ratios are higher in Korea, operating che text), this difFerence underestimates the gap betweencosts 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 determiningteachers' vage and in controlling costs. Sec Tan and 11. Howevcr, since the mid-1980s, public savings inMingat (1992), pp. 28-34. Argentina, Brazil, Chile, Colombia, Costa Rica (since

1983), and Mcxico (in 1987) have incrcased in response5. 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, maythat limit the private provision of educational services, have avoided large and systematic bankruptcies, whichPsacharopoulos, Jimenez, and Tan (1986) present esti- surely helped the solvency of financial institutions.mnates of private and social retrns from education. SeePsacharopoulos (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 prudentialingtheir failurc to acoount forabiityand forqualiy ofCed- regulations and capital adequacy requirements has beenucation. Neither he nor others, however, question the evidenr in Indonesia, where low capital requirements forexistence 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 byfects 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 in8. 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 ofschooling of the school-age cohort (essentially an aggre- 15. Stiglirz (1993a) notes that in economies in whichgateof enrollment rates) is the independent variable. For governments do nor take an active pro-competitivea sample of fifteen Asian and Latin American economies, stance, there are a limited number of firms in the bankingthe expenditure variable was insignificant. See Tan and sector. Furthermore, even when there are a large numberMingat (1992). of banls, e&ctive competirion (if viewed from the per-

spective of the receiver of the financial service) is likely to9. 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 markerkthe one hand; universal and near universal enrollmrient There are differences in the kind of infbrmation that isrates 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 specializationconsistenr with the shift in the strucrure of production among financial institurions confirm this view.

- and exports to more technologically sophisticated andskill-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-incomeeconomies. 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 thethe evidence concerning the negative effect of share issucs basis of its excellent financial performance (Gordonon 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 with18. 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 frombeen 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 ofmost of these bonds arc government guaranteed. Firms investmenr outcomes exceeding investment efforts amonghave begun to issue bonds, because governmenc has the five HPAEs in the sample. Malaysia is perhaps the leastbegun to control the debt-equiy ratios of large corpora- representativc in the sense that during the 1970s (and uptions (an internal World Bank report). Many of the cor- to 1982) the nominal investment share cxceeded the realporate bond issues may be simply a form of bank investment share. All five economies, however, had realfinancing 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 relativecapital goods prices in 1988-89.

21. More substantial long-term lending to industryhas 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 GDP22. 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 andhave 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 pricesDevelopment 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 achieverapid accumulation. But high levels of physical and human cap-ital formation are not a guarantee of economic success, as themodem 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 theHPAEs had total factor productivity growdt that exceeded that of morethan 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, contributedmuch more to output growth than would be predicted on the basis ofthe 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 workHPAE governments have relied primarily on market mechanisms to guideallocative decisions in both labor and capital markets. Labor markets inthe 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 lessthan in most other low- and middle-income economies (as we demon-strated in chapter 5). Government efforts to direct credit have generallybeen undertaken within the framework of moderate interest rare subsi-dies, and financal institutions have been given the final decision on thecreditworthiness of investment projects. This has meant that loans sup-plied through directed credit have been monitored and generally havebeen repaid. The relative prices of capital and labor have broadly re-flccred their relative scarcities.

The HPAEs have also used the international economy more effectvelythan many other low- and middle-income economies. Each of theHPAEs, except Hong Kong, went through an early phase of protection ofimpo rt-subsituting industries. But these policies were later modified byreductions in import controls and tariffs, combined with strong incen-tives to export (see chapter 3). This opened up much of the modemeconomy to international compeiition and introduced world prices asperformance standards, not only for exports but also for tie intermedi-atc and capital goods used in export production. As a result, domesticprices 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 directedcredit to priority sectors and sometimes to specific firms. Where theseprograms have been effective in -allocating resources to high-yieldinginvestrnents-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 developmentof specific knowledge- and capital-intensive industries. We test the rela-tionship between industrial policy and productivity-based catching upand conclude that industrial policy has generally nor been successful inchanging industrial structure or raising productivity, What, then, hascontributed to the iPAEs' apparendy superior performance in adoptingand mastering international best-practice technologies? We argue thatthe combination of compedtive discipline and well-firictioning factormarkets 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 internationaltechnology markets. Hih and rising levels of htiman capital in theHPAES, especialy the superior educationd atminment and cognitive sklof the industrial labor force, helped to ensure that these new processesand equipment were used and adapted effectively. Thus export orienta-tion and high human capitl formed a virtuous circle; exports raised thereturns 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 andwages, 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. Utilizinghumnan 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 tounderstand why public policy might be used, to help markets function

-- - - ; 0 ~~~~~~~~~~~~~~26t -

,'',' e!'kCLE

better. In the case of labor markets, however, coordination failures arelimited, and information, although not perfectly symmetric betweenworkers 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, thereis a presumption tiat, ifleftto themselves, labor markets would performtheir coordinating (allocative) role quite well.

Two conditions must be fulfilled for a growing supply of educatedlabor to be utilized in high-retum activities. First, rhere must be rapidgrowth of labor demand relative to supply and, in particular, of demandfor skilled labor. Second, the labor market must perform well. It must beefficient, flexible4 and responsive to changing conditions so as to ensurethat 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 thatrapid educatonal expansion may result in labor market problems dtat

erode the benefits of human capitml accumulation and have a seriouslyadverse impact on the growth of output and wages.

In our assessment of why human capital in Fast Asia appears to havebeen efficiendy utilized and why changes in labor market outcomes havebeen so fuvorable, we focus on two facors: the dynamics of labor supplyand demand and labor market performance. We also examine the inter-actions between these two factors. Then we consider whether there areany common denominators in thae policies of East Asian governmentsthat 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 realwages (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 thereslt of slower growth of supply and more rapid growth of demand forlabor.2 For a given rate of growth of labor demand, the slower the rate ofincrease of labor supply, the bigger the increase in wages likely to resultfrom competitive pressures in the labor market. One measure of thedampening influence of labor force growth on the rare of increase of realwages was obtained for a cross section of twenty-two economies in the1970s; a 1 percent increase in labor force growth resulted in a 1 percentdecline 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 11Increase 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 economiesgrew, 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 romughlythree times that race. Only in East Asia have populaton growth rares, andthe 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 EastAsia markedly reduced the rate of growth of the school-age populationto levels well below those in other developing regions, thereby easing thefinancial burden of maintaining education enrollment rates. The earlydemographic transition also reduced, with a lag, the rate of growth ofnew entrants into the East Asian labor force. The annual rate of laborforce growth during the 1980s was 2.6 percent in Sub-Saharan Africaand Latin America and 2.2 percent in Soudt Asia. In East Asia the ratewas 1.8 percent (see table 6.1). The labor force growth rare in Singaporeis only 0.8 percent, about the same as that experienced by the high-income economies in the nineteenth century. In Sub-Saharan Africa therate of labor force growth continues to increase. In Est Asia, despite in-creases in the particpation rates of women, labor force growt rates havebeen 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 1.8Indonesia: . -~ -; ; - -2.4 - 2.2t'rKc, Rep. of. 2.7 1.-9Malaysia - 2.9 2:6..Sinore 1.9 0.8Thailand -; .- 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 beenslower in East Asia than in other developing regions, but the growth oflabor demand has been faster. For the period 1960-90, the rates ofgrowth 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 becausegowth of employment has been unusually slow esewher Widt the ex-ception of the 1980s, during which output and employment stgnaredbecause of the macroeconomic adjustments made necessary by debt-service problems, the pace of industrial sector employment growth inSub-Saharan Africa, Latin Arnerica, and South Asia was more than dou-ble the pace in the currently high-income economies during the perod1880-1900 (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 hasbecome 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 in1980 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 inthe 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 33.11983 32.1 - 45.6 33.1-1984 32 - 46.5 33.71 l985*. :-- 34.3ta;; - - - |--.46.5 34.21986. 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 technologicallysophisticated, skill-intensive goods as rapidly rising wages of unskilledlabor 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 EastAsia? Since the demand for labor is derived from the demand for output,the explanation must be sought, in the first instance, in those factorsthat contributed to East Asia's rapid output growth: high rates of savingsand 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 hasbeen more labor demanding in East Asia than in other developing re-gions. As we shall see, the dynamism of agriculture and export-pushstratcgies contributed to rapid increases in the demand for labor and torising 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 bearsthe responsibility for melding the changing needs of the economys pro-

265

iR_AC LCE

ductive apparatus with the changing skills and preferences of labor forceparticipants (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 toshift 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 toneeds, improving resource allocation across firms and industries andthereby 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 unpalatablein the short run or to particular groups (for example, wages below whereworkers believed they ought to be, educated unemployment, or the em-ployment of educated workers in lower-level occupations). A relativelyhigh level of efficiency in the allocation of labor was achieved by allow-ing wages and employment to be determined largely by the interactionof those supplying and those demanding labor services, rather than bygovernment 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 forlabor, whereas elsewhere there was a greater tendency for wages to bepushed up artificially. Earnings growth was determined more by thegrowth 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 insome sectors to rise well above what workers could earn elsewhere in theeconomy, 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 thegrowth 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 labormarkers are highly segmented, workers with the same ievel -of skill earnnonagricultural wages that exceed agricultmral wages by 150 percent

266

USING -R

(Fields 1992). Still larger intersecroral wage gaps have becn observed inEast 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 wagesin the nonagriculturil sector (see figure 6.2; also see Ranis 1992). Thesmall 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 thatit reduced the incentive for educated workers to conduct a lcngthysearch 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 toprovide more high-wage jobs than justified by the derived demand forlabor.8 Thus, the share of the public sector in the increase in total wageemployment in recent decades is a crude indicator of the magnitude ofpublic sector surplus labor, and hence labor market segmentation.9 Astable 6.3 shows, in all the Sub-Saharan Afrcan, Latin American, andSouth Asian economies for which data were available, employment inthe public sector grew more rapidly than wage employment in the pri-vate sector. The median share of the public sector in the increase in totalemployment 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 EtAsian 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 publicsurplus employment has been kept in check and stronger evidence thatthe scope Sor lobbying for make-work jobs has been limited. The needto 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 thepublic enterprise deficits that they would finance.

- In- Malaysia the share of the public sector in the increase in tota wageemployment is as high as in other developing regions. The increase in

267

1Figum 6.2 Industrial Wage GapsWage 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 wageemplymentLnAmerica aznd C"bantbe:

* -: ~~~19733 1.4 00. 0.3 92CostaRica ' 1~~973-43. :6 , 28353

Panama *':, :1963-82 . ~7.3 ,.,18 2. 45Peru ~~~~1970-44 * 6.61- 061.4

Trinidad` :1970-84 4 12 ,195

Venezuela ;1967-82 ~~~~~51 3Ž4 '. 37 27Middle ase and Not nfipc

Egjp . , . 1~~966-76f 23-. . 103n*nj.I

tA -c

Gh~~~~ina ~~196048-734 -5.93-4.,Keny 196381 64.2. . 67

Tanzania ~~~196246. 6.1' --3.8' 1.619LZambia -1966-080 7.2 -r0943

K India .~~~~~~1960-80 42213.2 7SnI 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 MalayMajority. 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 thatsewed 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 generallymore willing to accept flexibility of wages around a rapidly rising trendbc&ause downward adjuLstment implies a: slower rate oinrease rtethan 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 enterpriseperformnance, 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- oreconomywide unions and to ensure that wage bargains were set at theenterprise level. We have suggested chat the rapid growth of wages con-tributed to the explanation of why, in contrast to odchr regions, wageswere not maintained above market-dclaring levels in E-ast Asia. An alter-native, or perhaps complementary explanation is that governments, incollaboration with employer repressed laibor markets.

Wage-seuingpractices One manifestation of the link between rapidgrowth and wage flexibility is the apparent willigness of workers, espe-diaily in the northeastern economies of Asia, to accept wage structuresthat include a substantial element of-performance-based bonuses. This'practice links compensation to enterprise, rather than wholly individuaperfbrm2nce. It is comnmon for current cash compensation to have twocomnponents-a base wage and a bonus (Shah and Mathur 1992). Tothe extent th at "special cash payments" are more responsive to changingeconomic circumstances than are base wages, thiis arrangement couldhelp 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 positivelyand significantly with such indicators of aggregate economic conditions

as cnrporate profits. By contrast, there was no significant relationshipbetween 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 oftotal current compensation (15 percent or more in Japan and Korea) orof corporate pois(40 percent or more) rend to be large, the elasticityof bonus payments with respect to profits is small. In Korea, most bonuspayments have become contractually linked to the base wage, and inJapan 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 (wagerepresion) or restraint on organized labor (labor repression) with indi-

rect restaint on, wages. Wage repression was dlearly nor a necessaryingedintin -the East Asian recipe for economic sucesThgorn

7-7 --

USING RESO

mernts of Hong Kong, Japan, Korea, and Taiwan, China,- did not represswages, ant yet in these economics output, exports, and ultimately wagesall increased at rates well above the international norm. Theseeconomies were adept at adjusting to changing comparative advantageas labor abundance gave way to labor scarciy. At the other extreme, thegovcrnmnent of Singapore repressed wages, at times quite severely, andthe consequent costs of labor shortages appear to ha've been high, lead-i ing the government to abandon. the policy (Fields 1992; Freeman1992). 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 thesame pace as real GDP. Nevertheless labor relations in Korea have beendeteriorating in the 1990s, so that labor relations problems serious

enough to threaten economic growth have arisen, partly in reaction torepression of unions (Freemanl 1992). In Singapore during 1972-79,

r real wages increased at less than one-quarrer the rate of increase of realGDP. This divergence was the consequence of the governmenes aggres-sive, and quite open, intervention on the side of wage repression in themarket for labor1 2 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 atighter 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 inFields 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 ofoutput, 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 awayfrom 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 thelabor market, the govemment of Hong Kong is among the least inter-ventionist in the world. Despite labor laws that are favorable to unionformation and action, union membership in private sector fims is low;unions engage in relatively litde collective bargaining and are generallyconsidered to be "weak." Perhaps because wages were rising rapidly inHong Kong, workers did not fcel the need for collective action. HongKong 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 andeconomic costs imposed by labor repression (Freeman 1992).

Japan's expenence with labor unions is another counterexample to thebelief that East Asian success required goverrnment repression of labor. TheJapanese example may be the more telling of the two because in Japan, incontrast to Hong Kong, union membership is high and unions are, at leastpotentialy, quite strong. Yet unions did not extract economic benefits fortheir 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 hanwell 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 earingaccountecd fior a higher proportion of investinent finance than orherwisewould have been the case, redung reliance orn underdeveloped capitalmarkets. The probable net effects, therefore, were higher levels of invest-ment, greater compeitiveness in international markets, and faster rates ofgrowth of output, employment, and ultdmately eanings. 14

Superior labor market performance in East Asia thus contributed tosustaining the extrordinary dynamics of labor demand we documentedabove. Moreover, by not allowing the price of labor to become inflatedrelative to the price of capital, wage restraint encouraged the use of morelabor-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 capitaLEfficient use of skilled labor also contributed to the high levels of

publicsector savings in most of the HPAEs. If the govenment capitulatesto 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 arelikely to divert scarce savings fiom productive investments, lowering therate of capital accumulation and stcadily eroding the grwth potential ofthe economy.15 For a time, Malaysia was an East Asian illustration ofthis point, but there are manry more examples outside the region wheregovernments 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 asubstantial 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 ofmakei-work jobs impedes the process of hunan capital deepening that iscen-tral to education's contibution to economic growth. Had the labormarket been allowed to functon without intervention, these sameworkcrs would have entered lower-paying occupations, and the consc-quent mcrease in the educational level in those occupations would havecontributed 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 theoil crises of the 1970s, was ofien quicker and less painful in the HPAEs

(Mazwrndar, forthcoming). Ifthepriceoflabordoesnotadjustquick-lyandsmoodily 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 greaterin 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 Marketsand AllocaioDn

-T-L MARKETSSPERFORM THEIR ALLOCATIVE FUNCTION IN

-ways: they aggregate savings, allocate funds to competinga_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 inchapter 5 that contributed to the deepening of financial sectors alsocontributed to better capital allocation. In addition to private capitalmarkets governments in East Asia, like governmecnts in many developingeconomies, have attempted to allocate credit directly to high-priorityactivities. Elsewhere, directed-credit programs have been catastrophicfailures. In some HPA-Es-parricu.larly Japan, Korea, and Taiwan,China-these programs have caused less damage to capital allocationand 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 thendescribe the nature and magnitude of directed-credit programs and as-sess their performance. We show that, where these programs appear tohave 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 strictperformance criteria were used to ration fumds. These preconditionsfor 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 6.3compares East Asia's ratio of M2 (rime plus demand deposits) to GDP,which is often used to measure the depthi of the financial sector, withthe M2IGDP ratio for a group of low- and miiddle-income economiesand the OECD econom-ies. The relatively greater financial depthi in EastAsia far surpasses most developing economies and compares weli evenwith the OECD economnies.. A broader meaure of financial depth, theratio of financial liabilities (M3) to GDP?, shows a similar trend.

Korea is unusual in that its ratio of M2 to GDP? is low comparedwith many developing economies at similar levels of per capita m-come, and these ratios have inacrased more slowly than Korea's incomegrowth would predicc. The largest incremental change was evidentafter the inte-rest rate adjustments in 1965. However, M2 fails to cap-ture the presence of a vast nonbank financial sector. The rapid growthof 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 100M2/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. andVenezuela. Earopc and Middle East indudes Algeria, Egypt, Greece. Morocco, Portugal. Syria.

Tunisia, and Yugoslavia. Sub-Saharan Africa includes Cameroon, Congo. C6te dt'lvoire, dheGambia, 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 andLevine 1993).

Fnancial assets in industrial and developing economies, inclingEast Asian economies, are held largely in deposits with banks and otherfinancial institutions and to a very limitedt extent in government or pivate securities (World Bank 1989c). The public policies described inchapter 5 that have- emphasized the development of banks and othersavings institutions may partly explain this composition, but there alsoare 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 theRole of Government

The lack of perfect information that characterizes all markets is mostserious 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 notedin chapter 5, bond and equity markets are slow to develop. This placesincreased 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 cannotdiscern perfecdy the probability of default, they cannot simply aUlocatecredit to the highest bidders-those who promise to pay the highest in-terest rates. Rdther, they must choose among agents willing to pay thegoing rate those that seem most likely to repay the loan. This allocativcprocess is, in effect, a form of rationing. Moreover, it gives banks somecontrol 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 bymore than lust interest rates. The namre and amount of collateral be-come key. Yet despite efficiency-enhancing marker responses, capitalmarkerS remain imperfct. Market power by banks may raise inrermedi-ation margins, resulting in monopoly rents for the banking sectorBanks also face incentives to misallocate resources; underfunded banksrend to finance excessively risky projes. Social returns on investmentmay differ markedly from the private remrns to investors, and socialperceptions of the riskiness of investments may differ substantially frompnvate 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 bankingspreads ro limit monopoly rents. While these measures are by no meansunique to East Asia, the HPAEs have been unusually active in exercisingthem. In addition, most HPAs have gone a step further and attempted..to direct credit to specific sectors, industries, and even firms. We discussthis 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 avoidexcessive lending in some areas or to increase lending in others, direcdyinfluence the allocation of credit by banks. In chapter 5 we argued thatEast Asian govemmcnts bave emphasized the safety of banks throughboth regulations that monitor and enhance the solvency of banlks andmeasures that avoid financial disruptions. These have positive effects onresource allocation. Solvent banks have fewer incentives than under-funded ones to take excessive risks in lending. Solvrent banks can alsolend more.

Portfolio requirements to enhance prudent lending-such as limitson lending to real estate projects- -also may guide credit away fromsome 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 suchlending increases the riskiness of banks' portfolios but also that it divertslending away from manufacturing. Thailand, for example, does notspecifj how much of their portfolio banks should lend to industry, butit provides lending guidelines that favor lending to industry and agri-culture and limit lending to other activities, induding real ,-r e. andsecurities.

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 wereheld in publicly owned commercial banks. In addition dte governmentsofjapan, Malaysia, Singapore, and Taiwan, China, have discretion overa large (and growing) prop ,rtion of total deposits iri postal savingsand/or provident and insurance fi^nds. Some governments have alsobeen able to complement those resources with budgetary allocations orborrowing 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) ofJapan 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 originateprimarily 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. 71960 -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 8551981-85 ~~~~~~ ~ ~~2.2 NA61

_1971-75- N' 44308½ 41976'-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 comeriealbanks 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 4 J pbi* * ~~~~~~postalsavings ~provident awntrance commercdialad

*SingPapre1971-75 2.1- 21.6 -23.7

1976409. 30.0 39.419814~5 -1.0- 34.6 456'

1986-90 ~~~~~12.7 34.9- 47.6

-. 1971-75- 10.8 A67978.71976-80 ~~~~~-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 -13.81976-80 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 havesimultaneously reguated deposit and lending rates and consequentlythe 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 (seetable 5.7). Given the rapidi growth of dhe financial sysems in the HrAsand 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 isnotable that, in Indonesia and Thailand, regulated intermediation mar-gins were substantially higher than those after liberalizarion of entiy andinterest 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 litdefrom other developing economies. However, HBPAEs, particularly Japanand Korea, have uscd unusual mechanisms that have helped to increaseloan repayment rates and overall success of directed credit. They havelimited 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 ofdirected-credit interventions. First, governments can direa credit tospecific firms or ind ries. Second, they can direct credit on the basis ofbroad 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 amongethnic groups. HimAE governments have engaged in all three.

Ta?eting firms or indutdes. Japan cuinng its postwar reconstructionand Korea during the 1970s directed substantial credit to specific sectorsand firms, mostly in heavy and chemical industries (Has) (see table6.5). Japan's FILP accounted for about a third of new equipment lendingin the 1950s. Between 1973 and 1981, Korea's policy loans were about60 percent of the total loans of its deposit money banks. Most ofJapan'spriority lending targets through the early 1960s were industries associ-ated with large optimum scales and increasing reurns to scale; so wereKorea's during the HCi period. Indonesia and Malaysia, conversely, haddisappointing experiences with credit interventions targeted to specificindustries in the 1970s and early 1980s and abandoned the schemes infavor of more functionally directed credit Hong Kong and Thailandhave not been active in using credit instruments to push selectedindustnes.

In contrast with many other develop'ing economnies, where directedcredit often went to public and quasi-public enterprises, direcred creditin 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. , iV4965. 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" 19.48:12'.~~~~

~~~~~~~~~* .. ;,..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 enterprisesin 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 bankprovided about 75 percent of its industrial loans to fbur state-owned en-.terprises. UJhfortunatelyv, t-hese public enterprises often did not performwell economically or fina-ncially. In East Asia, Indonesia, Malaysia, andSingapore have directed credit to state enterprises, but the proporions oftotal acrdit were not p tversistently Ehigh the parastatals tended to perfbrmbetter financially, and thae intLerest rate subsidies were nor large (except inIndonesia).17 1Korea 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 subsidizedexport credits, often through the central bank redis onSt te (see

table 6.6). Exporters from all industries' have had access. Small andmedium-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 throughthe 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 percentagepoints lower than that applicable to general trade bills).During this period, the HOy directed 10 percent of its loans toexport finance (which financed 50 percent of toal exporr-related bank Icnding).

Kara, Rep. of Offers concessional rediscounting of prreshipment creditbased on Ictters ofcredit (tcs). The differential betwecn loanrates of this cxport credit and general loan rates was between3 and 20 percentage po;nts until 1982, when this difFerencewas eliminated. In 1984 the dififrcntia was restored ar 1.5percentage 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 hasrediscountd export bills fbr postshipmcnt financing

Since 1979, the ECR refinances peshipmenr export acedits atsubsidized rates for cligible exports, based on lecters of crediLA domestic leter of credir was also introduced to enible thefinancing of indircct expores.

Singapore In 1975 the government established a rediscount facility forexports and re-exports.

The Export Credit Insurance Corporation provides insurancefor exports by Singapore.

Taiwan, Cbina Offers firms short-tcrm preshipment loans based on LCs.After products are shipped, firms are cligible for loans basedon LCs. documents agait acceptance of payment, and ship-ment documenrs. In addition, the Central Bank provides aspecial discount rate to designated domestic banks for exportloans, 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 loasbetween 1970 and 1980, and it dropped to I tO 2 percentagepoint 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 reprcssedfinancial 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- adisci&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-iutvesm5 nt 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 perveitnrernrt > 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 inIndonesia, where they fiiled and were discontinued, and Korea, wherethey have been more successful. As we argued in chapter 2, the size ofthe subsidy was probably less relevant to firms, especially to small andmedium-size exporting firms, than was the automatic access. Allocationwas 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 verilfrperformance. 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 smalland mnediun-scale enterprises. Small and medium-size enterprise lend-ing through specialized development banks is a growing proportion ofJapan'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 hasdirected 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, whichhave few other directed-credit programs, provide subsidized credit formass housing.

KeepingSubsidies to Directed CmditSmalL Compared nith most other de-veloping economies, the subsidies for directed-credit programs in mostHPAEs 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 alimit on subsidies. Singapore also ensured that the Development Fundfinanced by the pension funds was self-financing. It was also the casethar, with the exception of the HMc period in Korea (1973-81), the moreprudent 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 typeof 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

FIPAESIIndonesia, 1981-83, liquidity credits -1.7-4.0Japan, 1951-60 0.5-3.0 3.1-4.6Korea, Rep. of, 197040, industry -2.7 2.9

cxporrs -6.7 2.9Taiwan, China, 1980-89, industry 1.9-3.9 4.6

1984-85, exports- 1.5 4.6

QOter developing economiesBrazil, 1987 -23.5Colombia, 1981-87, industry 1.5 13.5India, 1992 - -2.5-4.0 7.0Mexico, 198748 -24.0 6.0Turkey, 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 higherreal interest rates than in other developing region&s Given capital's highinternational mobility, we conclude that interest rate subsidies in theHPAEs were lower than in other regions.' 9

The HPAEs generally have smaller shares of their credit systems subjectto 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 directedcredit as a proportion of nongovemment credit was 54-68 percent inthe late 1980s and I990s (an internal World Bank report). In Brazil,about 80 percent of the total average stock of credit was directed in1986-87. Colombia's directed credit to indusay in 1984-87 was on av-erage 56 percent of total credir. By contrast, in EastAsia, Korea's policyloans in the HCI period were more than 50 percent of total credit, andJapan's never exceeded 15 percent of total credit. Indonesia briefly had48 percent of its total loans in priority liquidity credits in 1982, but itreduced these rapidly following the liberalizations of 1983 and 1991.

Moreover, dircted credit is becoming less importanL WhiLe earlierefforts of govremments in East Asia concentrated on developing the

285

' S!WNMIRACLE

banking system and guiding its lending decisions, more recent policieshave been directed at deregularing financial markets: Japan started in the1970s 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 thechapter). The majority of HPAEs deregulated their interest rates duringthe 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 Asianeconomies has been gradual and was meant to minimize disruption inthe manufacturing sector.

Institutional Mechanisms for Selection and Monitomng. In contrast ro most

other economies that have rried but fiiled to use directed credit, Japanand 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. Moreimportan-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 acorrespondingly low level of loan losses.

In Japan, officials working within government sectoral guidelinesoperated 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 betweendevelopment and commercial banks ensured continuous monitoring ofthe performance of the borrower and enabled development banks totake early action if loan repayment was in arrears. Coupled with Japan'soverall economic success, effltive pre-appraisal and monitoring re-suited in very low loan losses. The Japan Development Bank, the maingovernment 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-termnlending and had a more diversified portfolio. The JDB's superior per-formance may be partly explained by its greater reliance on collateralsecurity, 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, discussedin chapter 4, supporred this close consultation (Cho and Hellmann1993). Enforcement of government goals--primarily the imperative toexport-was achieved by rewarding successful firms with continuedcredit support while reducing credit to less successful firms. Such con-tinual and pragmatic review and adjustment of policy implementationreduced, but did not always eliminate, the moral-hazard costs of gov-ernment intervention.

Other HPAEs fiall along a spectrum between Taiwan, China, which hasutilized mechanisms similar to Japan and Korea and has had similar suc-cess, arnd Indonesia, which was plagued by the problems discussed belowand has largely discontinued directed credit.

In contrast with Northeast Asia, credit allocation decisions in manydeveloping 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 thatcould be obtained from subsidized credit were a strong incentive to cor-rupt practices. In smaHer economies, credit allocation to specific sectorsbecame credit to individual firms, because of the limited number ofcompanies 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 largeand politically powerful firms or state-owned enterprises were charac-terized by weak monitoring and follow-up on debc recovery, producinghigh loan losses and eventally large-scale failure of directed-creditprograms.

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 financialmarkets, combined with prudent restrictions on lending by banks forreal 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 developmentis 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 ofcredit? Here tie record is mixed. NWe conclude that credit directed toexporters was a positive factor in accelerating growth. All East Asianeconomies provided automatc, generous access to credit (or creditguarantees) to exporters with varying degrees of subsidies. Japan,Korea, and Taiwan, China, went a step further they used exports as acriterion to determine whether priority firms, regardless of the marketfor the actual priority investmenr, continued to receive credit.21 Othermore specific directed-credit programs had both positive and negativeresults. In general, the economies with strong civil services and highlyprofessional public financial institurions-Japan, Korea, Singapore,and Taiwan, China-apparendy had some success in implementingcredit policies to promote priority industries and firms. Key to this suc-cess was the use of economic performrance criteria and the monitonngof 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). Thestudy 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 financialsystem, accounting for less than 10 percent of total loans. The studyidentified three characteristics ofJapan's policy-based lending: relianceon 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 forchanneling funds to the targeted sectors by selecting individual firmsaccording to purely economic criteria.

By comparison, Korea's use of directed credit during its economictakeoff in the 1960s and 1970s was more extensive, and interest ratesubsidies were larger. As in Japan, the governmnent in Korea consulteddosely with businesses and banks, attempted to monitor perfbrmancedosely, and relied on a competitive business environment to provide anobjective measure of performance. Accordingly, policy-based lendingprobably contributed to the rapid expansion of Korearn indusy, partic-ularly manuFactured exports (Cho and Kim 1993). Most of the directedlending to priority industries in Korea translated into investments inequipment. 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 creditswas substantial-and likely to increase as development proceeded. Di-rected credits burdened banks widt nonperforming loans, interferedwith 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 rateson deposits, thereby funding the subsidies for loans to corporations. Inaddition, relatively easy credit in international capital markets andKorea's strong relationship with two key lenders, Japan and the UnitedStates, 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 opennessof capital markets in most developing economies and the heightenedcompetition in international capital markets suggest tiat financial re-pression and directed credits are now less viable options for developingeconomies. 2 2

To argue tlat directed-credit programs in general reached their tar-get groups and that loans were effectively monitored does not establishthat they contributed positively to growth. To do this, microeconomicevidence is needed to establish two facts: government lending pro-moted investment and did not simply displace investment from othersources, and the social returns on the incremental investment ex-ceeded private returns, justifying the targeting of specific sectors. Suchevidence is rare-even for the northern-tier HPAEs-but two recentstudies attempt to address these questions for a sample of firms inJapan.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 highpotential for externalities due to innovation and learning-reaches thefollowing 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 grossinvestments in fLxed capital assers.

* Government lenders withdrew credit once firms had earned accessto private credit markets and, in turn, extended credit to new bor-rowers rather than preexisting borrowers, consistent with (thoughcertainly niot concdusively establishing) the daim that credit wasbeing allocated according to objective lending standards, ratherthan on thc basis of political considerations.

* The characteristics of firms receiving increasing government credirindicatc that assistance was targered toward firms most likely togrow, 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 increasinggovernment financing.

The authors conclude: "Possibly because of the institutional structuof these lending programs, it appears that economic logic rather thanpolitical capture motivates policy-based lending decisions. Govemmentcredit is withdrawn quicldy from seasoned firms, is targeted towardgrowing, R&D intensive firms, and produces substancial increases ininvestment 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 andHimmelberg's conjectare. Horiuchi and Sui (1992) examine the role ofthe 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 consistedof 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 privatebanks, suggestng that JDB's loans did not crowd out private banklending.

* JDBs loans to firms preceded the firms' increased borrowing fiomprivate banks and improved the access to credit of firmns that werenot 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 encouragedcomplementary lending from private financial institutions with a conse-quent net increase in investment. Moreover, policy-based lending wasmost successfiul wlhen it was directed at nontraditional borrowers. TheJDB's loans stimulated greater lending by private banks to firms not affil-iated with main banks but did not have the same impact on firmsaffiliated with main banks.

Another study of credit programs in Japan and Korea based on ananalysis of the objectives and outcomes of the programs adds an addi-tional element, noting the positive role of coordination achievedthrough govrmment lending programs. During the 1950s the bulk ofdirected credit in Japan was given to "basic industries"-electricity, ironand 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 subsequendypromoted industries whose expected spilover effects on the economywere large, for example, machine toolS (JDBIJERI 1993). During the1970s, Korea channeled most ofics directed credit for fixed investmentsin manufacruring to the HCI industries, among them such producergoods industries as steel and machinery. Using a framework not dissim-ilar to the contest framework oudined in chapter 2, the study reaches asimilar conclusion: the coordinating role of government in Japan andKorea has diminished over time, and hence the effectiveness and desir-ability of using credit programs to improve coordination among privatefirms are diminishing as well (Cho and Hellmann 1993).

But many other directed-credit programs, even in the HPAEs, did notmect their objectives. In fact, the track record of most directed-creditprograms in Indonesia, Malaysia, and Thailand was so poor that theirgovernments, 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 andmarket interest rates and smaller volumes of directed credit relative tothe size of the financial system meant that financial market disrortionsin the HPAEs, even from unsuccessful attempts to direct credit, were notgenerally 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 ofthe 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. Withneutral incentives, it was argued, resources would flow to sectors inwhich the economy was most internatonally competitive (Corden1971), and gains in technical efficiency (or "X-efficiency") would resultfrom a more competitive environment (Nishimizu and Page 1991).Studies in the late 1960s and early 1970s convincingly demonstratedthe extreme bias in incentive structures against exports and agriculturein most developing econonmies and the resulting high costs of import-substituting industrialization.4 Recent theoretical work has arguedthat, where market power, economies of scale, learning, or externalitiesare 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 studiesof the costs of protection, provides very litde evidence conceming theimportance of these factors or the potential benefits to grwwh of depar-tures from trade neutrality25

Proponents of trade neutrality and intervention both dee the recordsof the high-performing Asian economies as evidence supporting theirviews. As we noted in chapter 1, international trade is important in allthe HPABs and is the factor most consistently correlated with their suc-cess. Balassa (1991), Krueger (1993), Hughes (1992), and others arguethat openness to international trade, based on largely neutral incentives,was the critical factor in East Asia's rapid growth. Conversely, advocatesof 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) duringtheir periods of rapid growth. They argue that the HPAE govemmentssuccessfully intervened to change comparative advantge (Amsden1989; 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. Duringtheir heavy and chemical industries programs, Japan and Korea were themost active HPAEs in promoting individual industries and sectors. Sing-apore and Taiwan, China, have also actively provided incentives fortednological upgrading. Malaysia had an HCI progrm reminiscent ofJapan's and Korea's, while Indonesia has attempted to leapfrog fromlabor-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 mightjustif government efforts to promote specific sectors. We then describethe trade and industrial policy regimes in the HPAEs and compare theirdomestic 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 promotespecific industries generally did not increase economywide productivity.On the other hand, the evidence shows that broad government supportfor 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 usefulto review some of the arguments specifically as they apply to trade andindustrial policy 526 Among the many reasons for discrepancies betweensocal and private returns cited earlier, four stand out:

X Interdendent investments and economies oficale. Increasing returnsfom scale and capital market imperfections may mean that investmentsthat could be internationally competitive at optimal scales will not beundertaken. This is especially true with large, interdependent projectsfor 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 coordinatingthc actions of buycrs of technology and trying to increase competitionamnong sellers, governments can appropriate more of the surplus associ-ated withl the transfer of technology than they otherwise could (Stiglitz1993a).

e Pecuinay externalities. Pecuniary external economies arise if, as thesize 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 andgreater specialization of individual skills. When economics are small,currcnt prices may not convey adequate information about prospectivelower 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 productleading to a better or cheaper good than is available internationally. Inthis case, the source of the externality is the nontradability of some typesof inputs or knowledge-otherwise the improved method or productcould be obtained from international suppliers.

* Learning. Externalities related to learning have traditionally beenidentified as important sources of mark-et fai]ures in developingeconomies. When firms gain knowledge of production from other firmswithout incurring cosrs, real externalities are present.28 Because of in-complete appropriability of knowledge, individual firms may spend lesson obtaining production knowledge than is socially optimal.3 9 Exter-nalities due to learning may also be conferred on other firms by the firstentrant. These indude the demonstration that the sector is physicallyand economically feasible and the leakage of information on technologyand 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 neededto lower costs and become viable.30 Industrial policy advocates take theaTgument a step fiuther. Because they associate learning with capital-and knowledge-intensive industries, they advocate a rapid shift inindustrial structure toward these activities, even if they are not intema-tionally competitive at present levels of scale, knowledge, and fiactorprices. In this view, the short-mn aflocative costs of establishing intema-tionally uncompetitive industries will be outweighed by the long-runbenefits 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 amplydemonstrate, prorection does not enisure that the promised learning andeconomies ofscale actually materialize.

Trade Policies in the HPAEs

Most IHPAES began industrialization with a protectionist oricntationand 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 traderegimes: they granted exporters duty-free imports oF capital and intcr-mediate goods while continuing to protect consumer goods. Exportprices were set in the international market and were often substantiallyless 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 ofcost 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 industrializationstrategy, similar in many dimensions to those of Argentina, India, andocher less successfid economies. As late as 1968, effective rates of pro-tection (ERPs) in Japan were still quite high and exhibited the cascadedpattern from raw materials (low) to consumer products (high) that isrypical of most developing economies (see table 6.8). Unlike manyimport-substituting economies, however, there was surprisingly highprotection of machinery (final producer goods), confirming other re-search indicating that the Japanese atuthorities engaged in an intensiveeffort 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 andKiyono 1988). Quite high levels of protection were afforded sectorssuch as iron and steel and nonferrous metals as late as 1970. Protectionin 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 andWestphal (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 -3.0Producer 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 7.6Final 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*m 1 :*1963 1968 1972

Manufacturin 32.3 24.2 14.4Tattiles 54.3 282 .18.6

Spinning 27.1 12.5 I.0Weaving -44.6 30.5 15.5Products 72.8 32.8 22.4

Wood products 14.0 25.6 16.1Pperandpulp 9.7- 18.0 11.0Publishing -16.7 1.0 . -0.9Learher and rubber products . 30.9 21.8 12.3Chernicals 33.4 17.7 8.8Perroleum and coil products 19.5 14.5 . 7.1Nonmeallic minerals 22.2 15.7 8.1Iron andsterl 30.1 30.0. 17.1Nonrfrous metals 30.4 34.1 22.1Metal products 13.8 -19.9 9.9Machinery 36.7 20.0 7.7.

Genemal machinery 23.0 14.5 8.7Electrical 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 andnontariff barriers. Table 6.9 presents a summary measure that combinesthe effect of nominal tariifs and non:ariff barriers.33 The figures spanthe 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 theRepublic 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 92.4V. Paper, paper products, and printing

and publishing 39.5 54.7 54.5 78.4 88.8 90.8VI. Chernicals, perroleurm, coal, rubber,

and plastic products 47.9 57.1 58.4 65.9 75.2 80.0VII. Nonmetallic mineral products 41i. 77.0 76.8 89.1 89.2 91.7VIII. Basic metal products 44.7 73.3 74.8 86.2 90.1 92.0IDC Meral protucts, machinery, and

eqwupment 41.0 59.3 55.4 63.6 69.7 77.8X. Other manufiacturing 31.8 39.5 38.7 58.2 65.4 76.1All 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 76.8Heavy 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 bythe current price valuc of dom:sric producton for respcrive yer

Sowcan:Sakong (1993).

liberal by 1975. Nevertheless, even by 1983, when Korea's success hadbecome an established fact, most sectors were still protected by somecombination of tariff-s and nontariff barriers. While Korea utilized a va-riety of instruments, especially export targets and rebates, to ensure thatexporters faccd international prices for their tradeable inputs, there wasconsiderable 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. Aslate as 1972, a significant percentage of items were subjecc to nontariffbarriers, and two-thirds of potential imports were subject to nondinaltariffs in excess of 30 percent. As late as 1980, more than 40 percent ofimports 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 thatwhile 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 tradestrategies 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 sufficientlyhigh 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 exceptHong Kong and Singapore, domestic prices in these economies arecloser 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 88Bev:rages 20 38 -22Tobaco 125 - -26Textiles - 58 15Apparel 400 45 6Wood products 33 38 82Furniure 40 84 43Paper and paper products 140 66 29Printing and publishing - 26 -9Industrial chemicals 230 32 121Peroleum refineries - 22 5Rubber products 60 129 14Plastic products 265 312 163Iron and sted 84 63 289Fabricated metal products 40 26. 30Nonelectrical machinery 1,600 89 19Electrical machinery 130 4 12Transport 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 Yearspercent)

Seaor 1975 1987 1990

Food, beverages,andtobacco 336.2 122 124Textiles, cloth, and foonvear 231.8 102 35Wood products -12 25 33Paper and paper products 873 31 20Printing and publishing - -

Industrial chemicals 28.4 14 13Petroleum refineries - -1 -IRubber and plastic products 426 57 48Ironaandstcel 18.2 13 10Fabricated meaul products - - -

* Nonelectrical machinery - -Electrical machinery - -

Transport equipmentNon-oil : -1 1

-Cement 63.6 60.2 53.6Manu&crmnrng Total

(eclduding oil sector) 74.1 68 59

-Nor ailabloNan The cflicrfie rare of protection is the pecennge by wcb value added at

domestic prices exceeds valuC added arworld pricesSowr.cBhmacharyaaand Pngestu (1993).

jusred for the presence of nontariff barriers are lower in the HPAES thanin 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 inother regions.

Table 6.13 presents one of the few systematic attempts to comparenominal taiff rates for a broad range of developing economies (Erzanand others 1989).34 The authors have also atrempted to adjust for theeflc of quantitative restrictions on levels ofprotecion. (We have addedcomparable data for Taiwan, China, and have computed the averagenominal tariffi for Korea, Malaysia, Thailand, and Taiwan, China.35)Nominal tariffs in the HPAEs are lower than for any other grouping ofdeveloping economies except the island economies of the Caribbeanand 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. AprilSecor 1971 1974 1978 1981 1983 13984 1984 1985

AveragesAll 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 66.3Textikl prmducts - - - 248.5 144A 127.8 108.0 118.4Leacherppmducs - - - 151.3 199.7 199.0 167.0 152.7Wood products - - 60.7 53.7 55.0 49.7 62.0Paper and pulp - - - 39.5 43.0 43.0 39.0 53.5Chemical products - - - 54.3 52.9 51.9 46.6 44.5Rubber products - - - 48.7 46.0 46.0 39.3 42.0Other nonmetal products - - - 99.8 102.7 102.3 89.8 108.5Metl products - - 633 62.0 61.9 53.9 70.9Machinery - - - 14.1 23.6 23.6 21.0 29.3Consumer 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 pricesSwxrc Brimble (1993).

Table 6.13 Average Tariffs and Para-tariffs, by Region 1985

Tranip Ceatral Soudt North Saharan West Otler DeveAlpg Allpara-ta f Caribban America America Africa Africa Asia Asa PEAt' aglow

Tarffi

Economiesunweighlted 16 23 34 29 32 7 36 19 26

Economies

-import-weightedb 17 24 38 30 35 4 22 17 24

Tariffpluspars -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 domesticprices. Figure 6.4 shows an index of outward orientation based on in-ternational comparisons of price levels and price variability for the HPAEscompared with other regional groupings (Dollar 1990). The HPAEs as agroup are more outward oriented than other regions; their relative pricesare doser to and more consistently related to international prices. Ofcourse, there is substantial variation among the HPAEs. As would be ax-pecred, Hong Kong, Malaysia, and Singapore rank in the top nvodeciles 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 thetop third. Japan, Korea, and Taiwan, China, rank in the fifth and sixthdeciles, below such developing-economy compararors as Brazil, India,Mexico, Pakistan, and Venezuela. This is consistent vith rhe evidencethat the three northern HPAEs intervened far more frequendy and sys-tematically in their economies than the southeastem HPAEs (Dollar1992). While any large, multi-economy effort at real price comparisonsis subject to methodological and empirical criticism, the results arebroadly indicative and consistentwith other evidence: EastAsias relativeprices of traded goods were closer on average to international prices thanother developing areas.

Openness to Forin Technolo

An important fctor in East Asia's successful productivity-basedcatching up was openness to foreign ideas and technology Governmentsencouraged 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 ISabSaharan 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 economieswith imporc-substituting industrialization strategies, even whcn protec-dion was practiced with respect to the domestic marker, the search forand absorption of foreign technology was encouraged iWhile Japan andKorea set obstades to DPI, they were hospitable to licensing, thougheven 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 tolocate appropriate foreign investors. Malaysia has aggressively sought ex-port-oriented DFI, particularly from Japan (see box 6.2). None of theHPAEs was hostile to the establishment of local buying offices by inter-nadional purchasers, an important source of production and marketngknowledge. 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 chemical2investment Was weloMe in Malaysia. Gonenbbe rduc mel proucs, andpolicy -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 echangeand 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 cacThe 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 -kgaporet nanfrlly afraed tiicafti&and bfisb&ffrtedrd hertcnooyr kceepu

302

USING RESOUPRCEts;EPfl

of knowledge of international best practice was a reflection of the viewthat 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 suspiciousof 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 accessto 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 aysihabulupnalso 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 cxword'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 largerdomestic markets, created a similar situation by encourging DFI in pro-

duction for acport and in those sectors where substantial technologytransfer could be anticipated. In contrast, the large import substitutioneconomies, such as Argentina, Brazil, and Mexico lured investment offoreign 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 asdefined 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 wasmore 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-incomeeconomies 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) havegrown unusually fast The sector's share of manufacturing value addeddoubled in Japan and Singapore, nearly tripled in Indonesia and Korea,and quadrupled in Malaysia. More surprising than the importance ofgrowth 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-economycomparisons. 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- ' ;drEismjf 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-predictedrole in manufacturing in most of the HPAEs (excepr Indonesia andMalaysia). 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 availablefor 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 valueadded 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, itseems unlikely that the measured growth rates of TFP could be attribut-able to cyclical phenomena or growing capacity utilization of initiallarge investments.3 6

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 :-519 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 d1563 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 0D419B1 - 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 rareTaiwan,

Japan, C-/ina,Sector Kirea 1960-79 1966-6

Food- 7.30 -1.76 2.0'Bcverages 7.90 0.0Tobacco 13.40 -Apparel - 1.98 10.5Tcxctiles 10.70 0.47 7.6Lcather 12.60 1.03 _Shoes - 1.03Wood 9.40 2.81 o3bFurniture 12.10 1.74Paper 8.20 1.44 2.3cPrinting 10.70 -0.18Chemicals 13.10 3.36 3.3Pecroleum -0.30 _3 55d O.ObRubber 1 1.40 1.02 63bNonmetallic minerals 2.80 - 2.4Basic metals - - 7.2Iron and-steel 3.70 134Metal products 7.60 3.41 4ANonelecrical machinery .00 - 230 67'Elecncal machinezy 10.70 537Electrical equipment - - 7.1Transport equiptnent 11.20 432 2.7Prccision instrumencts - - 11.0Plastic products - o.92f: 0.0Oilher manufacturing 7.50 -1.76Average 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 respect3 7 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 carefyl 7 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 seenin 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 helpthe firms realize static and dynamic economies of scale. mmTt allocatedforeign 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 Japanesefirms. Relatively high levels of protection were used as part of a strategyto 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 ofmanufactured 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 selectedaredit, protection, and limitation of entry into specific sectors (Padc andWestphal 1986). However, the Korean government promoted individualfirms 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 andexports induded making direct and indirect inputs to exports available atworld prices. In Korea, the selectively promoted sectors were the heavyand chemical industries: iron and sreel, metal products, machineiy, elec-tronics, and industrial chemicals. The motivation fbr these appears to havebeen both strategic-to increase defense capability-and economic-toshift to capital- and technology-intensive sectors in anticipation of a lossof competitive advantage in labor-intensive sectors.4t The costs of theHa: 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 thegovernment 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 effortto 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 internationalprices, 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) notesthat a guiding principle in the 1960s may have been the existence of

. ........ ~ ~~~~..

Box 63.31Te HCIDrive:Costsof lidterventionTHE-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, overseasr CaCed wit intervenCiin. Govcrnment intervention in-:X-- construcrioh, and shipping indusries in' th mid-.- ',.cureddiecthefosts thform ofsubsidies to strateic 1980s, nonperfbrming lo of conmrcial bankssectors 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:iectidon. 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 DrCThepornmoUooftea scows spporteda ti process, rnb. 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:urtbtheUpfipjwo 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<collecned)>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 93tpercent 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 tcpiceit 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 interventionin many sectors, there are no studies of effective rates of protection orsubsidy 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-technologysectors.

Malaysia. In the 1980s Malaysia experimented with a heavy andchemical industries drive similar in focus to those of Japan and Koreabut dependent on public investment. The Look East policy adopted in1981 was, apart from its sociocultural aspects, an explicit attempt to em-ulate the heavy industrialization efforts of Japan and Korea, which Nvereregarded 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-intensiveprojects 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 ofthese activities, die government actively promoted joint ventures withprivate (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 venturewith 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 theHiCOm program. Poor management and low profitability also forced the

treasury to bai out loss-making enterprises, a practice it could not sustainunder 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 beenp'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 interestsfiMrs existing Mi th mid9Os, about a hundred had, have 'frequently been more imporrant.-lthan thebeen 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,thertelecc}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 l

6sing 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 bb- , 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, themore.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~ uitovrpro 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 reionON,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 havingthe potential for technological learning, but, as we pointed out in chap-ter 3, it is difficult to discern any sysrematic effort to change industrialstructure from the pattem of promotion of the BOI. Indonesia has usedpublic investment in an effort to move roward high technology indus-tries. Indonesia's technology development efforts represent an cxtremecase of state-supported technological upgrading, but the public sectorresearch facilities and strategic industries have fev links with the privateindustrial sector. One experiment with leapfrogging into high technol-ogy industries-Indonesia's attempt to create an aircraft industry-isdescribed 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 seaCoralcomposition of output are largely rnarket-conforming, industrialpolicies must have failed irh at least one of their objectives: to guideindustrial development along paths thar it would not take if itwere guided by market forces.

* Second, what were the rates of productivity change in industry inthe HPAEs, and what were their sectoral patterns? If rates of pro-ctucivity change in industry are low overall or in promotedsectors, prima facie, industrial policy did not meet its productivity-enhancing objective.

In both cases, our answers lead us to conclude that industrial policieswere 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. InKorea, for example, despite the government's extensive efforts to speedthe 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 prediaedin 1988, a substantial increase relative to international norms from1968. 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 inSingapore, which, like Korea, has a government that aggressively inter-

Box 6.5 Indonesia's TurbulentLeap into High Technology

* SECr1oRAL-. TARGETING TO ACHIEVE RAPID PRODUCTIN .

change is a risly business-paTilarly when publicly funded firsare pemiedm to ily on protecred domestc mardratherdian bingsubJece 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 hasabsorbed $1 billion in goverment. fiud since its: esitablishient in1979 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<pductio& Moreovcr, the company hasihad lide lexposure to the

:-worid mker-.Of the 230 aircraft(30 hlic:opteri, 100fixl:-wing,r-. that IPiNhas manufacrured sincerits inaiion, -Cl

.beensold to Indonesian airlini<the miiaiyrhe pmoli and th&stame-ruoi in

Foreign sales are uniikely to imptrove.While other aircaaftaimanu-&u-- r&s Jcihaive been zibusy, foginhnological ianc&,to%Iinkrhe

-time from design compleonoplucon, i is- iesiligXto~ gDIt alone. Its majorpwd@ forthis decadan dendeilydesigned fifyaurboproppi rs .nor. fr el u.

-1995,f ieryearseim designicoxpiieon By h..for slmalar ar f&alr~adywedwill have xseral hern en t -ranrsall pncing their uairdfrCaggressively. rr cn ill affztpnC

arompyr srbu195'hle, m ostsrc:,o4i-:cate that IPTN-was.nor pfiptirabWlewhen capitaL coAstsarleen into ac-

COUUL Noduogin xhompP2a,nys pinCrtormancsnce rhenindic-te3thgat the sztuandiibs iniproveI: . . >-

313

:AN MIRACLE

vened through a variety of mechanisms to promote capital- andtechnology-intensive industries. Here, the importance of textiles relativeto 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 fromrwelve to fivc times its predicted size.

Appendix 6.1 describes a simple rest of the impact of industrialpolicy on the basis of the factor proportions theory of comparativcadvantage (Hecksher-Ohlin-Samuelson). We assume that given therelative 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 statisticallythe pattern of growth of the industrial sector in each economy withits pattern of factor intensities. We argue that if industrial structure isderermined primarily by market forces, low wages and low capital in-tensity at the beginning of the period of active industrial policy willpredict 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 bethe case. Indeed, the opposite should be true; high wages (a measureof human capital intensity) and high capital intensity at the end ofthe promotion period should predict changes in industrial structure.

We have performed these tests on each of the Four Tigers plus Japanfor periods relevant to their industrial promotion efforts, using two-digir isic classifications of industrial sectors (see table A6.2 in appendix6.1). 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 inseac 'I shares of value added occurred in lower-wage or lower value-addeC_ per worker sectors. In Korea at the two-digit level, sectoralgrowth was broadly market confbrming in terms of traditional factorintensities. The results for Japan are incondusive. In Hong Kong thereis no relationship between our explanatory variables and patterns ofproduction. In Singapore, for the period 1980-89, output grew morerapidly in more capital- and knowledge-intensive sectors, supportingthe 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. Butthey suggest that, despite governmenr intentions, the manufacturingsector 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 goodnews and bad news for advocates of industrial policy in the productivityperformance 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 havederailed 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 beenhigher in promoted sectors. Japan may be an exception. Between 1960and 1979 chemicals and the metalworking machinery complex have un-

usually good TFP performance Jorgenson, Kuroda, and Nishimizu1987). Japads industrial srructure differs from international norms inthese sectors and exhibits quite high values of the share of value added

in total manufacturing. These industries are those that observers usuaDypoint 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 avariety 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 ironand steel was accompanied by quite low TFP performance between 1966

and 1985; textiles and clothing conversely, had very high rates of TFpgrowth. 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

1 STN MI R A CL E

The three sectors that exhibited the greatest expansion in the share ofvalue added-apparel, electrical equipment, and metal producs-areall characterized by below-average wages, capital-labor rarios, and laborproduct. Moreover, there is no statistical relationship between wage orcapital inrensity and productivity change at the sectoral level. In fact, thehighest sectoral rates ofTFP change are recorded in textiles and apparel.

Recent sectoral TFP growth rares for industry in Malaysia from 1973to 1989 showa similar pattern ofgreatvariabilicyin cheTr growdt ratesin promored sectors, from high in nonelectrical machinery and paperand 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 thesectors promoted under the HICOM drive and rapid productivity growth.

Overall, the evidence that industrial policy systematically promotedsecrors with high productivity change is weak InJapan there is some sup-port for the assertion that TFP growth was higher in selected seccors, whilein 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 thevery high overall rates of TFP change in manufacuing yield mosdy neg-ative results (see appendi 6.1). The main reasons for manufactringssuccess in Japan, Korea, and Taiwan, China, lay in the high general ratesofT"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 quicldydoser to international best practices. We termed this "productivity-based catching up." We believe that rapid growth of cxportsi a result ofthe 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 humancapital contribute to the productivity performance of the East Asianecononies? Most explanations of the link between TFP growth and ex-ports emphasize such static factors as economies of scale and capacityutilization. While these may account for an initial surge of productivitysoon after the start of an export push, they are insufficient to explaincontinuing high TFr growth rates. Rather, the relationship between ex-ports and productivity growth may arise from exports' role in helpingeconomies 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 humancapital 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 thebasis of productiviry change due to such domestic efforts as plant reor-ganization, the cumulative magnitude of productivity growth overnany years is most unlikely to have been a result of purely domestic ef-forts.42 It stretches credibility to suggest that the large cumulative effctsof TFP growth in Japan, Korea, and Taiwan, China, could have beenachieved by the plant floor innovations proposed as important sourcesof productivity growth at lower TFP growth rates (Pack and Page 1993).Were that the case, it would be difficult to explain why these gains faroutstrip productivity increases in the industrial economies during theirown rapid-growth periods, when TFP growth was attributable largely todomestic 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 inquire imperfect markets. Often the markets do not exist-some knowl-edge is simply not sold, because its owners fear that licensing or direccforeign investment will eventually Ieak the knowledge to future com-

petitors. Even where markets exist, they are likely to be characterized bybilateral monopoly, so that a variety of difficulties confiront finns tryingto 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 oneif knowledge is obtained at a lower cost, a real extemality if the knowledgeturasfer is more efficient. For example, the transfer may be more beneficialif 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 forknowledge and permits its acquisition through a variety of mechanisms.These include:

m Purchase of new eqiupment. The purchase of new equipment is astraightforward method of obtaining new technology insofar as it is em-bedded in equipment The two main impacts of exports in this case arein providing the competitive pressure for firms to make such purchases(if the equipment is used to produce exports), and, more generally, inearning the foreign exchange to finance the purchase of machiner)y

a Direct#iriezt investment The firms that generate much ofthe world'snew technology are reluctant to part with it. They perceive that the bestuse of tieir new knowledge with respect to developing economnies often isthrough exporrs of producrs to them. In East Asia, however, most of theincoming DFI has been intended for production of exports rather than asa 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 suchgrowth often reflects good macroeconomic management Moreover,econornies with significant export growth have generally provided moreinfrastrucure useful to exportes than internally oriented economies44DFI makes crucial production and marketing knowledge available to de-veloping economies. It permits them to begin manufacmring along theworld's best-practice producton fiLnction by substituting foreign physicaland human capital for absent local factorsfi5 As local labor learns from thepresence of best-practice knowledge and equipment, knowledge rends tobecome diffiLsed through labor mobility and informal contacts amongmanagers. Even when the investing firn attempts to hinder such diffu-sion, signicant knowledge ranfers are inevitable.6

* Technology licensing. licensing exsting technologies, both rights toproprietary equipment and derails about production pocesses, offers de-veloping economies substantial opportunities for improving dteir levels ofbest practice. During the 1950s and 1960s, Japan benefired considerablyfiom 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 withthe domestic sarch expenses avoided. However, there is some evidence,and a growing subjective sense, that arz's-length licensing is decreasing asan 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 theduration of the license. Particularly in R&D-intensive sectors such aschemicals, machinery, and electronics, firms are increasingly unwilling tolicense technology-, they believe royalties provide an inadequate return foractions that may impair their own long-term competiuveness. This createsan 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 theyobtain access to the licensee's own technology or to its manufacturingskills. Given asymmetries in knowledg. the best test of the potentialparner's abilities is its performance in export markets. Moreover, thefact that a firn has exports and the requisite technical and commercialskills to produce them improves the bargaining position of the licenseeSales of licenses contain a significant element of rent (Pack 1993c), sothe cost of licenses will be lower where the seller perceives the strengthsof the purchaser as signaled by exports. Finally, where a mark-er doesexist for technology licenses, the foreign exchange to pay for fees is moreikely 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 oldmachine designs and production technology. When exports are basedon comparative advantage, precisely these conditions are met. HongKong, Korea, Singapore, and Taiwan, China, all began rapid industrial-ization in labor-intensive and low-technology manufacturing. Theequipment was readily purchased, and the production knowledge wasavailale at low cost from engineering publications, trade literature, andindependent consultants.47 The simpler equipment was conducive toliocal productivity-enhancing improvements, often on the part of bluecollar workers (Ranis 1973). Moreover, in labor-intensive industries icwas easier to acquire and absorb information from customers withour alarge stable of educated engineers.

* Infibmwonfrom custorme. In the presence of imperfectly tradedinformation, 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). Buyerswant low-cost, better-quality products from major suppliers. To obtainthis, they transmit tacit and occasionally proprietary knowledge fromtheir other, often OECD-economy, suppliers. Such knoNvledge transfersare more common in simple production sectors such as dothing andfbotwear 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 wagesmade 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 havereceived 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 foral/the postgraduates employed in the industry of Taiwan, China (Pack1993a). This source of knowledge of international best practices be-comes more important as changing factor prices dictare a shift to morecapital- and technology-intensive sectors in which higher-level skills areneeded 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 considerableproportion of R&D has been devoted to improving exports and reduc-ing production cosrs. There is no evidence that the gap between R&Dand 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 ofhuman capiml supplement and reinforce one anothec It is doubfl that tieHPAEs couldchave made as prductv use offoreign knowledge and importedcapital 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 managersand technicians in individual plants in inward-oienred Lain Americaneconomies to improve the productiviLy of eitig capital smckwith internalinnovations 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-performingAsian economies in the form of cheaper and more effective knowledgetransfers would have undoubtedly been less productive had there beenfewer skilled workers to facilitate their absorption, while the HPAEs'rapid increase in education levels reflected in part rising private rowardsfor 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 inchapter 1 to look at the impact of trade and indtLstrial policies ongrowth (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) andmeasures of export performance.51 Openness to the world economycaptures not only effidencr gains induced by the need to remain glob-ally competitive but also the likely allocative benefits of having relativeprices dose to international prices.

Introducing these two trade-related variables substantially increasesthe explanatory power of the cross-economy regression. The effect ofopenness is positive. Economies with limited relative price distortionsgrew more rapidly.52 We choose as our measure of export performancetwo indicators-the average share of manufactured exports in total ex-

ports and the share of manufacrured exports in GDP.5 3 Manufacturedexport performance is strongly correlated (at the 1 percent level) withhigh rates of per capita income growth. When the share of manufac-tured exports in totl exports is introduced together with the opennessindex, only the latter is significant. Conversely, when the share of man-ufactured exports in GDP is used, the openness index remains positiveand significant. One possible interpretation of these results is that a highconcentration of manufactured exporcs relative to total exports, ratherthan openness, contributes relatively more to productity change in across-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 79Intercept -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 0.0003exports, 1960-85 (0.0001) (0.0001)

Averac manufactured exportslGDP. 0.0011 0.001a oi1965-85 (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 inTFP growth rates in terms of relative income, educational attainment (asmeasured by the average stock of education per person), openness, and

our measures of manuFactured export perforrnance (see table 6.18). Theeducation stock variable, while positive, does not explain variatons in TiFP

growth among economies. This is appropriate, since we have measuredTFP 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 positivelycorrelated with higlher rates of T PP growth.

We also find some evidence of.- positive interaction between the shareof manufactured exports in total exports and in national income and thestock of education. The coefficient of the interaction term between thesetwo variables is positive but not significant at conventional levels, and dteexport 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 0.0032manufactured 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 . 0.0284manufactured 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 educationraise 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 anumber 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 rapidexport growth is correlated with the pattern of productivity change; ax-porting sectors have higher seoral rates of rF growth. Wei (1993) usescity 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 productivitydifferences 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 understatedthe variation across sectors. But the revisionists who see in the HPAEs'

success evidence that highly targeted industrial policy worked overlookthe fac that the calculations on which borh relied wer- static, that firmsbased their decisions not only on current levels of protection and factorprices but on the certain knowledge that thy would need to compete inthe fuiture. Effective rates of protection may shape short-term tactics butthe long-term strategies of currendy protected firms are more likely roreflect 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, sinceall bur the very dull knew that their turn to export would come soonerthan 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 alleconomic actity would be judged, even firms benefiting from higher-than-average rares of protection in the domestic markcet understood thatin 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 wereimportant 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 andtheir 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 selectiveinterventions 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 waslimited differences between international relative prices and domesticrelative prices in the HPAEs. Market forces and competitive pressuresguided resources into activities that were consistent with comparativeadvantage and, in the case of labor-intensive exports, laid the founda-don for learning international best practice and subsequent industrialupgrading.

Does this mean that selective interventions played no role in EastAsia's superior growth? Our conclusion is that selective interventionswere neither as imporant as their advocates suggest nor as irrelevant astheir 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-creditprograms were less successful Where rhey appear to have resulted in al-location of credit to high-yielding activiries, mainly in Japan, Korea, andTaiwan, China, it was because of careful szieening and monitoring ofprojects and an orientation toward the private sector. Where they failed,governments limited the damage by offering more limited subsidiesthan the credit programs in other developing economies. Finally, wehave conduded thac industrial policy, with the possible exception ofJapan, did not alter industrial structure or patterns of productivitychange. 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, andindustrial 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 IndustrialPolicy on Productivity Change

G OVERNMENT POLICIES THAT CORRECT MARKET FAILURESshould 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 whichthe intervention occurs. Subsidization of knowledge acquisition in tex-tile production will mainly benefit other textile firmns as knowledge seepsout to them or trainecd engineers go to other textile firms. Some knowl-edge in textiles may benefit dosely related sectors such as dothingthrough 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 growthare desirable as aweak rest ofthe efficts of intervention. Productivitygainsin rhe entire industrial sector or the entire economy attributable to otherfactors might mistakenly be ascribed to the intervention, even if it had ac-

tually reduced productivity growth. Undoubtedly, the productivity gainsftom correcting various market Milures may diffse to many sectors, burmost should have a major impact in the initial sector of intervention.

Empirically, many of the intersectoml interactions will be measured asintrasectoral at the two-digit level. Improved production engineering inspinning may improve TFP in weaving. Both, however, are induded in theEwo-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 industrialeconomies demonstrates that the major beneficiaries are dosely relatedsectors, 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 whoargue for the benefits of selective intervention and then offer as evidenceeffects 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 internationalnorms, 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 isgross domestic product, and vcl; is total population (Chenery 1960).The equation attempts to capnure dte influence of demand elasticitiesand dte evolution of supply. POP factors in scale economies, although ina world of free trade this may not matter.5 6 The equation is purely de-

scriptive and has no normnative content Indeed, a given pattern mightindicate that all economies follow some initial leader and undertakepolicies to emulate its sectoral evolution. Despite its flaws, the equationprovides 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 bySyrquin and Chenery (1989) to derive the predicted sectoral shares ofvalue added in GDP. Table 6.15 presents the results for eight branches ofmanufacmring, showing the ratio of actual, VI, to predicted, vp, sharesof gross domestic product for various sectors. Thus a value of 1 repre-sents confbrmance t the international norm, values less than 1 indicatethat 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 havelarger overall manufacturing sectors than international norms based onthe 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 therough 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 eachsecror's level of total fitcror productivity, A,0 by the sector's share invalue 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 productivityof existing sectors, logAi.> log Ai r- or the growti in th.e value-addedshare of these sectors, vi,> z-, 1 whose productivity is growing.57 As willbe seen belowv in table A6. 1, small differences in die rate of growth ofsectoral TFP, A*i,, lead to large cumulative differences in Ai,

As the text of chapter 6 points out that on the basis of internationalcomparisons two sectors were generally overrepresented in rhe industrialsectors of the HPAEs: metal products and machinery, and textiles and apparel. Only the former sector was promorecd in HPAEs that used indus-trial policy. We define two types of sectors, those fiavored by governmentpolicy, f and those subject to benign neglect, h. The implicit model ofthose who believe that industrial poLicy has had an important effect isthat 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- ortechnology-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 onManufacturing-Wide Growth of TFP

TFPgrowh, :TFPSrow*,,Economy actua vae-ad weighn -adIjal wegkh .

Korea, Rep. oE 1966-85 6.7 6.1Japan, 1960-79 2.3 1.9

NeArWeightod by value-added shae datwould have prvaied ifthe meal productsand macduneysectorhad conformneddt redictp d by txcquatuion estimated byChenery 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 aresult 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 termsof edquation 6.1, Vf1 increases more than woulc have been "normal"from its initial level, vf Thus, ir is necessary to demonstrate thatthe sectors to which resources flowed were not those that wouldhave grown on a simple comparative advantage basis. If, however,the sectors that grew1, 1> 'M' are characterized by low wages, orlow 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) canbe 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 existingones. In terms of (a), the sectoral structure of output is radicallychanged by the growth of modern sectors that are introduced athigh levels of TFP, and this raises average manufacturing-wideproductivity.

Because the HPAEs were such successfhl exporters, predicting the impacaof selectve interventions on the sectoral pattem of production depends toa considerable extent on one's preferred theory of international trade-Ricardian, Hecksher-Ohlin-Samuelson, neotechnology, and so on. Todemonstrate the impact of selectve intervention, it is necessary to posit acounterfictual model. This is obviously exceptionally difficult, and to berigorous requires a general equilibrium model that in turn has difficulties ofits own, such as con-crdy specifying sectorl production fimctons andlearning patterns. Again, the "natural" evolution of an economy also de-pends on one's preferred theory of international trade- Nevertheless, aweaker 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 sectorscan be predicted by the factor endowments theory of comparative ad-

329

;M-IRIARACLE

vantage, or Hecksher-Ohlin-Samuelson (Hos). Although large numbersof addirional explanations of the patterns of crade have been put forwardin the last two decades, many of them are actempts ro explain intrasectortrade and specialization arnong products that broadly require the sameFactor 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 FourTigers and rheir size in the GDP, the changing strucaure of production inthese 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 restedwhether the sectoral evolution of successfil Asian economies has differedfiom what would have occurred had there been neutral policies. Neverthe-less, showing that many instrumenms were selectively employed is not thesame 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 independenrvariables, xP, are the wage per worker at the beginning, zv5, or end of dteperiod, 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 workershould measure the skill intensity of workers in each sector. Value addedper worker should reflect both skill and capiral intensity and is a proxyfor technological complexity. Although value added per worker and thewage rate are likely to be correlared, both are used, as there is a possibil-ity of divergence. Governments also may have employed one or theother as indicators in targeting sectors. Wage per worker may be trans-

lated as "good jobs,' while value added per worker is a natural measureof "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-incomeeconomies 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 manynewer 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 amongindustrial economies than the evolution of the sectoral structure of pro-duction of developing economies whose production structure is likely toreflect differences in factor proportions.

If intervention has been of great importance in promoting the expan-sion of capital- or knowledge-intensive sectors and, eventually, affectingtrade patterns, the data should show dvjldwb> 0 and/or that do1.Jdva6>0. An alternate test of the impact ofselective promotion is to testwhetherthe sectors that expanded exhibited high capital intensity or wages at theend of the period. A result fvoring the interventionist view would beehi-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 A6.2shows the sign of the regression coefficient and its significance for eachperiod for each economy. The years chosen reflect a distillation of rheexisting literature for each economy on significant intervals duringwhich intervention was practiced. It is not possible in all cases to utilizethe ideal interval. In Japan, for example, data from before 1953 varyconsiderably in definition from those after 1953.

These are clearly very simple tests of the determinants of structuralchanges in production but capturc the spirit of the opposing views. Thestatistical results must be employed carefully. Even if it is slhown thatthere was a significanr negative correlation between 9i and w1, support-ing the HOS view, some of the individual sectors that grew most may stillhave been high-wage sectors. Thus, we will examine some individualsectors 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 negativecoefficients 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 ExplainingChange in Value-Added Share of Sectors

Economy Period Wu WE VB VE

IHeng Kong 1973-80 - + - +

1980-88 + + + +

1973-88 - + - +

Japan 1953-631963-73 + 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 leveLNbOwtWSmZ 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 morelabor-intensive seccors increased during each period On the face of it this

result is surprising. Quite apart from policies of selective promotion, thiswas 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 isthat 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-averagelabor 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 rapidgrowth in TFr, and the cost-reducing effect of productivity change stim-ulated expansion of output by more than the output growth induced incapital-intensive sectors by the decreasing relative cost of capital andskills 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 ofthe 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 amongsectors. Moving away from the anecdotes about individual chaebol, thequantitative 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 nosignificant coefficients. At least a simple version of HOS does not work,probably nor surprisingly as Japan by the earliest year considerea in thetable, 1953, was sufficiendy advax±ced so that intra-industry rather thaninterindustry trade would have become an important determinant ofthe sectoral producton structure. Scale economies or a neotechnologyexplanation of production and trade patterns are more useful at thisstage of development. Going back to 1945, which the data do not per-mit, would not change this picture much insofir as this earlier periodwas one of reconstuction, in some cases of firly advanced sectors thathad brken quite important before and during World War IL A simpleHOS story would be interesting to test for the period from the Meijirestoration onwards. Once an economy is in a world of technology- andscale-economy-based development, the requisite conditions exist toallow 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 diffcrentpatterns of development. In Hong Kong there is no relationship be-tween our explanatory variables and patterns of production. In thissense it is similar to Japan but derived fiom different causes. The slowgrowth of capital stock per worker in Hong Kon& partly reflecting largeimmigration, 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 inthose sectors that at the beginning and end of the period exhibitedgrearer value added per worker, supporting the view that the Singa.-porean authorities intervened to encourage increasingly capital-intensivedevelopment. 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 andvalue-added variables for Taiwan, China, are negative, although insig-nificant. Again, this implies that the relative size of more labor-intensivesectors increased during each period, albeit weakly, at the two-digit level.Growvth was market conforming; government intervention did litde toalter 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 particularfor Korea, during, 1973-80, the years of greatest intervention, the mostrapid growth in sectoral shares of value-added occurred in lower wage orlower 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 undertookmeasures such as the provision of social overhead capital and technicalservices such as the Korean Institute for Science and Technology, whichfacilitated the 'neodassical' tansformation.

Singapore, in contrast to Korea, confirms the casual empiricism ofmany 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 andlabor productivity than the laggards. There is no evidence of a factor-proportions-based expansion. In contrast, the evolution of sectors inJapan 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, andtrade 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 hieenormous importance of its trade. There was, however, relatively litdevariation in the dependent variable, the change in the sectoral shares ofvalue 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. Theissue here is whether given the observed values of sectoral TFr growth,A*i, the unusual evolution of the sectoral pattern of production in someof the economies had a significant quantitative impact on overall TFPgrowth.

Going back ro equation 6.1, the analysis focuses on whether thechanging strucure of vj, given the pattern. of A*j, contributed to morerapid growth. The value of A* is taken to be exogenous. Of theeconomies considered here, detailed sectoral growth rates ofTFP are onlyavailable for Japan and Korea (surprisingly, sectoral TFP growth rates arenor available for Hong Kong or Singapore), but these are two of thethree economies in which activist industrial policies were employedmost 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.At i in most sectors. From these studiesa number of patterns can be identified that are broadly consistent withone another. There is no simple correlation between promoted sectorsand 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 answerthis we first calculate the manufacturing-wide growth of A* using theobserved 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 theirpredicted value, VP, on the basis of intemational norms as a share of sec-torwide value-added and, reassigning the residual value added to allother sectors equally. Thus, instead of accounting for 38 percent ofmanufacturing value added, sector 38 accounts for 14 percent in thiscalculation.59

335

SIRANL MIRACLE

The result is shown in row 1 in table A61. The actual secrorwidegrowth rate, A* was 6.7 percent, the recalculated one 6.1 during a periodin 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 havebeen attributable to selective intervention, its rare ofm'FP growth was notsufficiently above that of other sectors to make a large contribution tooverall sectoral TFE growth. The major reason for Korca's manufacturingsuccess lay in high individual values of A;. for most secrors in mostperiods.

Japan. As in the case of Korea, a considerable literature argues thatJapan's selective industrial policy was critical for its industrial success.Nevertheless, many Japanese analysts are skeptical of its effects, as aremany analysts from other economies.60 The existing literamure indicatesthat many of the critical acts of the Japanese government with rcspectroindustrial 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 toanalyze with systematic quantitative evidence rather than anecdotally.61Chemicals and the metalworking machinery complex have unusuallygood performance. These industries are those thar observers usuallypoint to as having received significant governmenr support, indudingefforts to stimulate productivity growth.

A calculation, similar to that for Korea, to determine the impact onA*for the entire manufacturing sector of the unusual size of sector 38can be done for Japan. Table A6.1 shows rhe actual growth of A* for1960-79 and the estimated growth, had the sectors constituting 38been at the international norm, 24 percent of manufacturing valueadded 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 rateof manufacturing value added of 8.7 percent per annum. If selectivepromotion is interpreted to mean that only a few sectors rather than allare aided, Japan's growth rate would have been only slighdy slower with-out an industrial policy. It is possible that industrial policy significantlyincreased the values of Ai in sectors other than those promoted due toone or more externalities. While this may be the case, it is likely thatmost externalities occur within individual sectors or in closely relatedones. 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 betweenTFP Change and Trade Policies

T HE ECONOMY-SPECIFIC TFP GROWTH RATES CALCULATED IN

chapter 1 allow us to make an alternative test of the relationshipbetween 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 withphysical capital, human capital, and labor as inputs; RGDP is GDP relativeto the United States in 1960; PRIM and SEC are the primary and secondaryschool 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 betweenthese variables and rates of TFP change derived from our productionfiunction estimates. The regressions control for other sources of variationin TFP by introducing a -variable that has been shown in previous workto 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 Weil1992; 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 madepossible by the diflerence between average practices and best practices.As we nored in discussing the cross-economy regressions, ir rime-seriesanalyses of aggregate coral Factor productivity growth, estimates of TFPconsist of two components-intraseccoral TFP growth and the impact offactor reallocation among sectors. Thus, the shift variable RGDP60 maycapture the reallocation effecc of structural change on TFP in addition toproductivity-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 theelastcity of output with respect to human capital. We have argued thatprimary 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 wouldthen be included in the estimates of TFP change. We attempt a similarrest for secondary education.

Openness to the world economy is another imporcant can.dlidate forexplaining rapid TFP growth. The theoretical case for this view rests notonly on allocative efficiency bur also onl extemalities assoiated withtrading activities and on "X-efficiencyf gains from creating a more com-petitive environment for domestic industry. Several previous studieshave found such effects, induding Harrison (1991), Dollar (1991), andThomas 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 andproductivity change. We choose as our measure of export performncetwo indicators-the average share of manufactured exports in toWtl e-ports and the share of manufactured exports in GDP.62When the shareof manufa-tured exports to cotal exports is introduced into the basicspecification, it is positive and highly significant (at the 0.01 level). Theshare of manufictured exports to GDP is significant at the 0.05 level inthe basic spedfication; how ever, when the degree of openness or theHPAE 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 economicsb1960-70 1.1 3.8 3.9 2.0

a. Developing economics include the counties of Arica with the exception of SoudhAfricr Latin Ameica; Asia with die exception of China, the Denocmaic Republic ofKoDrea, Japn; Mongoli:. and Viet Nan; and Oceania with the exception of Australiaand 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-wgnla 7

Jfrg fly 19* 6kTn sop innrvmewidn Tudiduifr r&d hum 6W awi 9IS-Tbeerrcrk 39Q, DatEd mknatu /ort..r he irans. fer-i5

nenuis dtpma .iln - undo th eoniumNMmforfrwva.x. Yeq ib cadmftgeablkxd liauibsnbe fRn HlU5et to Hls 1in litAdimison 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 etnrmaBen Us tanRg Uqunc d ITden imoatnspipsad qBiz.Wl .tobmU mnndma%ima ams nidby 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 Fumeswjydryc-nnpa atn Usn RadA Reuian LiCe Oft C:Gti atm - ltea CRn Onindelesun. iehuuansiiYwEmk aaben 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-m25d5f-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_nes 1 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 mCa,ndial hajneaaMrt

AtktSmckindfis ma wading

Ms-Tkofiac un.mjrfTey 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 shiauiauWhi.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, offfkb loUtes aam.csa11icMFumnl bmil kw Inured aainsO 50, &Ldqwia p-cer. wma- dqiWmnc anvocrUe9i maumpnamu and aural ofInkik&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 ogle57&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 biWngto1 UL CLCruStbaku 6, poad rtheakingAadZ5l'Manl FonmanceGimpmap wd n f Act .1M. SoL:i..tm,.lacu lc&g guidolina

-)flCrmmrulbmk-alminhunced Bar iuqMaCCaKrhia. andactig awtdiinmsu.darda&afiaaCd,hainnd itLadIng 1atehrodcednmanlhe. 1:&ibdla c ue,pWcctc .aitardind rnamisffi.aeW upcidg an Linac199-. (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. Bankouitn 7 aa 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 a

1a!,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 />sindl 3 aflhlsc 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*riwaiinlocalboaokomps.Poretbasik

madlbomgarvhow, nuk(hdtacnve tehdi,19 poite sula,

re... aim. 1950( Lirbclsd C at. Ckit AttaInhit 1 ban mw IQiEahlsbehrx T.iwa o19t D. SsxhiS ad Exitnu b Crwnuiona194 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 aadahidinpia - 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.dindopMMimdaanigfI.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

Notes1. 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 amongworkcrs 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 firms3.2 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 of4. 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 inThailand. Indonesia and Malaysia are forecast to grow at 9. Clearly, a high share of the public sector in wagehigher 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 allIn 1989 it was 30 percent in East Asia and 36 percent in its employees.Latin America; in 1986 it was 46 percent in Sub-SaharanAfrica. 10. In more than one-third of the economies, private

sctor employrnent actually decreased, implying that the5. 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 inskill 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, incducated 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) andand geogaphic areas as though they had a fine regard for chapter 4.marginal costs and benefits. Ihough behavior consistenrtwith rational decisionmaking on the part of individuals is 12. See Fields (1992), who nores that until 1972 unionsa 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, haspation for which they are qualified. Restrictions on free- asserted, "Singaporc is possibly the most interventionistdom of cntry arise from such sources as nonopsony government in the capitalist world." It is irteresting to notepower, discimination, government interventions, and that the same instrument-the National Wages Council, aworkers' 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 begrowth 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 in13. 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 coalassociated 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 ofthough he recognizes that this generalization is not an ad- Japan's directed-credit programs was lcss than 2 percencequate dismissal of the argument that in some economic of total invcstments in manufacturing between 1961circumstances, 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 subsidies14. 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 crediLa 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 competitionuates. 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 Japanese17. 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 warrantedwithdrew this directed credit scheme in 1986. when there is a significant discrepancy between private

and social benefits, when the investmenr risk of particular18. 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 (OECFfderal govemment lent direcdy to public financial and 1991). Use of policy-based lending rather than othernonfinancial 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 the16 percent in 1980 and 9.8 percent in 1989. These loans main constrainc facing new or expanding enterprises iscarried real interest rates, which were mildly negative to their access to extemal finance at reasonable terms andslighdy positive, compared with the substantial negative conditions. Directed-credit programs involving smallreal races of many other economies (Caprio, Atiyas, and subsidies overcome this consuaint, but Japanese officialsHansen 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 inthe maintenance of macroeconomic stability to minimize the long run, it would pay for the firm to encounter lossesdistortions in incentives and on effrive monitoring to today. But this counteragument runs into problems: it isensure 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 (seeical 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 the24. Among the best known of these arc Balassa and long-run marginal cost.

others (1971), Little, Scitovsk,y and Scotr (1970), andKrueger and Bhagwai (1973). WorldDevtdopmentReport 32. In the early and mid-1970s, nominal tariff rtes1991 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 Packand Westphal (1986) and Itoh and others (1988). World 34. It would be more desirable to compare efnectiveBank (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 systematicdevelopment projects. cross-economy studies based on comparable definitions

udilize evidence for only one year in the 1960s and early27. 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 duty28. 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 chement for trade policy interventions only if they allow paper was writren in 1985, the TP estmares for bothgoods 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 sufficiendyshortsufficient to justifjr intervention. A socially successful in- period so that increasing capacity utilization or one-rimetervention depends on whether the present discounted benefits from scale economies could have accounted for avalue (Pnv) of fumtre producer surplus exceeds the PDV of substantial share of unexplained growth. One study ofthe 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 veryimproved 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 inplained growth. 1975-84 and demonstrated the viability of new exports.

Moreover, much of the investment was in relatively small37. 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. Sincelabor 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 introducingOkuno-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 ofon nominal tariff rates, not implicit price comparisons. spillovers from DFI, see Blomnstrom (1989) and HaddadGiven the presence of quantitative restrictions, they are and Harrison (forthcoming).likely to be minimums.

47. See Rhee and Westphal (1977) fbr evidence on the40. 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) forTaiwan, China.

41. In Korea and Taiwan, China, the initial spurt inexports was in labor-intensive products as predicted by 48. On Taiwan, China, see, for exanple, Wall Stretthe static theory of comparative advantage. These, how- Jounmal June 1, 1990, and Liu (1987). On Korea, seeever, are also the sectors dtar Dollar and Sokoloff find ex- Westphal, Rhee, and Pursell (1981).hibited greater TFP growth between 1963 and 1979 inKorca. 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 improvcinitial 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 of42. 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 fromquality 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 the43. 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 buris 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 (tamenand Summers ( 991), Mankiw, Romer, and Weil (1992), together, the two amr significantly diffrent from zero at theand Pack and Page (1993). 0.01 level). Where rhere is a high degree ofmulricoilinearigy

berween the variables, the coefficients on Lhe interaction52. 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 recentlyand international relative prices (1990). This measure been revived by Murphy, Shleifer, and Vishny (1989).uses Summers and Heston's international comparisons ofprice 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 anof tomal exports were also employed with similar but less initial disequilibrium in marginal productiviEies amnongrobust 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 Tigers54. This conclusion is reinforced by the arguments as they ypically had rdatively comperitive labor markets

that low distortions relative tO international prices canarise from having a significant share of the economy oper- 59. Put differendy, it has been shown earlier thataring 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 ratherthe openness index, than 36 percent; thar is, the share would have been con-

stant ar the 1968 level. If factors had been allocated to all55. 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 interactionterm and the export term. The Ftest rejects the null hy- 60. See Itoh and others (1991), Komiya and otherspothesis that, taken joindy, the coefficienrson thethreevari- (1988), Saxonhouse (1983), and Trezise (1983).ables (education, manufacwured exporrsftotal exports, andthe 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 Pragmatismin a Changing World

HIS BOOK HAS SHOWN THAT THERE IS NO SINGLE

East Asian model. Rather, die eight high-performingAsian economies have used different and changingsets 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 ALLeight 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 subsequentchapters 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 generallyconservative in avoiding inflationary financing of budget deficits, andsome HPAEs, such as Singapore and Taiwan, China, have run sustainedbudget surpluses. While Indonesia, the Republic of Korea, and Malaysiaacquired considerable foreign debt, their debt to export ratios werelower than other developing economies, so creditors never lost faith andthey never faced a fuil-blown debt crisis requiring rescheduling. More-over, budget rules and legislation requiring char sovereign debt be usedfor development expenditures helped to ensure that borrowing fundedinvestnent rather than consumption. The successful record of low in-flation and sustained growth of the HPAEs reinforces the view that thereis 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 coupledwith flexible exchange rate management were used to keep exchange ratemovements aligned with dchanging structures of trade protection and in-flation differentials with trading parmers. The HPAEs avoided strategies ofmacrocconomic stabilization that stressed the role of the exchange rtc inbreaking inflationary expectations. Rather, exchange rate adjustmentswere supported by expenditure-reducing fiscal measures. While Indone-sia, Korea, and Thailand have had periods of real appreciation, they didnot try to maintain a fixed exchange rate regime in the face of inflationrates much greater than those of their trading partners.

The HPAES responded to external shocks more quickly and effectivelythan other developing economies. East Asian policymakers have beenmore adept at managing fiscal contractions-cutting expenditures andincreasing revenues in a growth-oriented framework-to attain viabledeficit targets. In contrast, governments elsewhere committed to fundinglarge 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 resortedto cutting public investment.

In each HPAE, a technocratic elite insulated to a degree fiom excessivepolitical pressure supervised macroeconomic management. The insula-

348

POLICIES CI

tion mechanisms ranged from legislation, such as balanced budget lawsin Indonesia, Singapore, and Thailand, to custom and practice in Japanand Korea. All protected essentially conservative macroeconomic poli-cies by limiting the scope for politicians and interest groups to derailthose 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 weresubstantially above those in other developing economies. Firms thereforehad an easier time upgrading the skills of their workers and mastering newtechnology. 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 beoriented toward vocatonal skills. Some HPAEs have also been unusuallylarge-scale importers of educational services, particularly in vocationallyand technologically sophisticated discplines. In addition, enterprise-level training, much of which was subsidized, accelerated technologicalacquisition.

The benefits of dosing the gap between girls and boys in primary andsecondary school enrollments have been subsmntial, even in economiesin which female labor participation rates are low. Better educated modt-ers raise betrer educated children, complementing public investments inbasic education. The rapid increase in the educational attainment ofwomen 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 twocrucial functions. First, they encouraged financial savings. Second, theychanneled savings into activities with high social returns. There wasgreat variety in the policies and institutions used to accomplish these

349

_RALLE

ends, but several flmdamental approaches were important in all eighteconomies. 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&ctorin encouraging financial savings. All of the HPAES had periods, usuailyearly in their development, when rcal intcrest rates on deposits werenegative. But in general real interest rates were more stable and moreconsisrently positive in East Asia than in other developing regions,where high inflation has often meant that volatile and highly negativereal 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 thecreation 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 speculativelending. Although the HPAEs had no explicir deposit insurance, theypromoted confidence in the banking system by devising rescue pro-grams as needed, sometimes, as in Japan and Korea, pressuring strongerbanks to absorb weaker banks. In addition, many of the HPAEs encour-aged savings by creating postal savings plans to lower transaction costs ofsavings 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'sbanking sector structures, as well as some measures of bank efficiency(profitability and borrowing-lending margins), suggest that banks in theHPAEs have been at least as efficient as those in other developing regions.Even so, innovation and efficiency undoubtedly suffered from limits oncompetiton. Pardy in response, HPAEs gradually have opened theirbanking systems to increased domestic and fbreign competition. Gov-ernments have also promoted bond and equicy markets to increase thedepth of mamring financial sectors.

Several HPAES supplemented commercial banls-and met the needfor long-term capitl-by creating specialized dvelopment banks.These banks fiuthered the growth of the financial sector by establishingprocedures for project financing and monitoring that commercial banksthen 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 developingcconomies. Flexible labor markets and capital markets with positive realinterest rates meant that wages and interest rates reflected more closelytrue scarcities of labor and capital. Thus resources were generally drawninto labor-intensive production in early stages of development, shiftingto capital- and knowledge-intensive activities as physical and humancapital deepened.

A combination of competitive real exchange rates and moderate pro-tection levels meant that domestic prices of traded goods were closer tointernational prices in the HPAEs than in other developing regions. Thelarge weight of exports in total output, duty-free imports for exporters,and extensive domestic competition meant that all the HPAEs had higherproportions 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 theirrapid growth periods. Hong Kong, Singapore, and the Southeast AsianHPAEs, 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 oftechnology are reluctant to share it, even for a fee. However, the successof 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. Somedeveloping economies, such as Argentina and India, have restricted cap-

35'

0 A: -X1RACLE

ital goods imports to promore the growch of domestic machine-buildingindustries. In these cases, however, domestic machinery generally fallsshort of international standards. Such machinery becomes yet anotherhigh-cost import substitute that inhibits exports.

Limiting the Bias against Agriculure

Agriculture has played a more central role in the H1IAEs than in manyother low- and middle-income economies. Wide adoption of GreenRevolution technology, high investment in rural infrastructure, and lim-ited direct and indirect taxation of agriculture meant that rural incomesand productivity rose more rapidly in East Asia than in other regions.

As in other economies, agricultural sectors in the HPAES were a sourceof capital and labor for the manuficturing sector. But in East Asia theseresouroes were generally pulled into manufacturing by rising wages and re-turns, rather than squeezed out of agriculture by high taxes and stagnantor declining relative incomes. As a result, rural-urban income differentialswere 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. HPAEcivil services range orom the highly meritocratic and insulated bureau-cracies of Japan, Korea, Singapore, and Taiwan, China, to the lesseffective and less insulated public administrations of Indonesia andThailand. Nevertheless, each of these economies has a core of techno-cratic managers. In Indonesia and Thailand, their scope is limited tomanagement 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 felshort of those for equivalent positions in the private sector, but theywere 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 andinsulated fromn political interventions. In particular, recruitment andpromotion were merit based. Third, public employment was accordedhigh status. These factors improved the quality of the bureaucracy, dis-couraged corruption, and created an esprit de corps among civil servantsthat 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 thesecouncils, and of the more informal mechanisms in the other HPAEs, areimpossible to measure systematically. It is likely, however, that econom-ically they improved coordination among firms and improved the flowof information between businesses and government Politically theyhelped 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 grantspecial 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, assessingperformance 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 ofmacroeconomic 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 ofimplicit taxation of households and firms-for example, through thestructure of protection or interest rates controls. One of the main char-

353

MIRACLE

acteristics of interventions in the HPAEs is that in general they havebeen carried out within well-defined bounds limiting the implicit orexplicit costs. Thus, price distortions were present but not excessive,

interest rate controls generally had as benchmarks international interestrates and were binding at positive real levels. Explicit subsidies werealso kept within bounds. Given the overriding importance that each ofthe HPAEs ascribed to macroeconomic stability, interventions thatthreatened to undermine thar policy fimdamenmally, were modified orabandoned-for cxuample, the heavy and chemical industries drive inKorea or the heavy industrialization push in Malaysia. These limits tointervention 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 madepossible by good findamentals or detracted from it is the most difficultquestion we have tried to answer. It is much easier to show that theHPAEs limited the costs and duration of inappropriately choseninterventions-itself an impressive achievement-than to demonstratecondusively that those interventions that were maintained over a longperiod accelerated growth. In this book we have evaluated three sets ofpolicy interventions: the promotion of specific industries or industrialsubsectors, directed credit, and the export-push strategy. We concludethat promotion of specific industries generally did not work and there-fore holds little promise for other developing economies. Directedcredit has worked in certain situations but carries high risk. Theexport-push strategy has been by far the most successful of the threesets 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 morerapid 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. Industialgrowdt tended to be market-confbrnning, and productivity change wasnot significantdy higher in promoted sectors. Although governrents inthese four economies were undoubtedly trying to alter industrial structureto 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 withmarket 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 intermediateand capital goods. Moreover, in Japan especially, but also in Korea, in-stitutional mechanisms that stressed cooperative relations and exchangeof information betveen firms and government introduced a largcamount 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 seemingsuccess of industrial policy in these three economies probably rests noron picking winners-that is, on the adroit selection of which industrialsubsector 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 competitiveframework-in short, contests-may have lessened investment coordi-nation problems and resulted in earlier and better investments in whatultimately 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 beenimproved-in Japan and Korea by the combination of repeated rela-tionships between business and government and the use of exports, notdomestic GDP growth, as the yardstick by which the success of the in-dustrial strategy was measured. In other HPAES, such elaborate contestswere 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 leapfroggingin Indonesia.

355

.X4 N-MIRACLE

Repressing Interest Rates and Directing Credit Worked Somefimes

All HPAEs controlled interesr races on deposits and loans during muchoF their rapid growth. Low real interest rates on deposits and loans in-creased corporare equity by transferring resources from savers (mostlyhouseholds) to borrowers (mosdy firms). At the same rime, by limitingthe spread between deposit rates and loan rates, govemments preventedbanks from earning excessive rents. On the downside, such controls un-doubtedly discouraged banks fiom taking on riskier, nontraditionaldients.

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 influencethe beihavior of firms. In this section we discuss why directed credit wasrelatively 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 mostdeveloping economies.

Interest rate subsidies were smaller in the HPAEs than in other devel-oping economies. In several HPAES, government-directed credit servedprimarily to signal commercial banks and private investors which sectorsenjoyed govemment support Banks and private investors then selectedprojects within those sectors and offered them additional funding at

commercial rates. Moreover, credit was frequently allocated accordingto export performance. Our evidence leads us to condude that creditprograms directed at exports yielded high social returns and, in the casesof Japan and Korea, other directed-credit programs also may have in-creased investment and generated imporcant spiliovers. Even withinEast Asia, however, when interest rate subsidies were large, or exportperformance 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 throughthe early 1980s took place in a world economy in which it was possibleto close domestic capital narkets. Since the early 1970s, developingeconomies have become ever more integrated with the global capital

markeL While this has reduced government? already limited abiity torestrict capital flight and speculative capital inflows, it has also giventhem access to less expensive capital and offered them better possibilities

356

POLICIES AND:P

to hedge external shocks. At the same time, increased integration hasconstrained policymakers, who are finding it more and more difficult tornaintain interest rates at variance with international markets.

The developing economies' growing integration with global financialmarkets can be seeni in the evolurion of their stock markets. Until themid-1980s, foreign investors' access to domestic securities in developingeconomies was often restricted. Since then, fourteen of the twentylargest developing-economy stock markets have become substantiallymore 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 1971and freed domestic interest rates in 1981. Since then, interest rates in In-donesia have followed international rates.

Although capital outflow restrictions are generally ineffective, theydo hinder the movement of some institutional capital, for example, do-mestic pension fimds. They may also hinder the flight of household andindividual savings in instances when the transaction costs of sendingcapital abroad are greater than the incentive for doing so. Because thesebarriers do marer, developing economies continue to be less open fi-nancially than they are conmmercially. This can be seen in the ratio ofgross capital flows to CDP compared with the ratio of exports-plus-imports to C;DP. For the aggregate of all developing economies, theformer is only 13 percent, while the latter is 45 percent.

Nonetheless, the limited impact of capital restrictions in deterringcapital flight has persuaded a growing number of developing-economygovernments 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 byassuring investors they can expatriate their money at will. Even wheregovermnents have not liberalized restrictions, the globalization of theeconomy, particularly increases in trade and advances in transport andtelecommunications, is rapidly eroding the isolation of closedeconormies.

Under these conditions, the choice for developing-economy policy-makers is not between financial integration with the global market andisolation. Rather, it is between two-way integration that permits out-flows and thus facilitates inflows, on the one hand; and de facto one-wayintegration that consists primarily of unauthorized capital flight, on theother. 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 ratesand contest-based direcEion of credit-may have contributed to rapidgrowth in such economies as Japan, Korea, and Taiwan., China. Bur eadiof these economies had competent and insulated bureaucracies andbanks 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 theirfinancial markets to the outside world. Few developing economies todayhave the institutional resources to consistendy impose performance-based criteria for credit allocation. Moreover, the capital markers of alleconomies are becoming more and more integraced& Thus, successful re-pression of interest rates and contest-based credit allocation may only bepossible for the very few developing economies with strong institutionsand 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 broadbased and successful application of selective interventions. Furthermore,of the many interventions tried in East Asia, those associated with theirexport: 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 earlyexport growth-the broad pro-xport stance characceristic of all eightHPAEs remains viable.

Trade policies in all the HPAES (except Hong Kong) passed through animport-substitution phase with high and variable protection of domesticimport substitutes. In all cases, however, policies that strongly favored theproduction of import substitutes to the detriment of exports were aban-doned, and HPAE governments adopted strategic pro-export policies thatestablished a free trade regime for exporters and offered a range of otherincentives for exporrs. This export-push approach provided a mechanismby 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 waysin the HPAES. Hong Kong and Singapore established free trade regimes,

358

POLICIES AN

linking their domestic prices to international prices; the export push wasan outcome of the very limited size of the domestic market coupled withneutral incentives between producing for the domestic or internationalmarket. Both economics made export credit available, although they didnot subsidize it, and Singapore focused iEs efforts on attracting foreigninvestment in exporting firms.

In Japan, Korea, and Taiwan, China, incentives were essentially neutralon azverge between import substitutes and exports. But within the tradedgoods sector, export incentives coexisted with substantial remaining pro-tection of dte domestic market. Export incentives, moreover, were notneutral 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 toexpor- In Korea, firm-specific export targets were employed; in Japan andTaiwan, China, access to subsidized export credit and undervaluarion ofthe 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 isnormnally ascribed to neutral exposure of both import-substituting andexporting 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 theircivil services and public institutions were largely staffed by competentand honest civil servants, and second, because rirms and bureaucratsknew that there Nvas a single yardstick for performance: exports.

Export targets provided a consistent yardstick to measure the successof market interventions. When protected sectors interfered with theexports of other sectors, exporters could successfilly seek redress. Forexample, domestic content rules imposed on foreign direct investors inTaiwan, China were suspended when they interfered with exports. Theemphasis on exporc competitiveness gave businesses and bureaucrats atransparent 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 inputsinto 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 isroo often directed at serving a protected domestic market and hedged byrestrictions 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 worldeconomy of the 1950s to the mid-1970s. Economic growth in the in-dustrial economies was rapid, barriers to trade were declining, andglobal efforts to spur free trade under the General Agreement on Tariffsand Trade (GArr) were the order of the day. Times have changed. Eco-nomic growth has slackened in the mature economies thar traditionallyprovided the main markers for developing-economy exports. Regionaltrade groupings are becoming more and more prominent. Industrialeconomies 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 theexport-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 thatmarkets will be adequate and open and thar an export push, therefore,remains possible. However, global realities will limit developingeconomies' abiliry to adopt rhe more interventionist instrumcnts of ex-port promotion. The experience of both the northern HPAEs, with theirmixed export-push regimes in which protection of the domestic markercoexisted widt incentives or requirements to export, and the SoutheastAsian 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 wouldbenefit from providirg specific incentives to manufactured exports.Modest subsidies to exports could be linked, for example, to the biasagainst 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 firthertilt toward exports can be justified on welfire-economic grounds, evenif the trade regime is neutral between import substitutes and exports.This might argue for functional export promotion measures such asguaranteeing 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 assuremarket 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 late1970s and 1980s, along with the continued rapid growth of exportsfrom the northern HPAEs, indicates that export drives can succeed evenrwhen industrial-economy markets are not expanding rapidly. In con-trast with the success of earlier export drives, which depended primarilyon increased consumption in the industrial economies, the later driveshave depended on consumers i industrial economies buying goodsmade in Asia instead of goods made elsewhere. If industrial-economymarkets remain open, developing-economy exports can continue to ex-pand in this manner.

Contrary to concerns in some high-income economies about aflood of exports from the developing world, these exports remain tinyrelative to the size of the major, industrial economies. In 1980,developing-economy exports of manufactured goods accounted forjust 2.4 percent of total manufactured goods consumption in the Eu-ropean Community (Ec), North America, and Japan. By 1988, despitethe success of Asian and other exporters, this had grown to just 3.1percenL Even if exports from all developing economies had grown asfast as those from Korea during this period, their total share would srillhave 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 currentlycontemplating an export-led expansion, could safely assume that de-mand for their products is infinitely elastic This is even true for HongKong'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 tojust 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 forceddeveloping-economy exporters to rely heavily on industrial-economymark-ets. Since then, formerly closed economies in Asia, Sub-SaharanAfrica, the Middle East, and Latin America have responded to debtproblems, slow growth, and the need to raise fresh capital by becom-ing more and more outward oriented-and therefore more open toimports. They have assumed greater obligations under the GA1T andthrough numerous regional trading arrangements.' Under these cir-cumstances, the potential for Soudt-South trade has increasedconsiderably.

But while global markets are growing, they are also becoming moreand more competitve. Each economy seeking to emulate the EastAsian axport-push model will encounter difficulties because manyother economies are trying to do the same. For example, Collins andRodrik (1991) estimate that Eastern Europe and the republics of theformer Soviet Union will increase their share of world trade from 10 to23 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 openmarkets under the auspices of the GAIT (see box 7.1). The other is thegrowing role of regional trading blocs, which present a major opportu-nity for increased global trade (see box 7.2). Who provides the mnarketif all economies push exports? The dramatic growth of trade withinAsia 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, productionin Asia has become more and more international, with capital goodscoming 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 worldhavc 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 ambitioagenda. Even at this late dare iris impossible to know how successfillthe round will be. Success could offer inip&rmnt-ropportuiities fbr:developing-economy exports, particularly in sucb arcas as agriculturealand tropical products. Progress in eliminating the MuiltifibreArrangement 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, develoneconomies.; -

--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, willimposc-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 seven 1 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 ECdes are easily overstaie. Unlikc-their predessors,m..arkets under schem such as thc gene ed S

th: en 5 wblcs 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 andbring their practices more closely in l ne with those in industrialeconomies. Those that attempt to use more interventionist versions ofche export-push model will risk retaliation from industrial-economymarkets or punishment under the GATT. For some policy options, suchas financial repression, increasing integration with global markets willlimit the scope for intervention.

Over the years, developing economics have daimed and received ex-emptions from GATT rules (Whalley 1990). By invoking Article XII, tosafeguard 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 developingeconomies have been free to implement trade policies of their choice atbome while benefiting from the openness of industrial economies. Butas GATT shifts from a narrow focus on tariffs to a broad approach thatseeks agreements on rules, developing economies will find it more andmore difficult to hang onto exemptions. Under Uruguay Round pro-posals, rules governing intellectual property, subsidies, countervailingduties, and trade-related investment would eventually be the same for

industrial and developing economies. Although developing economieswould have up to eight years to bring their practices into line, the tradeobligations of developing economies clearly are increasing. Changing

rules about dumping illustrate the case. Industrial economies are in-creasingly applying antidumping and countervailing duty measures todeveloping economy exports (Lawrence 1991). These measures willmake it more difficult to follow the example of Japan and Korea, whichprotected domestic industry in order to subsidize exports.

Developing economies that seek access to the emerging regionaltrading blocs may find that it comes tied to increased obligations thatcould 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 itsprovisions 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 infantindustry 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 pushhas 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 theNIEs 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 consistentwith developing economies' emerging obligations for market aco- rs andlimited subsidies. Improved institutions-mak-ing customs services,duty drawbacks, and fiee trade status for exportcrs and their supplierswork well-are all consistent with agreements on intemational tradcand arc effective mechanisms for export promotion. Access to exportcredit, 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'ssuccess. We found that the diversity of experience, the variety of institu-tions, and the great variation in policies among the high-performingAsian economies means that there is no East Asian model of rapidgrowth with equity. Rather, each ofthe eight economies we studied usedvarious 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 thatthe HPAES used reflected sound economic fundamentals: they enhancedthe working of markets, helped prices communicate information aboutrelative economic costs, and fostered competitive discipline.

Some economies-notably Japan, Korea, arid Taiwan, China-wentbeyond 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 becauseof highly unusual historical and institutional circumstances. The govern-ments in these economies were more successfi than most in combiningthe benefits of economic cooperation with the benefits of vigorous com-petidion. They did this by creating contests with exports as the yardsick ofsuccess, with subsidized credit and other government-controlled favors asprizes 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 incentiveswithout a clear link to economic perlormance and strong institutionalsupport. 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 neglectagriculture; 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 orattempting to leap srages of technological development will generally

fAil; strongly negative real interest rates and large subsidies to borrowersdebilitate the financial system; and directing credit without adequatemonitoring and selection of borrowers distorts allocation. Finally, we

found that a successful export push, whether it results from an openeconomy and strong econornic fundamentls, or from a combination ofstrong findamentals and prudently chosen interventions, offers- high

economic gains. Of all the interventions we surveyed, those to promoteexports 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 inscrumentsand institutions to achieve the objectives of continued growth withequity. In many cases these reforrns involve reducing& modifring, orabandoning policy instruments thar were judged to have succeeded inthe 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 providefurther valuable lessons on how successful policy instruments shift overtime, as the relative roles of markets, the public administration, and theprivate sector change in response to economic and social development

The experience of the HPAEs broadens our understanding of the rangeof policies that contribute to rapid growth. it also teaches us that will-*ingness to experiment and to adapt policies to changing crcumstancesis a key elemenit in economic success. What we have not discovered fullyis why the governments of these economies have been more willing andbetter 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 challengefor 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 orderto make a wider range of policy options available.

Note1. 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 Mexicosoildate more outward-oriented strategies. Subregional that aims at reducing tariffs to zero by 1998. Whilearrangements are flourishing. In Latn America, for exam- progress coward formal arrngements witrin Asia haveple, these include the North American Free Trade Area; been less apparent (although the free trade area within theMercosur (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 backgrundstudies, outputs of related World Bank-sponsored re-search, economic and sector work carried out by Bankoperational staff; and the errensive academic literature

_L on the East Asian economies. Citations in the text andfootnotes refer to specific sources used. The principal sources and specificcontributions to individual chapters arc discussed below. Lists of back-ground and other commissioned papers and refrences for works cited areat the end of this note.

Many members of the World Bank Staff assisted in the design of theresearch 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, AjayChhihber, 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, EmmanuelJimenez, Christne Jones, Koji Kashiwaya, Homi Kharas, Miguel Kiguel,Elizabeth King, Michael Kline, Brian Levy, Samuel Lieberman, WillMartin, Deepak Mazumdar, Saha Meyanathan, Vikram Nehru, ArvindPanagariya, Martin RIavallion, D. C. Rao, Anandarup Ray, Setsuya Sato,Andrew Sheng, Lyn Squire, Vinaya Svaroop, Shekhar Shah, Jee-PengTan, 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 andmaterial indude: Alice Amsden, New School for Social Research; MasahikoAoki, 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, StanfordUniversity; David Lindauer, Vassar College; Kazushi Ohkawa, InternationalDevelopment 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, Universityof Rochester, Helmut Reisen, OECD; Hilton Root, Sranford University; 11Sakong, Institute for International Economics, Seoul; Herman van der Tak;Shujiro Uraca, Waseda University; Rovert Wade, Institute for AdvancedSrudy, Princeton; and Toni Yanagihara, Hosei University

Chapter 1The 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 versionof the data contained in Summers and Heston (1991). A summary ofthe results of the resch projcct are contained in a symposium issue ofthe Journal of Monetary Economics (forthcoming). Additional data are

drawn from the World Bank Economic and Social Data Base (BESD) andfrom the World Development Report data base. The TF estimates weremade 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). StanleyFischer generously provided his country-specific estimates of TFPgrowth, 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 Sciffand 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 incomedistribution. The discussion of private investment draws heavily onMadarassay and Pfeffermann (1992). Box 1.1 draws on Birdsall andSabot (1993). Box 1.2 draws on Pack (1993) and Pack and Page (1993).

Substantial research on the economic tmnsition in China has beenconducted 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 rapidgrowth of China's coastal provinces. Preliminary outputs of that researchwill be available in the fall of 1993.

Chapter 2This chapter draws primarily on Pack (1993), Page and Petri (1993),

and Stiglitz (1993). The description of contests also draws on Leipzigerand Kim (1993), Uy (1993), Okuno-Fujiwara (1993), and Ito (1993).WorldDevelopment Report 1991 provides the basis for the discussion ofthe market-friendly approach. Box 2.1 draws on World DevelopmentRe-port 1991. Box 2.2 draws on a paper prepared by Nancy Birdsall andJohn 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 andDevelopment 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); andSalleh, 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 Easterlyand Ross Levine. The concept of "export push" is due to Coln Bradford(1990). Villiam Easterly, Lant Pritchett, and Michael Walton providedmajor 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. Box3.4 was prepared by Varuni Dayaratna.

Chapter4Chiah Siow Yue, Oh-Hyun Chang, Chaianan Samudavanija, Hadi

Soesastro, and Bruce Tolentino provided information and analysis fortheir countries. Hilton Root and Mhary Shirley made imporrant contri-butions. Discussion of the role uv small and medium-size enterpriseshac drawn heavily on the results of a research project, Support Systemsfor Small and Medium Enterprises in East Asia, directed by Brian Levy.

37'

R&fl(c OTE

Chapter 5Material on human capital accumulation is drawn primarily from

Birdsall and Sabot (1993). Robert Gassen provided the material onvocational training. Data and analysis on savings and investment aredrawn primarily fiom Uy (1993), Stiglitz (1993a, 1993b), and Sriglitzand Murdock (1993). YoonJe Cho, Dimirri Vitras, and Andrew Shengmade 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 TaxAdministration 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 byLeora Friedberg. Andrew Sheng provided the material on which box 5.3is based. Box 5A was writren by Jennifer Keller.

Chazer 6The 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. Thesections on industrial policy and exports arc based on Pack (1993) andPack and Page (1993). Data on axport performance are drawn from theBESD and World Development Report 1991 data bases. Discussion at the1993 Canegie-Rochester Conference on Public Policy were helpful inimproving the sections on export extemralities and growth.

Box 6.1 was written by Brian Levy. Box 6.2 was prepared by LeoraFriedbeg. Box 6.3 draws on a recently cornpleted internal World Bankreport on Indonesias industrial seaor. Box 6.4 was prepared by LeoraFriedberg. Box 6.5 was prepared by Danny Leipziger.

Chapter 7Discussions 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 andthe Developing Countries (World Bank 1993a). Vinod Thomas made anumber of helpful comments.

372.

BIBLI9CAi

Background Papers Samuudavanija, Chaianan. "High Speed Growth and HighPertfrmance in a Technocratic Polity: The Thai Case."

Behrman, Jere R., and Ryan Schneider. "An International Shah, Shekhar, and Subodh C. Marhur. 1992. "SourcesPerspective on Schooling Invesrment in the Last Quar- of Labor Markecr Flexibility in East Asia: A Review ofter 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 RoleHuman 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 Economiesomies: Part I, Insulation Mechanisms and PublicSectr-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 andJapanese 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 LessonsMarkers 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: PublicEast Asian Model." Policy and Economic Development?

Murdock, K., and Joseph E. Sdiglim "The Effect of Fi- Thomas, Vinod, and Yan Wang. Government PoLiciesnancial 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 JapanHigh-Performing Asian Economies."

Pack, Howard, and John M. Page, Jr. "Accumulaton, Er- These papers were presented at a Seminar on Public Poliyports 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 Postwarvelopment 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. "Productiviryioka. "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.: JohnsEvidence from the Postwar Period." Hopkins University Press.

Okuno-Fujiwara, Masahiro. 'Government Business Rela- - 1991. Eco nomic Polcies in the Pacifx Area Developinguionship 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 WorkingMedium Enterprises in Japan.' Paper 285 5, Cambridge, Mass.

Yoshikawa, Hiroshi. 'Demand-Led Growth and Income _. 1990. "Govenmment Spending in a Simple Model ofDistribution: 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. 'IncecmaonalComparisons of Educational Artainment." Paper pre-

Aheuwal, DcherOJud. 1985. dssa . sented at a conference on How Do National PoliciesNew 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 Dosity 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 Someduction 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 GrowthD.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 PullsStntutreandProdiuctiviy in Developing Counzries Bal- undcr Trade Liberalization." joUrnal of Internationaltimore, 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 andand Mvledium Industrial Exporters and Their Support Spillovers London: RoudedgeSystems." Report prepared for the World Bank, Policy Blumenthal, 1. 1970. "Savings in Postwar Japan." EastResearch 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 CorporateWorld 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. NewExchange Control Regimes. Cambri"-e, Mass.: York: Oxford University Press.Ballinger Press, for the National Bureau of Economic Boneili, R. 1992. "Growth and Productivirv in BrazilianRescarch. 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 Productivetext, 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: Mythston, 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. Thirddon 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 Garysticute 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 and1993. "Underinvestment in Education: How Much EastAsia. Princeton, N.J.: Princeton University Press.Growth Has Pakistan Forgone?" Paper presented at Brimble, PeterJohn. 1993. "Industrial Development andthe Ninth Annual General Meetng of the Pakistan Productivity Change in Thailand." Ph.D. diss., JohnsC..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 Yorkin 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 Theopment 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 AlternativeC ic, Caridge:, Camidg e, Uni Press. SInterpreton of Japancse and Korean ExperiencesCarroll, 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 ondion." Paper presented a1t the Carnegie-Rochester Th Efetvns*fCei-Plce nEs saPublic 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 ThailandKorea, 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 Industrialford 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 SouthChenery, 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. Virgindon: Oxford University Press. Islands, May 7-10.

-. 1974. Trade Poicy and Economic Welure. London: Dekle, Robert. 1988. "Do the Japanese Elderly ReduceOxford 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 theThe Est Asian Miracle. World Bank, Policy Research Japanese Experience." Background paper for The EastDepartment, Washington, D.C. Asian Miracle. World Bank, Policy Research Depart-

Dahlman, Carl J. 1990. "Technology Strategy in the ment, Washington, D.CEconomy 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 ofSector 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 EconomiesCase Study in Thailand.' Industry Series Paper 24. Really Do Grow More Rapidly: Evidence from 95World 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 ofTechnological Dependence ro Technological Devel- Producivity Groth in Soudl Koreai MaencEcturnngopmenr: 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: GrowthChina. Leson of Eas Asia. World Bank Country RegressionsMeettheGangofFour." PaperpresentedStudy. 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, andUnited Nations Industrial Development Orgization. Lawrence H. Summers. 1993. "Good Policy or Good

Luck? Country Growth Performance and TemporaryDe 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? WorldQuareer4journal ofEconomics 10C6(2): 445-502. Bank, Washington, D.C., February.

1993. 'How Robust Is the Growth-Machinery Easterly, fiiam, and Sergio Rebelo. 1993. "Fiscal PolicyNexus?' 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 NationalBank, 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 SectorInternational 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., PublicSectortiviry 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 ofGrowth, 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 LaborNations Industrial Development Organization. Contribure to Economic Success? Labor Relations

Edwards, Sebastian. 1988. Exchange Rate Misalignment in and Markets in East Asia.' Background paper for TheDeveloping 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 ofEconomics 39(1): 31-57. Japanese andf!nternanoonalEeonomies1(2): 168-94.

Elias, VictorJ. 1990. "The Role oFTotal Factor Produc- Frischtak, Claudio R., Bita Hadjimichael, and Ulrichtivity on Economic Growth.' Background paper for Zachau. 1989. Competition Polcesfor Ina¼strializingthe Worlk Development Report 1991. World Bank, Countries. World Bank Policy and Research SeriecOffice 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.: Harvardciency in OECD Countries." In R. Caves, ed., Indus- University Press.tialEEfficiernyinSilNaionss Cambridge, Mass.: Mrr Ghose. T. K 1987. "The Banking System of HongPress. 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 StaffKorea, 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. "TechnologyResearch Department, Washington, D.C. Policy." In Ryutaro Komiya, Masahiro Okuno, Ko-

Fischer, Stanley. 1993. "Macroeconomic Factors in taro Suzumura, eds., IndustrPolicy oflapan. NewGrowth." 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." PREply, 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'son 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. 1984 a. 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 Householdment? Evidence From Morocco." Indus-ry Series Savings Rate So High? A Litcrature Survey."JounalofPaper 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 PopulationDirect Foreign Investment? Evidence from Panel Dara and Other Factors." The Eronomic Strdies Quarterlyfor 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 SystemsIncendvesJfrInvestment in Developing Countries. New Supporting Technical Advance in Indusry. The CaseYortc 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 NationalArchive, Review of WorldEconomics 125: 522-46. University.

ed. 1990. Textles Trade and the Developing Coun- Hulten, Charles Jr. 1989. "The Embodiment Hypothesisins: 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 InstirutionalTwo 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 Yorklgar. 1992. 'Thc Effort of Financial Liberalization on Academic Press.

379

India. Various issues. Reserve Bank ofindia Bdlletin. JDB/JERI (Japan Development Bank and Japan EconomicIndones;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 ofMobility from Central Government .o Public Corpo- Thailand's Economic Development." Working Paperration." 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 ContainerStatistics Washingronj D.C. Terminal, Sports Toro." Paper prcsented at a World

rsu(Indonesian Economists' Association). 1990. The Un. Bank Conference on The Welfare Consequences ofdemandng 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."joumalofJapAkiyoshi 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 Researchbridge, 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 MajorPolicy offapan New York: Academic Press. Count7is. Boston: Kluwer Academic Publishers.

Itoh, Moroshige, Kazuharu Kiyono, Masahiro Okuno- Kawaura, Alihiko. 1991. "Identification of the Size andFugiwar, and Kataro Suzumura 1988. 'Industrial Pol- Direction of the Post-War Credic Allocation inicy as a Corrective to Marker Failures." In Ryutaro Japan. World Bank, Financial Sector DevelopmentKomiya, 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 YorkYork: 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 EastAsianProfits 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 Eduational1970s: 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 StatisticsEconomic 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 StrategiesGrowth 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 DevelopmentWorld 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 Reviewsonne 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 ThinkingCapital 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 PoliticalSchooling 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 SouthAmerica, 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 inGrowh? 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 Incomeversity 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, Kualaing Paper 54, Queen Elizabeth House, OxfordlCen- Lumpur. Institute of Strategic and Internationaltro 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, PolicyModel.' 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 PossibilitiesTraining Modes in the Republic of Kora? Education of Emulation." Working Papers, Asian Employmentand Trainiing Series 6. World Bank, Education and Programme, ILO-ARTEP, BangkolkTraining 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 Developingand Su-Yong Song. 1992. The Distribution ofIncome Countnies. London: Oxford University Press.

-and WeaLrh in Korea- Washington, D.C.: World TBandh 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 ProductivityAssessment.? World Bank Economic Review 7(1): Change-Evidence from Chile" Policy Research6585. 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 ofversiry 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 theEconomnic Developmenr in Comparative Perspec- Rules of the Game." Policy Research Working Paperive." 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 ofSaving Rare and Economic Growth: Consumers' Financial Repression in an Economy with PositiveStimulus ro Purchase Durable Goods." Paper prepared Real Races." Background paper for The EmFt Asianfor 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 YorlcCorporate 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 DevelopmentAdjustment 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 Compensationcoming. 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 ReformSavings 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 117Honor 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.CDepartment. 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. "TotalWorld 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 andin 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 AnnualChange 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-PerfirmingAndres Sapir, eds., Trade Theory and Economic Refonn: Asian Economies." Background paper for The EastNorth, 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. "Industrialchochiku 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 CreditHoken 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 Asianmum, 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 offord, 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.CChenery 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 enHolland. 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 andPolicy 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 InvesunentPerforming Asian Economies." Background paper for in Education: A Global Update." Policy ResearchThe ast Asia Miracl. World Bank, Policy Research Working Paper 1067. World Bank, Larin America andDepartment, 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 paperfor The East Asian Miracle. World Bank, Policy Re- Rosenberg, Nathan. 1976. Perspectives on ecohnoloVsearch 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 andnomic 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 Perspectivavestrment 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 OxfordDeveloping 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 DhevanLong Run Growth!. journal of Poitdcal Ecotnomy Manatan 1993. Growth, and Srcr rat

9(3) 50>21 ~ ~ ~ ~ ~ ~ . Meyanathan. 1993. Growab, Eqity, andf SDxrd99(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 GrowthWashington, 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 irTf[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: ResultsBanks? 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 ofSynthesis 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 witheds., 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 anEconomics." 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." TechnicalIthaca, 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 TheorymenctinEght Asian Counris. London: Maanilan. ofEconomic6Growth." QuareeryJornalofconomics

Shah, Anwar, ed. Forthcoming. Fiscal Incentives for In- 701)65-9.Tvetcmentt in Devloig Coun Ne York Oxford -- 1957. "Tehnca Chang and die Agreae Pro-Uesnivrersity Press ducton Function.' Reizew of Economics and SttsticsUniverilty Prem ~~~~~~~~39(3): 3 12-20.

Shah, Shekhar, and Subodh C. Madiur. 1992. Source Soon, Tek-Wongand Suan Tan. 1992. Public Policyof 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 theMirrd, 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 DevelopingShea, 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 PresYung Chul Park, cds., Financia Development Japan Srinivasan, T. N. 1992. "Datbas for DevelopmentKorea, 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 Theton, D.C. Korean Heavy and Chemical Industry Drive."

_. 1990. "The Art of Bank Resuucturing2 Isues and Research reporr Harvard Institute for InternationalTechniques." 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 Chinznancial 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 inPolicy." 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 Inrious 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 TliailancLin Taiwan. Taipei: Institute of Economics, Academia Thomas, Vinod, and Yan Wang. 1993. "GovernmentSinica. 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.Newtures, 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 MajorPaper 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 forTollison, 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 AsianPublishers. 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 atOECD. 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 dieTokyo. 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 DevelopingNew York. Countries. New York: Oxford University Press.

_. Various years, b. Eneg Statistics Yearbook. New Wei, Shang-Jin. 1993. 'Open Door Policy and China'sYork. 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 CameKorea 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 theHomewood, 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 ThailandWorld 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 Bankcial 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: ChoosingYoung Son, and Ezra F. Vogel, eds., Korec The Strains between Impect Altenzetives. Cambridge, Mass: Mrrof 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, andEast Asian Success?" Annab oft heAmerican Academy of North Africa Region, Washingron, D.CPolitical 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 KotaroCounties. 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 ReformRepor- 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 theannual hIder ofPublications, which contains an alphabcical tide list and

indexes of subjects, authors, and countries and regions. The latest edition isavailable 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] $ ~~~~~~~~~ClHINA6r ~ ~ ~ ~ ~ N r - 1)

C't ,.. -6r- j,

>~~~~~~~~~~~~~~~L , _-. - r

j2kmmx~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

%L:0 be fI/ t

. - - - D D ;, ;; - .. ~~~~- *.

SfDN I

Ais,~ ~ ~ ~~~~~~- 1