S ~ ~~~~~~~~~.

''

Cll ( rOS

s~~~~~~~~~~~~F f)

:.;] IX
grX~~~I
,0

fib~~~~~~~~~~~~~~~~~~~~~-

Lul=lpph. Jakarta ; z - .r.

* *NWLI

_' d ;. 0 X\ - .• I) 0 0
6S -.MF
:6

C)~~~C
The
East Asian
Miracle
EconomicGrowth
and PublicPolicy

Publishedfor the World Bank

CxCORDUNIVERSrrYPRS
OxfordUniveirmPressw

OXFORD NEV YORK TORONTO

DELI BOMBaY CALC;UJrA MADRAS KARACHi
kUAA LUMPUR SINGAPORE HONG KONG TOKYO
NAIROBI DARES SALAAMCAPE TOWN
MELBOURNEAUCKLAND

andasidad companiain
BERLINIBADAN

0 2993 7lmriarmnd Ba&nkfor

Reomtruawn
wdDevaop mewl OPJE U'2LD HAAh
1818 H Sreet N.W.
Mathinog D.C 20433. USA

Pub hed by Oxfrd UniAnityPrs, fINC.

200MAdiwrAvemd4New York N. Y 10016
Oxfordis a rqwra bradeark of OxfordUnirsiyPreaz
A(1rgth raerd Noparjrh5pub&icatum 6be reprw l .woredina rrtrudqsfem
or Drwmm&mc4
in any farm or by any meas, deamn, merhank4 pumwpy4
rewonmg&
ore.z se,iw etvepdorpermrnuonfxf0uirdUriviy Prm

Mamftiazdn she Unisd s /Ameria

Fon paning&Spemner 1993

Ciogun
ph s-At the wp. a ricefud inJava, Indoniaig;c resy
ofMatricrAiser.
At
r beraomI, the por ofPuan, Rephfir ofKar=cowty ofJindre &pjDA
The map tnappars o n the izsiderova, wbib shws t6c eight eom wigs diiwed in Mis
bok wsprqmred soyfortke onvenie ofthe reader the dres-arions andpresramt
ofmaeial in it d wet imply the owrr ofrmy opbun warsoener on thepavt of th
World Bank its affiiner or isr Bad ormember conries onwMtming the egl as1 0f ay
ewonM,
mq rroj cirt or or /t&e anthrifia xref, or rcwemning rthedeimiwdon
qfirs boitames or in natiolaie

rfCwFs Caragtoing-Publadon Dar

The EasAswaw mirade :omwgrvww5
w =d; bicpAl
p. cn.-(MordBankpdiy r=* reponw)
Incdaw r 3uietpieal
ner
,IrN 0-19-5209934I
L. AsEa;rnee-Ecwmicpli 2- tE a-Econidc cendo-,
3. Asi Sawi&s-Ecenamicpole 4. Ask CitiomS
L. WorldBank f. Se,ie
HC4625.A275 1.993
338.95-c20 9S33046&

ismv1G20-0851

Torprepaper anom theAmeria Natnax SxdrdfrPawzw*w wf

P1pfr1Jiwd MreLw
ialr Z39.48-1984
Foreword

EBATESON THE APPROPRIATE ROLE OF PUBLIC POUCY IN

economic development have occupied policymakersand
scholarssince the study of developing economnies bcgan in
earnestat the dosc of World War IL.The successof many of the
economiesin EastAsiain achievingrapidandequitablegrowth,oftenin
the contextofacivist publicpolicies,raisescomplexquestionsaboutthe
relationshipbetweengovernment,the privatesector,and the markern
Seningly,therapidlygrowingeconomiesin EastAsiausedmanyof the
same poficy instrumentsas othcr developingeconomies,but with
greatersuccess.Understandingwhichpoliciescontribucedcotheir rapid
growth,and how,is a majorquestionfor researchon developmentpol-
icy.ForthesercasonsI announcedat the timeof the 1991AnnualMeet-
ings of the Board of Govemors of the World Bank in Bangkok,
Thailand,that our DevelopmentEconomicsVicePresidencywouldun-
dertakea comparativestudy of economicgrowthand public policyin
East Asia.
This volumeis the summary ch. appears
as the firstin a seriesof PolicyRcsearchReports,which intendedto
arc
bringto a broadaudiencethe resultsof research on developmentpolicy
issuescarriedoutbystaffof theWorldBank.Asreportson policyissues,
we intend that theyshouldhelp us to take-stockof what we know and
cIealy identifirwhat wedo not know;theyshouldcontribureto the de-
bat in both the academicand poli-y communitieson appropriatepub-
lic policy objectivesand instrumentsfor developingeconomies;and
they shouldbe accessibletX nonspecialiscs.Becausetheysunmarize re-
search,wealsoanticipatethat PolicyResearchReportswill provokefar-
ther debate,bothwithinthe Bankand outside,concemingthemethods
usedand the conclusionsdrawn.
What doesthisreport tell us about the EastAsianmiracle? The re-
seawhshowsthat most of EastAsiasextraordinarygrowthisdue cosu-
perior accumulation of physical and human capita. But these
economiesWerealso better able than most to allocatephysicaland
 buman resourcesto highlyproductiveinvestrents and to acquireand
mastertxchnology.In thissensethereis nothing'miraculous"aboutthe
EastAsianeconomics'success;eachhas perfirmed theseessential func-
tions ofgrowthbetter than mosrother economies.
The eight economicsstudied used very differentcombinationsof
policies,firomhands-ofTtohighlyinterventionist.Thus, themisno sin-
gle 'East Asianmoder of development.This divcrsityof experiencere-
inforcesthe view that economicpolicies and policy advicemust be
country-specific,
ifthey areto be effectivec
Buttherearcalsosomecom-
mon threadsamong the high-performingEast Asianeconomies.The
authorsconcludethat rapidgrowthin eachcconomywas primarilydue
to theapplication of a set of common, market-friendly economicpoli-
cies, leadingto both higheraccumulationand berter allocationof re-
sources.While this condusionis not strikinglynew,it reinforcesother
rcsearchthat hassressed the essentialneed fordcvelopingeconomicsto
ger the policy fundamentals igIL The researchalso firther supports the
desirabilityof a two-trackapproachto developmentpolicyemphasizing
macroeconomic stabilityon onehand and investmentsin peopleon the
ocher. The importance of good macroeconomnic management and
broadly based educational systemsfor Easr Asia'srapid growth is abun-
dandlydemonstrated.
The reporr alsobreaks some new ground. It condudes that in some
economics,nainly thosein NortheastAsia,someselectiveinterventions
contibuted o grwth. and it advancesour undersmadingof thecondi-
tions required for interventionsto succeed. The authors argue thar
whereselectiveintervcntionssucccededthey did so becauseof th es- o
sential prcrequisites.First, they addressed problems in the functioning
of markets.Second,they took placewithin the contex of good,finda-
mentalpolicies.Third, theirsuccessdependedon the abilityof govern-
merts to establishand monicor appropriateeconomic-performance
criteriarelated to the interYcntions-in the authors' terms, to create eco-
nomic contests. These prerequisitessuggesr that the institional con-
text withLinwhich policies are implemented is as important to their
successor failureas thepoliciesthemselvcs,and the reportdcvotessub-
stantial artention to the institutional basesfor East Asi-a rapid growth.
While thesefactorshelp to explainwhy apparentlysinilar policies
did not succeedin mranyothereconomies,the report alsoleavesunan-
sweredmany importantquestions.The market-oricntedasj ctxsofEast
Asia!spolides can be recommendedwith few reservations,
but the more

vi
institutionally demanding aspects, stich as contest-based interventions,
havenor been successfullyused in other settings.Noneconomic factors,
induding culture, politics, and history, are also important to the East
Asian successstory. Thus, there is still much to be learned about the in-
teractions between policy choices and inscitutional capability and be-
tween economic and noneconomic factors in development. Work in
theseareaswill continue beyond this report.
The support of the Government ofjapan for the researchprogram on
the high-performing Asian economics is gratfilly acknowledged.The
report is a product of the staff of the World Bank, and the judgments
made herein do not necessarilyreflectthe view of its Board of Directors
or the governmentsthey rcprescnt.

Lewis T. Preston
President
The World Bank

Augur 1993

vii
Contents

7TeRear Team xi
Ack3wwkdgmnae xiii
Defmiufons xv

Overview-rThe Making of a Miracle

The Essenceof the Mirale: Rapid Growth with Equity 8
Policiesfor Rapid Growth in a ChangingWorld Economy 23
Note 26

1 Growth, Eqrity, and Economic Change 27

Rapid and SustainedEconomic Growth 28
Declining Income Inequality and Reduced Poverty 29
Dynamic AgriculturalSeaors 32
Rapid Growth of Exports 37
Rapid DemographicTransitions 38
High Investment and SavingsRares- 40
Creatng Human Capital 43
Rapid Productivity Growth 46
Appendix 1.1:Accounting for Growth 60
Appendix 1.2: What Do Tess of Cogniive SkillsShow?- 70
Notes 76

2 Public Policy and Growth 79

PolicyExplanations 81
The Functional Growth Framework 87
Notes 103

3 Macroeconomic Stabiity and Export Growth 105

Pragmatic Orthodoxy in MacroeconomicManagement 106
Creatng an ExportPuLsh 123
Appendix 3.1: Economricand PoliticalTimedines 148
Notes 156

ix
4 An Insitutional Basisfor SharedGrowth 157
AchicvingLegitimacythroughSharcdGrowth 158
theEconomic
Insulating Technocracy 167
WooingBigBusiness 181
Notes 189

5 Strategiesfor RapidAccumulation 191

Explaining HighHumanCapitalForrnation
EastAsia's 192
ExplainingEast Asia'sHigh SavingsRates 203
ExplainingEast Asia'sHigh Investment Rates 221
Appendix5.1:GrangerCausaliyTestsforSavings
RatesandGrowthRates 242
beween
Noteon theRelationship
Appendix5.2:Technical
InterestRatesandGrowth 245
Notes 255

6 UsingResourcesEfficienty:Relyingon
Madretsand Exports 259
Explaining Use
EfficientResource
EastAsiais 259
UsingtheMarkenLaborMarketsin EastAsia 261
theMarkctFinancialMarketsandAllocation 273
Assisting
MarketTrmde
UsingtheInternarional Policy
andIndustrial 292
HowManufactured Productiviy 316
ExportsIncreased
Appendix6.1:TestingdheImpactofIndusaial
Change
Policyon Productivicy 326
berween
Appendix6.2:TestsoftheRelationship TFP
ChangeandTradePolicies 337
Notes 342

7 Polices and Pragmatismin a ChangingWorld 347

FoundationsofRapidGrowdh-Gettingdhe
Fundamentals Right 347
CreatingInstitutions
toPromoteGrowdi 352
in Markets
Intervening 353
Note 368

Note
Bibligraphic 369
The Research Team

T HIS POLICYRESEARCH REPORTWAS PREPAREDBYATEAM LED

by John Pagcand comprisingNancyBirdsall,Ed Campos,W
Max Corden, Chang-ShikKim, HowardPack,RichardSabot,
Joseph E. Stiglitz,and MarilouUy. Robert Cassen,WilliamEasrerly,
RobertZ. Lawrence,PetcrPetri,and L-nt Pritchettmade majorcontri-
butions.LawrenceMacDonaldwasthe principaleditor.Casestudiesof
the seven devclopinghigh-performingAsian economieswerc under-
takenunder the directionof DannyLeipziger.The tcamwasassistedby
MariaLtiisaCicogniani,VaruniDayaratna,LeoraFriedberg,Jay Gon-
zalez,JenniferKeller,MayKhamis,SoniaPlaza,MyriamQuispe,Carol
Strunk, and AyakoYasuda.The work was initiatedby LawrenceH.
Summersand was carried our under the generaldirectionof Nancy
Birisall.
The editorial-productionteam for the report was led by Alfred
Imlhof The support staffwas headedby SushrmaRjan and included
MilagrosDivino,Jan-MarieHopkins,Anna MarieMaranon,and Linda
Oehler PollyMeansgavegraphicalassistance.BruceRoss-Larsonand
Meta dc Coquercaumontprovidedadditionaleditorialsupport The
Mapwas designedby jeffreyN. Lecksel.MiklaIwasakicoordinatcd
workin Japan.

XI
Acknowledgments

REPARATION
OF THIS REPORTDREWON SLVERALOTHER RE-
lacedresearchprojects.Studiesof theJapanesemain bank sysrcn
and civilservicewerccarriedout under thedirnctionofHyung-Ki
Kim.Yoon-JeCho and DimirriVittasdirectedthe projecton the effec-
tivenessof crcditpoliciesin EastAsia.BrianLevydirectedthe projecton
supportsystemsforsmalland medium-sizeenterprisesin Asia,andJohn
Pagecoordinatedprojectson lapaneseperspectives on publicpolicydur-
ing the rapid growthperiod, and tax policyand tax administrationin
EastAsia.
Manyindividualsinsideand outsidethe WorldBank providedvalu-
able contributionsand comments.(Specificacknowledgmcntof their
efifrts is madein the BibliographicNote at dceend of the book.) Par-
ticularthanksare due ro the followingseniorofficialsof the economies
studiedwho reviewedand comment"don an earlierdrafc Seung-Chul
Ahn, Isao Kubota, Shijuro Ogata, J. B. Sumarlin,Teh Kok Peng,
Chilao Tsukuda,Tan Sri Dato Lin SeeYan,and CesarVirata.Vinod
Thomascomnmented on severaldraftsand coordinatedthe commentsof
his colleaguesin the EastAsiaand PacificRegionof the Bank.Com-
mentsbyCad Dahlmaan, KemnalDervis,MarkGersovitz,RalphW Har-
bison, Magdi kskander,Hyung-KiKim, Danny Leipziger.Johannes
Linn, GobindNankani,MiekoNishimizu,Guy P. Pfeffermann,D. C.
Rao,ant MichaelWaltoncontributedto the improvementof the book.
Preparationof the reportwasaidedby seminarsorganizedby the In-
donesianEconomistsAssociation,the EconomicSocietyof Singapore,
the Monetary Authority of Singapore,StanfordUniversity,and the
WorldInstitutefor DevelopmentEconomicReseareh.The financialas-
sistanceof the Govemrnment ofJapan is gratefullyacknowledged.

...
Definitions

EconomyGroups
OR OPERATIONAL AND ANALYTICAL PURPOSES. THE WORLD
Bank's main criterion for dassifying economies is gross national
F product(GNP) per capita. Everyeconomyis dassifiedas low-in-
come, middle-incomc (subdividedinto lower-middleand upper-mid-
dle), or high-income. Other analytical groups, based on -:gions,
exports, and levelsof extemal debt, are also used.
Because of changes in GNP per capita, the economy composition of
each income group may change from one Bank publication to the next
Once the dassificationis fixedfor any publication, all the hisroricaldata
prescnted are basod on the same economygrouping. The income-based
economy groupings used in this sEudyare defined as folows
* Low-income economiesare thosc with a GNPpcr capit of $635 or
less in 1991.
* Middle-inme ecoomziesare those with a GNP per capita of more
rhan $635 but lessthan $7,911 in 1991. A fijrther division,atGNP
per capita of $2,555 in 1991, is made between Iouer-midlk-in-
cow and upper-midl-income conomies.
* High-incomeeconomiesare those with a GNP per capit of $7,911
or more in 1991.
* Word comprisesall economies, including economies with sparse
data and those with fewer than 1 million people.
Low-income and middle-income economies axesormetimesreferred
to as developingeconomies. The use of the term is convenient; ir is not
intended to imply that all economies in the groupare experiencingsim-
ilar development or that other economies have reached a preferred or
finalstage of development. Classificationby income does not necessar-
ilyreflectdevelopment status.

'V
 Groups
-Anabtical
F OR ANALYIICAL PJRPOSES, THISSTUDY
into severalregions,defined as fIllows
GROUPS
ECONOMIES

*High-perfirmingAsian economies(HPAEs),led byjapan, arc iden-

tified by severalcommon characteristics, such as very rapid ex-
port growth. The HPAESare subclassified roughly according to
the duration of their succcssfiulrecord of economic growth:
.
vurTheF
: igersusuallyidentifiedas Hong Kong, thc Republic
of Korea, Singapore, and Taiwan, China, have been growing
rapidly for decadesand havejoined or approached the ranksof
high-income economies.
* The newly industrialidZWgeconomies (NIEs) are Indonesia,
Malaysia,and Thailand. They havejoined the group of HPAEs
more recently,within the last two decades.
* EastAia comprisesall the low- and middle-income economiesof
East and Southeast Asia and the Pacific, east of and including
China and Thailand.
* South Asizr comprises Bangladesh, Bhutan, India, Myanmar,
Nepal, Pakistan, and Sri Lanka.
* S&zb-&harnAf5ca comprisesall econoniies south of the Sahara
including South Africa but ecduding Mauritius, Reunion, and
Seychelles,which are in the OtherAsia and islandsgroup.
* Europe Middle F."t and North Africa comprises the middle-in-
come Europan economies of Bulgaria, the former Czechoslova-
kia, Greec, Hungary,Poland, Portugal, Romania, Turkey,and the
forrner Yugoslavia,and all the economiesof North Africa and the
Middle East, and Afghanisan.
m Latin Americaand the GCaibcancDmprisesall American and
Caribbean economiessouth of the United States.

The regional grouping of economiesin occasionalparts of the text or

tables maydiffer from that used in the main tt of this study,described
above.Such variations are nocedwhere dthy occur.
DataNotes
* Bilion is 1,000 million.
* Trillionis 1,000 billion.
- Dolars arecurrent U.S. dollarsunless otherwisespecifiecd.
* Growthratesarebasedon constant price dart

Historical data in this study may differ from those in other Bank
pubLicationsbecause of continuous updating as better data become
available,becauseof a changeto a new baseyear for constant pricedata,
or becauseof changesin economycompositionof income and analyti-
mlgroups.

Acronymsand Initials
CPI Consumerpriceindex
DFI Direct foreign investment
GDP Gross domestic product
GNP Gross national product
rlPAEs High-performing Asian economnies
1MF International Monetary Fund
ISIC Intemational Standard Industrial Classification
mI Ministry of Trade and Industry Japan
NIs Newly industrializing economies
OECD Organization for Economic Cooperation and Develop-
ment (Australia,Austria, Belgium, Canada, Denmark, Fin-
land, France, Germany, Greece, Iceland, Ireland, Italy,
Japan, Luxembourg, Netherlands, New Zealand, Norway,
Portugal, Spain, Sweden, Switzerland, Turkey, United
Kingdom, and United States)
PPP Purchasing power parity
R&D Research and devclopment
TFP Total factor producdivity
UNDP United Nations Development Programme
UNESCo United Nations Educational, Scientific, and Cultural
Organization

-n.
 Overview: The
Making of a Miracle
RECORD OF HIGH AND
AST ASIA HAS A REMARKABLE
sustained economic growdh.From 1965 to 1990 the
twenty-threeeconomliies of East Asiagrew fasterthan all
other regions of the world (figure 1). Most of this
achievement is attributable to seemingly miraculous
growth in just eight economies: Japan; the 'Four
Tigeser-Hong Kong, the Republicof Korea,Singapore,and Taiwan,
China; and the three newly industrializingeconomies (NiEs) of South-
east Asia, Indonesia, Malaysia, and Thailand. These eight high-
performingAsian economies(HPAEs)are the subject of this study.*
Selectinganysecofeconomiesand attemptingtoundesand theorigins
of their successfulgrowth are necessarilyarbitnaryprocesses.'Botswana,
Egypt, Gabon, and Lesotho in Sub-SaharanAfrica have also been
among the world's top growth performers in the past two decades, as
have suchdiverseeconomiesas Brazil,Cyprus,Greece,and Portugal(see
figure2). Why focuson eight economiesin East Asia?In part the choice
reflectspopular interest;it has become common to see referencesto the
"AsianEconomicMiracle., In part it reflectsrecentartention by the aca-
demic and development policy communities to the relationship be-
tween public policies-which some authors haveargued have a number
of common threads in the eight economies,especiallyJapan, Korea,

*Recendy China, particularly southern China, has recorded remarkably high growth
- ates using policies that in some ways resermblcthose of the HPAEs.This vtay significant
development is beyond the scope of oursudy, mainly because China's ownership seruc-
. ure, methods of corporate and civil govcrnance, and reliance on markets are so differ-
ent from the those of the HPAEs,and in such mapidflux, that cross-economy comparison
is problematic. We touch on China's recent development in chapters 1 and 3. Te eco
nomic transiion in China is thesubject of current reseach by the Policy Research De-
parment ofthe World Bank (see Bibliographic Note).

-- : ~~~~~~~~
 5 1jRRL CLE

Figure1 AverageGrowthof CNPper Capita,1965-90

East Asia
| _ ; _____
HPAEs

EastAsiawithoutHPAES

South Asbi

Midae EastandMeditermean..

Sub-SaharanAfrica

OECD ec__ , ;

andCarlbbean
LatinAmedcm

0 1 2 3 4 5 6
GNPpercapit growthrate (percent)
Skwre World Bank(1992d).

Singapore, and Taiwan, China-and rapid growth. And in part it re-

fleas the belief of those involved with this study that the eight
economies do share some economic characteristics that set them apart
from most other developing economies.
Since 1960, the HPAEs have grown more rhan twice as fist as the rest
ofEast Asia, roughly three times as fast as Latin America and South Asia,
and five times faster than Sub-Sahann Africa. They also significandy
outperformed the industrial economies and the oil-rich Middle
East-North Africa region. Between 1960 and 1985, real income per
capita increased more than four times in Japan and the Four Tigers and
more than doubled in the Southeast Asian NEs (see figure 2). If growth
were randomly distmbuted, there is roughly one chance in ten thousand
that successwould have been so regionally concentrated.
The HPAEs have also been unusually successfiulat sharing the fiuits of
growth. Figure 3 shows the relationship between the growth of gross do-
mestic product (GPD)per capita between 1965 and 1990 and changes in
the Gini coefficient, a statistical measure ofthe inequality ofincome dis-
tribution. The HPAEsenjoyed much higher per capicaincome growth at
the same time that income distribution improved by as much or more
than in other developing economic% with the exceptions of Korea and
Taiwan, Chna, which began with highly equal income distributions.
The HPAEs are the only economics that have high growth anddedining

2.
Figure2 Changein GDPper Capita,1960-85

Taiwan,China
Indonesia
Hong Kong . -

Singapore
Rep.of Korea
Japan
Matta
Lescotho
Egypt
CYPMUS . o
Gabon
Greece
Brazil
Syria
Portga :wA

Malaysia
Yugoislavia
China __ _ - - - ~ -

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

Norway
Itay
Spain ~-.-.-.---
Israel
Jordan ___

Germnany.. .

Mexico
Pakistan
Colombia .Middle 20:
UnitedStatbes
United Kingdom ----- ________- _

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

Venezuiela .-. ..

Keniya . -.

Peau
Argentina
Uganda___
India __ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _

Benin__ _ _ _ _ _ _ _ _

Senegal
Haiti

Uberla
Sudan
Somalia I___
Zaire_ __ _

NIgerfia _

Afghardstan ~-*
Midi '~~~~~~~~Bottom20
CentralAfricanRepublic
Ghana . ..

Guyana ___ -

Madagascar .-

Chad__ _ _ _ _ _ _ _ _

Zambia ------- ---- H A s--

Angola_
Mozamibique - I
Kuwalt __ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

-2. 0 1. 2 3 45
PercentagechangeIn GDP

Soarae:.
Summers and Heston (1988).
 -RACLLE

F.gum3 Change in Inequityand the GDPperCapita GrowthRate

Change In avenage Glni coeffcient (1980s minus ISGOs)

0.2-------- ----
I - . RankingbyGIni coefficient1980
* Most equal third ofsample
0Q15 - ------------ - U Middle thibd
A Least equal thlrd

0.05 -- - - - - - - . - - - - - - - - - . . - - - - - - - - - - ~ -- - - -

A Chilet Brazil A

0- ; .Rep- of Korea

Argentina

______________
!Mexico
______A_Colom
v Colombla
bia _ _ _ _ __
.Taiwan. _ _ _ _
China
_ _ _
40.05 - _ _
Thailand

* Venela . Malaysia Srigapore

* Philippines
-J1- Hong
Indonesa
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:Figurc3plots the relationshipbcetwenaveragcper capit incomcgroth and changesin the decadeavege ofthc Gini cocfficientfim
rde 1960sto the 190s; a negativenumber indicatesthar incomebecumelss concentraredL The decadeavcge is used becausedata areavailable
for different years in diffacmr conornies the decade average
fbr the 1960s begins with dat from 1965.
SowrerWorld Bankdam.

inequality. Moreover, the fastest growing East Asian economies, Japan

and the Four Tilgers,axethe most equal.
As a result of rapid, shared grovth, human welfare has improved dra-
matically. Life expecancy in the developing HPAESincreased fiom 56
years in 1960 to 71 years in 1990. (In other low- and middle-income
economies, life expectancy also rose considerably, fiom 36 and 49 to 62
and 66 years, respectively.)In the HPAEs, the proportion of people living
in absolute poverty, lacking such basic necessities as dean water, food,
and shelter, dropped-for example, from 58 percent in 1960 to 17
percent in 1990 in Indonesia, and from 37 percent to less than 5 percent

4
 in Malaysiaduring the same period. Absolutepovertyalso declined in
other developingeconomies,but much less steeply,fiom 54 to 43 per-
cent in India and from 50 to 21 percent in Brazilfronm1960 to 1990. A
host of other socialand economic indicators, from education to appli-
ance ownership,have also improvedrapidlyin the HPAEs and now are at
levelsthat sometimessurpassthose in industrialeconomies.
'What causedEastAsia'ssuccess?In largemeasurethe HPAEs achieved
high growth by getting the basics right. Privatedomesric investment
and rapidly growing human capital were the principal engines of
growth. High levelsof domeLic financial savingssustained the HPAEs'
high investment levels.Agriculture, while declining in relativeimpor-
tance, experienced rapid growth and productivity improvement. Popu-
lation growth rates declined more rapidly in the HPAEs than in other
parns of the developing world. And some of these economies also got a
head strt because they had a better-educated labor force and a more ef-
fecrive system of public administration. In this sense there is littdethat
is Kxmiraculousrabout the HPAEs' superior record of growth; it is largly
due to superior accumulation of physical and human capital.
Fundamentally sound development policy was a major ingredient in
achieving rapid growth. Macroeconomic management was unusually
good and macroeconomic perfornance unusually stable, providing The
essential firnework for private investment. Policies to increase the
integrity of the banking system, and to make it more accessible to non-
traditional savers, raised the levels of financial savings. Education poli-
cies that focused on primary and secondary schools generated rapid
increases in labor force skills. Agriculural policies stressed productivity
and did not tax the rural economy excessively.All the HPAEskept price
distortions within reasonable bounds and were open to foreign ideas
and technology.
But these fimdamental policies do not tell the entire story. In most of
these econoniies, in one form or another, the government intervened-
systematically and through multiple channels-to foster development,
and in some cases the development of specific industries. Policy inter-
ventions took many forms: targeting and subsidizing credit to selected
industries, keeping deposit rates low and maintaining ceilings on bor-
.rowing rates to increase profits and retained earnings, protecting
domestic import substitutes, subsidizing declining industries, establish-
big and financially supportng government banks, making public
investrmentsin applied research, establishing firm- and industry-specific

5
 r1A!N.
-M IRACLE

export margets,developing export marketing institutions, and sharing in-

formation widely betweenpublic and private sectors. Some industries
were promoted,while otherswere not.
At leastsome of thescinterventionsviolate the dicturn of establishing
for the privatescaor a levelplayingfield,a neutral incentivesregime.Yet
thesestrategiesof selectivepromotion were closelyassociatedwith high
rates of private investmencand, in the fastest-growingeconomies,high
rates of productivitygrowth.Were some selectiveinterventions,in fict,
good for growvh?
In addressingthis question, we facea central methodologicalprob-
lem. Since we chose the HPAAEs for their unusually rapid growth, we
know already that their interventions did not significantly inhibit
growth. But it is very difficult to establish statistical links between
growth and a specificinr.w..tion and even more difficultto establish
causalityrBecause we cannot know wrhat would have happened in the
absence of a specificpolii); it is difficult to rest whether interventions
increased growth rates. Other economies attempted similar interven-
tions without success,and on averge they used them more pervasively
than in the HPAEs.Because the HPAEsdiffered from less successfiil
economiesboth in their doser adherenceto policy findamentals and in
the manner in which they implementedinterventions,it is virtuallyim-
possibleto measurethe relativeimpact of findarnentals and interven-
tions on HPAE growth. Thus, in attempting to distinguish interventions
that contributed to growth from those that were either growth-neutral
or harmfil to growth, we cannot offer a rigorous counterfactualsce-
nano. Instead, we have had to be content with whar Keynescalled an
&essay in persuasion,"based on analyticaland empiricaljudgments
Our judgment is that in a few economies,m ly in Northeast Asia,
in some instances, govenment interventions resulted in higher and
more equal growth than otherwisewould have occurred. However,the
prerequisitesfor successwere so rigorotusthat policynakers seekingto
followsimilar paths in other developingeconomieshave often met with
failure.What were theseprerequisites?First, governmentsin Northeast
Asia developedinstitutionalmechanismswhich allowedthem to esab
lish dear performance criteria for selective interventions and to monitor
performance- Intervention has taken place in an unusually disciplined
and performance-basedmanner (Ainsden 1989). Second, the costs of
interventions, both explicit and implicit, did nor become excessive.
When fiscalcosts threatenedthe macroeconomicstabilityof Koreaand

6
 Malaysiaduring their heavy and chemical indusrries drives, govern-
ments pulled back In Japan the Ministryof Financeacted as a checkon
the ability ofthe Minisrryof IntcrnationalTrade and Indusry to carry
out subsidypolicies,and in Indonesia and Thailand balanced budget
lawsand legislativeproceduresconstrained the scope for subsidies.In-
deed, when selectiveinterventionshavc threatenedmacroeconomicsta-
bilty, HPAE governmentshave consistentlycome down on the side of
prudent macroeconomicmanagement.Pricedistortionsarisingfrom se-
lective interventionswere also less extreme than in many developing
economies.
In the newly industrializingeconomies of Southeast Asia, govem-
ment interventionsplayed a much less prominent and frequently less
constructiverole in economic success,whileadherenceto policy finda-
mentals remained importamt These economies'capacity to administer
and implementspecificinterventionsmay have been lessthan in North-
eastAsia.Their rapid growth, moreover,has occurred in a very different
internationaleconomic environment from the one that Japan, Korea,
and Taiwan,China, encountered during their most rapid growth.Thus
the problem is not only to try to understandwhich specificpoliciesmay
havecontributed to growth, but also to understand the institutionaland
economiccircumstancesthat made them viable. Indeed, the experience
of the Southeast Asian economies, whose initial conditions parallel
thoseof many developingeconomiestoday,may prove to have more rel-
evanceoutside the region than thar of Northeast Asia.
The book isorganizedas follows:chapter 1 describesthe distinguish-
ing characteristicsof the East Asian economic mirade, rapid growth
with equity and uscs economic models to attempt to account for this
growth. Chapter 2 reviewspolicyexplanationsfor East Asia'seconomic
successand introducesthe frameworkthat wewill use throughout to ex-
plore dte relatonship berween public policy and economic growth.
Chapter 3 discussespragmatism and flexibilityin the formuation of
policiesthat led to two impormnt characteristicsof the HPAEs'economic
performance:macroeconomicstability and rapid growth of manufac-
tured exports. Chapter 4 disLcusses the role of institutions. Chaprer 5
looks at the role of public poliqc in the HPAEs' unusually rapid accurnu-
lationof physicaland human capital,whilechapter 6 analyzesthe means
usedto achieveefficientallocationof resourcesand productivitygrowth.
Chapter 7, in condusion, assessesthe successof EastAsian policesand
their applicabiliy in a changingworld economy.The remainder of this

7
-' .--
- V IICACLE

overview parallels the organization of the book, highlighting the central

arguments and conclusions.

TheEssence
oftheMiracle:
Rapid
Growth
withEquity

T HE EIGHT HPAEs ARE HIGHLY DIVERSE IN NATURAL RE-

sources, population, culture, and economic poliqc What shared
characteristics cause them to be grouped together and set apart
from other developing economies?First1 as we noted above, they had
rapid, sustained growth between 1960 and 1990. This in itself is un-
usual among developing economies; orhers have grown quickly for
periods but not for decades at such high rates. The HPAEsare unique in
that they combine this rapid, sustained growth with highly equal in-
come distributions. They also all have been characterized by rapid de-
mographic transitions, strong and dynamic agricultural sectors, and
unusually rapid export growth (see chapter 1).
The HPAEsalso differ fiom other developing economies in three fac-
tors that economists have -traditionally associated with economic
growth. High rates of investment, exceeding20 percent of GDP on aver-
age between 1960 and 1990, indcuding in particular unusually high
rates ofprivate investment, combined with high and rising endowments
of human capital due to universal primary and secondary education, tell
a large part of the growth story. These factors account for roughly two-
thirds of the growth in dte HPAEs.The remainder is attributable to im-
proved productivity. Such high levels of productivity growth are quite
unusuaL In fact, productivity growth in the FHAEsexceeds that of most
other developing and industrial economies. This superior productivity
performance comes from the combinadon of unusual successat allocat-
ing capimt to high-yielding investments and at catching up technologi-
cally to the industrial economies.

PublicPoliciesandGrowth
What was the role of public policy in helping the HPAEs to rapidly
accumulate human and physical capital and to allocate those resources to
high-yielding investments?Did policies assist in promoting rapid produc-

8
tivity growth?There are severalexplanationsfor EastAsiassuccess.Geog-
raphy and culturewere dearly important; however,dthy do not entirely
account for the high-performingeconomies'success,as the presenceof
unsuccessfuleconomiesin the same regionattests.Among the varietyof
policy explanations,two broadviewshave emerged(seedcapter2).
Adherentsof the neoclassicalviewstressthe HPAEs'successin gettingthe
basicsright. They argue that the successfulAsian economieshave been
better than othersat providinga stablemacroeconomicenvironmentand
a reliablelegalframeworkto promote domesticand internationalcompe-
tition. They also stress that the orientation of the HPAEs toward inter-
national trade and the absenceof price controlsand other distortionary
policieshave led to low relativeprice distortions.Investnents in people,
education,and health arelegitimateroles for govenrnment in the neoclas-
sicalframework,and its adherentsstressrhe importanceof human capital
in the HPAEs'success.
Adherents of the renisionistview have successfillyshown that Easr
Asiadoes not whollyconform.to the neodassicalmodel. Industrialpol-
icy and interventions in financial markets are not easily reconciled
within the neoclassicalfiamework.Somepoliciesin some economiesarc
much more in accordwith models of state-leddevelopment.Moreover,
while the neoclassicalmodel would explaingrowth with a standard set
of relativelyconstantpolicies,the policy mixesusedby EastAsian econ-
omieswere diverseand flexible.Revisionistsargue that East Asiangov-
ernments "led the m.narke"e in critical ways. In contrast to the
neoclassicalviev, whichacknowledgesrelativelyfew casesof market&il-
ure, revisionistscontend that markersconsistendy fail to guide invest-
ment to industrieschatwould generatethe highestgrowthfor the overall
economy.In East Asia, the revisionistsargue, governmentsremedied
this by delibemtely"getting the prices wrong"-altering the incentive
struccure-to boost industries that would not otherwise have thrived
(Amsden1989).
The revisionistschoolhas providedvaluableinsights into the history,
role, and extent of East Asian interventions, demonstrating convinc-
ingly the scope of governmentactions to promote industrial develop-
ment in Japan, Korea,Singapore,and Taiwan,China. But, in generalits
proponentshave not claimedto establishthat interventionsper se accel-
cratedgrowth. Moreover,aswe shallshow,some important govemment
interventionsin EastAsia, such as Korea'spromotion of chemicalsand
heavyindustries,havehad little apparentimpacton industrialstrucmrt.

9
 -4iE$AS,kAN MIRACLE

In other instances,such as Singapore'sefflortto squeezeout labor-inten-

sive industries by boosting wages,policieshave dearly backfired.Thus
neither view fiullyaccounts for East Asia'sphenomenal growth.
View. In describing the policies associated with
The Market-Frendly
rapid growth, WorldDevelopment Report1991 (World Bank 1991b) ex-
pands on the neodassical view, darifyringsystematicallyhow rapid
growth in developingcountries has been associatedwith effcctivebut
carefully limited govcrnment activism. In the 'market-friendly" strategy
it articulates, the appropriate role of government is to ensure adequate
investments in people, provide a competitivc climate for private enter-
prise, keep the economy open to intemational trade, and maintain a sta-
ble macroeconomy. Beyond these roles, the report argues, governments
are likely to do more harm than good, unless interventions arc marker
fiendly. On the basis of an cxhaustivereview of the experience of devel-
oping economies during the last thirty years, it condudes that attempts
to guide resource allocation with nonnarket mechanisms have generally
failed to improve economic performance.
The market-friendly approach captures important aspects of East
Asia'ssuccess. These economies are stable macroeconomically, have high
shares of international trade in GDP,invest heavily in people, and have
strong competition among firms. But these haracteristics are the out-
come of many different policy instnuments. And the instruments cho-
sen, particularly in the nordteastern HPAES,Japan, Korea, and Taiwan,
China, sometimes induded extensive government intervention in mar-
kets to guide private-sector resource allocation. The success of these
northeastern economies, moreover, stands up well to the less interven-
tionist paths taken by Hong Kong, Malaysia, and more recendy In-
donesia and Thailand.
AFunctional Appmachto UndemtaingGrwth. To explore these varying
paths to economic success, we have developed a framework that seeks to
link rapid growth to the attainment of three functions. In this view, each
of the HPAEs rnaintained macroeconomic stability and accomplished
three fiunctions of growth: accumulation, efficient allocation, and rapid
technological catch-up. They did this with combinations of policies,
ranging from market oriented to state led, that varied both across
economies and over time.
We classifjrpolicies into two broad groups: fundamentals and selec-
tive interventions. Among the most important fimdamental policies are
those that encourage macroeconomic stability, high investments in

I0
human capital,stableand securefinancialsystems,limitedpricedisror-
dons, and opennessto foreigntechnologySelectiveinterventionsin-
dude mildfinancialrepression(keepinginterestracespositivebut low),
directedcredit,selectiveindustrialpromotion,and trade policiesthat
push nontraditionalexports.We try to understandhow government
policies,borh fundamentaland interventionist.may havecontributed
to fasteraccumulation,moreefficientallocation,and higherprnducriv-
ity growth.
We maintainas a guidingprinciplecharfor interventionsthat at-
temptto guideresourceallocationto succeed,theymust addressfilures
in the workingof markets.Otherwise,marketswould perform the
allocationfunctionmore efficiently.We idenrifya dass of economic
problems,coordinationfailures,whichcan lead markersto fail,espe-
ciallyin earlystagesof development.Wethen interpretsomeof the in-
terventionistpoliciesin East Asiaas responsesto these coordination
problems-responsesthat emphasizedcooperative behavioramongpn-
vatefirmsand clearperformance-based standardsof success.
Competitivediscipline is crucial to efficientinvestment.Most
economiesemployonlymarket-based competition.Wearguethat some
HPABs havegone a step furtherby creatingconteststhat combinecom-
petitionwiththe benefitsof cooperationamongfirmsandberweengov-
ernmentand the privatesector.Suchcontestsrangefromvery simple
nonmarketallocationrules,such as accessto rationed credit for ex-
porters, to very complex coordination of private investment in the
government-business deliberationcouncilsofJapanand Korea.The key
featureof each contest, however,is that the governmentdistributes
rewards-oftenaccessto caedit or foreignexchange-on the basisof
performance,whichthe governmentand competingfirmsmonitor.To
succeed,selectiveinterventionsmust be disciplinedby competitionvia
eithermarketsor contests.
Economiccontests,like all others,requirecompetentand impartial
referees-that is, strong institutions.Thus, a high-qualitycivilservice
that hasthe capacityto monitorperformanceand is insulatedfrompo-
liticalinterferenceis essentialto contest-based
competition.Of course,a
high-qualitycivilservicealsoaugmentsa government's abilityto design
and implementnon-contest-based policies.
Our frameworkis an effortto order and interpretinlformation.We
are not suggestngthat HPAE governmentsset out to achievethe func-
tionsof growth.Rather,theyusedmultiple,shiftingpolicyinstruments

II
 RAC MIRACLE
LAN

in pursuit of more straiglhtforwardeconomic objectivessuch as macro-

economic stability,rapid export growth, and high savings. Pragmatic
flexibilityin the pursuir ofsuch objectives-the capacityand willingness
to changepolicies-is as much a hallmarkof the HPAEsasany singlepol-
icy instrument. This is well illustrated by the great variety of ways in
which the HPAEs achieved tvJ important objectives:macroeconomic
stabilityand rapid export growth (secchapter 3).

Achieving
Macroeconomic
StabiltyandExportGrowth

More than most developingeconomies,the HPAESwere characterized

by responsiblemacroeconomicmanagement.In parricular,they gener-
ally limited fiscal defcits to levels that could be prudently financed
without increasinginflationarypressuresand responded quicklywhen
fiscal pressureswere perceivedto building up. During the past thirty
years, annual inflation averaged approximately 9 percent in dtese
economies,comparedwith 18 percentin other low-and middle-income
economies.Becauseinflation was both moderate and prediaable, real
interest rateswerefar more stable than in other low-and middle-income
economies. Macroeconomic stability encouraged long-term planning
and privateinvestmentand, through its impacton real interest ratesai, 4
the realvalueof financialassets,helpedto increasefinancialsavings.The
HPAEsalso adjusted their macroeconomicpolicies to terms of trade
shocksmore quicklyand effectivelythan other low-and middle-income
economies.As a result, they haveenjoyedmorerobustrecoveriesof pri-
vate investment.
Many of the policiesthat fosteredmacroeconomicstabilityalso con-
tributed to rapid export growth. Fiscaldisciplineand high public sav-
ings allowedJapan and Taiwan,China, to undertake extended periods
of exchangerate protection. Adjustments to exchangerates in other
HPAEs-validatedbv policies that reduced expenditures-kept them
competitive,despite differentialinflationwith tradingpartners.
In addition to macroeconomicpolicies,the HPAEsused a variety of
approachesto promoting exports.All (exceptHong Kong)beganwith a
period of import substitution, and a strong bias against exporEs. But
each moved to establisha pro-export regime more quicldy than other
developingeconomies.First Japan, in te 1950s and early 1960s, and
then the Four ligers, in the late 1960s,shified trade policiesto encour-
age manufacturedexports.In Japan,Korea,and Taiwan,China, govern-

12
 ments establisheda pro-exportincentivestructure that coexistedwith
moderatebut highlyvariableprotectionof the domesticmarker.A wide
varietyof instrumentswas used, includingexportcredit, duty-freeim-
ports for exportersand theirsuppliers,exporttargets,and tax incentives.
In the SioutheastAsian NIEsthe export push came later, in the early
-1980s,and the instrumentswere different.Reductions in import pro-
tection were more generalizedand were accompaniedby export credit
and supportinginstitutions.In theseeconomiesexportdevelopmenthas
relied less on highly selectiveinterventionsand more on broadlybased
market incentivesand direct foreigninvestment.

Building
theInstutional
Basis
forGrowth
Some economistsand politicalscientistshave argued chat the East
Asian miracleis due to the hiighqualityand authoritariannature of the
regionsinstitutions.They describeEastAsianpoliticalregimesas "devel-
opmental statee'in which powerfultechnocraticbureaucracies,shielded
from politicalpressure,deviseand implementwell-honedinterventions.
We believe developmentalstare modelsoverlook the central role of
government-privatesectorcooperation.Whle leadersof the HPAEshave
tended to be either authoritarian or paternalisdtithey have also been
willingto grant a voiceand genuineauthorityto a technocraticeliteand
key leadersof the privatesector. Unlike authoriruian leadersin many
other economies,leadersin the HPAEs realizedthat econornicdevelop-
ment wasimpossiblewithoutcooperation(seechapter4).
ThePrinciple
ofShardGrowth. To establishtheir legitimacyand win the
supportof the sociey at large,East Asian leadersestablished the princi-
ple of shared growth, promising that as the economy expanded all
groupswould benefit.But sharinggrowth raisedcomplexcoordination
problems.First, leadershad to convinceeconomicelitesto supportpro-
growth policies.Then they had to persuadethe elitesto sharethe bene-
fits of growth with the middle class and the poor. Finally,to win the
cooperation of the middle classand the poor, the leadershad to show
them that they would indeed benefit from future growth.
Explicit mechanismswere used to demonstrate the intent that all
would have a shareof future wealth. Koreaand Taiwan,China, carried
out comprehensiveland reform programs;Indonesia usedrice and fer-
filizerpricepoliciesto raiseruralincomes;Malaysiaintroducedexplict
wealth-sharingprogramsto improvethe lot of edtic Malaysrelativeto

'3
 - IRfLMIRACLE

the better-off ethnic Chinesc; Hong Kong and Singaporeundertook

massivepublic housing programs;and in severaleconomies,govern-
ments assistedworkers' cooperativesand establishedprograms to en-
couragesmall and medium-sizeenterprises.Whatever the form, these
programs demonstrated that the goverment intended for all to sharc
the benefitsof growth.
Cualmg EnvironmentTo cacidecoordination prob-
a Business-Fiendly
lems, leadersneeded institutions and mechanismsto reassurecompet-
ing groups that each would benefit from growrth.The first step was to
recruit a competent and relativclyhonest technocratic cadre and insu-
late it from day-ro-daypoliticalinterference.The power of thesetech-
nocracieshas varied greatly.In Japan, Korea, Singapore, and Taiwan,
China, strong, well-organizedbureaucracieswield substantial power.
Other HPAEShave had small,general-purposeplanning agencies.But in
each economy,economic technocratshelped leaders devise a credible
economicstrategy.
Leaders in the HPAs also built a business-friendlyenvironment.A
major element of that enviromnenrwas a legal and regulatorystrucmre
that was generaly hospitable to private investment. Beyond this the
HPAiEshave with varying degreesof successenhanced communication
between businessand government.Japan, Korea,Malaysia,and Singa-
pore have establishedforums, which we call deliberation councils, in
whichprivatesectorgroups are invited to help shape and implemnentthe
governmentpoliciesrelevantto theirinterests.In contrast to lobbying,
--whererules are murky and groups seek secret advantageover one an-
odter, the deliberation councils are intended to make allocation rules
dear to all participants.
UsingDeliberation
Council.In Japan and Koreatechnocratsused delib-
eration councilsto establishcontestsamong firms. Becausethe prvate
sector participated in drafting the rules, and because the processwas
transparentto all paricipants, privatesector groups becamemore will-
ing participantsin the leadership?s
developmentefforts.One by-product
of these contestswas a tendencyto reducethe privateresourcesdevoted
to wastefulrent-seekingactivities,thus making more availablefor pro-
ductiveendeavors.Deliberationcouncilsalso&cilitatedinformationex-
changesbetwren the private sector and govecnment,among firms, and
between management and labor. The councils thus supplemented the
market'sinforrnation transmissionfunction, enabling the IPAEs to re-
spond more quickly than other economiesto changing markets.

'14
 communicationhavenot bem sta-
Institutionsofbusiness-government
tic in theHPAJ.The roleof the deliberationouncilis changingin Japan
and Koreato a moreindicatveand consensus-building role,alongfinc-
lines.InMalaysia
tionalasopposedto industry-speific thecouncilsappear
in importanceandscope.In Thailandthe formalmecha-
to beincreasing
nismsof communication havegenerallybecnusedto presentbusinesses'
positionstogovermentand to reducetheprivatesector'ssuspicionofgov-
development,
ermnent.In institutional asineconomicpoliymakingEast
Asiangovernments havechangedwithchangingcircumstances.

AccumulaingHumanand PhysicalCapital

Drawingon the strengthof theirinstitutions,EastAsianeconomies

used a combinationof fundamentaland interventionistpoliciesto
achieverapidaccumulationof human and physicalresources.Funda-
mentalsincludedsuchtraditionalgovernmentobligationsas providing
adequateinfrastructure,
education, and securefinancialinstitutions.In-
terventionsinduded mild repressionof interest rates, state capitalism,
mandatorysavingsmechanisms, ofrisk(seechapter5).
and socialization
Capiti. The East Asian economieshad a head start in
Bilding Human
termsofhumancapitalandhavesincewidenedtheirleadoverotherde-
velopingeconomies.In the 1960s,levelsofhumancapitalwerealready
higherin the HPAEsthan in otherlow-and middle-income economies.
Governmentsbuilton this baseby focusingeducationspendingon the
lowergrades;firstbyprovidinguniversalprimaryeducation, laterbyin-
creasingthe availabilityof secondaryeducationsRapid demographic
transitionsfacilirated
these effortsby slowingthe growth in the number
of school-agechildrenand in some casescausingan absolutedecline.
Dediningferdlityand rapideconomicgrowthmeant that, evenwhen
education investment as a share of GDP remained constant, more re-
sourceswere avilable per child. Limited public fimding of post-
secondaryeducation focused on technical skills, and some HPAEs
importededucationalserviceson a largescale,particularlyin vocation-
ally and technologicallysophisticateddiscplines.The result of these
policieshasbeena broad,technically inclinedhumancapitalb±vewell1-
-stited to rapideconomicdevelopment.
.IHPAEeducationpoliciesalsocontributedto moreequitableincome
distributions.To be sure, initialconditionshelpedto set up a virtuous
circle:initiallowinequalityin incomeand educationledto educational

I5
.:~~~~~~~~~~~~~~;n

11
:t4iSjA-N MIRACLE

expansion, which reinforced low inequality. In addition, by focusing

Figu 4 SavingsRatesof HPAEs
andaSelected Eiconomies,19704
HPAEs
I -
0 2.0
'Percenta,ge ar GDP
Nenar.Finlind,uds AustriaBthee
um.
Dcrnmark, Finland,Fmncc,theE:ederol
Republicof Gcmanybebore reunificaion,
Greece, Iceland,Iredand.Imly, Luxem-
bour, dte Netherlands, Norny, Portugal.
Spain,Sweden,Switxrland, and the
UnitedKingdom.'Orie indudesthese
&-veoping cconomies: ArgenonaBrazil,T
Carenina
developin eColombia. SEl,
Chile,Cdlombia,COredivoire, Egyp4,
Ghana,India,Mexico,Morocco,NiScria,
Pakistan,Pem,SriLnka, Turkey,
Urugua.
Urgan y Ycnzuels the
the fone YugosIiaI
Venezuela. upEra
and Zaire
andHeson (199l).
SawreuSummer
spending on primary and secondary education, and leaving demand for
tertiary education to be largely met by' a self-financed private system,
governments served large segments of the population that otherwise
would have lacked accessto education.
bcreasingSavniugandImnestneuL The HPAEsincreased 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 HPAEshave 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 wverestrong prudential
regulation and supervision, limits on competition, and institutional re-
forms In Japan, Korea, Malaysia, Singapore, and Taiwan, China, postal
savings systems lowered transaction costs and increased the safey of sav-
ing while making substantial resources available to government. These
initiatves promoted rapid growth of deposits in financial instiutions
(seefigure4)
(se fiur 4)
Some governments also used a variety of more interventionist mecha-
nisms to increase savings. Singapore and Taiwan, China, maintained un-
- - 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-
20 30 40 called luxury consumption. Whether these more interventionist measures
to increase savings improved welfare is open to debat On one hand,
making consumers save when they would not have otherwise imposes a
welfare c -r. On the other, the benefits are apparent in the rapid growth of
these ecovomies. Savings, forced or not, generaed high payoffs based on
consistendy high rates of return to investments. In conwnt to other
economies that have used compulsory savings, such as the fonrer Soviet
Union, welfare costs were dearly offset by substantive benefits.
The HPAEsencouraged investment by several mans. First, they did a
better job than most developing econom>es at creating -infastructure
that was complementary to private investment. Seco.d, they created an
investment-friendly environment through a combination of tax policies

I6
ivoring investmentand of policiesthat kept the reative pricesof capi-
malgoodslow,largelyby avoidinghigh tariffson importedcapitalgoods.
These fiundamentalpolicieshad an important impact on pnvate invest-
ment. Third and more controversial,most HPAEgovernrents held de-
posit and lending rates belowmarketdearing levels-a practicetermed
financialrepression.
Japan, Korea,Malaysia,Thailand, and Taiwan,China, had extended
periodsof mild financialrepression.To be sure, increasinginterestrates
from negativeto zero or mildlypositivereal rates and avoidingfluctua-
tions (by avoiding unstable inflation)encouragesfinancialsavings.But
becausesavingsare nor very responsiveto marginal changesin positive
real interestrates, HPAEgovernmentswere able to mildlyrepressinterest
rateson depositswith a minimalimpact on savingsand to passthe lower
rates to finalborrowers Becausesaversweremostlyhouseholdsand bor-
rowers were mostly firms, this resulted in a transfer of income from
householdsto firns and in a changein the form in which savingswere
held, from debt to corporateequity.
Holding down interest rares on loans increasesexcessdemand for
credit, which in turn leadsto rationingof credit by the governmentit-
selfor by privatesector banksworkingwith goverrunentguidance.This
heightensthe risk that capitalwillbe misallocated.Thus there is a trade-
off betweenthe possibleincreasein investmentand the risk that the in-
creasedcapitalwill be badly invested.There is some evidencethat in
Japan, Korea,and Taiwan,China, govemmentsallocatedcredit to activ-
ities with high social returns, especiallyto exporcs.If this was the case,
there may have been benefiEsfrom mild financial repression and
government-ided allocation; mirroeconomic evidence from Japan
supports the view thar accessto governmentcredit increasedinvestment
(seechapter 6).
Generally, financial repression is associated with low economic
growth, especiallywhen realinterestratesare stronglynegative.But test
of the relationshipbetweeninterest rates and growth in Japan, Korea,
and Taiwan,China, do not show the negativerelationshipbetweenin-
terest rate repressionand growEhfound in cross-economycomparativ
studies (seechapter 5). Whilewe cannotestablishcondusivelythat mild
repression of interest rates at positive real levelsenhanced growth in
northeastAsia, it appaently did not inhibit it.
Finally,some governments,especially in the northeastemAsian tier,
haveencouragedinvestmentby spreadingprivateinvestmcent risksto the

'7
s;~ A? -om
~~~~~~~~V- vLRC
. ZiN [.IA
y
RC
LE

public In some economiesthe govern-mentownedor controlledthe in-

titutions providinginvestmentfunds,in others it offeredexplict credit
guarantees,and in sdll others it implictly guaranteedthe financal via-
bility of promoted projects. Relationshipbankingby a variety of public
and privatebanking institutionsin Hong Kong,Japan, Korea,Malaysia,
Singapore,Thailand, and Taiwan,China, involvedthe bankingsectorin
the managementof troubled enterprises, increasingthe likelihoodof
creditor workouts.DirectEd-creditprogramsin Japan, Korea, and Tai-
wan, China, signalcddirectionsof governmentpolicyand providedimn-
plicit insuranceto private banks.

EffcientAllocationand Pmductity Change

Somepoliciesthar favoredaccumulationin the HPA,s,induding fi-

nancial repression and the socializationand bounding of risk, could
haVeadversely afFectedthe allocationof resources.Similarly,industrial
targetingcould have resulted in extensiverent-seekingand great ineffi-
dency.Apparentlyrheydid not The allocadionalrulesfolowed by H4PAE
govemments-particularly the devicesusedto shift marekeincentives--
are thereforeamong the mostcontroversialaspectsof the Eat Asiansuc-
cessstory (seechaprer 6).
Like policies related ro accumulation, policies affecting allocation
and productiviy change fall into findamental and interventionistcat-
egories.Labor marker policies tended to rely on fundamentals, using
the market and reinforcing irs fiexibility.In capital markets, govern-
mentsintervened systematically,both to control interest rates and to
direct credit, but acted within a frameworkof carefulmonitoring and
generallylow subsidiesto borrowers.Trade policieshave indcudedsub-
standalprotection of local manufacurs, but less than in most other
developing countries; in addition, HPAEgovernmentsoffiet some dis-
advantagesof protection by activelysupporting exports. Finally,while
interventionsto support specificindustrieshave generallynot been suc-
cessfil, the export-push straregyE-themix of fundamental and inter-
ventionist policies used to encourage rapid manufacured cxport
growth-has resulted in numerous benefirs,including more efficient
allocation,increasedacquisitionof foreignrechnology, and more rapid
productivitygrowth.
Feible LaborMarkets.Govenment roles in labor narketsin the suc-
cessul Asian economies contrast sharply with the situation in most

I8
 other developingeconomies.HPAEgovernmentshavegenerallybeen less
vulnerableand lessresponsivethan other developing-economy govern-
ments to organized labor's demands to legislate a minimum wage.
Rather, they have focused their efforts on job generationseffectively
boosting the demand for workers.As a result, employmentlevelshave
rnsenfirst, followed by market- and productivity-drivenincreasesin
wage levels. Becausewages or at least wage rate increaseshave been
downwardlyflexiblein responseto changesin diedemand for labor,ad-
justmenr to macrocconomicshockshas generallybeen quicker and less
painfiul in East Asia than in other developing regions.Rapid adjust-
ments helped to sustain growth, which in turn made more rapid real
wage growthpossible.
IHighproductiviryand income growth in agriculturehelpedto keep
EastAsian urban wagescloseto the supplyprice of labor.In contrast to
many other developing-economies,wherethe gap betweenurban and
rural incomeshas been largeand growing,in the HPAEs the incomesof
urban and rural workerswith similarskilllevelshaverisen roughlyat the
same pace; moreover,the overallgap betweenurban and rural incomes
is smallerin the HPAEsthan in other developingeconomies.
In Sub-SaharanAfrica,Latin America,and South Asia,wherewages
in the urban formalsector are often pushed up by legislatedminimum
wages and other nonmarket forces,urban wage earnersoften have in-
comestwice their counterparm'in informalsectors.In contrast, the gap
between the formal and informalsectors in EastAsia is only about 20
percent. Smallerincome gapscontribute to overallsocialstability,thus
enhancing the environmentfor growth.
andfllocatio. Most HPAEs influencedcredit allocation
CaprtalMarkets
in threeways:(i) by enforcingregulations to improveprivatebanks'pro-
ject selection;(ii) by crearingfinancialinstitutions,especiallylong-term
credit (development)banks, and (Li) by directingcredit to specificsec-
tors and firms through public and privatebanks. All three approaches
can be justifiedin theory,and each has worked in some IIPAES.Yeteach
involvesprogressivelymore governmentinterventionin credit markets
and so carriesa higher risk.
Goverrunent relationships with banks in the HPAEshave varied
widely.In Hong Kong banks are privateand regulatedprimarilyto en-
sure their solvency.In Indonesia, Malaysia,Singapore,and-Thailand,
banks areprvately ownedand exerciseindependentauthorityoverlend-
big. While governmentshave broadlyguided credit allocationsthrough

I9
 isACLWMIRACLE

regulations and moral suasion, project selection is generaUlyleft 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, tightlycontrolledthe allocationof creditby public
commerial banks.
Each of the HPAEsmade someattempts to direct credit to priorityac-
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-sizeenterprises
were trgeted sectors in Indonesia, Malaysia,and Thailand. Taiwan,
China, has recently targeted technological development. Japan and
Korea have used credit as a tool of industrial polic34organizingcontests
through deliberativecouncilsto promote at various timesthe shipbuild-
ing, chemical,and automobileinduscries.
The implicit subsidy of direcred-creditprograms in the HPAEs was
generally small, especiallyin comparison to other developingecon-
omies, but accessto credit and the signalof goverunment support to if-
vored sectorsor enterpriseswereimportanL In Korea,the subsidyfrom
preferentialcredit waslargeduring the 1970s,resultingin a largegap be-
tween bank and curb market interest rates. This gap has dedined
sharplyin recentyea, as Koreahas shifted awayfiom heavycreditsub-
sidiesto selectedsectors.In Japan imnplicitsubsidiesweresmall, and the
direction of credit may have been more important as a signalingand in-
surancemechanism than asan incentve.
AlthoughEastAsiasdirected-creditprogramsweredigned to achieve
policy objectives,they neverthelessinduded strict performancecriteria.
In Japan, public bank managerschoseprojectson basiceconomiccrite-
ria, employingrigorous credit evaluations to select among applicants
that fell within governmentsectoraltrgets. In Korea, the govemment
individuallymonitored the large conglomeratesusing market-oriented
criteriasuch as exportsand profirabiliy.In some cases,majorenterprises
that filed to meet these testsweredriven into bankruptcy.Recenras-
sessments of the directed-creditprograms in Japan and Korea provide
microeconomicevidencethat directed-creditprogramsin these econo-
mies increased invesnnent, promoted new actvities and borrowers, and
were directed at firms with high potential for technological spillovers.
Thus these performance-baseddirected-creditmechanismsappear to
have improved credicallocation, especiallyduring the early stages of
rapid growth (seechapter 6).

z0
 Directed-creditprogras in other HPAEShaveusuallylackedstrong,per-
fbrnance-basedalloation and monitoringand thereforehavebeen largely
unsuccessful.Evenin the norther-tier economies,the increasinglevelof
financialsectordevelopmentand theirincreasingopennessto international
capimlflowshavemeant that directed-creditpograms havedeclinedin irm-
portance,as the economnies haveliberalizedtheirfinancialsectors.
Openness to ForeigTechnology.The HPAEshave activelysought foreign
technologythrough a varietyof mchaisms. Allwelcomed technology
transfersin the forn of licenses,capital goods imports, and foreign
training. Openness to directforeigninvestment(DRi) has speededtech-
nologyacquisition in Hong Kong, Malaysia,Singapore,and, more re-
cently,Indonesia and Thailand. Japan, Korea and, to a lesser extent,
Taiwan,China, restrictedDFI but offierthis disadvantge by aggressively
acquiringforeign knowledgethrough licensesand other means.
In contrast, odter low- and middle-incomeeconomiessuch as India
and Argentinahaveadopted policiesthat hindered the acquisitionoffor-
eign knowledge.Often they have been preoccupiedwith supposedlyex-
cessivepricesfor licenss They have refised to provideforeign exchange
for trips to acquire knowledge,been restictive of DFI, and have at-
tempted prematurelyto build up their machine-producingsectors,thus
forgoingthe advancedtechnologyembodiedin imported equipment.
PrmofingSpecificlidusies. Most East Asian governmentshave pur-
sued sector-specificindustrial policiesto some degrce.The best-known
instancesincludeJapan'sheavyindustrypromotion policiesof the 1950s
and the subsequentimitation of thesepoliciesin Korea.These policies
induded import prorectionas well as subsidiesfor capitaland other im-
ported input& Malaysia,Singapore,Taiwan, China, and even Hong
Kong have also establishedprograms-typically with more moderate
incentives-to acceleratedevelopmentof advancedindustries. Despite
these acions we find very litde evidencethat industial polices haveaf-
fected either the secroralstructure of industry or rates of productivity
change. Indeed, industrial structures in Japan, Korea, and Taiwan,
China,have evolvedduring the past thirtyyearsaswe would expect ven
factor-basedcomparativeadvantageand changingfactor endowments.
It is not altogethersurprisingthat industial policy in Japan, Korea,
and Taiwan, China, produced mainly market-conforming results.
Wile these govemmentsslectively promoted capital-and knowledge-
intensiveindustries,they also took stepsto ensurethat they were foster-
ing profitable, intermationally competitive firms. Moreover, their

21
 I rRAA
CVLIACLE

industrial policies incorporateda large amount of marker information

and used performance,usually export performanc, as a yardstick. Ef-
forts elsewhereto promote specific industries without better informa-
dion exchange and the discipline of intemational markets have not
succeeded.This has been the case with the ambitious industrial policy
programsin Braziland India, and with the morelimited but alsodisap-
pointing effortsto build an aerospaceindustry in Indonesiaand to pro-
mote heavyinduswtries in Malaysia.
ExportPush:AWinning MixofFundamentals andInterentionsOne combi-
nation of fundamental and interventionist policies practiced in the
HPAEs has been a significantsource of rapid productiviy growth: the ac-
tive promotion of manufactured exports. Although all HPAEs cxcept
Hong Kong passed through an import-substitution phase, with high
and variableprotecrion of domestic import substitutes, these periods
ended earlierthan in other economies,typically becauseof a compelling
need for foreign exchange.In contrast to many other economies,which
tried to preserve foreign exchangewith sticter import controls, the
HPAEs set out to earn additional foreign exchange by increasing exports.
Hong Kong and Singapore adopted trade regimes that were dose to free
trade; Japan, Korea, and Taiwan, China, adopted mixed regimes that
were largey free for export industries. In the 1980s, Indonesia,
Malaysia, and Thailand have adopted a wide variety of export incentives
while gradually reducing protection. Exchange rate policies were liberal-
ized, and currencies frequently devalued, to support export growth.
Overall, these policies exposed much of the industrial sector to interna-
tional competition and resulted in domestic relative prices chat were
doser to international prices than in most other developing economies.
The northem-tier economies-Japan, Korea, and Taiwan, China-
halted the process of import liberalization, often for extended periods,
and heavily promoted exports. Thus while incentives were largely equal
between exports and imports, this was the result of countervailing sub-
siclies rather than trade neutrality the promotion of exports coexisted
with protecion of the domestic market. In the Southeast Asian HPAEs,
conversely, govermments used gradual but continuous liberaization of
the trade regime, supplemented by institutional support for exporters,
to achieve the export push. In both cases governments were credibly
committed to the export-push strategy; producers, even those in the
proteced domestic market, knew that sooner or later their time to cx-
port would come.

22
 East Asia'ssectoral policies were usually geared toward export perfor-
mance, in contrast to the inward-oriented policies of less successful de-
veloping economies. Japan, Korea, Singapore, and Taiwan, China, all
relied on economic performrancecriteria, usuallyexports, to judge success.
For example, in Taiwan, China, the government suspended domestic-
content requirements that interfered with the exports of foreign in-
vestors. In addition, sectoral policies were closely monitored and
frequently adjusted. Thus, many of East Asia's "industrial upgrading'
programs of the lare 1970s and early 1980s were substantially modified
or abandoned when they failed to produce satisfactory results. Using the
export rule meant that even programs of selective industrial promotion
were indirectly export promoting.
Manufacmred export growth also provided a powerfil mechanism
for technological upgrading in imperfect world technology markets. Be-
cause firms that export have greater access to best-practce technology,
there are both benefits to the enterprise and spilloversto the rest of the
economy that are not reflectedin market transactions. These information-
reated extemalities are an imporrant source of rapid productiviy
growth. Both cross-economy evidence and more detailed studies at the
industry level in Japan, Korea, and Taiwan, China, confirm the signifi-
cance of exports to rapid productivity growth.
These experiencessuggest that economies that are making the transi-
tion from highly protectionist import-substitution regimes to more bal-
anced incentives would benefit from combining import liberalization
with a strong commitment to exports and active export promotion, es-
pecially in those cases in which the pace of liberalization is moderate.

Policiesfor RapidGrowthin a
Changing WorldEconomy

5 IT>71AT ARETHE BROADLESSONSOF SUCCESSIN THE HPAEs?

'sKI/Their rapidgrowthhad two complementry elements. Firz,
vv getting the fundamentalsright was essential. Without Jhgh
levels of domestic savings, broadly based human capital, good macro-
economic management, and limited price distortions, there would have
been no basis for growth and no means by which the gains of rapid pro-
ductvty change could have been realized. Policies to assist the financial

2-3
 -XIAIEMI RAC LE

sector capture nonfinancial savings and to increase household and cor-

porate savings were cental. Acquisition of technology through openness
to direct fbreign investment and licensing were crucial to rapid produc-
tivity growth. Public investment complemented private investment and
increased its orientation to exporcs. Education policies stressed universal
primary schooling and improvements in quality at primary and sec-
ondary levels.
Second, very rapid growth of the type experienced by Japan, the Four
Tigers, and more recently the East Asian NIEs has at times benefited
from careful policy interventions. All interventions carry costs, either in
the direct fiscal costs of subsidies or forgone revenues, or the implicit
taxation of households and firms, for example, through tariffs or inter-
est rate controls. Unlike many other governments that attempted such
interventions, HPAE governments generally held costs within well-
defined limits. Thus, price distortions were mild, interest rare controls
used international interest rates as a benchmark, and cxplicitsubsidies
were kept within fiscally manageable bounds. Given the overriding im-
portance ascribed to macroeconomic stability, interventions that be-
camnetoo cosdy or otherwise threatened stability were quickly modified
or abandoned.
Whether theseinterventions contributed to the rapid growth made pos-
sible by good fundamnentalsor detracted from it is the mosc difficult ques-
tion we have attempted to answer.It is much easier to show that tie HPAEs
limited the costs and durtion of inappropriately chosen interventions-
itself an impressive achievement-than to demonstrate condusively
that those interventions maintained for a long time accelerated growth.
Our assessment of three major uses of intervention is that promotion of
specific industries generally did not work and therefore holds litde
promise for other developing economies. Mild financial repression com-
bined with directed credit has worked in certain situations but carries
high risk. Export-push strategies have been by far the most successfil
combination of fundamentals and policy interventions and hold the
most promise for other developing economies (chapter 7).
But are these approaches feasible in the early 1990s? While limited
repression of interest rates may have contributed to ovemll higher rates
of investment in Japan, Korea, and Taiwan, China, these three north-
eastern economies undertook their initial growth spurts-and their
most sustained and forceful repression of interest rates-during a period
when it was possible for a developing economy to dose its financial mar-

24
 Icetsto the outside world. Furthermore,strong bureaucraciesand a gen-
emlclimate of government-privatesectorcooperation meant that their
restrictionson capitaloutflowswere more effectivethan similarrestric-
tions in many other economies.In today'sincreasinglyglobal economic
environment,fewgovernmentshavethe abilityor desireto closetheir 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 repressinginterest rateswithout provoling capital
flight is sharply narrowed.However,in some exceptionalinstances,very
mild financial repressionof short duration to increasecorporate equity
remainsa viableoptiOII. This has been the case in Malaysia,which has
wide open financial markets but nonethelesssucceededwith very mild
financialrepressionfor more than a year.
The export-push strategy appears to hold great promise for other
developingeconomies.But the conditions of market accessunder the
General Agreement on Tariffs and Trade (GArr), and odlier trading
arrangements,will hamper developingeconomies'use of policiesviewed
as unFairin major industrial-economymarkets.Subsidiesto exports and
directed-creditprograms linked to exports are not generallyconsistent
with the GATrand may therefore invite retaliation from trading part-
ners. Furthermore, like financialrepression,thesehighly directed inter-
ventions require a high level of insttutional capacity now lacking in
most developingeconomies.Fortunately,many powerfulinstrumentsof
export promotion are nor only within the institutionalcapacityof many
developingeconomiesbut remain viable in today's economic environ-
ment. Creatinga free trade environmentfor exporters,providingfinance
and support servicesfor small and medium-sizeexporters, improving
trade-relatedaspects of the civil service,aggressivelycourting export-
oriented direct foreigninvestmnent, and focusinginfrastructureon areas
that encourageexportsare allattainablegoalsthat are unlikelyto provoke
oppositionfrom tradingpartners. Indeed,some or all of thesehave been
part of the export push in Indonesia, Malaysia,and Thailand. These
three economies,the most recent participantsin the "economicmirade,'
may show the way for the next generationof developingeconomiesto
followexport-pushstrategies.

The phenomenalsuccessof the HPAEs is alreadyinspiringattemptsat

imitation.We haveshown that the HPAEs used an immense variety of

25
 ' _Ni1ACCLE

policies to achievethreecritical functions of growth: accumulation,al-

location, and productivity ,rowrh. The sheer divenrsiyof these policies
precludesdrawingany simplelcssonsor makingany simplerecommen-
dations, exccptperhapsthat pragmaticadherenceto the fundamentalsis
central to success.These market-orientedaspects of East Asia'sexperi-
ence can be recommendedw;ithfew reservations.More institutionally
demandingstrategieshave often filed in other settingsand they clearly
are not compatiblewith economic environmentswlherethe fuindamen-
tals are not securelyin place.The use of contestsin Japan and Koreare-
quired competenrand insularedcivil servants.In parts of Sub-Saharan
Africa and Latin .Amnerica, and elsewherein Asia where such institu-
tional conditions are lacking, activistgovernment involvementin the
economyhas usuallygone awry.So the fact that interventionswere an
element of some East Asian economies'successdoes not mean that they
should be attempted everywhere,nor should it be taken as an excuseto
postpone neededmarket-orientedrefonn.
The successof the HPAEsbroadensour understandingof the range of
policies that are consistent with rapid development. It also teaches us
that willingnessto experiment and to adapt policies to changing cir-
cumstancesisa key element in economicsuccess.In the followingchap-
ters we explore more fully the contribution of fundamental and
interventionistpoliciesto East Asiis remark-ablegrowth, and the crucial
role that institutionshave playedin their evolutionand application.As
we shall see,maakinga miracle is no simplematter

Note
1.Japan, which has been firmlyin the ranks of indus-
trialeconomicsarguablyforallofthiscentury,mayat first
seemto be an inappropriatesubjectfor study.However
manyof the policyinstrumentsusedby theJapanesegov-
emmnentduring the period of rapid growth, 1950-73,
have been used subsequentlyby devdoping economies.
Thus, notwithsranting Japan'slongerhistoryof modem
economicgrowrh,it may providesome usefulinsights
into the relationshipbetweenpublicpolicyandgrowth.

26
Growth, Equity, and
Economic Change
ml .HE EIGHT ECONOMIES OF OUR STUDY ARE HIGHLY
diversein natural resources,culmre, and political in-
stitutions. Japan, unlikethe others,wasalreadya rela-
tively mature industrial economy at chebeginning of
the postvar period. Moreover,the eight differ in the
degree of government intervention in the economy
and the manner in which their leaders have shaped and implemented
poticies.Korean policymakers,for example,have intervenedheavilyin
industrial,labor, and credit markets,whilepolicymalers in Hong Kong
have been consistentlyhands-off.
Despite many differences,however,the eight economieshave much
in common. In a number of ways, their postwar experience distin-
guishesthem as a group. Their most obvious common characteristicis
their high averagerate of economic growth. During the same period,
incomeinequality has declined, sometimes dramatically.These two
outcomes-rapid growth and reduced inequality-are the defining
characteristicsof what has come to be known as rhe Easr Asian eco-
nomic miracle.
The eight economies share six other characteristicschat set them
apart. Compared with most other developingeconomies,all have had:

* Morerapidoutputandproductivity growthinagriculture
- Higher rates of growthof manufactured exports
* Earlier and steeperdedines in ferility
* Higher growth rates of physicalcapital,supported by higher rates
of domesticsavings
a* IHigher initial levels and growth rates of human capital
* Generallyhigher rates of productvity growth.

2-7
 * A C LkkYI&CE

These charaaeristics are all related to their rapid, more equitable

growth. Someare sourcesof growth, some are outcomesof growtl, and
some are unique features of HPAEgrowth, but most fall in more than one
of thcse categories.

RapidandSustained
Economic
Gmwth

T HE EIGHT HPAEs GREW MORE RAPIDLY AND MORE CONSIS-

tendy than any otler group of economies in the world from
1960 to 1990. They averaged 5.5 percent annual per capita real
income growth, outperforming everyeconomyin Latin Americanand
Sub-Saharan Africa (except diamond-rich Borswana). Another East
Asian economy, China, has grown 5.8 percent a year since 1965 and
could stake a claim to join the ranks of the HPAEs.1
Figure 1.1 shows the relationship between income level relative to the
United States in 1960 and per capitaincome growth for 119 econormies
during the period 1960 to 1985. Developing economies were noc catch-
ing up with the advanced economies; more than 70 percenr of the de-
velopingeconomicsgrew more slowlythan the high-income-economy
average. 2 More disturbingly, in thirteen developing economies, per

capita income actuallyfell. Growth among the eight HPAEsis quite dif-
ferent. Their growthrates are significantlyabove the high-income-econ-
omy average.Unlike most of the rest of the developingworld, the HPAEs
were catching up to the industrial economies.Hong Kong,Japan, the
Republic of Korea, Singaporc, and Taiwan, China, were particularly
notable.
Other developingeconomieshave grown ist fr severalyears,partic-
ularlybefbre the 1980s, but few others have sustainedhigh growth rates
for three decades.3 Figure 1.2 shows the growth rates in per capita in-
come for 119 economiesin two periods, 1960-70 and 197045. The
11 that achievedrapid growth during both periods are in the northeast
corner. Of these, fiveare East Asiau successstories: Hong Kong,Japan,
Korea,Singapore,and Taiwan,China Japan'sshift from exatremcly rapid
growth in the 1960s to rates more typical of high-incomeeconomiesin
die 1970s is apparent. The other threeIIPAEs-Indonesia, Malaysia, and
Thailand-all show acceleratinggrowth,with highergrowth rates in the
secondperiod than in the first Indonesiais one of only three economies

'-28
 GROWTH, EQUITY. AND ECONOOM

1.1 GDPGrowth
Fhgure Rate,196085, andGDPperCapitaLevel,1960
CDPgrwth rate (percent, average, 1.960.85)

Taiwan,China i
~*Rep.f 'aHongKong
Kor
6 - *Rep.-f Ko --

ma~~~~~aa
lballand 111i

hOOh8SIB*
i j -

2-----O---- -*--_ -- *A

0 **. S ...... . ... S -. ..-

*S U HIgh4I.come econondes
-2 -PEs ----- - -* HPAEs
S Other developingecononies
*~~~~ I

. I III
0 0.2 0.4 0.6 038 1.0 1.2
elativepercapitaGDP,1960 (percentag of U.& GDPpercapita,1960)

Neo-This figuroplotsthis egession equation:GDPG = 0.013 + 0.OG2RGDP60 - 0.061RGDP602. N = 119: R7 = 0.036.

(0.004) (0.027) (0.033)
SorseSumnmersand Hestn (1991);Barnf(1989);Word BankdanL

to mnovefrom the bottom to the top of the distribution of growth rates

between the two periods.

DecliningIncomeInequaliyand ReducedPoverty

HPAES HAVE ALSOACHIEED UNUSUALLYLOW AND DE-

lininiglevels of inequality, contrary to historical experience and
contemporary evidence in other regions (Kuznets 1955). The
positive association between growth and low inequality in the HPAEs,
and the contrast with other economies, is illustrated in figure 1.3. Forty
economies are ranked by the ratio of the income share of the richest fifth
of the population to the income shareof the poorestfifth 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

RatePerisbence
GDP growth rate (percent), 1970-8S

- Sin.s .oret Taiwan, China

6%-
6$|- _.!._ ._ d_>_______ : * @ ~~~Singapore**
*Mai*a.siaRep. of Korea

. t *Thland Japa
4% -w*,*f

0%~~~~~

0 I

*. | U High-Income economies
4%- * *HPAEs
a Otherdeveloping economies

-6%-- !
.4% -2% 0% 2% 4% 6% 8% TEM
GDP growth rae (percent), 1960670

Nne: Boxesare seventy-fifthpercenrilcof growdtratcsineach period.

S&natvSununcrsand Heston (1991);Barn (1989);World BamkdatL

identifies economies with high growth (GDP per capit greater than 4.0
percent) and low relative inequality (ratio of the income share of the top
quintile to that of the bottom quintile less than 10). There are 7 high-
growth, low-inequality economies. All of them are in East Asia; only
Malaysia, which has an indexcof inequality above 15, is exduded.
When the East Asian economies are divided by speed of growth, the
distribution of income is substantially more equal in the fast growers
(Birdsall and Sabot l993b). For the eighr HPAEs, rapid growth and de-
clining inequality have been shared virtues, as comparisons over time of
equality and growth using Gini coefficients illustrate (see figure 3 in the
Overview). The developing HPAEs dearly outperform other middle-
income economies in that they have both lower levels of inequality and
higher levelsofgrow&h Moreover, as figures Al .7-A1.9 at the end of the
chapter show, improvements in income distribution gmerally coincided
with periods of rapid growth.

30
 GROWTH. EQUITY. AND ECONOX-rO

Figure1.3 Income
inequalitandGrowth
of GCP,1965!89
GOP growtb per capita (percent)

Rep.of Korea * Botswana

7 . .
Taiwan.China
Singapore*. Hoagon

' Gabon

5 ---------- apan---------- - --
] e Indonesie .M . , ,

i Th.lland :- Malaysia

_ _ _ _ _ ;___ Brazil
. _*I
IItaly
* Austria - ;.

.Spain
o France
* Belgium
United Kingdom AuS ilia i* Colombia
e Sri Lanka
Switzerland' * iine Mec,
Ia Pakistan '0Kenya

. India * Venezuela
1-._ lMalaw_ _ _ .A.t--
*NeChIIe Argenttna'j
Bangladesh *PBolivia * *eruge!tn

- Coted'lvoErr
* Maurita
nia
*. Ghana I
a Sudan *

-1 * -- - - - .- . __--...

Za~~mbia

$_ _ _ _ _ I
_ _ _ _ _ _ _ I _ _ _ _ _

0 5 10 15 20 25 30 35 40 45
IncomeInequality
N-et Incomeinequalityis measured
bythe radoof thcincomesharesof the richesc20pexnr andthe poorest20 percenrofthe popularion.
Sm:n. WorldBankdacaL

31
 ECLE

Twoqualificationsshould be noted herc.First, some studiesof Korea

have shown increasinginequality in recentyears; however,most of this
is due to dte rising value of assets, particularly land, rather than in-
creasedvariation in incomes. Second, reductionsin inequalityin Thai-
land havebeen relativelyminor comparedwith chosein the other HPAEs,
although Thailand's performance is still better than that of most devel-
oping economies.
Givenrapid growthand dedining inequality,theseeconomieshave of
coursebeen unusuallysuccessfulin reducingpovertyTable 1.1 compares
the declinesin poverry,definedas the inabilityto attain a minimal scan-
dard of living (World Bank 1990b), in some HPAs and other selected
economies(theperiodvariesdependingon data available).ncreasesin life
exptcy have alsobeen largerthan in any other region(seetable 1.2).

Dynamic
Agricultural
Sectors

T iYPICALLY,
tof
AS AN ECONOMY DEVELOPS,AGRICULTURE'SSHARE
he economy dedines. The six HPAEswith substantialagri-
cultural sectors-Indonesia, Japan, Korea, alaysia,Thailand,
and Taiwan, China-have been mak-ingthis transirion more rapidly
than other developingeconomies.4 But the decline in the relativeim-
portance of agricUlture in the HPAEsis not becauseagriculturehas lacked
dynamism.Acrossdevelopingregions,agriculture'sshare of output and
employment has declined most and fistest where agicultral output
and productvity have grown the rnost (see table 1.3). From 1965 to
1988, growth in both agricultnualoutput and agriculmtalproductvity
washigher in East Asiathan in other regions.Many factors contributed
to the successof agriculturein theseeconomies.Land reform (notablyin
Korea and Taiwan, China), agriculural extensionservices,reasonably
good fastructure (especiallyin the former Japanese colonies), and
heavyinvestmentsin rural areas(notablyin Indonesia) all helped.
East Asiangovernmentshave activelysupported agriculatralresearch
and extensionservicesto speed diffilsionof Green Revolutiontechnolo-
gies. Their substantial investments in irrigation and other rural infia-
struct hastened adoption of high-yieldingvarieties,new crops, and
the use of manufacturedinputs, such as fertilizerand equipment, to cul-
tivate dthne In Taiwan, China, during the 1950s, 45 percent of rhe

32
 GROWTH, EQUITY.- AND ECCOiL

Table1.1 Changesin SelectedIndicators

ofPoverty
PTerntage ofpoprdauion
below thepoverty line N mber ofpoor (millions)
F.rst Last Fis - Last Percent
A&vnenq Year year yar aAnge 'ar year chang

HPAEs"
Indonesia 197242 58 17 -41 67.9 30.0 -56
Malaysia 1973487 37 14 --23 4.1 22 -46
Singapore- 1972-82 31 10 -21 0.7 0. -71
-Thailand: ` 1962-86 59 26 -30 16.7 . 13.6 -18

Ot,ers
:Bra4YAb
:- - 0 1960-80 50 21 -29 :36.1 25A 29.6.
Wolombia -1971-88 - 41 25 -16 8.9 7.5 -15.7
:. CostaRica 197146. 45 24 -19 0.8 . 0.6 -25
C6reds'Iire 1985-86-1 30 31 1 -3.1 3.3 .6.4
I-ndia 1972-83 54 43 -9 311.4 315.0 1
Morocco- 1970-84. -43 34 -9 . 6.6 7.4 :12..
-R;
Pikismn0;
.-. > . 1962-84 . ; 54 23 -31 26.5 21.3. -19
Sri-Laik 1963-82 37 27 -10 3.9 . 4.1 5

:NoeThiis ableuseseconomy-specificpovertylines Officialorcommonlyusedpovertyrlineshavebeenusedwhenav2iblec In other

C3SC5die poveyie hasbeensetas 30 percentofean incomeor expeaditrc.Theenge of povertylincs,epressed in tems ofewendi-
. perhouseholdmemberandin tersofpursingpowrpari (P) dollars,isapproximnatdy$3004S700aycarin1985ececptfor-
CosaRic (S960),-Malaysia (SI 420), ind Singapore(S860).Unlessothenise indicatd, the cble is basedon expexliure perhousehold.s
- ember. - - - - - -- -- -- -:- - - :-..
- a. Measures for thse enties useinmcmccAherthanexpenditure.
b.Mcasurs fir dtse entriesare by hhold aier than by householdmemnber.
Smwz World Bk (1990b,tabics3.2 and 3.3). -

growth of agriculure was due to rising productivity, much of which re-

sulted fiom government programs (see Ranis 1993).
Information on the allocation of public investment betveen rural
and urban regions is limited, and it is difficult to make good compar-
isons among economies, but available data suggest that the HPAEs have
allocated a lager share of their public investment to rural areas than did
other low- and middle-income economies. Of critical importance in this
respect has been the build-up of infrastructure-roads, bridges, trans-
portation, electicity, water, and sanitation. Table 1.4 shows that there
has been a more even balance between rural and urban public invest-
ment in sanitationand water &cilities in Indonesia, Korea, and Thailand
than in other developing economies. The data on rural electrification
also suggest that the HPAEswith rural sectors have, on average, more ef-
fectively provided electricity to rural areas. Since the early 1980s, elec-

33
 I :1R.ACLE

Table1.2 LifeExpectancy
at Birth,1960and1990

Lif expeeane, 6a)

birthnh
:-&onoalrfreon -1960 1.990
-;HPAEs- :,--
HongKong 64 78
Indonesia 46 59
-K(orea,
Rep.of. 53 772
Malaysia 58 71
Singapor . 65 74
Th-
Nailand: - -: . ;-. -: - - 52 .68
: OtherAsW: .0 .62
-Chin. 43 -69
In&dia .47 58.
-Sub-SaharnAlea . 43 52
Lad Ame7ia and Cribbean . 54 . 70.
:Low-Inome econom .s : 36 - 62-
-Middle-lacnc
ieconomies . 49 66 -
-Industnaleamenimie 70 77:.
N- TAherCgionalpavsarweghreda.; - -
-tExdudes Chinaandndi:L
SountWorld Bankdra.t --

tricity has been universally available in the rural areas of Korea and Tai-
wan, China. Malaysia and Thailand have made great strides in rural
electrification. Endonesiahas not done as well, but even there die relatve
disparity between the urban and rural sectors is smaller than the dispar-
ity in econorries with approximately the same per capita income (Bo-
livia and Liberia) or the same population (Brazil [see table 1.4]).

Table 13 GrowthRatesofAgculturalIncome,LaborForce,andSharein Output,1965488

(percent)

* - Decline,Average anased'growth rv&eof

1965-88l. A - t,i d Atcrds
:,,-;,: -irc,--: 196
.;-:-;:e 198 kiborfirce
A'd'-:.-.:'..' rru8,,,A "
t SouhAsa ,3 2. ' <2., 6 1.l9U.::
4
:' asrArn 41 22 : :-t _ - 2.2
:~xree..LLL2jt.... ~~c:~;'nA2, 34 21~~3 LO)

34,., S :10" ~.. 37 2 17

34
 GROWTH. EQUITY. AND ECC

Equally important, however,were the typically low levels of direct

and indirect taxationon agriculturein East Asia. During the past three
decades, dozensof governmentsin other regions, eager to promote m-
dustrial growth, have funneled surpluses from agriculture to industry
through taxes,food price controls,and pro-industryallocationsof pub-
lic investment. Less overtly, govemments have fivored manufacturers,
and hurt agriculture,by overvaluingcurrenciesand protectng domestic
industries tharmanufactureagriculturalinputs and the goods purchased
by rural households.The exchangerte chatresultsfiromrestrictionson

TableIA Comparison
ofRuralandUrban
Public
investnent
ARnral-urbandipadrics in accessto gervies, 1987-90
(100= rural-urbanpaitj)
-c y - : . - . Watrr - S- ti don
S PA Es
Korea,Rep. of ;54 101
Thailand -26 1- 102
Indoncsia -168 113
OtcrAia - - :- 64.6 38.6
Latin America 58 4665
SubS- bnibwAfrica 43 30

.- Prcatage of rrurl adnduran pop ukonser&d by ecaicity

co~namy Ur Rura
' PAEr
Indonesia,-1984 39 10
:M.a:ilays.;1;983 - : -:0: g 5- 85 55
Thailand,1984 . 78 :40
-OrhlerAsik:

-Bangradash, 1981 :20 2

:India, 1981 <25 15
-S -Lakna,j982 -.. 35 8
- LeninAme-ic
*~ArcndinaJ
1982- ->95 : 5
Bolivia, 1981: 72 . 9
.. Bra;Ail,
1981 :-.i
N :0 0- - :
.......... >95 19.

;C6tc -dvoir, 1981 93. : 20

i iberia,1982 ^ ~-- . -~ 86 ; ; 4 ~0~
- Senegk1982 . * ; 83 . .12

Soercertoppanel:
UNDP fariousyis);Bottompan. Muasii
('r (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 HPAEswith
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
andGmwt intheAgricultural
Sector
.7=,,,_T; r--,.
Rep.of Kore,

Malaysia

-------w------
Pakis-tn'
_, ,
w __. ......... _. _---- --- -'i-~~R 19709~~
PSilippines K. -. . . _ 1 j

Sri lanka '::H-~

_ _ +_
_ _ _-1_
_j___________________________ _ __I___._________

! : I I *; : I ! | I :~~~I
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
Levelof piotection (percent)

Rep.
of Korea

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

Thaland . --

PsidalsiU =96 i
___ ____ I ___ ~~~~~N1970s
I - 3----
197ks
PhEippines . . I k

Sri lanka IE~ .~ ~ ~~ ~ ~~

~~~
l_ _ _ _ _ _ __ _ _
1 ~ ~~~~~~~~~~
I

-2 0 2 4 6 8 10
Agrutural growthmte (percent)

Sour= Schiff(1993).

36
 GROWTH. EQUITY. AND EC_OO

agricultural sector receivespositive protection. Thailand's taxation of the

agricultural sector was similar to South Asian levels in the 1960s and
1970s but fell in the 1980s while taxation in South Asia was rising.
We are nor suggesting that the HPAEsavoided transferring any re-
sources from agriculture to industry. Rarher, in contrast to ocher devel-
oping economics, they transferred fewer resources. Moreover, transrs
were often voluntary-financial savings, for example-and were nearly
always limited to levds that did not choke off agricultural gwth.

RapidGrowth
of Exports

A NOTHERSTRIKINGFEATUREOF THE HPAEsHAS BEEN THEIR

superior export performane This is reflected in their steadily
Ilk rising share of woldd exports (see table 1.5). As a group, the
HPAEs increased their share in world exports from 8 percent in 1965 to
13 percent in 1980 and 18 percent in 1990. Manufactured exports have
provided most of this growth. From 1965 to 1990, Japan emerged as the
world's biggest exporter of manufactured goods, increasing its share of
the world market fiom nearly 8 to almost 12 percent. In the 1970s and
1980s, the locus of growch shifted to the Four Tigers, whose share of
manufactured exporrs grew nearly fbur tmes faster than Japan's (see
table 1.5). Beginningaround 1980, the three SoutheastAsianHPAEs (In-
donesia, Malaysia, and Thailand), which had been historically depen-
dent on commodity exports, recorded a sirnilar but so far smaller surge
in manufactured exports.
Some analysts have, with hindsight, attributed these achievements to
unique cultural and geographical circumstances. But there was litde evi-
dence at the outset that East Asian economies would achieve such spec-
tacular results. In the 1950s even trade optimists were export pessimists
and did not anticipate that Korea's exports would grow four times as fist
as world trade during the next thirty years (see, for example, little 1982).
One obvious efFec of rapid export growth has been a marked in-
crease in the openness of thcse economies, the share of exports plus im-
ports in GDP (see table 1.6). Malaysia has been heavily tade oriented
throughout its post-colonial history while Korea, Thailand, and Taiwan,
China, had average trade propensities untdl 1965 but have since moved
wel above intemational norms. The trade orientation of Indonesia,

37
 I1RAc IrE

SelectedEastAsianEcononies,
Table1.5 ExportPenetration, i965-90
Share in wuorldeports S-are iL developing-economy xporm
Eeonq
- :1965 1980 1990 1965 -1.980 1990
-Totalmg,parn

Japian -7. 5.0 9.0 -7

Fourflgcr- 1.5 3.8 6.7 6.0 13.3 33.9
SoutheastAsian NEs - 1.5 - 2.2 2.4 6.2 7.8 12A
HPAE subtoal- 7.9 -13.1 18.2 12.2 21.1 56.3
All developing economies .242 28.7 198 100.0 100.0 100.0
World 100.0 100.0 100.0 NA NA NA
: --i - a-u.-:;

Japan> -.
-- =.- ;- 7.8 11. 11.8 -
Fourr rsa - 1.5 5.3 7.9 13.2 44.9 61.5
SoutieasAsian NIEsb- 0.1- O 1.5 1.1 3.8 12.0
HPArEsubtotal 9.4A 17.3 21.3 14.- 48.6 73.5
* AIIdevc1opingeconomies 11.1 11.8 12.9 100.0 100.0 100.0
*Wodd . 100.0 1;00.0 -100.0 NA NA NA -
-Nor available.
NA-.Norapplicable.
-ai Republicof Korea.Hong Kong,Singapore.and Taiwan. China.
b. Indonesia.Mala a. anlThallu:xL
: ort U.N.Trade Systemsdarn.

w-ith its vast economy and only relatively recent export drive, is only
sightly above the world average but growing.

RapidDemographic
Transitions

THE DEMOGRAPHIC TRANSMON FROM HIGH TO LOW BIRTH

and dearh rameSbegan in Europe and North America with the
.1. Industrial Revolution, and it took nearly two centuries to com-
plete. That same transition is occurring much faster in the developing
world. In virtually all economies, deati rates have fillen substantially in
the postwar period; in ali eccept Sub-Saharan Africa, birth rates have
also fllen significantly (see table 1.7). However, compared with other
developing regions, the transirion to low fertility began sooner in East
Asia (in the 1960s in the north and in the early I 970s in the south-and
of course even earlier in Japan) and has gone firther.

38
 GROWTH, EQUITY, A-ND E

Table1.6 RatioofTotalTradeto GDP

Ecnousn" ; EO1970 1980 1985- 1.988
HPA&
'Hong Kong. 1.50 1.52 1.78 2.82
Indonesia- . - 0.25 0.46 : -0.38 -0.42
-lorea,Rep. of: - 0.32 0.63 :0.66 0.66
Japan 0.19 '0.25 0.23 0.11
Malysia 0.89 .1.00 0.85 -1.09
Singapore 2.12 3.70 -2.77 3.47
-Taiwan,China -.0.53 0.95 ...0.82 0.90
Th1ailand 0.28 049 0.44. 0.35
SubR
-SSI,arnAfr-k - 0.24 : 0.30 0.27 0.45
5out6Asia -- ; ; :; 0.17
-: .-; - -0.11
0.16 0.19
A-men-Amc'sa
ad Caribbema 0.20 025 0.22 .0.23
N..-eTotlrade .-valueofxportsandvalueofiimpomr
dividedbygrassdomesdc
- -.- produce :- -. - -t'
-.SowrwWodd'Bandata. Taiwan,China,vaniousissues.Nadionial
Accounts
Nations).
SMtaddcs:AnalysisofMainAggegates, 1988-89 (Unietd -

During the period 1965-80, all the developing regions of the world
experienced a marked decline in crude death rates (see table 1.7). The de-
cline in most economies was 30-40 percent and did not vary much
among regions. There was, however, substantial regional variation in the
extent to which declines in birth rates held in check the potentially ex-
plosivegrowth in population from the rapid mortality decreases. In Sub-
Saharan Afriicaduring this period, the declines in birth rates were around
2-10 percent; in South Asia, 10-30 percent; and in Latin America,
3040 percent. The sharpest declines were in East Asia: 40-50 percent.
As a result, the rare of population growth declined in all the East
Asian economies, in some cases quite sharply. For example, in Korea it
fell from 2.6 percent a year in 1960-70 to 1. I percent in 1980-90; in
Hong Kong, from 2.5 to 1.4 percent; and in Thailand, from 3.1 to 1.8
percent. In Latin America, fertility dcdines were also sufficient to reduce
population growth rates, though generally not to the levels observed in
East Asia.-In South Asia the picture is mixed, with fertility declines suf-
ficient to reduce the rate of population growth in Bangladesh but insuf-
ficient in Nepal or Pakistan. In Sub-Saharan Africa, the combination of
sharp declines in death rates and modest declines in birth rates resulted
in an acceleration of population growth-for example, in Ghana from
2.3 percent a year in 1960-70 to 3.4 percent in. 1980-90, and in Kenya
from 3.2 to 3.8 percent.

39
 Table1.7 TheDemographic
Transition
(percent)
Changein -ange in Aventgeannul
cr-de crude grovtt of
birth . eadl popadation
rante, rate,
Eonomy
- 19650 1965-80 160-70 1980-90
EaswAsia
* HongKong --52 -54 2.5 1.4
Indoncsia -40 -55 2.1 1.8
. Korea,Rep.of - -54. - 2.6 1.1
Malaysia -25 -58 2.8 2.6
Singapore 45 -16 2.3 2.2
Thailand -46 -30 3.1 1-.8
LatinAmen'ca
Bazil - -31 -36 -2.8- 2.2
Mexico -40 -55 :3.3 2.0
Peru -33 -50 2.9 2.3
Venezueda -31 -38 3.8 2.7-
Sub-SaaranAfri-.
Ghana -6 :-2 23 -3A-
Kenya -13: -50 -3.2 3.8
SierraLeone -2 --29- -1.7 2.4
Tanzarta -2 -22 2.7 5.1
SouthAj~a
Banglad -27 --33 2.5 2.3
India -33 , -45 2.3 - 2.1
Nepal -13 -: 42 - 1 . 2.6
.-Pakisran. I -13 -43 2.8 31
-Nor avaibbk.-
So5Wrrw
WorldBa dara -- - - :-

HighInvestment
andSavings
Rates

P HYSICALINVESTMENTINCLUDESALLOF THE ECONOMY'SOUT-

Jput that is not eitherdirecdyconsumedor usedup in the pro-
ducuionof othergoods.Machines,buildings,and infrastructure
are physicalcapital,bur elements ofworking capital,suchas inventories,
are also important. Economiststraditionallyhave viewedinvestnent as
one of the driving forcesof economicgrowth. In a dosed economy,sav-
ings is the only sourceof invesrmenr,and the two, by definition, must
be equaL But in an open economy, investment can be financed by bor-

40
 GROWTH, EQUITY. AND ECOQi

rowing from abroad as well, that is, with foreigners' savings. Even so, in-
adequate domestic savings will eventually pull down investment rates,
either direcdy or through constraints on the continued build-up of for-
eign liabilities, which must eventually be repaid firomdomestic savings.
Between 1960 and 1990, both savings and investment increased
markedlyin the HPAEs, outstripping -',e performanceof other develop-
ing regions (see figure 1.5). Savings rates in the developing HPAEswere
lower than in Latin America in 1965, but by 1990 they exceeded Latin
America's savings rates by almost 20 percentage points. Investmenc lev-
els were about equal in Latin America and East Asia in 1965; by 1990
East Asia's investment rates were nearly double the average for Latin
America and substantially exceeded the rates for Scuth Asia and Sub-
Saharan Africa. The HPAEsare the only group of developing economies
in which savings exceedsinvestment, making them exporters of capital.
When we compare the HPAEs individually to all 118 economies for
which investment data exist, the picrure is more complex (see figure
1.6). During the period 196045, the HPAEs' investment levels were in

Figure£5 Savings
andInvestnent
as a Percentage
of GDP
Gross
domesticinvestrnt

Developing
HPAEs

LatinAmericaandCauibbecn

SouthAsla

Sub-Sahaan
Africa

Gross
domesticsawings

HPAEs_-_
Developing _ _|
I --_____________________________
- i .1-...-&-----

LatinAmericandCaribbean

SouthAsia

Africa
Sub-Sahoran

: - . g. e10 20 -30 40
N-eo:The regionalavages arcunweiglhted.
SepurcWorld Bankdarn.

4'
 RAC L E

InvestmentRateasa Pementage
Fgut 1.6 Average ofGDP,1960-1985, Level,
andGDPperCapita 1960
AverageInvestmentrate (percentageof 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' ! . * Otherdevelopingeconomies

0--

0 0.2 0.4 0.6 0.8 .0 2.2

Relativeper capitaGDP(percentageof U.S. GDPpercapita, 1960)

Note.Regressionequation: 16085= 10.125+ 59.12ORCDP60-51.88IRGDP602.

N = 119: R2 = 0.295.
*(1383) (10-344) (12.593)
Swrre: Summersand Heston (1991); Bam (I989): World Bluk dat.

the top quartileof all low- and middle-incomeeconomies,but sincein-

vestmentgenerallyriseswith incomes,this is not remarkable. 5 What is

remarkableis their high share of private invesrment.Figure 1.7 com-

pares averageprivate and public investmentshares in GDP for 1970-89
in the developingHPAEsand other middle-incomeeconomies.Private
investment is about 7 percentage points higher in the HPAES than in
other middle-incomeeconomies.It rose from about 15 percent of GDP
in 1970 to nearly 22 percent in 1974, then declinedand held at about
18 percent between 1975 and 1984. Private investment contracaed
sharplybetween 1984 to 1986, reflectingthe global recession,then re-
coveredby 1988.6 In contrast, private investment in other low- and
middle-incomeeconomies has remainedrelativelystable at about 11
percent of GDP.

.427
 GROWTH. EQUITY, AND

The story is differentwith respectto public investment.In the 1970s

overalllevelsof public investment did not differmarkedlybetween the
HPAEs and other developing economies; during the decade public
investment rates in all economies rosefrom about 7 to 10 percent (see
figure 1.7). But during the 1980s the HPAEs and other developingecon-
omies diverged.In other economies, the fiscal contraction of macro-
economicadjustment was reflectedin lowerpublic investmentraces.In
the HPAEs,conversely,public investmentshares actually rose between
1979 and 1982 and then remainedat a levelnearly4 percentagepoints
above their 1970s average.Only after 1986 did they begin to dedine
towardhistoricallevels.In short, in strikingcontrast to elsewhere,pub-
lic investmentin 1980-37 in these East Asian economieswas counter-
cyclicalto the reduction in private investment.

Creating
HumanCapital

I-N NEARLY ALL THE RAPIDLY GROWING EAST ASIAN ECONOMIES,

the growthand rransformarionof systemsof educationand training
during the past three decadeshas been dramatic The quantity of ed-
ucation children receivedincreasedat the same time that the qualityof
schooling,and of training in the home, markedlyimproved.Today,the
cognitiveskill levelsof secondaryschool graduatesin some East Asian
econormiesarecomparablero, or higher than, those of graduatesin high-
income economies(seeappendix 1.2).
Figures 1.8 and 1.9 present a stylizedsummaxyof the resultsof re-
gressingprimary and secondaryenrollment rates on per capita national
incomefor more than 90 developingeconomiesfor the years 1965 and
1987.7Enrollment rates are higher ar higher levelsof per capita income.
But the HPAE'senrollment rates havetended to be higher than predicted
for their levelof income. At the prmary level,this was most obviousin
1965, when Hong Kong, Korea, and Singaporehad already achieved
universalprimary education,well ahead of other developingeconomies,
and even Indonesia with its vast population had a primary enrollment
rare above70 percent8 By 1987, EasrAsia'ssupenroreducarionsystems
were evident at the secondarylevel. Indonesia had a secondaryenroll-
nmentraceof 46 percent, well aboveother economieswith roughlythe
same level of income, and Korea had moved from 35 to 88 percent,

43
 jA C L EitACLE

Fgure17 PublicandPrivateInvestnent
Private Investment/GOP(percent)
25

.5

-D- HPAEs(excludingJapan) Other LMlEs

0
1969 1974 1979 1984 1989

Public Investmfent/GDP
(percent)
20 - ------

5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I
10

-U111-
FPAEs(excluding Japan) -4- OtherLMIEs
0
- II ' } t i ~~~~i
; ! i . ; 1 : i I I i i , I

1969 1974 1979 1984 19

Note: LMIEs
= low- and middle-income cconomics.
So=eueWorld Bankdami.

maintaining its largeleadin relativeperformance-Only in Thailand was

the 28 percent secondary enrollment rate well below the income-
predicted 36 percent and the 54 percent mean for middle-income
economies.9In recentyearsThailandsweakeducationalperformancehas
been felt, as seriousshorages of educatedworkershave begunto threaren

44
 GROWTH. EQUITY. AND ECC

Firm 1.8 Crss-Ecnony Regressin for Prmaiy Enrolment Rates, 1965 and 1987
Pdmatyenrollmentrates
140' - -

:120 .. ......... .
120-- -- .n lindonesia (1118V
Brazil(108) Singapore(1053 [37.0] HongKong(106) Brazil(103)
U [21.9] * ~~~~~~~~~~~~~[-7.5]
ioo Rep.of Korea-(101)u -
[29.2,1 HongKong(103) - - _- i7aysb(102) Rep.of Korea(101)
. Malaysia(90) [15.61 Tbai95) [4.3] [-7.8
Thailand(78) [17-9_ [7.01
8 [10.8]

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

*Bangladeh(4.9) ,Pakistan (52)

40-its . 1965 - -.
[O79]1 9 7
4-Palkistan (40) 6$2........... . ------- 197---
1-24.5]
20-- Lw .
0 250 500 750 1,750 2,000 2,250 0 500 1,000 1*500 3,500

Per capita Income (198 U.S.dollar)

Nater Figures in parentheses arc enrollment rates brackretd numbers show rcsiduaLs
Sourre: Bchrnan and Schneider (1992).

continued very rapid growth. In part as a function of their success in in-

creasing enrollment, the East Asian economies have also been faster to
dose the gap between male and female enrollments (see box 1.1).
A common, though imperfect measure of educational quality is ex-
penditures per pupil. Between 1970 and 1989, real expenditures per
pupil at the primary level rose by 355 percent in Korea- In Mexico and
Kenya, expenditures rose by 64 and 38 percent, respectively,during the
same period, and in Pakistn expendituresrose by only 13 percent be-
tween 1970 and 1985 (Birdsall and Sabot 1993b). These dramatic dif-
ferences reflect mosdy differential changes during the period in income
growth and in the number of children entering schools, both of which
favored the East Asian economies. A somewhat better measure of school
quality is the performance of children on tests of cognitive skills, stan-
dardized across economies. In the relatively few international compar-
isons available from such tests, East Asian children tend to perform
better than children from other developing regions-and even, recendy,
be'tter than children from high-income economies.' 0

45
 & 4TIRACLE

Fimre
1.9 Regresin for SecondaryEnrollment
Cross-Economy Rates,1965and1987
Secondaryenrollment rates
100 - . -. ,U , ,~- , _- -; , . , -

. ¢ ,| I ', . . Rep.of Korea(88)

so - -- -- i

60 - -.. Malaysia(59)

Singapore(45) Indonesia1 _-
40 .119.5] >I BraA 9*-[18.
40~~~~~~~~ - - Rp're\35 --lt- ril3s

*!k [19.1] .- ~slhailand T (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
Percpl Income(In 198 US. dollrs)
NomtFigures
in parentheses
arcenrollmentraes;bracketed
numbesshowv
residuals.
SeurceBhrman andSchneider(1992).

How much a childleams isalsoinfluencedby the nature of the leam-

ing environmentin the childcshome-Again, children in the East Asian
economics had advantges in the 1970s and 1980s. Using a simple
index taking into account the mother's education and the number of
children at home, we estimate that the learing environment in Korea
during twenty yearswas enhanced 114 percent more than in Braziland
147 percentmore than in Pakistan (Birdsalland Sabot 1993b).

RapidPrdu Growth

OUTSITANDING
HE FINTAL CHARACTERISTIC
OF THE HPAFsIS
their rapid productivitygrowth. That labor productivity in the
HPAEsincreasedrapidly is self-evident per capita GDP grwth
dependsprimarilyon risingoutpur per worker Someof this increasein
output per worklerresultsfrom increasesin physicalcapital per worker
(for exarrplckmore machines)and some reults from increasesin human

46
 GROWTH. EQUITY, AND:,F-COX0

MT VT!"12 rT-77
Ze.
,i Mz
w,Was-,FAA
IF.AJRL'' 11AAHI'..MRICALLY-k6---,----!: eci c.- ene tv o cu!,adng:V
.9d. es Lrsuto - oyt i'an biin subiiiiia 7 4.
o y a=,;-.,us ygra uaUj;'..'ib;gdsj gender' J--- lop.1
Th d il iffii:ntii f6cyiii-,.,
in ca-d6i. n";'evc
1.o-- -rnenriv aryan sc6ndzrj'k:&
-1--'-'6d---'
c,-th ower ataun.-..,,.-,,,to-aiaineas-incomesmaeast
--- --bf thi Aia tir).Ji-I 965i at c pd-n-.----
yztdj,.,1mpr6ve, I the' 'c[ p

V V
rnaryp_ c ccorioriuq,-,gnder
,,;,.t-',"Jt)r,outst e.--homm`-
e,--.tlii i" The- e'--HPAB_ 'j!aI[70Wid-,- WM Swl"cr-,,E=ticxpected.,,-By-
e:gcn-':-'.,-',;;,' '1987'thi,&ur
pmg jan!eqonoa,Lies-f6i.W.W&6ca.
ot er eV 0 aM
r; ;cconomjqs,' splteZwr'u''iat,norms-,,ttl2t. put, greater--, achieveduniversaip6mary cation
I

C, :r
Wqq01L- ucation-ofs6diaahi, tisci; -Ov-iiiiirii'n'aungthe''gendir p-at'th'a-'t-1e'%'rib7SLLbi6iL
dvt: I ii ai C-d6cationl., dau
g gender-gaps,M'Mm um mtescontinik46"
10 ,i'. E. -'. r -;- " -, - -. - , -- -. 4.-, -"s. I. .......... Sl

M e; ra er. C CXCCPtI0r4.-- in-liow-tncom"C'

t Ca-1r0m-1 SUCCCSS
'. -,; Z:. - , -
PUS 0r-t1MVCrsJid[7-, P-M
Is
7-,C ...
'yucauon Sphe'thi! m c3q%=on- o e,,
f G61 oni ive4rgu -64c4,,th
7 S31S 64i =&T.1fij,'-1991 :F6v
qarrip
lecon6nues, GNPpq.qLpibi '$500'-%
0'ff 'I. A -in aTsmoolv -o
m Miienr, etwccwiboys
.AD"Mak-Yea 6f'Edudfialfl ges-2 6fita- i pio,in, fs-,--=,a, as,'no-r4n
p
-.44 In!;1965, 1-ir '(6c,cq 6 Thi
nAmi
gcn..d&,gaps-,-avtqe-k'di
Iiidcc'd'-"
-16iill: increastrig.second aryschool-pp
e;M as;mu .2-C
Ij 'an
C5 70, rhl --v-166'.1-970icl -,q80s!2,S1t-.1S
ey opmg.-wo &
-Ccon6mli SUCCCSS Ilh laL%ngj
cntsto .98 PV

V4.1 gn !cr:ga,p
-wJ-`6ii6,
4-. -jk er 1) Ja ii I
capim-41 -6
A-siadeconornies.5nour,stu Z.1
fi

capit2l. (for cample, better schoolecl workers). In addition, however,

somc may resulr fiom greater efficicncy, fiom changes in producdon
practices that result in greater output from the same stock of physical
and human capital. We caU the increase in productiviry thar cannot be
accounted for by measured increases in inputs total factor procluctivity

47
 r1;CTACLE

(TFP) growth." Among its sourcesare bectertechnology,

betrer organi-
zation,gains fiom specialization,and innovatons on the shop floor.
As we demonstrate below,about two-thirds of East Asia'sextraordi-
nary growth is attributable to rapid accumulation;that is, to unusually
rapidgrowth in physicaland humancapital.The remainingthird of ti-s
growth cannot be explainedby accumulationand is tierefore actribut-
able to increasedefficiencyor TrP.This is largerelativeto other econo-
mics, both absolutclyand as a share of output growth, and therefore
partly explainswhy theseeconomieshave been catchingup wriththe in-
dustrialeconomies,whilemost other developingeconomieshave not. In
box 1.2 we reviewtwo broad interpretationsof the relationbetween ac-
cumulation and output growth and the link between productvity
changeand technologicalcatchingup.

EvidencefromCross-EconomyRegressions
of perCapitaIncomeGrowth

In this section we use two statistical techniques to examine the rela-

tionship betweenaccumulationand output growth Our first model is a
cross-economyregressionin whichwe estimate the relationshipbetween
the rate of rcal per capita incomegrowth and the share of investmentin
GDP,and two measuresof educationalattainment, controllingfor dte
rate of growth of the economicallyactive population, and the relatve
gap between per capitaincome and U.S.per capita income in 1960 at
1980 U.S. dollar prices.'2 Our measuresof educationalattainment are
primary and secondary school enrollment rates in 1960, the area of
human capiral accumulationin which the HPAEsshoweda substantial
differencefrom other low-and middle-incomeeconomies.
The results of applyingour statisticalanalysisto 113 economiesare
summarizedin table1.8.The estmatesindicatethat fcror accumulation
matters. Investmentin both physicalcapitaland schoolingcontributesig-
nificantlyto economicgrowth.'3 An increaseof 10 percentagepoints of
GDPin the rate of investment-about the differencein privateinvestment
rates between HPAEsand other low- and middle-income economies-
would raisethe growdt rate of GDPper capita by 0.5 percent An increase
of 10 percenmgepoints in the primary or secondaryschoolenrollment
rarewouldraiseper capitaincomegrowthby 0.3 percent
Beyondaccumulation of physicaland human capitl, initial income
also has a significant relationship with per capita income growth.

48
 GROWTH, EQUITY, ANDE'CO.'i

M77 i 7,
x ,Zrl%evjc,;i-.Ii_-.Ip"-."-
_2__ A

)NOMISTS, ROAD, mmmenr umanan

1-4

accumuw- er:
onj'am Murpur
VC.'.
-w-V h ton;t,&I
.,Jf.

rexten
Flj e --o 3K
3Vlcwlls"_b"l -
V4,1.
_T'.
yIC um2n ppi ajt
ait
r=en
ini
p nonluesi- irk*, -o'wth' an
A ysi
&pfil- &72ACt
1.06
".w -,,-,,.amo4ccon6n3jcs_cr cromki&m
rz-k iE66-Er; esame, 6i&bffinva= ccononiic,_-or
-4 gan,=qon;,-jnnovau
nonuc-.,gr -0 o ogyarc o arm y.14666i
MR assem- 'lik" r
P tLon iAiii '446 31i
ccum non-, cKami)
t
_-llFAE:fa ii'ibilbimer vict-, ruo M
have
"d zin 4 "W o .. ........
CoDrifth lporpcmicnVor economc gm.
M, Om ccum
07Ww"Ah
con-._t-,,, -,-tc oil ficior-
_9M Y
Z.- Q
---- 2ssum rec
i=cs 0 ,p LW2'CX (SOIW49-5.7)
N
-nu e-mcoifi 0 f -T,
P,P'. J:i
_7
ll,z, Om r coiiiii.econoun v
Oj_ Cj;V;.
onspi c
aupajmong de
,-,'2ni t
-J V 19P
crcmeat&jj06, ;_2 t
'gro
er,
6ibasji-,tVoutpu poon 0
0 gt
.4;_
f
ItaL
A

econo V4
n

OD

'p t
Sd'

49
 i-MIRACLE
L

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

?%e 7 aWe to acqwre technologyhniore~ =Žie4iy; daiinical efnaen4 chan ge- Bo6htho y -andempi ic
prgres&more rapidlj than induiial-- economics. evidence:leadusto, ncludechat tol&cr pruc-
U-pnid4crrybasedcatcbiiig:
K>Thisiis radbe- up..
i" a vy growthl
; thdeindusirialecmomiioicilrer
< . mov dc
j-- r'BuLiiiwrnadbsn4a1 be fpraiceis uiseif amoving, t. technologicacg the-. m
targetTow fiictor.producvity cbangeii; el.igh t
-nonal e practic4Whileinflow-and middle-icome
i--- OM '-Co .-di Z = -sc toieto, oc
g5wrkXChenry 1986;B.n an 192 m iic j(Pacad P 11993XForpductrv-
-liI;pm~~~~~~n
a &uncV~A9)--4 l .nocr;
i, pendix
in resents acformiantan4is"oEfi rea~onow-bs
; - >~~~~~~~~~
pracice, whitWChe
bcwce
slp temi
ni#eicrsKn r
"ltec-noloejcl
~do
O& 4'
banmcemanon
iia bhs'd
-
mov4iidj
-moIn
ccInomits, cam-b
Ass*
occr,
zzG*
gaicing on, 6te
'lo-Cand

9; >.~movemenrs)owardbestphiice,whicWweolraech'- -. must .: .- Ceposk,iv---:

Economies that were relatively poor in 1960 grew significandy fister

than relatively rich ones, controlling for levels of education and invest-
ment So although, as shown above, poor economies did not do better
on average than the high-income economies-income levels were not
converging-this was partially because poor economies invested less.
But an economy at 50 percent of the level of U.S. per capita income in
1960, and at average levels of education and investment, would have
grown about 2.1 percentage points more rapidly thian the United Stares.
We call this conditional convergence, because economies with low rates
of investment and school enrollment would not catch vp despite the
"advantages" apparently offered by being relativly poor.
This conditional convergence can be interpreted as a measure of the
gains realized as a consequence of moving from lower to higher techno-
logical levels, or "catching up" (see box 1.2). But rhe relative income
variable may capture more dtan the productivity-based catch-up effict
that some initially poor economies benefited fromr. One of the impor-
tant empirical regularities found in the early literature on structural
transformation (Kuznets 1959; Chenery 1960) is the discrepancy in the
average product of labor between traditional (agriculture) and modem
(industry) sectors at low income levels.A dominant share of producriv-
ity growth in low-income economies can be attributed to intersectoral
realocation of labor from agdriclture to industry (Pack 1993c). Thus
the initial income variable may also be capturing the reallocation effect
of this structural change on per capita GDPgrowth.

50
 GROWTH, EQUITY, AND ECO:4I-

Table1.8Basic
Cross-Economy
Regression
Results
(dependent variable: average rate of realper 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)
RelativeGDi' to U.S.,1960 -0.04O -0.0293~*-0.0320
(0.0118) (.15)(0.0110)
Priuary enrollment,1960 0.0264M 0.0233 0.0272
(0.0065) ~ (0.0062) (006)
Secondaryenrolmenr, 1960' 0.0262 0.0160 0.0069
* (0.0139) (0.0132)- (0.0131).
*Growth of population, 1960-85 0.1015, 0.0201 0.0998
(0.2255) ~~~(0.2095) (.03
Averageinvestmendr/CD.
1960-85 0.0578 0.450.0285
(0.0224) (0.0211) (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 cachceIllStandard
crror is in p crnhses
t.Also indiudeTun".isiand
South.AMica.
S6an WoildBanksnffcsdnnrcs.

Armnedwith these statisticalresults,we can examinetwo facrs about

the role of accumulationin growthin the HPAES.First,our estimatesof
the parametersof the cross-economygrowthequation enableus to do a
simple "accountingfor growth"in the HPAEs.Table 1.9 shows the pro-
dicredcontributiontogrowthofinvestment,humancapital,population
g'rowrh,and relativeincome for eac-hH-PAE.It alsoshowsthe share of ac-
tald growthpredictedby thesevariablestExcept for Hong Kong (44 per-
cernt),60 percent or mare of the actual growth rate in the HPAESis
predicted by the accumul-ationof physicaland human capital,initial in-
come levels,and population growth, ranging as high as 87 perce-ntfor
 _ E I R-A-CL E
FML~"xN

Table1.9 Coutnbution
of Accumulation
to theGmwthof theHPAEs,
196045

11- os
,~~~~~~~~~~mm Iha . w . , ; r_ f +:'
fl,.~A,
P0-
-aaf hiv -- fa-d - aeihsh ral -.. &vh If-d rmd
6
Pear Mg. .6 PedSI MlUe prial 5W. pen&tu ugaqh pia
*VWzh& mlal (vi uik(si gush (rf-) guhw. s

Ian~~~x - - -~40676
. .X7e 4270
.e70 (.26)
{.Nz 476
wu78
402 (.1)
-141 4710
.1.1 4.151 473
-W48 (-191
17-191
car US.2560 4.030 -ISe (7) - 4 32 (.242 -2.16 (1 +X4 (-
-f-mapa,Imau 15%6 O-.Jf 20o (a5) 1.77 -9) .72 (sw 243 10
S 5 md . 19S0i
-cmXular 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 lex57 (4) 2.10 (.1)la
M .4 3
:~ ~ ~ ~~~~~J
-L0N12 -1q -- It1._D.108 {1918 D 1.40 Lin --
- AaW Ewd.A 69 5.7 =.6P 5
- U6 P- M& - 2A 223 - 4.69 30

gn
F- pdiicS (44)(

* .vAwM cmibdn wpz.n vacsylwdwwAaeesI .iwnyim

XI nbdal a enIa dwbEnul(nr dc bienps. flqnJul rend Grcceyqlsdwwm
JaUsihIa- yr. -:dbc .pi .m.bl-dn.

Malaysia.On average,about two-thirds of the observedgrowth in the

HPAES is predicted by our model.
Primaryeducation is by far the largestsinglecontributor to the HPAEs'
predicted growth rates. Between 58 percent (Japan) and 87 percent
(Thailand) of predicted growth is due to primary school enrollment.
Physicalinvestment comes second (berween35 and 49 percent), fol-
lowed by secondaryschool enrollment. Japan's high secondaryschool
enrollment in 1960 makes a particularly strong contribution to its
growth (41 percent), more than investment, while the laggardsin sec-
ondary enrolhmentrates, Indonesia,Malaysia,and Thailand, have the
smallestproportion of their predictedgrowth attributable to secondary
enrollments(lessrhin 15 percent).Investment,conversely,is most im-
portant in Hong Kong,Indonesia,Malaysia,and-Singapore,whereit ac-
countsfor more than 40 percent of explainedgrowth.
Second,we are also able to look for regional patterns in growth rates
that are unexplained by physicaland human capitl investment. Con-
trolling for their performancein education and investment, and initial
income,the HIPAEshave a significantlyhigher rate of growth than all
other economies.Latin Americaand Sub-SaharanAfrica, in contrast,
have significantlylower underlyinggrowth rates (about 1 percent),con-
trollingfor the same variables.Thus, the expectedgrowrh rate differen-
tial between the high-performing East Asian economies and
Sub-SaharanAfrica or Latin America,even if they had the same accu-
mulation and initial income, is nearly3 percent.
To assessthe contribution of the HPAEs'superior accumulation rel-
ative to other groups, we can predict growth rates in four groups of

52
 GROWTH. EQUITY. ANDND

4.~~~'~ .I ., - : .fw C.gi7 -~i X .Jimut.-

.=: '- i. ::, . :.

; ,- -;,.,,.- .r., ,^-

c>*d*.s_ - p- G- *rr
... P -. 70. (-4q .0.7 j19J - 7 (.)3 a('
die 4NB
-. *
-"I' '' a.77 -2- i) .L4 442. t 21)
- '-;)- QOb 25 .S (69' 231; .%t Om72
0,0284:
:.f -X - 0 ,- . 1141 - W;t2 (tv 04
Om oil (231
.. . lwt uJ
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

'-R' , .- .(6) IG -
-; S Z

economicsbased on level of investment,primary and secondary

school enrollment, relativeincome in 1960, anrdpopulation growth.
The differencesin actual growth rates betweenthe HPAEs and other
groupscan bedividedinto the predicteddifferences(dueto groupdif-
ferencesin accumulationor initial income)and to an unexplained
residual(seetable 1.10).Accumulationexplainsonly someof the dif-
ferencein per capitaincomegrowthratesbetweenthe HPAEsand ocher
groups.For example,between1960 and 1985 averageprimaryand
secondaryschoolenroUmentratesand investmentrates in the HPAEs
were belowthose in the oECD, leadingus 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 benweenEast Asia and Latin America or Sub-
SaharanAfrica.Betweenthe HPAEsand LatinAmerica,34 percent of the
predicted difFerencein growth rates is due to higher investment levels
and 38 percent to higher carollment rates. Far and away the major dif-
ferencein predictedgrowthrates btween HPABs and Sub-SaharanAfrica
derivesfrom variationsin primary school enrollment rates.Investment
accountsfor only about 20 percent of the diflerenceand is offset by the
oonditionalconvergenceadvantageof Sub-SaharanAfrica and its more
rapid population growth. Educaion is the main theme of the story of

53
 ACLE

Table1.10 Accounting
for DifferencesinGrowthDueto Differences
in Accumultion:
TheHPAEs, Latn America,Africa,andthe OECDEconomies
: . . .:; -: ff : f - u X : -. : : ;~~~~

-- AI -- .- -.
..Pa r -e . : - K- ;UAaSt Afie.- O
; V Ab~~~- twst .lar 41. OEC :AAim-0D

G4CPIZdISIW
I CtLS..9l60 -013- 4tE574 y
0a7m7 :
IrInay.enmIm-.-. 1960 -:D612 . - - X4774
774
- r,tuy a9uIbnnv.1910
- -0- - . O. .£. - - - 0*3 2 - 2427 .- 66 . .
Aq.whFdpla. 156045 M.U095
.0*015 -(. 4003 l
*-- PAEn Dl. 39.--AS Q25
- *73 . .I3: :4o :
iHr:E.
am. I.3 .
- b- mtME *fa - : -:w et - . . 40 Irk

-DiPkmr ad posh .
dLmImtMIbimriutjwdkm3cW.rhto
: ;--T M,mtilhedSehthcrCn.lnfoawsehnwairlEa
hkm6dundanJSwhA-. c :. Irk & m.uqe wdwin nrlm EbnkrX ,:
.wsqewmbldlucnlueln DNbuuhudaSdnb.wbmebn& - :

the diferences in growth between Sub-Saharan Africa and the East

Asian high performers.
WhIiatis most striking, however, is how lirde we are able to account
for differences in growth rates between the HPAEsand other economies
on the basis of conventional economic variables.14 We are able, in the
end, to prcdka only about 17 percent of the actual difference in growth
rates between the HPAEs and Latin America. We do somewhat better be-
tweenSub-Saharan Africa and the HPAEs, predicting about 36 percent of
the difference. Controlling for their superior rates of accumulation, the
HPAEs stil outperform while Sub-Saran Africa or Latin America -
derperform the statistical relationship between accumulation and
growth, leaving much of tie regional difference in per capita income
growth unexplained (even though a large fiacton of HPAE success is ex-
piained).1 5 They have been apparendy more successful in allocating the
resources that they have accumulated to high-productivity activities and
in adopting and mastering catch up rechnologies.

Evidencefrom Estimatesof TFP Gnxvb

change captures both of these important aspects of productivity

iTP
growti. Te is estmated in a neoclassical fiamework by subtracting
from output growNtthe portion of growth due to capital accumulation,
to human capital accumulation, and labor force growth. In keeping with
this framework we estimate TFP change for eighty-seven economies

54
 GROWTH, EQUITY. AND ECONONIIMC 9C1H
.

iup,2
*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ pwe pfnwi j .,.

.ad = wu,f t- _ li2 .. ,s''0' Rel-" i - -"' '';'

UFAA.

A.Mm ~~~~A4i.ORD fm.

0.17 (;0) 130 H -1- rs ;- - - (

136) 0.2 , ()
,,;:.-. 4b , - ., ,h M-. P11
(4). -013
CM! (11i2 .- L9 .< . ;
15i200 - (I 4

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

using a produc-ion function esdimatedwith cross-economydata. (The

techniques are discussedin appendix Li..) Figure 1.10 plots the result-
ingrTr growth rstimates for 196049 against relativeincome levelsin
1960. There is greatva-riancein the rates of n.' growth among low-in-
come economies.A number of develo'pingeconomicsshow higher rates
of TEP growth tha industrialeconomies,consistentwith the possibility-
of largecatching-upgains.Many others,however,have lowor even neg-
ativi rates of productivitychange. The East Asian economiesstand out
sharply, withahigh absolure levels of TF. Five HPAEs-Hong Kong,
Japan. Korea,ThAand, and Taiwan,China-are in the top decile.The
other three HimaEs-Indonesia,Malaysia,and Singapore-aze closer to
TF' growth rates in high-incomeeconomies (about 1.5 percet), but are
still in the top third of alldeveloping economrues.
Presumably,in the high-incomeeconomiesmost ofthe estimated'r
growth is due to advancesin bestpractice,which explainsdthir relatively
compact distribudtonofTFP hrates around 1.5 percent a year,and
the tendency for IT? growth to dedine with rising income among the
high-income coonomies(see figure 1.10). In low- and middle-income
econozPrwst, howevr, changes in *rTFPmust rcflect more than technical
progress,othierwisewe would never find negativeTF growh ratev.We
have already argued that TFP growtd for low- and middle-income
-economlies conains an element of catching up to (or fing behinde)
bT-pracice technologies.(The relationship between technical change
and catching up is discussedin appendix 1.o1.)Bu TFPgrowth rates in an

55
Figure1.10 TotalFadeor Growth,1960489, andGDPperCapitaRelative
Pmodndivty to 11.S.
GDP,1960
1FPgrowth(percentparyear)

Taiwan.China
* ~~*Hong
Kong

*Japan
*Rep. of Korea
Thai!rd
3 -

2 --

Indonsia*Malaysia U

* *5

-2~~~

-2

0 0.2 0.4 0.6 0.8 1.0 1.2

Ratioof GDPto U.S.GDP,1960

Source.WorldBankdama.

one-sector, cross-economy estimate will also contain an element of al-

locative efficiency; economiecschar allocate physical and human capitcal
to low-yielding investments will have low or negative estimat-ed TFP
growth .rates. The dclarest -demonstrations of this ane the thirtee-n econ-
omies with negative output growth. and positive accutmulation. Thus
our estimates of TEP growth answer the following question: on the basis
of the average efficiency with which physical and human capital are used
in the world economy, does accumulation over- or underpredict income
growth? The answer is that for most low- andt mididle-income
economiesit overpredicts,while for the HPAEs it underpredicts.'
Under very restrictive assumnptions(see appendix 1.1), we are able to
offer some speculations concerning which of the H-PAEswe.recatching up

56
 GROWTH. EQUITY. AND ECONOMICŽC

to best practice between 1960 and 1989. To do this we assume that the
elasticities of output, which should be used to calculateTIP change,are
those that apply to high-income economies only. These more alloca-
tivelyefficienteconomieshave elasticitiesofoutput with respectto both
physical and human capital that are higher than those for the whole
cross-economysample. We subtmct the averagerate of TFP change for
the high-incomneeconomies, which we have associated with movements
in international best practice, from TFP change estimated using those
higher elasticities to get an estimate of technical efficiency change. Using
this metlhod, Hong Kong (2.0 percent); Japan (1.0 percent); Taiwan,
China (0.8 percent); and Thailand (0.1 percent) are the only HPAEs
catching up to international best practice. Korea (-0.2 percent) was es-
sentially just keeping pace with technological progress in the high-
income economies, while the investment-driven economies of
Indonesia, Malaysia, and Thailand were falling behind international
best practice ac rates of 1.2 percent (Indonesia) to 3.5 percent (Singa-
pore) a year (see table A1.3 in appendix 1.1).
When the HPAEs are contasted with other developing regions, how-
ever, their ability to keep pace with intemational best practice seznms
somewhat more remarkable. Using the same method, we have estmated
rhe average rate of technical efficiency change for Latin America (-1.4
percent) and Sub-Saharan Africa (-3.5 percent). Against these bench-
marks, all the HPAESexcept Singapore stand up well in their ability to
keep pace with the world's shifting technological frontier. 7
Although the absolute magnitude of TFP in HPAEs is higher, so is
growth. Is the proportion of growth due to TFP high? On the basis of an
analysis ofsources of growth forasample of economies, Chenery (1986)
devc-cped a typology of the contribution of TFPgrowih to total output
growth by income. A typical low- and middle-income economy had a
relatively small contribution of TFPgrowth to total output growth, be-
tween 10 and 20 percent. High-income economies, conversely, derived
about 30-50 percent of dteir total output growdt from TFPgrowdt.
Figure 1.11 largely confirms this pattern for low-, midcle-, and
high-income economies. Only seven of the fifty-nine non-HPAElow-
and middle-income economies have contributions of TFP growth ex-
ceeding 33 percent, while for the high-income economies it was much
higher. The HPAEs fill into two distinct groups: investment-driven and
productivity-driven economies. The investment-driven economies-
Indonesia, Malaysia, and Singapore-conform to the developing-econ-

57
 &EAAS4
ksA N M I RAC L E

Figure1.11 TotalFactorProducti Growthand Part of Growth

Dueto
Gmwthof FactorInputs,196049
TFP Growth1960-89 percentpar yer)

v ,\ s s~~~~~~ 10%
4. ' 8%

4. 4. 4. 4.

4 '. 2966 . ". ': . 3[

.4 .4 '4 4.npan *Ta"n.Ch1na

2 . 4. . . 64 4.an

- 0% .* Hong Ki

D4 .4 - 4. .4 . -

'2%' .4 . .4 1E 4
-4~~~~~~4
r _
0 4
~ -
N .4
.4. 4 ''
.
*.
4.__ '

j Highincomeecon om. . *

HPAEs %
| Othiordevelopingecornomles ^ * \

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

Note:Dashcd
line:s tcprcscnc: toul 2vcngF GDP growch raus, 19650 89. Soli rays reprewnsi
EhecontributionEotoclJgrowthby TFP growth.
S#tirees.Nchruand Dh2rcshw:ar11993):World B3ankd2m.

omy pattemnwithalowTFPcontriburion. Conversely,theproductvity-

driven economies-Japan, Korea, Hong Kong, Tliailand, and Taiwan,
China-are more unusual and look-more like industrial economies,
widi a highicont-ributionof TFP (above33 percent). China'srecent
growth (seebox 1.3) shows some of these characterisEics.
The EIPAEsare unusualarrong developingcconlomies becauseof the
relativelyimportantrolc of TFP growth.Evenso, we find, consistent
=: our previous findings with cross-economyregressions,that be-
~~~~~~~~~with
tween60 and 90 percentof rtheiroutput growthctcrivres
from accumu-
lation of physicaland humanacapital. Producdvirychange has been
higher than in other clevelopingeconomicsand is importat to the East
Asiansuccessstor. But it is nor the dom-inantfactor.

''- s8
 GROWTH, EQUITY, AND O

,~~~~~~7
-: - .

-..- -1 EN.FECNMI(.;OWFI.'EAST..o;
lnim.est:-an-astoundiiien3-... G..P.. da
;:.tc~~~~~CtW Ati:wud-btcmle
dout.'i cosdeaontf,: 1 men smc 17.ase ii. rkia improvement: in
,111a'S;! i I;anoa45!"26
r-. i-.o;. -, - - o- a~v>s i--
...- 7 ] ' -
,jp, =;^ Y: ;--tic t
;!. '@-' - - C6ns1ump'non.ldf-I
S . -. on..of.-
---- m4.
B-soxn TnheChnese M ac em e.- :-
NO--.
ASSESSMENT tECns-Ofi
OFapn=enGNOMiC OdmsIlas T nvest aatu n 30erearonspnit op.
i y
su'' A .. -aea.
-?-ea;.e; y .n-ca...r o"' .& ow- .. ub
1, Asia.ewofd: :. v_of e19i8_8 oft metZ' sinledo
17hasoa see a 1978:t.v
lnlfunaln
akdimrvmnt in
-b&&mpledy,tcu.wir-thourconsideadolsm.he
71- 1 4 4-49 f4..r.; .
_..;ie-.t ze. ^ ,5m ;9 ,P;A,E_
g~~~~~~~~~~~~~~~1gtCnM-;.. t- -, - ,/~- t.o.

mwlgg -wcm<tj~
ra
fsh -te.- -; ,:gr.d.m7
- ;...s wlde .a>mos',-s -o tuc,:2E, -rpDi
.tame~~nt pronne: fyn gntao
!-
-~~~~~
Ghina's
.9.s
_. , =-r {to 't;---
'v
'-..
WrtL - a ->- gpal
x- i ..
T
>; - l _ JWiflg'staiildsgo
f r. thia.za,utOr19
comaierdmds.
t

e.orublihgnedpopuation oFalleigh-sbonomf
mC the
- l sour;& W:nimal2'peror inp iesebqdir als-
-,: ; -j
%AAuciMAdannbaawaL.I
4
.4<. . .;¢;>t;!-<,.+<-=.<-2.
.&zt
. !,
>,'4-
CWfiyP
,
w;.i-heiAEwdbledYfroi
:^W3- . ,- -s
198 5
'"-f ;!n ;i'9 >-.- 4'
't? Ona
Jdp termeo d an'
be-at eo p
;'ba
and evemen government-fbszeehxci& slXRoiinsgwing
..recorde.av-nna
. Bt'.*.urdiW4oremteonents gewdoflocDeta;ose
d. GNP- ud gnll;6roucy 'g tliey
non ti ar>a.a-.
lled,c-;l
ton' and ie Cinacrnoeuhhiiiicsr.'bili'ty,
'health,
w.:erenr'n~lso0,o -~
:onowicjmeŽifChina .,China --- ~~~gecrber
wdt 2 jrcenr
televi ia ksit
nuiFxceibk4f%lorX ihcreincrTpjancs
?r 9.leibusexiosnS.-'
i½Scr&Did1t0 -r.carmEa-relm- .t
' .;: eoetsn: 1d16&arnvtlper. alddyinnually
-'n ill aulEzdiD
S
eds- wmgatn,oterenmdyrai&y.avo&hlltaed
t- fro<i;vasmessof,CltinE.lIiaks
B-, ass, tonenili7
1979
-t zessment ph-:h^S
'&rf or.ii iah&isi1u.4;>pcrct}en.~R-<-
t,;saviing ftrnEm
Asih'soCre Uk8 Jx:o5
l ta baarmcsinvsted
T in.ms.
H'Fi in 1990.-uds,-,
inXrw nsip-nd
growdi-n;
Af&0MF1982wjt 198~Fo bier-retreii.chmin
aeti~se.Czi
il
tharS.-
comc.cauamlgyihsEhEr
Droblcmanc. 8iehi
rcM'4&44eii& - mg
3)1125 h~iidM suFpn4dealbi.tiowih
c anL
ndshi
v-~~~~~~~~~~~~~~~frerLNqs
gnrou,
, ,-wzA;§
cottE
tro
{tsIea'te
Ex St*n
ovr the Ian.1,!
p n n iY-.U:-.-e.,;nomy;Tas ,e5Lrv Tt-
po*k-^nm~uvcs as
-- ixbiJqjig
rapid 4-breald&&ckfTt2r|n onpmentn,,T-i.rW
s -: p"-'u '. tt

-~ X-.-&WZ- i oftheecopnomies in-inour

-ikA'r.-.--oer stud_

-Ci s eee e&oiiozfies

r s, A i aeynwichiuecnvinwg 4- t_.h.._
-si
our-p dof
high terms
un rates reductionsind ininequali
captala a.nd~ W I ar-
wererrm
--
cmpa- od wheno %ro r ip
noive
- ... , - DeoDkallewedEom
Dvmosmce-EF &n2 :S, jDiledyLbr.i of
Ch-inniaksj
ofsms flSnr -C$Qeiiid oAtes
4.- ~~~~~~~~~~~~~~p~~~~~~6~~~~--Ie ; t,s. r..
.:y~~~~
~~ ,sfxpd nu"C-S

--. . W -- '

";- iebhn
comesstIi tbknaii&
chea ghptr,wth
ha et44 efiuiiveN~rpsgy grwhanx-du&v
mb

We havedescribedtheEast Asianeconomxies'
aIst..aa'~~~~~~~~~~~~~~~~~~~~~5 success
r in achieving .w

hlighrates of growth and redluctions in inequality- These were accompa-

mied by-.agriculturaltransbrmnaions, rapid dedinci, andmau
fertility
factured. export: growth; and accounted for by rapid accumulation of
physical'and human capital and by productivity growth. What policies
lie behind this success? In the next chapter, we examine alternative cx-.

59
 CLE

planations of the relationship between public policies and economic

success and particularly the ways that public policy can lead to improved
economic outcomes. In subsequent chapters, we explore the relevant
policies and their apparent links to the inputs and outcomes we have de-
scribed in this chapter

Appendxc
1.1:Accounting
forGrowth

IN THISAPPENDIXWE BRIEFLY DESCIBETHE METHODSUSEDTO

estimate the relationship between accumulation and growth using
both cross-economy and producton function methods. We also
compare our results with other resulrs available.

Cross-Econonm
Regressions

The cross-economy regressions employ the basic specification of

other cross-economy growth smudies(Barro 1991; Dc Long and Sum-
mers 1991; Dollar 1992):

(1.1) GDPG= f(L ED, LFG,RGDP6O)

where GDPG is the average rate of real per capita income growth using
Heston-Summers' measures from 1960 to 1985; INV is theaverage
share of investmenr in GDP for the period 1960-85; ED is a measure of
educational ataiunment LEG is the rate of growth of the economically
active population, and RGDP60is the relative gap between per capita in-
come in 1960 (at 1980 U.S. dollar prices) and U.S. per capita income in
1960.'; Mankiw, Romer, and Weil (1992) have recently demonstrated
that this specification corresponds to the transitional dynamics of an en-
hanced neodassical growth model with human capital.
Table Al. below reports our basic results on the relationship be-
tween accumulation and growth. The estmated equations compare fia-
vorably with other studies using similar specifications. The overall fir of
the regressions is good, and the coefficients of the variables are ofthe ex-
pected sign and are significant at conventional levels (0.05 level).' 9 We
have also estimated the basic relationship fir two subperiods, 1960-70
and 1970-85. The fit of the regression is markedly better in the earlier

6o
 GROWTH. EQUITY, AND,

TableA.1 Deterinants
ofGrowth
(dependent variable.-per capita CDPgrowtl)

Smp-pe rio 196 -965 .19160-70 1.970-85 1960-5 -1960-85 .196-55

Nwnberofobsennrtianr 1413 .11.3 11.3 61 "98 5
Intercept-, -0.0070. 0.0064. -. 16 -. 04 004 0.0243*
*(0.0079): (0.0092) :(0.0109), (0.0113) 0.8): (0.0094)
GDPrelativeto 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
AverageInvestmendGD,.1960-85' .00578* 0.1153"' 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)
2
Adjusted R * ~~~0.3480 0.3424 0.1921 0.36 0,1893 0.2614
Safl1ilM
sgtfintt the0.01II~
Staisdeallyiignicattth 0.5 lve
* NeCoffiien istopnumbermn eachceIL.Smiinkdanirrris
in parahee
Bankstaffestiniares.
SOUIrcZWorld

two periods, but in most respects t-hc'parameter estimates have similar

values and significance. The major exception is the coefficient estimate
on investment, which becomes insignificant in the period 1970-85.
Table AM. also reporcs on two e%frcs to examine the impact of
changing the basic specification. DeLong and Sununers (1991, 1993)
have argued that equipment investment,rather than'total investment, is
a superior explanatory variable for per capita income growth-When the
investment share in GDPis replaced with the share of equipment invest-
ment, the explanatory power of the regression is improved, and the co-.
efficient on the share of equipmrent investment is highly significant. Our
sample size is much reduced, however, and we condude that the reduc-
tion iniresidual variance is not sufficient to justifyra change in our basic
specification'. Because school enrollment may not be a good indicator of

Gi
 ,* 1IRA C LE

human capital accumulation, Barro and Lees (1993) measure of educa-

tion stock,basedon populationcensuses,can be usedin an alternative
specification.
Deviationsfiom the regressionare estimaresof labor productivity
clangethat cannot beattributedto accumulation-investmentin phys-
icalor hunan capiralor to the component of TFPchangeassociatedwith
relative income levels.The patterns of productivity change resulting from
the basic specification of the cross-economygrowth regressions can be
seen in figure Al.1. The figure plots the partial scatter of the relationship
between growth of output per capita and investment, controlling for
human capital and the component ofTFPchange related to relative back-
wardness. Per capita income is an increasing function of the share of in-
vestment. Observations are plotted relative to tle estimated regression
line, and the HPAEs are identified The estimated productivity perfor-
mance of the HPAEs is remarkably siillar to the pattern derived fiom the

Fgm Al.l GDPGrowthRate and AverageInvestment

Ouithagonal
componentof GDP,1960-85
Y.06 -----

Hong Kong* -
*Taiwan; China
Rep.of Korea*: * *Japan
0.02 - - indon ia - ngpo .
F i @ e ; e , s*! __

I* , . .halwidi
O002- --*
- - --

r ~~ ''*' 0

-0.04 _ . .. -

HPAEs
H*
* * OUuerd.velopIugeconomies
.0.06.
-15 -10 -5 0 5 i0 i5 20
Orthaganalcomponentof I/GDP,1960-85

Note:TheXand Ycomponents rc orriagouai on RGDPBO,PRIM60, and GPOPGO85,wherc RGDP = eal GDPpcr c2pim relativeco US.
GDcper 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 economiesare positiveoutliers.A dummy

variabletaking on the valueof I for HPAEs is positiveand significant.

HowWeEstimate
TFPChange
Suppose that every economy has access to an international cross-
economyproduction function of the form:

(1.2) Q =AF(K,E,L)

whereA is totalfictor productvity,Kis a measureof capital services,Eis a

measureof human capita endowments,and L isa measureof laborservices
in naturalunits Then growthof output per head can be representedas:

(1.3) q--a + s k-l) +s (e-l)

where lowercaselettersindicaterates of changeand sAand sEare the dlas-

ticities of output with respect to physical and human capital.The sham-
weightedgrowth rates of plhysicaland human capiml per head give the
contributionofaccumulationto growthin output per worker.TEPchange
can be then found as the residualof gowth of output per workerafterde-
ducting the contributionsof human and physicalcapitl accumulaton:

(1.4) a = (q-1)- s, (k-l)- SE (e-0.

Under assumptions of competitivefactor markets and constant returns

to scale,s,K
and SEare equal to the income shares of factors.Thus most
empiricalapplications of equation 1.4 estimate the output elasticiy co-
efficientswith income shares.2 0
Sinceincome share data are not availablefor most economiesin our
sample, we instead estimate s5 and sE direcdy using a simple, cross-
economyproduction fnmction.The data for this analysisindude a new
constant price capital stock data set (Nehru and Dhareshwar 1993).
Measures of human capital are incorporated in the specification using
Barro and Lee's(1993) measure of educationalatainment. We regress
annual log output growth on log capital growth, log human capital
growth, and log labor growth during the 1960 to 1990 period,
constraining their coefficients to sum to unity (that is, specilig the
production function to be Cobb-Douglas).We also indude cconomy-

63
 - RLIRACLE

specific dummy variables to estimate individual rates of TFP change for

each of the sample's 87 economies.
Table A1.2 reports the production function parameters and the esti-
mated TFPgrowth rates. One striking finding of our analysis is rhe low
elasticity of output with respect to capital from the wholc cross-
economy sample. This is not altogether surprising. As we pointed our in
chapter 1, there is a subset of developing economies that has positive net
investment and human capital growth rates but negative output growth
rates. In effect, the marginal product of both physical and human capi-
tal was negative in these economies, reducing the elasticity of output of
both physical and human capital in the production function based on
the whole sample. We have reestimated the production function on the
basis of only high-income-economy input-output relationships.These re-
sults and the estimated TFPgrowth rares are displayed in table A1.2. The
elasticity of output with respect to capital in the high-income-economy
production function rises to more conventional levels. TFP estimates,
particularly for those economies with rapidly growing capital stocks, are

TableA1.2 Elasticity to CapitalISKI,Labor(SL),

of OutputwithRespect
andHumanCapital(SH):FullSampleandHigh-Income Economies
Obsations SK rtaO SL (t-)m . SH (f-stat
Fullsample 2,093 0.178 10.895 0.669 6.411 0.154 1.49
High-incomeeconomics 460: 0.399 10.237 0.332 1.679 0269 1.476

Ruling toasaifacuerprodsadvi4ygtou*h
&ximafi r t HPAEs
TFPrwsA TFPwtb1
; - - - - 0 ; ; -0t;t(filsa5Jmpk,
. - (ldg-im v:j
-Eonomy/region. parpameer ates) .te
Hong Kong 3.6470 2.4113
Indoncsia 1.2543 -0.7953
Japan 3.4776 1.4274
Kore,.Rep.of 3.1021 0.2355
M2;ysia - 1.0755 -1.3369
sing2pore1 1.1911 -3.0112
Taiwa,China - 3.7604 1.2829 -
-ThaiLnd 2.4960- 0.5466
~+ - ~0
LatinAmerica ~ ; 0 0.1274; -0.9819
--SSub-Sahv fric' -0.9978 - --3.0140
a. Alsoincludes
TunisiaandSouthAfNic:-
Sor- World Binkda:L---

64
 GROWTH. EQUITY. AND ECONOlT

correspondinglyreduced.The pattern of productivitychange,however,

with the five productivity-driven H'AEs fiallingas outliers in the distrib-
ution of developingeconomicsis maintained. (SeefigureA1.2.)
It is possible to compare the consistencyof the TFP estimates pre-
sented abovewith other independent estimatesof TFP growth in a large
sample of economiesto test their robustness.Using a data basederived
from World Bank real product data and alternative estimates of con-
stant price capital stock, Elias (1990) calculates economywide TFP
growth rates for a subser of 73 economies using standard growth ac-
counting assumptions. The relarionship benveen his results and those
presented above are summarized in figure A1.3, which plots the scatter
of Elias's estimate of average TIP growth during 1950-87 against our
estimates. Fischer (1993) has also presented a set ofTFP growth rates for
a large sample of economies. A similar plot of his results against ours is
shown in figure A1.4.

FigureA112Per CapitaOutputGrowth
andCapitalStockGrowth,
196089
Percapita output growth (percent)

Taiwan,China
*Rep. of Korea__---i;
*S.ong Kong *Singapore

i @ Thailand *Japan _ -

4 _ - , -------- *MalaWsia-----'

* Indonesia *

is
0- MS
2 - '-__ _
26--~ ~ - 4i t' ; I - - - - Capita-growth
-- pL_ _ ---- U N4igh-Income
economies
- _ _ _

Se S *NPA~~~~~~~~~~~~~
tes eelpn economies
9--- ~ -- Fl
0D sa.k et .06@(2* rwh

0 510 1520
capita growth(percent)
-- High-IncomeeconomiesGDPgrOwth= 0.010 + (0.395*K growth)
- - - Fullsanuple;
GOPgrowth=0.0076.+ (0.23*K growth)

Soure.rWodd Bankdata

65
 MA.-IvIR
AC LE

Figure
41.3 Compadson
ofTotalFactor
ProductiNt
Cmowth
Estmates
TFPgrowthcamparator
(percent)(Ella)
4.

3 0*

2- ~~~~~~~~~~~~~~Singapore
-Taiwan. Ohina*

0--

a Hwighicomneeconomies.
-2 ----- A
* Otherdeveloping
econiomies

-3 -2 4. 0 ± 2 3 4
TFP growth,1960-99 (percent)

Snuru:vElias(1990a); World Bankdata

Both samples show sim-ilarlygreat variance in rates of TrP growth

among low-income economies, while the high-income economies are
rather closely distributed around a mean ThFPgrowth rare of about 1.5
percent a year. There is also great consistency between the relative rank-
ing of TEL?growth rates for the HPAES.The same productivity~-driven
economies-Japan, Korea, Thailand, and Taiwan, China (Hong Kong
is absent from Fischees TF' estimrnaes)-are in the upper quintile of both
distributions, with the except:ion of Thaiand, which under the Fischer
estimate is still in the upper half of the distribution but is not in the
upper half under the Elias esrimara-

TechnicalPrmgress
andTechnical
Efficiency
Change
As we pointed out in the conduding section of chapter 1, it is difficult
to disentangle the relative contributions of allocative mistakes, technical
progres, arnd technological catch-up to rm'growth, especiallyin a cross-
economy, one-sector setidng.One way of separating the concepts of tech-

66
 GROWT1, EQUITY. AND E£CCN

FigureAlA Comparison
of TotalFactorProductivity
GrowthEstimaites
lVP growthcomparator
(percent)(Fischer)

3 -- __ ---- -

-3--
-2 ---O
*
,:in
I~~~~~~~~~~~~~~~~~~~~~~~~~~~awn Ch_

O- ~~ ~ ~ ~ ~~ ~ ~ ~ ~ ~~~~~~~~~~~ econoiesn
Hlg-,h4ncldoesau
t.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Rp of Korea

/~~~~~~~~~~~~~~~~~~~~~ Whetdevlfopingeconomies

4 -3 -2 -1. 0 1 2 3 4
growth.1960-89 (percent)
TiFP
Sources:
Fischer(199,3);World Bankdata.

nlologicalprgress and changing efficiency in the use of technology istO

model formally the relationship between observed output and best-
prIacticetechnology;We address the relationshipbetween accumnulation,
productivity change, and growth using a simple neoclassical model.
Following Nishimizu and Page (1982), we define an internationlally
accessiblebest-pralcticeproduction function of the form

(1.5) Qf(W-=F7(t#]t

where Qf(t) is porential OUtpUt at bestpractice, and Z(t) is a vector of inl-

puts in natural units at time t. We assume that the frmncrionx
F(.) satisfies
the usulalneodlassicalproperties and that an appropriate aggregate index
of OUtpUt exists.
The best-practice function defines the "stare of the art" in the sense
that fulrther increasesin output at given levels of inputs can.not be
achieved without the introduction of new techniques. Firms can move
along pract fnctin, incrain theuse esult of accu-

-67
 MTRA CL E

mulation of inputs. The introduction and dissemination of new tech-

niques move the best-practice frontier and is tchnological progress as
defined by Solow (1956). Observed performance in a sample of
economies or firms reveals that few are at best practice.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 time
yields:

(1.7) Q (tOIQ( = (t)IZ(t) + Ft +u- ()(t)

where F and Eare the output elasticities of F(Z(t);t) with respect to in-
puts Z4-) and timnce and dotted variables indicate time derivatives.
Output changes in cquation 1.7 are decomposed into three main
elements. The first one gives output changes due to input changes,
weighted by the elasticity of output with respect to each input. This
is the component of growdt due to accumulation. The second ele-
ment is the rate of technological progress of the best-practice frontier,
and the last element, u-et), is technical efficiency change during
period t. While i; is always non-negative, u *t) can be either positive
or negative.
We define the rate of total flacor productivity change as the variation
Figure
A1.5 TheGrowth in output nor explained by input changes. Thus for any observation, i:
of InpuisandOutputs
Output-labor (1.8) TFPJ(t)= F+ u -(t).

A- is the sum of technological progress, measured at the frontier, and the

change in efficiency observed at the individual level.
fh These concepts are represented in figure Al.' 22 We assume con-
stant returns to scale in capital and labor. The international best-
:: fflz
qo f- - _ 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

figureA1.5.23Catch-up can be achievedby movingfrom a point such

as A to D, combining accumulation with a movement toward best
practice.
T!-,s reinterpretation of TFP diange is useful in understanding the
sources of rapid catch-up in technologicallybackward economies.In-
dustrial economies,which employ international best practice, are lim-
ited to rates of TFP change determined by the rate of technological
progress, zvi(t)= 0. Economies that do not employ best practioe can have
TFP growth rates exceeding the rate of technologicalprogress if technical
-efficiencychange is positive, icit) > 0. It is also possible for TFP change
to be negative, if technical efficiency change is negative and greater in
absolute value than technological progress. A rapid shift from average
practice to best practice-positive technical efficiency change-can pro-
vide a powerful engine of growth that is recorded as high rates of (TFP)
change.
There are two possible approaches to the estmation of the model
outlinecdabove. One wvouldbe to estimate the best-practice production
function aiLdderive estimates of both the rare of technological change
and the rate oftechnical efficiencychangc3 4 This approach is notsuited
to the cross-economy data available, and we do not employ it. An alter-
native is to measure TFP change direcdy by growth accounting. Since
measured TFPchange consists of both technological progress and tech-
nical efficiency change, we impose the strong assumption that techno- TableA1.3 TechnicalEfficienq
logical change, the movement of best praaice, is constant and does not ChangeEstimatesforthe HPAEs
2 5
vary across economies. Under this assumption all of the variance in I enbikid
rates of TFPchange derives from variance in the rate of technical effi- effdciency
ciency change. w --
lkEcoJomylrqov/g t-.1960L9
60-89
'
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 TrrPgrowth rates of Japan -. 9876.
those economieswith rnid factor accumulation may represent alloca- Ko Rep. of -0.2044
CM a~Mlaysia 1L7767-
Eive efficiencvmore 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 LatinAmeca- -- i :17
estimated TPP growth rate the average TFP growth for the high-income SubSdiaAfi 34539, -j
economies, our estmate of technical change.: The residual estimate of 1 a.AI inudesi,Ttnisiada
tedical efficiency changeis presented for the iihPP and other regional .. . -

grouping in table Al .3. - -

-:~ 0 ::~ ~ ~ ~ ~~~~~~~~~~~~~6

 MTRA CL E

Appendix1.2:'WhatDoTests
SkillsShow?
of Cognitive
Tests provide a convenient quantitative measure of diflFerencesin 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
areasnor tested may fail. Differences in the curriculum may also account
for some of the observed differences in performance (see Hanushek and
Sabor 1991).
Tlhesecriticisms suggest thart whenever possible, evidence firoma 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

Perfomingat or aboveEachLevelof the MathemacsScale,Age13 Years

of Children
TableAI.4 Percentage

- ::and simple ; two-step'

ab/s : Udrtns
-o.c-ptP - .-nria'
Ar -*U
,,,,m/nvnc _ m&mac!;,t,
|^s - ,'problems ! p0kn

-Biish b . 100 ;a - 6- 69 - :- 24- - 2

,Irelirdant, - 98-. 8t . 55
- e14 -. .
I ea;-.Rp- 100- --. . 95- 78 - . :4 5
Newf wkC lh 100 .5- ; -65 18 1
N-w Bf`Ilik. (Frenc) ;100 95 58 12 -.
J Onrado
JE~igIish~ .

.99
92 5 .16- 1

tquebe(Fwnch) - . ~100. 7 3222

:::-87 55
.tg :-78

70.
 GROWTH. EQUITY, AND'E-

mathematicsadministeredby the EducationalTestingServicein a num-

ber of different economies, including Korea.2 6 Students from Korea
ranked first in cach of the five categoriesof problems.The performance
gap between Korean students and the others, all of whom were from
high-income economies, was greater for higher-order skills.
The International Association for the Evaluarion of Educational
Achievement also administered mathematics tests to thirteen-year-olds
in 1980-82 and included children from Hong Kong, Japan, and Thai-
land among the children from twenty economics in the sample. Figure
A1.6 summarizes the results for algebra,which are very similar to those
for arithmetic, geometry, and measuremenL Japanese students rankled
first, students fiom Hong Kong were in the top half of the distribution,
and students fiom Thailand -werenear the bottom, at about the same
level of performance as students fiom Nigeria.
FigureA1.6 TestScores e:: AlgebraTests:Selected Economies

Swazland
Luxenboug _
Nigei _ ____
.-_______ ______ _ -. L
Sweden _

Thailand _
NowZealand --
England and Wales
Canada(Ontado) _ _;
UnitedStates
Scotland
Hong Kong I . j

Finland
Israel _.
Canada(Br Columbia) .
Belgium(French) i
Hungary ____

Netherlands - i
_ __ ____ * _*- _ _ __ _

Belghirn _ _ __

France
Japan _ .

0 20 40 60 80 100
Mom score
SouweHauvsheh and Sabor (1991).

7'
 - M 1RACLE

These results reinforce the conclusion that in some East Asian

economies cognitiveachievementlevelsexceedthose in other develop-
ing economiesand match, or exceed,those in high-incomeeconomies.
The results also suggest that just as Thailand has laggedthe rest of East
Asia witl respect to rates of enrollment in basic education,Thai chil-
dren also lagwith respectto levelsof cognitiveachievement.

FigureAL7 #-Lii andCW perCapitaGrowth

Coefficient Rate,1965-70
G1nlcoefficint
0.7 ;-

* F ' * Gabon

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

'Brazil
*Colonmbia
j Venezuei au Chile Malaysia Singapore
0.5 ,- -~- U~~~~~~~
-~~~s~Phiil;Dlr
-
U Philoianes
~ EZamba EHong Kang

Sudanm ThailandI
N~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
t Sudanu E Argentina *
0.4A--4--- *lnldia - *SrlLankau Olcoe-
I * Pakistan '

X3---X--i TaIan, CaJapanU

Chinau*Rek'rwa

0.2-1 - ---- ----- '----- -…

* O . , I I : ' ! ' :~~~~~~~~~*

-0.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
GDP per capita growth rate

SaurceWorld Bank darm.

72
 GROWTH, EQUITY. AND _|O

Fire Al.8 GiniCoefficient

andGOPperCapitaGmwthRalte,
197180
rlnl coefficlent
0.6 ----- - -enya . NCOI
IPeru iBrazil
EZambla Nepal CWIie MeVdcof enezuela

i I *Malaysia
I i j r EPhillippines * Singapore
I E~~~~~~~~~~~Horg
Kong
OA
i-
t-- . . .: , *ArgenU;tinda
,_,,_hdiaE Indong
0.4j - ------ Idnea
* Rep.of Korea
* Banfjadesh 1111halland
*~ L.~ Taiwan,China.

02 --1 -- -- |1apa-- i -

0.2-h

. I

0
-- , ,
.0.04 -0.03 -0.02 -0. 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
GDPper capita growthrate

Source World BankdataL

73
 MPI RACLE

Figre A1D9GiniCoefficientand GDPper CapitaGowthRate, 1981-90

Gilnlcoeficent

XMexiao EChile

Venezuela

0.4
!. -
Argentina*
Peru- - ;> _
alyi
_-ngapore
O
l-~----------EPeru-- * Philippines - - Ehilane - M Hong Kong
EThailand
Rep.of KoreaU
0.3 - ,donesiaE
--In--
r - _ --- **Talwan.-China
-

I~~ . ~ ~~~
.- ~ ~~~~~~~~~~
.

.I ~ ~~~~~~ '- -

U.1~~~~~~~

-0.02 -. 01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

CDP per capitagrowth rate
Soun:Wo rid Bank data

74
 GROWTH. EQUITY. AND EC

for GenderGapin PrknaryEnrollmentRates,1965and 1987

Regression
Fgure Al.1G Cross-Eonomy

* Pakistan
120.61 I I
I 1965
1987
987

di
*Pahistan
E Indonesia >-
6 ~~~~~~~~~~-~~~~~~I I
~~~~~[19.21
* [-5, I .-

E maliand 1nd
%ralaysna I i
i [-10.t6] [5.1] N '-
jRlof Korea A-ndnei
t as21 Singaporeu; \ [8 a

I - -~~~~ 14-31:3

. ^ Hong kng * \ ,I X 7 [-2.2]

0 2,50 2,0070 2,750 0 o °500 1,000 1,S0 3,500

P^ercapita Incomer(1958 U5S doflars)
Note Figue in bracketsshow resiEdua
Sowre: Behrnmanand Schneder (1992).

-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~[22
Figure M41.1Cross-EconomyRegression EnroilmentRates,1965and 1987
for GenderGapin Seconldaiy

. ,, } 1 ~~~~~~ ~ ~ ~ ~ ~- ~~~~1965
I, 1987
6, [7-7]- * -Rep.of KoreaK, 4 .a, i
a *__Indonesia
E37o.'j]§. . .
Figre 0 At1 250 Soo
CosEnoIy 750 RersinfD :1,750 2,000
edrGpin 2,M5 0
eodr SOD
noletRts 1,000 1,500
95ad 30500
di .L[.1, - *-61 J-

- - *P'akistani ,, ,
5ware.
S Bdhmnanand Schncidkr(1992).
IndonesaKo3g.
913 ~~ ~ ~ ~ ~ ~ 29 P.r cap6a 1h US. do laysa)
Mcame
PoakKontan
Not- F4urcs in brucker show
E
residual
~~~~'Thailand -J~~~~~~~~~~~~~ 18.2]I [-.2
.~~~~~~~~~~~~~~~~~~~16 .98
* 7077
a Sn apore__ _ _ _ _ _ _ _
E
* Av~~~~~~~~~~~~26
N Maayi
0 250
750 500 1,750 2,000 2,250 0 500~ ~~~~~~
1,00 ~~~~~~~
1,003,0

Per1.998
(In capitaU.S.
Inconie
dollars)Kore
in~brackets
NateFigures ~show~ ~
residuals. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~3.4
andBehaan
Scarce Schnidern(1992)

1[2.91 rigN~~~~~~~~7
 SI R A~~~~~XRC L E
Notes
1. As we pointed out in the Ovcrview,a strong argu-
ment couldbc made for indudingChina amongthc "mir-
ade econorniesof East Asia. We lhavenor done so for
three reasons: (a) China's recent very rapid growth has
been highlyregionallyconcentrated;(b) the experienceof
very rapid growth is more reccnt than in Japan and the
Four Tigers and than thar of Malaysiaand Thailand
among the southeast Asian NIEs;and (c)-and most
important-the nature of the economicpolicyframevwork
in China is sufficiendydifferent from that of thc other
eight economies,and is in such rapidflux, that it makesar-
tempts to understandthe links bctwen public policyand
growchon a cross-economybasiseven moreproblematic.
We lookatsome aspectsof China's rapidgrowthin the r-
mair.der of the book but have not tEratcdit as systenat-
ally as the other economies.
2. High-income economiesare defined according to
rclative incomie to the United States in 1960. They in-
dude Australia,Austria, Belgium, Canada, Denmarkc,
Finland, France, Germany,Iceland, Isrel, Italy, Luxem-
bourg, Netherlands, Norway, New Zealand, Sweden,
Switzerland, Trinidad and Tobago, United Kingdom,
and United States. Simple catching up by low-income
economies,with growth rates exceedingthose in indus-
trial economies, is defined as 'unconditional conver-
gence' in the new literature on growth dteory (see box
1.2) and would be expectedto occur only under very re-
stnricveassumprions,for example,equalsavingsratesand
technologicallevels. Below we exarminethe concept of
"conditional convergence,"catch-up controlling for dif-
ferencesin variablescontributingto growth.
3 This issue,termed"persistence,"
is exploredby Easterly to address different questions-equipment invesrment,
and others(1993),who find evenmorestiking evidenceof
the uniquenessofpersistentgrowdtamongthe HPAEs.
- not includedthesevariablesbecausetheyare not centralto
4. Hong Kong P-:d Singapore,which almost entirely our theme.This approach,whilegenerallyassociatedwith
lackagriculturalseaors, are exdudedfrom this discussion.
5. Staristicalanalysisindicatesno significant (in the
statisticalsense)differenceof theseeight economiesfirom fectLong-RunGrowdt, summarizesthe state of the art in
other cconomiesin their share of total investment,con-
troling for relativeincome levels.
76
6. This basic pattern is observed in fbur individual
economies-Korea, Malaysia,Singapore,and Thailand.
The pattern fir Indonesia difers; real private investment
dedined continuouslyduring the 1980sfroma peakof2O
percent ofCDPto a low of 13 percent in 1989.
7. Behrman and Schneider (1992). The regressions
control for a polynnomialin averageper capira income in
the relevanryear.The authors usedper capita CNP at offi-
cial cxchange ratesas the measurcof income.
8. The residualsfor Sri Lankaand Egyptare alsoabove
the linein 1965.
9. Despite increasing its secondaryenrollment rate
from 29 to 74 pcrcent, Hong Kong also fell wedlbelow
the predictedlevelin 1987. This is becaLse,at $15,000,
the per capitaincomeof Hong Kongwas sohigh that the
OECD economieswere now its comparators.
10. Birdsall and Sabot (1993b) cite Stevenson and
Stigler(1992),amongothers,who report resultsoftests in
citiesofJapan, the United Stares,and Taiwan,China.Ap-
pendix1.2 reviewsthe resultsofother tcstscomparingthe
cognitive skills levels of East Asian chidren to other
students.
11. See appendix 1.1 for a formal definition of TPP
growth.
12. This specificationis the simplest of a group of
cross-economyregressionsthat have been used in recent
years.These studies have also introduced other variables
De Long and Summers(1991); trade orientation,Dollar
(1992); endogenousinvestrment,Barro (1990). We have
restsof endogenousgrowth theory is alsoconsistentwith
neoclassicalassumptions (Mankiw, Romer, and Weil
1992).Arecenrconference,HowDo National PoliciesAf-
this field(Eastely and Schmidc-Hebbel,forthcoming).
 GROWTH,
13. We have also estimated the basicrelationshipfor
two subperiods, 1960-70 and 1970-85. The fit of the re-
grsion is mnrkedlybetter in the earlierperiod, but in
most respectsthe parameterestimateshavesimilarvalues
and significance.The major exception is the coefficient
estimateon investment/GDP, which becomesinsignificant
in the period 1970-85.
14. Easterly (1993) finds similar resultsfr the Four
Tigers.
15. Useof a variablemeasuringequipment investment
(DeLongand Summers 1991, 1993)enhancessomewhat
the explanatorypower of the regression,though in a
smaller samnple.An alternativeeducationvariable(Barro
san Lee.19) amasreo educationstock, . b a ron
and Lee 1993), mcasurcof educationStOCK,based on
population censuses,failsto perforn better than the p-
maty enrollmentrare.
16. FiguresA1.8 and AI.9 (in the appendixesto this
chapcr)r
chaptercompare
compareour <:simatcs
oursrmaLes
e .
of
of TFrr growh vvidirItwihTO. 22. Mankiw, Romer, and Weil (1992) also employ a
other independentlyderivedestimats for a largesample
of countries (Elias 1990; Fischer 1993). The pattem Of
pr.oductivitygrowth rates in figur 1.10 is remarkablyro-
- i ,^, ^ , . functionvAr. Aminishingreturnsto cpita.
busr to the specificationof the growth accountingequa-
tion and to the capital stock series used. IMF (1993),
contrasting Asia with other devcloping economies,
reachessimilar condusions, both with respect to the esci- fiunctionsand theirrelationshipto traditioralestimatesof
mated magnitudesof TFPchange and the relativecontri-
bution of TFP change to output growth. Thomas and
Wang (1993) and S. Edwards (1992) also reach broadly
similar results concerning the pattern of productivity
changein the HPAEscomparedwith other economnies.
17. Young (1992) finds similarlydisappointingresults
with respectto Singapore's Tr perfiormancc.Some cau-
tion is neededin interpretingtheseresults,however.Much
of Singpore's investmrentbetween 1960and 1990 was in
housingand socialinfrastructure,outputs of whichare no-
ntoiously
difficultto meas;:e. It is possiblethat we have: and acrosseconomnies.
thereforeundervalued the rate of growth of output and
hence the rate of rvp change. Similar derailedcriticsms
couldbe made for other economies,both HPAEand non-
HPAE,aLwour TFP resultsare rherefore best regarded(asin-
deedare the :ross-economyregressionresults)asindicatve
of broad internatonal trends. (On problens of estimates Progress,a test designedfor U.S. students. See Hanushek
- of cross-economyregressions,see Sriniasan 1992).
EQUITY. ANND-EC
18. The studies cited arc listed in note 13 above.We
hlavenot included these variablesbecause they arc nor
centralto our theme.
19. The magnitudeof the coefficienton investmentis
lowerthan thatreportedin Dollar(1992) for hissanmpieof
114 economies(0.113).The magnitudeof the coefficient
foreducationis abourthat givenin Dollar'sresults,and the
coefficientfor RGDP60 is essentially
similar to the results
obtained by Dollarand by De Long and Sumrmers(1991).
20. This literatureis brieflysurveyedin Nishimizuand
Page (1987).
21Farl(95)wsoefthfittoneted-
vergence between observed behavior and best practice,
whcheadtcnilicfcecyTrcisbnoa
There is by now a
which he cllled technical inefficiency.
large literature on technical inefficiency. Pack (1988)
summarizesmuch of rhis literature as it appliesto devel-
oping countries.
Solo
wtype neoclassicalmode. Proponents of endoge
nous growth theorywoull not acceptthe depictionof the
~~~~~~~production
23. A comprehensivereviewof the largeliteratureon
te measurementof best-practiceor frontier producton
the production funcdion is contained in Aigner and
Schmidt (1980).
24. This is the approach adopted by Nishinmizuand
Page (1982) and recently applied to high-income
economiesby Fecherand Perelman (1992).
25. This assumption,whilestrong,may not be fir fiom
the truth. Industrialsectorestimatesof TFP change in in-
dustrialeconomiicsgenerallyyieldacompactdistributionof
ras with a meanvaluenear 1.5percenra year,both within
This may thereforebe a good first
approximationofthe rateof technologicalchange.
26. The restswereadministeredto a random sampleof
2,000 students in each county. The test questionswere
drawn from the National Assessment of Educational
and Sabot (1991).
77
 R.TE 2

Public Policy and Growth

HAT ACCOUNTS FOR THE REMARKABLE

record of rapid income growth and declin-
ing inequality described in chapter 1? Some
observers attribute success to geography
and culture rather than economic or other
policies. But geography and culture cannot
explain it all; East Asia indudes Myanmar and the Philippines as well as
the Republic of Korea and Malaysia. Nor is there a simple policy story.
Of the successful East Asian economies, Hong Kong and Thailand have
had relatively little activist intervention in economuicsectors by their
governments, while Korea and Malaysia have had a lot.
in this chaprer, after examining the significance of geography and
culture, we suggest a simple framework for understanding growth in
the eight HPAEs. This framework explicidtlyallows for diffcrenc, bur po-
tentially successful, mixes of policies across time and across economies.
To set the stage for latef chapters, we then show how very different
policies can be used, if circumstances are propitious, to address the
three critical functions of economic management mentioned in chap-
ter 1-accumulation of resources, efficient allocation of those re-
sources, and productivity growth, that is, increasingly grearer output
for given resources.
How much of East Asia's success is due to geography, common cul-
tural characteristics, and historical accidenit Certainly some-but def-
initely not all. Ready access to common sea lanes and relative
geographical proximity are the most obvious shared characreristics of
the successful Asian economies. East Asian economies havc dearly ben-
efited from the kind of informal economic linkages geographic prox-
imity encourages, induding trade and investment flows. For example,

79
 S_ MIRWACLE

throughout Southeast Asia, ethnic Chinese drawing on a common cul-

tural heritage have been active in trade and investment. Intraregional
cvnomic relationships date back many centuries to China's relations
with the kingdoms that became Cambodia, Japan, Korea, Laos, Myan-
mar, and Viet Nam.
In South and Soutleast Asia, Muslim traders sailed from India to
Java, landing to trade at points in between, for several hundred years
befre the arrival of European ships. Thus tribute missions and tradi-
tional trade networks, reinforced in the nineteenth and twentieth cen-
turies by surges of emigration, have fostered elements of a common
trading culturc, including two lingua francas, Malay and Hokein Chi-
nese, thar remain important in the region today.
In our own cenrury, key Asian ports were integrated into the emerg-
ing worid economic system as the result of European military and trade
expansion. Cheap ocean transport and shared historical experiencesfir-
ther knit together a far-flung, culturally disparate region. U.S. assistance
in rebuilding Japan after World War 1i, followed by massive U.S. eco-
nomic assistance and military spending in the region rbroughout the
cold war, also helped to set the stage for rapid gowth. Positdveimpacts
included enhanced security, which enabled governments not directly in-
volved in the combat to focus on economic development Perhaps as im-
portant, the U.S. military's exensive purchases in Asia provided a ready
market for emerging export industries. Japanese industry receiveda sub-
stantial boost from the provisioning of U.S. troops in Korea; just as
some of Koreis biggest conglomerates got their start selling goods and
services to the U.S. nilitarv during the war in Vietnam.
Regional linkages fadilitated the adoption of imitative strategies, in
both public and private sector activity. Policy imitation-specifically of
Japans industrial strategy-was an explicit objective in Korea and
Malaysia. Korea borrowed Japanese rechniques for building large trad-
ing companies and directing the structure of industry; Malaysia fo-
cused first on developing heavy industry and more recendy on building
business-govcrnment relationships. Furthermore, the general model of
Japanese success undoubtedly impressed policnfakers throughout East
Asia, engendering a sense of confidence as well as providing models of
potential instruments of gowth.
Fmally, geographical proximity has facilitated capital flows, particu-
larly in the past decade, as Northeast Asian manufacturers of labor-
intensive exports moved their fiictories south to take advantage of lower

80
 PUBLIC POLI

wages. A surge of inves-tment (since the real appreciation of the Japan-

ese yen following the Plaza Accord)-flowing first from Japan and later
from Hong Kong, Korea, Singapore, and Taiwan, China-has con-
tributed significantly to the dynamism of Indonesia, Malaysia, and
Thailand. From the policy viewpoint, these linkages have been encout-
aged by the generally liberal treatment of foreign investment But even
where foreign investment policies have been restrictive, informal credit
and information networks have helped investors to move capital rela-
tivelyfreely.
In addition tO direct linkages and imitation, East Asian economies
may have benefited from positive regional externalities. Through their
earlier trade with Japan, Western importers had become farmiliarwith
Asian business, established networks for sourcing East Asian products,
and gained respect for East Asian quality. Later, U.S. trade policies im-
posed quantitative restrictions on Japanese products and created rare
opportunities for other East Asian producers to enter international
markets. Producers of garments, shoes, television sets, automobiles,
and other products in Korea and Taiwan, China, took advantage of
these episodes to establish lucrative market positions (Petri 1988).
Of course, regional cl'2racteristics alone cannot account for East
Asia'sremarkable success. If geography, history, and culture were an ad-
equate explanation, other economies would have little to learn from
Asia's success stories. Fortuntately evidence suggests that this is not the
case. Indeed, cconomies that are part of the same matrix of geography,
culture, and history as the HPAEs but followed different economic
policies-the Democratic People's Republic of Korea and the Philip-
pines are two widely divergent examples-have yet to share in the East
Asian miracle. We turn, then, to the heart of this book, an examination
of the policies that have shaped East Asia's success.

PolicyExplanations

TEMPTS TO IDENTIFY THE POLICIES THAT HAVE CREATED

A the East Asian economic miracle fall inco several broad cate-
gories. Below we present a summary of these, focusing on ihe
differences in interpretation that have arisen among observers con-
fronted with the same set of economic facs.

-8i
 SvA C MIRACLE
LSYXN

TheNeodassical
View

In the neoclassicalinterprectaionof EastAsia'ssuccess,the markettakes

center stagein economic lifeand governmentsplay a minor role. For ea-
ample,Wolf (1988, p. 27) found it 'a strikingfact that the few relatively
successfuldeveloping [economies]-Hong Kong, Malaysia,Singaporc,
the Republicof Koreaand Taiwan,China-have greatlybenefited from
decisionsand policiesthat limit government'srole in economic decision
making, and insceadallow markecs--notwithstanding their imperfec-
tions and shortcomings-to exercisea decisiverole in determining re-
source allocation."Similarly,Chen (1979, pp. 183-84) argucd that in
Japan and the Four Tigers(Hong Kong, Korea,Singapore,and Taiwan,
China) state intervention was largelyabsent. "Wha. the state has pro-
vided is simply a suitable environmentfiorthe entrrc -ne.iursto perform
tieir functions."
The neodassicalinterpretationof the axperienccof die HPAEs, and e;
peciallyof Koreaand Taiwan, China, presents a coherent and powerul
view of one path to successfulindustrialization.' In this view, govern-
ments in all of the HPABs provided a relatively scablicmacrveconornic en-
vironrment, characterized by limited inflatinn (except at times in Korea).
Real effective exchange rates rarely appreciated, and such episodes were
quickly corrected (S. Edwards 1988). Interludes of intensive import-
substituting industrialization in Korea and Taiwan, China, were brief
Manufacturers were thus able to concentrate on improving productivity
perfbrmance rather than coping with mpidly changing relative prices of
inputs and outputs.
Although neoclassical interprettions of the success of Hong Kong,
Japan, Korea, Singapore, and Taiwan, China, are ofren nuanced-and
advocates offer them with varying levels of consumer warnings-they
stress the benefits to the HPAEs of limited policy distortions in the for-
eign trade regime and domestic factor markets. In the neodassical view,
East Asia's export success is due to similar rewards for selling in the do-
mestic and foreign markets. iBecause variation in incentivesacross sec-
tors, as measured by the effc-tive rate of protection of value added, has
been limited, inputs havr-flowed to sectors roughly on the basis of static
comparative advantage, anti inrernational competition has provided the
impetns for cost disciplne and technological upgrading. Traded inputs
have been made available to exporters at international prices, and ae-
porters have iced an n.ternational price regime in making their deci-

-St
 PUBLtC P OTC P

sions. Finally,factor markets have been roughlycompetitive,so positive

real rates of interesthave prevailed,and thcre has been an absenceof du-
ality in the wage structure by size of firm or sector of production.

TheRevisionistVww

In the past six years,the neoclassirl interpretation of the sourcesof

rapid growth has been criticizedfor its lack of factualvalidity,at least as
it appliesto Japan, Kor=a,and Taiwan, China (Packand Westphal1986;
Amsden 1989; Wade £990). Advocatesof this new view, sometimes
dubbed revisionist,have systematicallydocumented that governments
in these three economiescxtensivelyand selecivelypromored individual
sectors. They have convincinglyshown that levelsof protection and the
variation of protection acrosssectorshas been greater than recognizedin
&>rthei neoclassicalinterpretations.2
Indeed, governmentsin each of these three economicsat times inter-
vened forcefiillyin marker Korea, for example, strongly encouraged
heavyand chemicalindustries by setting targetsand offeringa varietyof
financial incentives (Pack and Westphal 1986; Westphal, Rhee, and
Pursdl 1988).Japan promoted the developmentof severalweak indus-
tries in the first fifteen yearsafter World War II, offeringprotectivemr-
ifis and financial incentivesto encourage the introduction of advanced
technologyand establishingrationalizationcartelsto &cilitatethe exit of
inefficient firms?3Taiwan, China, used public investment in large-scale
manufacturing enterprises to ensure inputs for predominantly small and
medium-scaleexportng industries.
Moreover, capital markets were not free in these three economies.
Proponents of the neodassicalview have focusedon the reduced inter-
-vention,particularly in the 1960s,that allowedreal interestrates to shift
from negative to positive levels (McKinnon 1973). Buitwhile interven-
nion declined, it nonethelesscontnuedL In fact, Japan, Korea, and to a
lesser extent Taiwan, China, did not rely solely on markets to allocate
savings. Rather, they repressed interesr rates and directed credit in order
to-guide investments.
-The revisionist view, proposed by Amsden (1989), Wade (1989,
1990), and others, sees market failures as pervasive and a justification for -
governments to lead the narket in crtidcalways In this view, the experi--
ences of Japan, Korea,. and Taiwan, China, provide evidence that gov-
ernments can foster growth by "goveriing miarker and "getting prices

83
 ; * - RA: C ~~~~~L
E

wrong" and by systematically distorting incentives in order to accelerate

catching up-that is, to fadlitate the establishment and growth of in-
dustrial sectors chatwould 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."

TheMarket-Fdendly
View
WorldDevlopment Report 1991 (World Bank 1991 b), in a compre-
hensive attempt to describe the policies needed for rapid growth, fallsin
the middle ground between the neoclassical and revisionist views. It
condudes that rapid growth is associated with effectdvebut carcfully de-
limited government aciivism. In the itartica-
sMarrkea-tendly"suategy
lates (see box 2.1), not only do governments "need to do less in those
areas where markets work," namely the production sector, they also
"need to do more in those areas where markets cannot be relied upon"
(p. 9). The appropriate role of government in a market-friendly strategy
is to ensure adequate investments in people, provision of a competive
cdimate for enterprise, openness to international trade, and stable
macroeconomic management.
But beyond these roles, govemments are likely to do more harm hn
good.On the basis of an exhaustive review of the experience of most de-
-eloping economies, WorldDevelopment Report1991 condudes that in
-gneral governments have been unsuccessful in improving economic
performance through attempts to guide resource allocations by other
than market mechanisms. Attempts to guide resource allocation in
international trade, financial markets and labor markets have reduced
competitive discipline, guided resources into low-productivity and in-
ternationally uncompettive sector, and resulted in widespread rent-
seeking. In short, though market failure is an importrant impediment to
rapid growth, so is government failure-and governent failure can
have high costs.
A cental contribution of WorI Development Refort 1991 is the ar-
-gument that sustined growth :esults fiom the positive interaction of
four critcal aspects of econoomic policjr macroeconomic stability,

4.4
 PUBLIC POLICY AND GROWTH

Box-2.1MakingtheMostofMarkets;
IN-THE;AST;-TWEN YYEASA CONSENSUS.HA -;trenologi c nged 'produ y. TeRepor-
'aenergkmdaiongeconomists on the best approac' to .stressd at omesticpolCy. ou d epeennepre- -
economic dedoprnetn -. This zonsensus'was dp
-.-...
cissed'az. ngth:in WoMDev p'nt Rport 1991 -provide die infasuucure- and institutions'-at will
The-TSChalnge oDv ne (WdrIdBaiik.1? ' r 1199
Ib). ,-h''help he respond.i_
The Repor+thighlightedte iportance of a helty-' ' Oprness to ona tr invesren nd'
privarc:sector, which.-rss m investments in -ides has been cra in encouraging omnsticpo
,people,tamuchreduced role or.,goveiment, ope- du-r' toucutrcostsbyw inrdi holopos-;:r
-,nessroXd,'o eom ndvs w ) t rds Ce,oft anddeveopingnewand tcerpruct. By' esame:
K'world;,and; macroeononmc-stability TIh~s'idcas -token,3protonSr erc indusa ol
-'''av c#lled into;whAtis now cae te-hemark' d6elom decades.He -gon np '
rv.m
Y
9' rriL-
-;appro&; 0'-
' okvole. Injapan an.dKorea,foruinstace,gov~-
T eReporrcnotedtrhtareeconomicufronim. ement :agen and-indut asaton co' bo-
t investng mi peopplcan be high,bur manygovern-. rate to gahier and. disseminate i:normationon:.-J,.
'no..t- on] ar rd-'- tt anddeop quaity control'Ibrp& t.. '
-ntarget thcir snbapkpmropacrod micc foundtionoi. on ofthe
ecoanro,-in':',-Ale'rocr
Brazil2nd Pakistan,'f alone5 did' m,sost i ' oosihagormenrs
OOpublic ca
- derto; .unprovt.
'w hie sot
~fl indicators;i Ce.and .rovide,- as th neotpule.ot o. Wt%an
jamnaica,howevr to c CarS inprov eVe. i mnt ocr th o
times,ofKislogw.wth..Governmentsmust ais - overspens,
Mr
monetayryl
. terdui st
senoonowinaj
n fnniscordfiute,' '-
opnin
.-prov- th 4ini fhes To, hled by inflation,choni i
'''o% pvisios,
o for -recurrent: pen are
' inade- of y alssof
,thel o 'p
quam
? i, te rEsuldjiawasrefilu undemrutli2ation> Thdeconomies with a:bs-orof macn&o'c u"r>
Motsessfuuonlei nudhJai;Si-" iaiitprruesleionthaiidd ArgeCntn.'olivia
~~~iposr-,adoc~hv
- - < =ic salihdter~o
o' m on s~~~~~~~~~~~~~~~n
-sf-tabedili
.y>ir iel d hnOh&nu~rj
ItcP%%Y :avarae.a through rhen riors,- of~' ke,'..' doeia Koe,Maas; nTalidhiemn
KL.markeslcqi.k%versely; retuctionson competition' "keep.thefir rmacroononic:. polid&son.'I
gecLtod
(fo e5pe;Indsiadcensn ,
o nappopriate.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.,
course adteir-.bioaUednoic jerfbr~ceha
ihes ?P opeoi) .biddk'.b
> 'betarizlyo S

humancapital CtiaaCion, opennesstO nternational trade,and an en'n-

romnmenr
that encouragesprivate invrestmentand comjpetition. Effective
policyin one diension (suchas human capital f hrmation)
imprves.
the results from effectve policiesin others (such as opennessor mac
o
economic stability)t In this view,the success of many economies in East
Asia has been due tO teinforcing policy feedbacks. No single policy has
ensured success;strong and effictive policies in all four critincal
aEcas,and
over a sustained period, have been hkey. f - -p
. 'csumThe
marke-ffiencdly.
approach captures important aspects of Eastcr
Asia's succe:m.These economies are macroeconomnicallystable,
havevEry

Asia-ssuc=ss These .oniies are macroe-onomiay stabl, he v85,

... -, ,,. .-. X
' - ' ' 0' - :'~ ~ ~ ~ ~ ~~~8
 high leves of human capital, are thorughly integrated into the world
*,~~~~~~~~~~~vr tmc ocure
economy, and have very high levels of competition among firms. More-
oe,East Asian success sometmsocurdin spite of rather than be-
causeof market interventions. Koreas heavy and chemical industries
(Hcr) driveand Japan'scomputer chip push did not live up to expecta-
tions. Evenso, other interventionscombined with export targetsappar-
ently were consistent wick rapid growth: quota-based protcaion of
domestic industries in Japan and Korea; targeted industrial policiesin-.
dludingdirected 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 successesof these three northeast-
em economies compares favorablywith the successesof Hong Kong,
Malaysia, and more recently Indonesia and Thailand, where policy
choiceshave been less interventionist.

Pathsto Gmwth
-NMultip
As these three viewsshow,while it is easyto see how elementsof East
Asianexperiencecontributed to rapidgrowth, it is nor easvto identifyra
single recipefollowed across the region. An alternativeapproach is to
recognizethis diversityofpoliciesand to assesswhether and how various
mixes of policiescontributed to successfilimplementationof three cen-
tral functions of economic management accumulation,allocation,and
productivitygroth. 4 Thi4s approach recognizesthat policies,like tac-
---
tics, can and should vary depending on the situation, while the central
functions,which are crucialto development,musr alwaysbe addressed.
One of the hallmarksofeconomic policymakingin the HPAEswas the
pragmatic-flexibility with which govenmentstried -policyinstruments
in pursuit of economic objectives.Instruments that worked were re-
tained. Instruments that failed or impeded other policy objectiveswere
abandoned. Thus, the Korean and Malaysiangovernmentsreduced in-
tensive promotion of heavyand chemicalindusries when the fiscalcosts
and strainson the financial systemthrdatenedmacroeconomicstability.
Indonesia abandoned capital marker controls ancdattempts to control
Japan shifted from
interest rates when they werc found to be ineffeccive.
higly selective interventions to promote exports in the 1950s and
1960s to industrial policies designed tO reducetrade friction in the.
1980s(seebox 2A later in the chapter).Taiwan,China, gradualli priva-
tizel key public sector investments-madeto support the rapid growthof

86-
 PUBLIC POLI

exports by smalland mediutm-sizefirms.Singaporeshifted from a policy

of repressingwagesto one of promoting mpidwage growth, finallycon-
cluding by coordinatingthe increasein wageswith rising productivity.
And Malaysiashiftedfrom policiesthat resultcdin business-government
conflict to highly structured mechanisms to encourage business-
government cooperation. Chapter 3 discussesthe pragmatic flexibility
of HPAEgovernments in pursuing two goals, macroeconomicstabiliry
and growthof manufacturedexports.
Obviously, our three functions of growth-orientedeconomic man-
agemnentwere not objectives in themselvesfor policymakers in the
rapidly growingEastAsian economies.Instead,they are simplya frame-
work for analyzing how the varying mix of policies and instruments
chosen by HPAEgovernments contributed to or detracted fiom rapid
growdt.At the same time, we do not want to imply that growth was due
to mere luck in the HPAEs. Though policiesvariedover time and place,
the formulation and implementation of policieswere purposeful, with
an overridingobjectiveof 'shared growth." (This tcrm is describedand
discussedin chapter 4.)
In contrast to the diverse,changing,but purposefulmix ofpolicies in
the HPAEs, ocher economieshave either rigidlyadhered to policies that
were demonstrablyunsuccessfil-fbr example,extendedperiods of im-
port substitution in Latin America and Sub-Saharn Africa, policiesto
developdomestictechnologyin India, and state-led industrializationin
EasteemEurope-or have lackeda purposeful direction,as the historyof
adjustment pograomsin some economies in other regions attests. In-
deed, as we illustratein the next chapter, the flexibilityand pragmatism
ofpolicyrnakingin most of the HPAEs reflectedgovernments'willingness
to try new approachesto achieveparticularpurposes,discardunsuccess-
fuilones, and keep those perceivedas successful.

Th Functional
Growth
Frameworlk

F_ IGURE 2.1 PORTRAYS THE FUNCTIONAL APPROACH TO UNDER-

standing growth in the HPAEs The figure shows, in four columnns,
the interaction among two sets of policy choices (fiudamentals
and selective interventions); two methods of competiive discpline
(market and contest based); the three central fimctions of econonmic

87>
 Apprachto G
Figure2±1 A Functional :owh

Comnpetitive
Policychoices discipline Growthfunctions Outcomes
-~~~~~~~~~~~~~~~~~~~ Hihsa.g

Fundanmentabe Accnnmulatn Rapidand

sustained growth
Increasing human
. Stable macroeconomy *
capital * growthof
Rapid
* Highhumancapital *Hg,aigsexports

andsecure MdHgll Rapiddemographic

: Effective
financialsystems Metetransti
Export
isin agricultural
RFoapid
Um.t-igpricedistortions ~~~~~~~~~~~~~capital
I Radai ho
transformation
capital competiin
* Hlghhuman
Openness to foreign
SOpennessto foreign It
*technolrtY p Domesc . _

AFicultuial depelopment I Contit . ron industalizadon

- § ~~~pollicis v Alboatin

: Directedcredit credit-Reduceyfective use of

} . ~~~~~~human
capital in.
: SelectiveIntprventions- laborrnart
I rI\ * High reurts on
_ Ilmestmrent < >Equal Ineeme
I Export push
Contest4edtu clstributbon
I vz Financial repression
*Export credit . Reduced poverty
* Directed cedit >

I * Lfwestmnent . Improving sorlal

I * Seiectivem prormotion
coordination Productty change inditrs
* Information * PmduftJvpbased
InsUtUdons
W S exchange catchingup

Insulation
o Technocratic R tec_ &
____
0- . _ * H gh-ualfty civil service +_______________________.______________

- Monitoring I

managemenrt and tie outcomes of growth and equity discussedin 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
nmasteryof technology. Effective and secure financial systems helped to
increasethe 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 technologiesand innovate. Openness to foreign rechnol-

ogy was a major vehide for productivity-basedcatching up, and agricul-
tural sector policies that promoted rural development were cental to
both growth and improveddistribution ofincome.
The arrows indicate that the system has numerous self-reinforcing
feedbacks.For example,rapid growthand relativelyequal income distri-
butions contributed to the HPAEs' superior accumulation by increasing
savings-rates and generating largr and more efective investmentsin
human capital; they also contributed to superior allocation of human
capita by fostering labor mobility and marker determinaton of wages.
Similarly,they helped in the creationand protecton of instimtuions,par-
ticularlythe dcvi service,by reducingincentivesfor corruption.
The sixpolicieslisted as fumdamentalsare so defined in the sensethat
they affectthe attainment of growti functionsprimarilythrough market-
based mechanisms of competitive discipline. Three-macroeconomic
smbifiry,effectivefinancial.systems,and limited price distortions-assist
markets. The two others-high investment in human capital and open-
ness to foreign tecLnology-requireefficientmarkets to operate.
But the HPAEsalso went beyond the fundamentals, interveningwith
varying degrees of intensity to alter market incentives. The broadest in-
rervenrions were generous incentives for manuhactured cxports. (These
are discussed in some detail in chapter 3.) As part of this export push, all
the HPAESpermitted 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,
Whereinternational prices guided the domestic economy, massive pub-
lic housing programs helped to keep down labor costs. In Japan, Korea,
and Thailand, implicit or explicit export targets were used as the basis
for awarding access to foreign exchange, investment licenses, or credit
In Taiwan, China, large-scale public investments, induding in public
enterprises, along with extensive support senices to small and medium-
size exporters, were used to support the export drive.
Most HPAEs also controlled interest rameson deposits and bank lend-
-ing. These financially repressive policies enabled governments to ration
credit; favoring some would-be bonowers over others. In addition to the
importance of these directed-credit schemes to export promotion, some
governmentsalso directed credit to other activities, induding agricul-
T

ture, small and medinm-size enterprises, and, in Japan and Korea, large-
scale imvestments. By intervening in credit makets, governments

- f - - - - - - . -
~~~~~~~~~~
 selectively promoted industries, ownership groaps, and, in some in-
stances, individual firms. (These issues are analyzed in detail in chapters
5 and 6.) Some HP'AEchose very narrow promotion iargets. selectively
assisting knowledge- and capital-intensive industries and firms in hopes
of upgrading the economjs overl level of industrial technology: In
Japan andt Korea the governments at timnesconcentrated on promoting
private domestic investment in)heavy and c-hemicalindustries (Hcas).In
Ta-iwan,China (and to some extent in Indonesia), the public sector in-
vested directly in technologically sophisticated industries. Singapore and
Thaiand have attempted to attract foreign investment in capita- and
knowledge-intensive sectors. Malaysia initially undiertook public invest-
menus in HCI firms but has recendly shifted to a strategy of promoting
private investmnentmore akin to that of japan and Korea.
Some selective interventions went beyond helping markets perform
better. Rather, theyrguided and in some cases even bypassed markets. In
the following section of this chapter, we explain why, in light of market
failures, some interventions make sense even within the neoclassical
framework. We then describe how some of the interventions under-
taken by HPAEgovernments addressed these marker fihilures,effectively
(if not necessarily intentrionally) reducing the gap between social and
privat-egains.

MarketFailures:
TheCoordinatiom
Prohiem
A primary function of markets is coordination. The price systemnis a
mechanism by w-hich the production decisions of the myriad firms that
make up the economy are coordinated; for instance, they signal to inter-
mediate goods producers what outputs are required by final goods pro-
ducer.
Whe marets arc incomplete or missing. they cannot efn
this signaling function. Since the prices of outputs and inputs (including
wages) depend critically on what- other firms are doing, there ar-egreat
potential benefits ofsharing information. Even in well-developed market
economnies,information is conveyedi in many ways other than prices.
Tradejournals,tradeassociationmeehings, and informationnewsletters
areal inportant minsitutions
for the trsnsmissionof informiation.
AdamSmiteHs invisiblehand paradigr arguesohat achindcividual,
in
pussuinghis or her self-inters also maidmizesthe commonwelare;
cooperationis thereforeunne lssary-
-In rafiutymoderneconomiesare
characterized
by extensivecooperation.Amongfirmsimeconomis that

'90
 PUBLIC POLI'CY 44$Dt½N`

':belong to the Organization for Economic Cooperation and Develop--

ment (OECD),for example,formal and informalsharing of information
is common. Cooperation among firms is maintained whenever the gains
from cooperating outweigh the gains fromi cheating. Coope-rativerela-
tions are ensured because of benefits from maintaining a good reputa-
tion. If a firm lidls to cooperate, its reputation will be damaged, and
future business chat depends on its reputation will be lost.'
Institutional. arrangements for cooper-ation and information ex-
change in developing economics are weak-er than in industrial econo-
minis,yet the needs for these forms of coordination are undoubtedly
gre'ater. When economic change is.siow, making predictions about fu-
ture prices is relatively easy: a simple projection of the recent past will
do. But when economic change is rapid, predicting the future is much
more diffi'cult. There is often need for coordination. beyond what mar-
kers can provide; hence there may be a greater role for govermecnt to
create institutions and facilitate coordiination.
MissingInfonnatio and CreditMarkets. Among the markets most af-
fectedi by information problemnsare capita markets. Even in industrial
economies, equity markers generally do not finance much new invest-
ment, largely because of information asymmetries (Stiglitz 1993a). In
economies where'equity markers are weak- or absent, credit mecha-
nisms become the primary vehicle for raising capital and. diversifyring
and spreading risk- But credit markects-even those flee of interest rate
controls-are often characterized by credit rationing. Credit is seldom
allocated to the highest bidden. This is for an obvious reason: bidders
* are bidding promnises;those who bid the most may not be able to fulfill
teir promiss Lenders are concerned nor with the promised return
but the actual return. As a result, capital is allocated by a screening and
evaluation process that is quite different from the anonymnitythat char-
acterizes resource allocation in an idealized marker.
In some oftdie IHPAEs, governiments intervened aggressivelyto address
this problemnin the credit marker, going beyond the normal regulatory
functions and prudential supervision that help ensure consumer confi-
dence. For example, in Japan, Korea, and Taiwan, China, government
helped establish andt aggresively supported development banks and
other financial institutions,-and only later encouraged equity and bond
markets. In Japan a-ndKorea, government set up and managed directed-
credit programs,.and through explicit and implicit guarantees or other
forms of interventions, reduced the risks borne by investors.

A'
 ScaleEconmnies, andCiooperation.
Externalities, Coordination may yield
substantial benefits when large indivisibiliries in investment lead to
economniesof scale. One such case arises when investments are inter-
dependenc. For example, if a steel plant and a steel-uising industry are
needed concurrently, it does nor pay to develop the plant unless there
is a steel-using industry, and it does nor pay to develop the steel-using
industry unless there is a plant. If each awaits the other, nothing hap-
pens. Market failures due to incomplete markets, such as the absence
of capital and risk markets, exacerbate thi~ssituation. With such large-
scale investments, no single entrepreneur could amass the capital
required for both investments, and capital marker imperfections mean
that investors hiavedlifficuiltyobtaining the funds required. Moreover,
there are likely to be large risks, and the market provides no mnecha-
nisms by which these risks can be divested.
Economists have traditionaly found exteralities a persuative argu-
nent forsintervention in marker allocation.
di My of these xtemalities are
related to learninge spiovers associated with developing markets,
spillovers associated wit discovering what goods can be produced, and
spilloversassocated wevith the incomplete appropriability of tecnological
knowledge. In each case, fims undertaking a new actvity facelong-term,
noncollateralizable ivestments that are difficult to finance. Coordination
failures loom especialy large where "diFuse externalities exist and there
are mutiple linkags between
s993a).firms (Stiglitz 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 Exampliesof
hese diffuse e mrnalities indlue, for insance,dith development of hgh-
technology servicesand nontraded intermediare goods.
In the earlier developent liteature, the coordination problem aris-
ing from interdcpendenc investments, economies Ofscalesand exteral-
ities wasgiene agrea dal paromipnec tThis was the argument for
planning in the 1950s and 1960s. Such planning ulled in part because
it attempted to concentrate all rclevant infonadvon nmthe goverumenit
planning bureau. This was sinply inmpossible;infrroation is too dif
fuse, too complex. Plnmningministries were not adept at generating the
infornation needed to support investmnt dceisions; as a rexsultemny
planned investments filled. Moreover, because many of the planned in-
vestments were in monopolistic public sector companies, severe prob-
lenmsof economIiicperformance arose related to- corporate govemance
and market powec:

92.
 PUBLI 0 L14

Some HPAEs tackledproblemsof ;caleand externalitiesbysetting up

mechanismsthrough which businessand government could exchange
information and coordinateinvestmentdecisions.Japan and Korea,and
later Singapore and Malaysia,developed institutions and market rela-
tionships that facilitatedthis 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, emphasizinginforma-
tion and cooperation through formal and informal interactions 1 is
discussedin chapter 4.5

Creating Contests

Cooperation raises severalproblems, however.First, cooperativebe-

havior may become collusion,if firms act together to raise prices. Scc-
ornd,cooperation may inhibit competition,leading to managerialslack
or a more generallossof efficiency.Third, business-governmentcooper-
ation may encouragefirms to seek &vorsfrom government.
How did the East Asian economies that encouraged cooperation
avoid theseproblems?They combined cooperativebehavior-indcluding
sharing of information among firms and betweenthe private and public
sectors, coordination of investment plans, and promotion of interde-
pendent.investments-with competition by firms to meet well-defined Table2.1 Concentration
RtoinManufacturing
-conomicperformancecriteriaThey developedinstitutionalstructures Ratiosm in- -
in which firms competed for valued economic prizes, such as accessto -
- credit, in some dimensions while actively cooperating in others; in 7%em0cewhflia-. 1
short, they created contests (see box 2.2). Market-basedcompeiition J i8 S 56Z'.
K&R9:oi962A
*and contest-basedcompetition both included prohibition of monopo
lies, although the number of competing firms was sometmes small.'
EventhoughJapanand Koreahavetended to have high levelsof con- nI9 4;
. 0; f centration in their manufacturing sector (seetable 2.1), domestic com- KBraz41980 ; AtiZN
petition has usually been vigorous. The Japanesc government has
proceeded-on the assumption that competition among fewer, more '--

evenlymatched firms is preferableto having one large firm competing s4

with many smaller rivals,a principle that is well-recognizedin athletic .

- -competitions (NalebuffadSii193.v
Contest-basedcompeution included dear well-enforced rules, and
prizesforwinners.Table2.2 brieflydescribessome of the contestsorga- & ~~~~~~
nized by HPmEgovernments.6 The simplestand most widey used were

93
 b . A -RC-
R RA
CLE

j:~~~~~~~~~~~~~~~~~~~~~~~~~~~'t.-
' .. 'X47 '- ,.'. W .

Y'Oi'3m2Running
a Contest,
Rewais;,Ruies,n ?Rdees
. JT-H-lE,ECONOTC
CONASTS.IT GOVERNMENTS. Contestsin NortheastAsian.economicshave suc-
:i.japan,.Korea, and Taiwan,.Chi'Ta,-serup-amoi g ceeded pardy becauserthe prerquisits were right.
*flins
montheo'combime
sQi , J die-
b&eherts co-
wri;orfcohpetidod andian 1.:Preferential-
* .credit .and-Ibrdgn excchange:
access~to -d .
::Cj_ -!j;-- , -.>..;--:
'''.;
7 - .Y . .... ..............
.. !.;,. e5..,|
1_...-
ies-. land somchmes
coh:
montns arrl&hi'base,however,these :-centere 'caconomicpertormance,primarilawull -
rcontestis,anddiir-role-mthe economy resmble chl7-" understoodd:mpcrauve to ,expor; Referees,he gov--
ren,sprc gaes - officials-wh&hbave
.''ernmenc design'ed~andsprie
ti.- .O:ganiihgconrc
run!;a e:,oomv Esez-fi*i;
s ists a more pccaned way
r
than,el:'i.i r:i W. , -jt a
to . ,ihetonresrs
t5m
.'
-have:been
."'
gFnerally comperciuandfir.,
. been
- economy
.L rC nng.on
,, ssz-talre, justas nA same time,;parcipation has nortbeen ,;
-b'f , °p s r r:,Complcate,
s' - anatoq. c at prce cdrrtosit out'a, contest
pAovA g. !
KaA~ieana-p ayingfid Thul etunga-hcxiuo
IC ffi.l1iezi dobeeni nave not punished
phdinr igo.; In some in-
-; l easc., i-fa
tnAy-fied,
ie -fi;
L both'vk. . -.-srancessuchas Hondasshifr
s;-l--vm1 *i*o-*e-toau
toHau- -

ut7laeyp"' and-u&ec"oplay game,:e-. 6ioniobile' maniffctrang .at a time whenzfJapines

.p
:^- .: .
Bu--a-wdll-run.contesr, a well-mrri pgarta
luce ugb
; Xi can:erite6er nibec&r ul&by prpVidhigi i~ .-..offiaiaIs
consideredthe
eautnomotiv fieldsatruraced th i
0 X forcwii.c
peduiioh~nL
qX~~cop
ta es --icac:llE
.pnt ; cooper-!..
aidnducingpartipantsto ofz5nconfdrmss
or iLn haspald offandsomely
ffhtolerance
P.;-
as..
mnzatr
m
- l, - . CYb C .Cnet.
onttr.
' won't w cvcryw er,.4
-J , c:
steprpoi -- I

lioihjmstances
6I therearc.threcprerequisitesfor.. derainceot:uuois m o:dvec6mga.
. success:~
0~~~~S ; rewards,'rules,p
f -g _6 andr&rees.Rewairds
ec mutts`tic
be reoo econiorreine'tat:
-st .- ep.of~uak-clear orovtck.a.e
la' r,,,veJ-.;l-
-r- :- o-u-h/t i-.; .eliit .broa' .partiipationan
'r.ues.and,retre m . M e in
eA&rt&c .oCnbeutidon :Rulesr dearcut so -; economies with-strong
ebe government instirutions,an-
k .iW
,vih:fehvllb e
, >;6 otestu 'e-ing contst -. becd.es hardea, fs growand ther,:
-wardedL-ahcL.
+;:z whih
iy&.wardyAandL.who puiserh'FnulyL_ a
ch -om1 b~c-punssned*- powurncreascin icladon to.tat othe bureaucur"acy
...y
gamnes-- .:As Japaneseofficialswho ate nowjmoving4ay.fom.,
* . ay .w hsrneclespary :
.- cmcal.
,knosi;cet,aieraare 'conrescard a morefinuonal ow,
:..cd &h :pa1ntnzr niz thed_`hi -: ;pevs
rel- ghavdicoverd.runmng pakygaineTor".
-- -peten'to!-W/nLZhaeFidZ...-.&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 schemesvariedwidely,all madecreditavailableto
-:
. : - with confirmed export orders. For many small and medium-scale
~~~~~irms
p- ts n entrprises, export creditswere their only accesstO the fbral financialS
- .- (Levy1993).Monitoringperfomiance
:sector wassimple;firmswerere-i-
;quired to show evidence of export orders to receive credi
In Koreaand Taiwan,China, exportcontestsweremore elaborate.
Nor onlydid exportprojects er easieraccessto fids, but firms
char ex-s
had an easiertimegettingfiure loans.Becausesuch
portedsuccessfilly
creditwasfungibleand creditgenerallywastight, firmssometimesdil
vertedexportfinancingto higher-yieldingnonexportinvestments. its

'.4-
 f . - - -; ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~P
U B Lil-C. '2

exports became a performancecriterion for projects that were not di-

rectly aimed at serving the export market. In these cases monitoring
problems were greater, since both export perfomnanceatid the related
projects needed to be monitored.
A second type of contest used the power of government to grant li-
censes.Table 2.2 shows two of these contests,based on Japaneseexperi-
ence. In the first, the Bank of Japan created a contest in which
commercial banks competed to provide financial servicesin line with
governmentpolicy-for exampic,by opening rural branches-in hopes
of winning highly sought-afterurban branch licenses.The contest also
gave banks a strong incentive to comply with technicallynonbinding

Table2.2 Examples
of Contests
-It.w*

. . . d.t . ... . . ...

..-

_,p Voooehaaniernaef £ndn to;Fctdmi eX.einIses Cana bob a,nno - ea ditM in.
Noon ep..r1 t- niaja lmeroroulbe.C). 1. Nomadidnbvn UCaMmnd MaIqt3 bonge po.lpmeoe u ofkrlnnL, :
-Ctnin.
7.1..,;:-:-. lndrn±npons Cnrsqjdknm spnenndmqr an.tita,-vo'.4

~~~,;1* C. -
LC. &smk
unto . <' iI1 mOtmuanmatipn- deump
lftvo.epena&f
*11.CUU~~Lordto - r.h r
kl 7'_,>:
. 'JI,

RM11112NEW411W ~~~~~(siwtubmncnwCa)ocscaonI
t
; ~ zb a0 ~ 1 '' 9 * n-iai es i1 ' :
iR - : ...::............ : - ',
io an. .ed- . n, -; nt,htMepmpm- ( . o *a-;k.
u .w *tep u

* 6 F-v, ; _. to.m
n 'mut : dexpnplocmfpeypm a-
, ee.r ,e at ,inInn ,n pre
,pntoeedm ;seadeomnu
hav hymn -&hudue
An . (Ja gonuesu eem ankmpnda-I
,-ef , .C -: : cadSwt-e
pnepnder-idsbanwud.......................r-hsnnnccenahe
is etttitre'-'-\ '>biw ..... . ...
.,boglCaeg .- ,s'. t I, , '. , "'.'
7'-7'' '4 ' _ 'i

< b.;-.. : . . *, . ' Oneith obetves.deFl

tow (194141
Muhinewrder Pamnodom
kfim
Tuyo;e.ayMemrnao
onSk- B 3 DOz
e , -rr *

.~~~~~~~~~~~~~i~f pei, l dkmi

.;-t

danh"okma cmin UPbnie

n

:--- - : - . - . - . . Z 6 . : ..

'~~~~~~~ -0pcmr'ti156 7 ~~~~ '

.'' - 'J - , - 'M .7 ' - ' :,7:

.c . . ~~~~~~~~~~TegIdaPund.,,'-:'..' - :C - , : .9
: C, .. .r . . -. . .g . - .j--g. : . -"
t.,. ,- .. . yoteJvbgrnnueppn. ..... . esnndafdSepne&u
.¼ ,$ ,.-.-7.

'.-''' AYf 4Z~- , CL{~, 7hr.r I' S a ~ Ct~. 9

,e-,:
. -. -.. -. : ... ' ... .~ , . . :- - '.: , :,: -;5
 RA C LE

'Table2.2 (contfinued)

IN-

JtrP o u a u h d S i d h m :~ Tiw hnk

k ufj p uimppnwkteu
m hatd mh uu m u
lnabSInInxhlnhm.um
udncmhhAnadit . -,two?"

44-. ~~~~~~~~~~~~ . t~~~~~iflncrwimm"AS1

ain
ls kMhlngaahmla.

Am* rudpmiadl bmulmyrnL

* h~iwIUcDlvsnn
andvbmmrad. Miseh.a-. zdlIc.M
~kiW*rnnd lie. *. -~~~~~~~~~~~~~~ko (1d ~ wte d,nknm o"
A.)..
~~~~Th
~~~~~~~~~~
I
-...
. ..
, (dr
I~ ~ ~ ~ ~ ~~~~d iime rd rJrmmck
yKrjymn&
rW61

initu cm mm, dwhmmsd&dtuW

~~bmm.Kq~
.1 Damrmguhcflsp-dykdLn
p"c apwtnruaiiitli uilidwaI ade
.1 *lugemlmerpand ammafc SlpI.nsio.la
dto mul. ~~ uhidgl*p*lluuca
ddr£rtmu aaa aS .
~ Nlazm afirms SsAflCs.
lca
.,.

;'L, Iwam mas, huh alps ais .

-'hZ- .. tmunymhiepiRaIMacMftWOI.
PCf'Wi Wdurn 1
tapumumnnsuuchatmaecdlmsrniiza~~~rnaum.-
. '"Mr

.'Z uilobmalbr mapnhdma.* 'VJ

*1~~~~~~~~~~i (C- Mm*mm:armshV,.a lafaryolavcn i
- 2 ItS nheppamatn4talmUd.
tedon Amaik

C -', '. Nandi ladmim'strativ. guidoMancemf S

thcauirau ulending banksotbar failneydIm oco pl

it. Cumitla
cm dehpmimmmihqClr3lesc-usuanmkn.y hev-n hniclidstis r h 90 n 19 -s

admsinistrpatsivegudacpithei
of ledig:bakthatwas feqired
to complye
wirsthaguidaceiabou the spreatsdubetweedepositiand lni-rngrtes were-
nrepermitteorexpandt
t thesepoir
brnhnewrs (Yohin o19)
Duringtherapidgrothperiodthe acMtinsr ofdIneratonlt Trad
andonduisrysal(Mrr lieamnsedIt capacit Thpasioin sevral keytibindu-t
deries-usueall heavycandichemicaldinduestiesgI makt'hea195s ands1960s,
Dmflpimted entrydoandimrkestrce compewtinga irmpworts. ofimin-the pro-
tected
ndustrywerereuireds toscordnaeivsmnstopeeta
pcsite expansion.1offupacit ibutby
tyheeas equiedtn to ompete
796~
~ ~ ~~~~Frs htrcie etaprft let euaio"(us-et)wr
 PUBLIC POLI 0:

economies of scale or leaming permitted industries to cxpandexport

market sharcs.
Our final example is that of the coordination of investments in Korea
under the Ha drive. This is the most complex example we have of a con-
test, one that combined licensing of capacity, access to credit, and pro-
tection against competing imports with both physical and economic
targets fbr performance. Its economic success is still not clearly estab-
lished, and many observers have judged it a failure, but it illustrates very
well the use of contests.
At the heart of the HCI drive was the goal of changing industrial struc-
ture toward more capital- and knowledge-intensive industries through
coordinating public and private investments. Large private companies,
the "chaebol," bid on individual, large-scale investments-fior example,
electronics, shipbuilding, and machinery-for which thev '2ceived ac-
dusive licenses combined with generous access to credit. The rewards
consisted of quasi-rents due to import restrictions and restrictions on
entry, access to credit in a highly credit-constrained system, and govern-
ment support during business cycle downturns.
These incentives have been the source of disastrous economicperfor-
mance in other economuies.The check on poor perfornance in Korea
wasthe result of rwo performance-based rules. Fi-st, government estab-
Iished a timetable for the attainment of international competitiveness in
eachindustrial sector Firms that failed to maintain the pace of cost re-
cductionsin line with the international norms faced both political and
economic sanctions (Kim and Leipziger 1993). Second, because access
to credit, even at early nonexporting stages of the development of HCI
investments, was linked to the export performance of other products
produced by the chaebol, efforts to maintain export Competitiveness
across a wide range of products were encouraged.
MheRulesof the Came:Exports as a Yanistici Contests are only as bene-
ficial as their rules. If perfbrnance criteria channel effort into unpro-
ductive activities, there will bt little benefit to the economy.8 If rewards
are unrelated to effort, contests will not increase compe'tiion. There is
one common thrrad among all of the contests outlined above. Exports,
and especially manufactured, nontraditional exports, were the yard-
stick ainst which the success of other allocation decisions-for exam-
ple, credit allocation, domestic content requirements, and industrial
licensinwere judged. Accordingly, there was a high degree of com-
petiion among finns, in spite of the &fctthat in certain domains and

'97
 rttC LVRcE

at certain times they acted cooperatively.The export performancerule

was broadly shared among HPAE governmentsand is almost unique to
the East Asian economies.Chapter 3 illustrateshow 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 HPAESused apparent constraints on their ability to export to
create export contests to new markets.
Using exports as a performance yardstick generated substantial eco-
nomic benefits. A firm's success in the export market is a good indicator
of economic efficiency-a much better indicator, in faict,than success in
a domestic market. Export markets are likely to be much more compet-
hiivethan domestic markets. Even if the firm's success is based on find-
ing a niche in a foreign market, its contribution to the domestic
economy is still as large as if it had succeeded by developing new pro-
duction processes or othenvise boosting efficiency9

''-~~~~~~~* '~ '~t W;'X

r ~~~~~M ~R
5 . l'i
zmo.g -

irk'~~~~~~~~~~~~k
C}:*ahezspmrheJsccPIAaEasJencdtixar&icellzngSowoFuha*cm
u -~~~~~~~~~~~~s YE;uats~amcRA
'n.amamiarllot ortherslP.asy¶V
: ~~~~~~ p_no
~ ~ ~" S p8 __tij.;!str*

R X FrsRER>^*z.zbV.F ^gg;2ii5X;t> ~~~~~~~Isoii

f
, -. - Eiwa a~~~goyrmmen< tioo D or

amodo

- Zsa&UO%%OOi5UDnaV3flCPeSw
§9
V ,*>w~~~~M.14o<tjf, -- rSi,'22...{ reihogn. .r emuncess,U
e Sz >o c uc nt areusofdlereamorun
.'-~~~~~~~ ,g .'~4r~ -0
, . . gs,s BB,..N.X>.^*,>t S *.co
- : - - - lR a*t; M S~~~
.. . 3 ncollnpSaEpovtcotowDhsnRneurApmanceMDo6luonsqo
t>;I:6Sy a~~~~~~~~~~~~

p8~~~~~~~~~~~~~~~~~~~~~~~M
9s- oXBg,.sp ttXs +;X:j 8
 PUB LIC PO LI C.YAbrD4rO z

There are, of course,other advantagesassociatedwith exports.In the

processof entering international markets, firms learn a great deal and
nor only about the particularmarkets they are entering.There are spill-
overs related both to mnarketingknow-how and to producrion know-
how,topics to which we shall return in chapter 6. For instance,success
in the production of intermediategoods requiresproducing to standards
that are typically higher than those tat prevail within developing
economies.The contacts made in the processof exports may also be of
valuewhen the firm decides to enter related markets. It will know to
whom to turn to acquire advanced technology.
While from a socialperspective,successin exportingmay be a better
indicatorof whethera firm meritsadditionalfuindsthan successin selling
domestically,banks have (in the absenceof govemrnment prodding) typi-
callypreferreddomesticlending to foreignlending and for a simplerea-
son (Stiglitz 1993b). Banls are likely to be less informed concerning
externalmarketsthan they are concerninginternalmarkets,and from the
banks perspective,there isgreaterrisk associatedwith lendingfor export-
orientedprojectsthanwith lending for productsfor the domesticmarket.
HPAEgovernmentswere also lessheavy-handedthan some others that
haveattempted selectiveinterventions.Though they made mistakesof
judgment, they generallydid not forcedecisionson otherswho werewill-
ing to risk their own capital (The creationof the MitsubishiAutomobile
Company in) 965, rejectng m's gtidance ro refrainfrom enteringthe
production of automobiles,is a casein point.) Thisisone ofthe strengths
of decentralizeddecisionmalcingit providesinsuranceagainstmistaken
views being given too much dominance.
Runningthe Gamen TheRoleof Referees.Referees are also fundamental to
contests. Someone must enforce the rules, administer the rewards,and
discouragecheating. If one group of participants in a contest captures
it and turns it to their advantage, they will reap rewards without the
necessary effort. If some participants feel that others are excessively
favored,they will refuseto abide by the rules.
In the contestsdescribedaboveithe refereesvaried.For exporccredit,
the contest was largelyself-enforcing.Banks had the incentiveto moni-
-tor exportperformance,sinceexecution of the export orderwas essential
to repayment Banking supervisorsand customs authorities could also
.-,use their coercivepower to enforcethe rules. In the case of banldng in
Japan, the supervisoryauthoritiesof the Bank ofJapan were responsible
for running the con Tlheofficialsof mmwere the arbitersof indus-

.99
 RAC
LBE

trialpolicyinJapan,and in KoreaduringtheHCIdriveboththepolitical
leadership and theeconomicadministrationmonitoredperformance.

The.Liits and CoustaintsofCootests

Whatdetermineswhethercontestscanbe effiecively usedto promore
betterallocationand fitstergrowth?Thereappearto be two important
elementsthe relativebenefitsof coordinatedbehaviorand the institu-
tional costs of implementingnonmarket,contest-baredcompeiitive
discipline.
Wherethe benefitsof coordinatedbehaviorarenot great-for exam-
ple,in small,highlyopen economicswith goodentrepreneurialskills
and smallnontradedgoodssectors-the benefitsof coordinationof in-
vestment decisionsor sharingof informationare likelyto be small.
Internationalprices conveysufficientinfiormationfor marketsto be
efficient,and socialand privatereturnswillcoincide-This mayexplain
whyHPAEs suchasHongKongand Singaporehavenot foundit worth-
while to run contestsand establishthe types of highly structured
information-sharing mechanisms(deliberativecouncils)foundin Japan
and Korea,despite the high quality of their bureaucracies.
Evenin caseswherecoordinadionbenefitsare potentiallylarge-in
nontraded goods, externalities,or interdependent investments-
govermnents maynot be ableto enforceperformance-based coordination
rules.Taiwan,China,is an example.Despitebenefitsfiom coordinated
behaviorthat werepresumablypotentiallyaslarge asin Korea,and the
presenceofa high-quality-civil service,contestsofthe types employedin
Japanand Koreawere not used.This may have been due to Taiwan,
China's,differentindustrialstructure.With a largenumberofsmalland
medium-scale firms,in contrastto Japansand Korea'srelativelysmall
numberofmajorindustrialgroups,the institutionalcostsofattempting
to coordinateall but the simplestactivitieswereprohibitive.Instead,
Taiwan,China,addressedthe coordinatonproblemthroughpublicin-
vestmentin large,indivisibleindustries.
The absenceof effectiveconrests-exceptfor the exportcontest-r
the SoutheastAsianeconomies,Indonesia,Malaysia,and Thailand,can
alsobe interpretedasa caseof highinstitutionalcosts.Civilservicesin
theseeconomiesare norwellinsulatedfiompoliticalor economicinter- -
ferenceIt isnot surprising,then,that effirtsto coordinareeconomicac-
t-vity tirough selectiveinterventionshave been lesssucssfL The

*100-
 contests cithcr bave had poorly defined rules or have bcen bijacked by
-their pardicipants.
- These nwofactors may also explain why the reliance on contest-based
- :competition may change over time in the samneeconomy. Box 2.4 doc-
urnents the changing nature of industrial potigr in Japan, from selective
tOfunctional incentives and fiom extensive reliance on contests during
the rapid growth period tOmarket-based competiuon in the 1990s. In
-large measure this is due to perceptions among poliyrnakers that the
large benefits from coordination characteristic of the rapid grwth pe.-

4waiJ.wISYt~OFuOsDEt etdf6i-r.iu -- et"N'Vith. e eonos(ni

;- ;v,==auSmat
. luc>S as<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; 71 P1
¢2-.
l$stp;.,

%~~~~~~~~~~~~~~~~~~~-Wr -- -vS
-iissS t->.<,<-3
; V ; <gX*
jb, -%w,7
-t4i s H:
I- .
4,^_s 4~

gi
m usg£aousmfluestrXn^SDriomeg-¢w
yjmr&|w[1Sim;-W*t;S>-rt-4r4---R,-F,3
zC
litel\97osaaI-S-lsineznotXiiaN
g -;

98,' w-t
;;~-
s

i£M
'

C q
f £'e 2 J 8,'.

t- sovv
m g'4 g:(~~~~~~~~~~~~~~~~~~~~~~~~~

- )sD~~~~~~~ ~~~~ inreqmnemure,nnn;otdaterneationalreemisvaiSSnieEt

R ;uncustue,Xandh8D~silldadrmcdl
B
m&m
t n>- FSl.gr.-otXZ
^se-i>t>tzr~¢<<;Qv lRitFl-r|J-t
tua~an-dc
i;>snonaerDqns
-s LtR g2J;i-*; 9stX-lr -9Oro!E, 5fS=§tX

-g>l~-leimR^ow
f~~~~~~~~~~~1~l fZpno;MTp ZEnSLEool>qcao^adeeomnhpnhat
3;is iR 9¢+j=,-t%rin-
sQaxoamolaI4- ~~~~~~~
z+mmal DoMan;.was~~~~~
iC g>z. sF^ ,>:*
mos<Xsor;cell!loLeaElDr-u=¢Es{
.>+; e 5s- ,>:~"rn|y;E>.fi#.ana i1L t

~
$t05Ar
=>. > '>4XAS-rW3^DRbRX 2*^t ttti >,== 84tSitt 8 -u^'§.^&a> t J-8 *iX45r£W .+5=i 9s.+N
:-~ ~ ~~~~~~~~~~~~Pq. Drexnod7sw;:nEnUte,uedM
ms;n. iotlkhEdasaloyEt
- 'gS 95,>uW,-^Q* ............................... Rr".wrl,>,,oE.isi,4e,a.2*,sw044>
i'-z-'tt.Z'^:
[^sl^ttt>'}Y'-'^F4Ea
Z,oAaodnra:$rDvdcpcyTa-ytt blcne '-ru
ngnle.d>ndu l
5W.^grLtg'@eS=< @X<' 4cot.2-Xti.>C{vlsgt/4't.
ncg- =.E$*-.w'xjl2>(g,.<t,'.X>an
>i

: i Q. W ut9.
a = | 4 t , A |R ,e ; 4 g ' %M l*jE;^;s > f ->4n4:,.,- Yu*

Br
Ostbrecnsructo,lSycmDeesacm
n ea coinasoofiie
- or d- ots

: ~~~~~~~~~~~~~~~~~~~~~~io _
 A MI~RACLE

riod had diminished as-the economy grew and became more complex,
.;e xe_^,r,>,U.le.x-U..
-t i,_ and that the insttutional costs of administering contes had increased
as power shifted from government admininstmraors to enterprises. Simnilar
concerns appear to be driving recent changes in Korean industrial poli-
ces (Km and Leipziger 1993).

The searc hfrpolicyexplanationsforEast AsiA succss hasnotbeen

completey successfulLEach of the broad views of the relationship be-
tween public policy and rapid growth-neoassical, revisionist,and
markcet-friendlyadds important elements to our undcrstandingl but
none fuiilycaptures the compleity of public poiicy and rapid growth in
the Hfl'AE.We have proposed using a fuinctional approach to under-
standing the relationship between policies and gowth. In this vieew
policies contribute to t atteauunt
oh of tv hreecentml Sanctions:aci-
nulation, allocation, and produc growth-neocl. Markcr-orienitedpoi-
cies, the fundamentas, were widely used by al the HPAEsand formed
the basisfor their rapidhgrowth. More interventonist policieshave the
potenti to contribute to growth in caseswohrecthe addresseconomic
coordination problems,but to succeed they must combinethe benefits
of cooperation with competitivedisciplineby creatingcontests.
Effectiveimplementationof contestsdependedon two sets of&fictors
- almost unique to the northern HPAEs. The first is the competence,insu-
lation, and relativelackof corruptibilityof the public administrationsin
Japan and Korea.The historicaloriginsof theseeconomies'high-quality
economicadministrations and the mcchanisms that they used to main-
cain quality and limit political influence and corruption are the subject
of chapter 4. The second is the pragmatism and flexibilityof govern-
ments in the high-perfonning East Asian economies where contests
were tried and abandoned or continued dependig on their results.We
.examine this flexibilityin policy formulation in two importnt areas,
macroeconomic management and promotion of export, in the next
chapteL

102

Notes
1. Forgoodexamples,seeBalam(1982) Bhgwvati(1978),
Krueger(1978),and Litle, Scitovsky,and ScoEr(1970).
2. Moreover,some of the measurespresenord,such as
those of Itoh and Kiyono(1988) for Japan, are probably
minimumestmates insofaras theircalculationsof efieclve
ratesof protection(Em's)arebasedon triffdata, not on es-
timatesof nominallevelsof protectionderivedfrom com-
parisonsofdomestc and internationalprias. On Korea,see
PadcandWcstphal (1986)andtherefrencrscitedthercin.
On Taiwan, China,see Wadc (1990). On Japan, see Itoh
and Klyono(1988).Wade doesnot presentor dceestimates
of effeciveratesof protectionin Taiwan,China,for the pe-
riod afer 1969, while the other two presentevidenceon
ERPS forjapan and Korea Conversely,Wadedoesoffercare-
frfly documentedqualittivc evidenceofthe extentofselec-
tive intervention.Despite the frequent use of Taiwan,
China,as a modd of limitedintervention,thercareno esti-
matesof whichwe are awareof effectriveratesof protection
or dome resourcecostfor the periodafter1970.
3. See Yamawaki(1988) on the steel industry, Mutoh
(1988) on cars, Yamnazawa(1988) on textiles,Yonezawa
(1988) on shipbuilding, and Tanaka (1988) on alu-
minum refining.
4. A formal expositionof this concept is containedin
R%gand Peuri(1993).
PUBLIC. PO'LICyA i
5.Popular discussionsof the successofJapan and sev-
erl other economiesof East Asiahavestressedthe coop-
eranve relations between government and business,
betweenworkersand cmployers,and among businesses.
This phenomenon is sometimes rerred to as 'Japan,
Inc." or "Malaysia,Inc," conveyingan impressionof a
single-mindeddirecrionto economicactivity,promoting
the collectiveinuerescsof the economy.Clearly,dtis pop-
ular imge caggeractes.Like all governments, govern-
ments in EastAsialack the control t cnforcea common
set of goals.
6. The list is by no meansexhaustive,nIordoes it in-
dude examples of failed contests, of which there are
many.
7. See, for example,Kim and Leipziger(1993) for a
dicussion of the instiutional features of the Ha drive
and an evaluationof the successof the program.
8. Amsden (1989) in a seminal workloudines the na-
ture of perfiormancecritera used in Japan and Korea to
disciplineinterventionistpolicies.
9. Again,this is perhaps not quite accurate;there may
be spillovcrsfrom Iarning how to produce some com-
modity morecheaply.
103
M4acroeconomnic
StabIlt
an ort
ACROECONOMIC STABILiTYAND RAPID Ex-
port growth were two key elements in scarning
the virmous cirdes of high rates of accumula-
tion, ficent allocation, and strong p roductiv-
ity growth thiatformed the basisfor EastAsia's
suiccess. Governments achieved macroeco-
nomic stabilityby adheringto orthodox policyprescriptions-in partic-
ular, by holding budget deficits to levels that could -be prudently
fin2nced.They achievedrapid export growth through an export push-
a combination of orchodox,market-orientedmechanisms,mianylinked
toMacconomic stabilityeand omplementarypro-exportincentives.'
Policy initatives in these two areas sharedtwo features that furtier
illustrate the framework outlined in chapter 2: respect for certain
findamentals-for example,fiscalpridence and avoidanceof exchange
rate overvaluation;and quick and flexibleresponsesto changing eco-
nomic circumstances.Policiesto achievemacroeconomicstability and
rapid export growth differedover time and acrosseconomlies,partly be-
cause governmentswere responding to a changingeconomic environ-
ment and paryld because they wcre seeing efficient policy packgens
thdrugh trial and error.
We define macroeconomicstability to mean that inflation was kept
under control, intieal and external debt remained manageable, and
macroeconomic crists tat emerged were resolved quiddy, usuaiy
within a year or two. Short recessionsand policyadjustments to macro-
economuc stress somtimces sq the private scector.But these weare
transidonalperiods to newepisodes of rapid growth, quite unlike the
yeamof recessionandiuncertainty that have plagued.many other devel-
Opig coonomes,particularlyin Sub-Saharan Africaand Latin America.

1105
 Export push isa more complex phenomenon than macoeconomic st-
bility and the variety of policies used is greater. Mechanisms range from
broad, export-friendly measuressuch as avoiding an appreciated exchange
rate, which are esmployedin all the HPAEs,to government-mn export con-
tests, which have been used primarilyin Japan, the Republic of Korea, and
Taiwan, China. Universal export incentives, such as tax breaks and credir
guarantees for all exporters, have been the main instruments in Hong
Kong Indonesia, Malaysia. Singapore, and Thailand. Despite this diver-
sity of approach, as a group the HPAES are unique among developing coun-
tries in the attention they havc devoted to export promotion and the
success they have achieved. The second half of this chapter reviews the
evolution of export-push policies in each of the HPAEs.

Pragmatic
Orthodoxy
in Macroeconomic
Management

ry"IHE SHPA WERE MORE SUCCESSFULTHAN MOST OTHER DE-

veloping economies in keeping public deficits within the limits
-. the economy could absorb; as a result, ,hey were better able to
restrain inflation and manage both internal and external debL Low in-
flation and manageable debt in turn facilitated realistic exchange rates
and the avoidance of the appreciation that elsewhereundermined export
performance. When the macroeconomy did go awry, usually due to ea-
ternal shocks, govermments quiddy implernented orthodox solutions,
reducing the fiscal deficit and, when necessary,devaluaring the currency-
In contrast, many other developing economies have been less successful
in keeping deficits within bounds and have therefore had more trouble
managing inflation, debt, and exchange rates. As a result, policymakers
in these econonmieshave often had less room to maneuver when con-
fionted with a macroeconomic shock; perhaps pardy because of this,
their response has often been hesitant and ineffecrive The HPAEs
macroeconomic management superior is reflected in less severe imbal-
ances and generally lower variance in key indicators, including teal ex-
change rates, ral interest rates, and inflation.
To be sure, macroeconomic conditions varywidely among the HPAEs.
Malaysia and Singapore have long-run inflation rates comparable to
Switzerland7s,at below 4 percent, while inflation in Indonesia and Korea

xo6
 MACROEC'ON

exceedsthe South Asian averageof around 8 percent. Singaporehas con-

sistendy avoided fiscal deficits,while Malaysia'sfiscal deficit peaked at
18 percent of GDP in 1982. Hong Kong,Singapore,and Taiwan, China,
-do not borrow abroad, while Korea was the worlds fourth biggest
-debtorin 1980 and Indonesia'sforeign debt tripled during the 1980s.
Exchangerate regimeshave varied from rigidlyfixedto managed floats.
Arnid the diversiy, however,are some common themes. While some
governmentshave run substantialdeficits,none has financeda deficitin
a manner that destabilizedthe economy.The level of the deficit that is
affordable,and hence nor destabilizing,is specificto eacheconomy.It is
generallylarger the faster the rate of growth and the larger the pool of
privatesavings(both at home and abroad)relativeto privateinvestment.
In both these dimensions, the HPAEs have perforned better than many
of their developing-economycounterparts during the past thirty years.
Becauseof this, while some economieshave had higher inflation than
others, none has had to endure the very highi,debilitating inflation that
has troubled other developingeconomies.
The HPABs chose a varieryof macoeconomic policy paths becauseof
different economic conditions and preferences. All lay within the
bounds of prudent stabiliy, and wheneverthe macroeconomyappeared
to be in danger of movingout of control, swift action was taken to re-
store stability.This was true even when the source of macroeconomic
instability was policies intended to promote growth in the real economy.
For example,the heavyand chemicalindustriesdrive in Korea,to which
we shall return below,was modifiedwhen its adverseimpact on the key
indicators of macoeconomic stability,inflation, and the real exchange
rate became excessive.
In the following section we discuss how low deficits enabled the
successfulAsian economiesto keep key macroeconomicindicators rela-
invelystable, and we contrast their performance with the macro-
economic instabiliy that plagues many other developing economies.
Wethen rum brieflyto four HPAEcase smudiesthat illustte thevariey
of rapid and effectiveresponsesto macroeconomicshocks.

to Macroonomic
Adhering Fundameuias
In contast with many other developing economies, where boom-
and-bustcycles have cauised wild swings in macroeconomic indicators,
the HPAES have been remarkably successul in creatng and sustaining

107
 macroeconomic stabiliy. This has been a potent encouragement for pri-
vate savings, investment, exports, and growth, since the private sector
could count on relatively constant prices and interest rates. Here we
consider the HPAE's successfiulmanagement of four macroeconomic fin-
darnentals budger deficits. inflation, external debt, and exchange rates.
Keepi.g Budget Deficits
Manageable. The HPAES' budget deficitsare not
dramatically better as a group than other developing economies'. But
there are two distinctive things about the HPAEs. First, dtey almost al-
ways kept the deficit within the limits that could be financed without
macroeconomic destabilization. Second, these limits were higher than
in other developing economics because of the beneficial feedback fiom
other good policies.
International experience suggests that the macoeconomic conse-
quences of pubLicscaor deficits depend on how they are financed. 2 Ex-
cessive monetary financing of deficits leads to inflation; heavy
govermnent domestic borrowing drives up interest rates and crowds out
private borrwing; large external financing of the deficit leads to debt
aises The -PAFEs kept each type of financing within bounds, avoiding
the corresponding macroeconomic disease.
The means for restraining deficit financing have varied widely. Some
governments established institutional watchdogs, such as the currency
board in Singapore. Others took rule-based appmaches. such as In-
donesias balanced budget law and Thailand's exchange rate manage-
ment framework, which undl the early 1980s resembled a gold
standard. Sill others reed on the discretion of economic policymakers,
as evidenced by the macroeconomic adjustment process in Korea and
Malaysia.
Table 3.1 shows consolidated public sector deficits for the 1980s fbr
three HPAEsfor which there are good data compared with a sample of
OECDand developing econornies. As a percentage of GDP,Korea's bud-
get deficits were below even the OECD average. This helps explain why
Korea was able to keep inflation, extrnal borrowing and interest rates
within bounds. Malaysia and Thailand are more complicated. Thai-
lands budget deficits were about average for devdopi economies in
the 1980s, while Malaysiaiswere substantally bigger than average. Both
ran bigger budget deficits than such troubled economies as Argentina,
Brazil, Mexico, and the Philippines. Unlike these and other economies
that encountered difficulties, however, Malaysia and Thailand success-
fully financed their deficits.This was possible for the foilowing reasons:

io08
 MACROE.ON-O,

Table3.1 Consolidated
Public
SectorDeficits,
Selected
EastAsian
andOtherEconomies
-Averag pliRaikmo ng 4O
. . def20tcii
pcet 0>;0vj6wPdkevelpngncotrnries
'age:

*Eir .,~ ~ -.
: -
~~~~O ., -gi
^. ofDP~
D,a9r
'980-88- ; .- (
(I ehi
4stefx)
-.

Korea,Rep.of 1^89 ; ;u0 034- r

;Mn.Iaysia .* 10980' 6
.$Thai -nd e ' 5,8023

Auemge~~~ 6.39,

-viag. OECDeronomnies- : 2.82

:.
-- Oderieconomi&
~~ Az~~cndna ~~9.62-
.BIUI :4.02-

Yhilippines ~~~~~z.4.30
£. Sewtcasticr1y Rdgc -and S i -
Hcbbdl (lirrhcomi-g).

:* Frst, dierewasfeedbackfromhigbgrowh. Sincegrowthwashigher

in Malaysia and Thailand than in less successfil economies, a
higher budget deficit could be financed. Becausehigh growth in-
creases the demand for financial assets, Malaysiaand Thailand
were able to absorbhigher levelsof monetary financingwithout a
rapid risein inflation.Moreover,their rapid GDPgrowth raisedthe
levelof sustainabledomesticand externalborrowin (an economy
can borrow more for a given debt to GDP ratio when GDP is rising
rapidly). In contrast, many Latin American economiesfell into a
vicious cycleof low growth and unsustainabledeficis
a Second, there was fredbackfrom bigbfinancial savings.Savings
rates were high in Malaysiaand Thailand, and much of this sav-
ingswent into the domesticfinancialsystem(asopposed to realas-
sets or capital flight as in Latin America) due to the pro-savings
financialpoliciesdescribedin chapter 5. This firther increasedthe
demand fir money and other domestic financial assets, making
increaseddomestic financing of the deficit possiblewithout re-
sortng to inflationary financing. In Malaysia, the government's
Provident Fund mobilizeddomesuc savingsfor the goverrment's
use in noninflationary financing of the deficit

109
 ~~~~~~~LE
- - _~~~~~rIR C L E-

* Third, therewere low init debt ratios. In Thailand,the initial

levetof externaldebt to GDPwas very low, which meant that exter-
nal financingwas availablewhen needed.
Becauseof this, the HPAEs have avoided the inflation-inducingburscs
of money aceationthat afflict other developingeconomies.Figure 3.1
showsmoney creationasa ratio to GDP in Korea,Malaysia,and Thailand
and in three unstablecomparators-Argentina, Mexico,and Zairc. The
contrast is striking while money creation has been relativelyconstant
among the HPAEs,each of the comparatorsexperiencedtwo episodesof
rapid money creation when fiscal balancesdeterioratedor external fi-
nancing dried up. The impact on inflation has been equallydramatic.
maintaining Moderateto LowInflation.Unlike many Latin American
economies,the HPAEskept inflation from spinning out of control. Table
3.2 shows the low inflation in Malaysia,Singapore,Thailand, and Tai-
wan, China. Indonesia and Korea have higher inflation but sdll far
belowLatin America. Jnternationalexperiencesuggestsinflationbelow
20 percent, a levelnot breachedby any of the FPAEsduring their rapid
growth periods, can be maintained for long periods without generating
Table32 InflationRates macroeconomicinstability (Domrnbusch and Fischer 1993). Low infla-
- - . tuon is a corollaryof fiscalprudence: East Asian governmentsnever had
Aveiag;' K to rely heavily on the inflationt becauseteir deficitswerewithin fi-
*Econonre tt96iAOi nanceablelimits.
,.- HPAt. 75 i. . r East
.Several . Asian governmentshave made fornal commitments to
Hong 1' .ong 8.8 low inflation that constrained their options fr acivism. Historically,
tjQAKorei~iRe~of .; 1^2.2 Athe most important of these was the commitment to a fixed exdhange
&z^4J,l -I i , $3 rate; however,aU but Hong Kong have since abandoned this mech-
nism. Other self-imposed constraints on fiscal policies and borrowing
r c -;
v induding balancedbudget lawsand variousinstitutional checks,remain
--..
6 > :;. in place and have generaly contributed to fiscaldisciplineand low in-
- M-% lie-.............
:. - . fladion.(We discussrthesein chapter 4.) In general, HPAE govenments
.A ~ ~ .;
been strong enough to alter public spending and foreign borrow-
>^;have
2005l|wzX,
- ing asneeded,althoughin Thailandthishasbeena continuous struggle
^ .
* i s .>^7^ . (Warr and Nadhiprabha 1993).
-1.1 Reasons fbr the commitment to low inflation vary widely but are
- - -: rooted in the recenteconomichistoryof each of the HPAEs.In Indonesia
and Taiwan, China (and to a lesserextent Korea), aversionto inflation
MiK& ld^y^ l grewout-of traumatic inflationaryspiralstiat accompaniedeconomic
and politcal crises.In Malaysiaand Singapore,the successof colonial-era
i
__*>'1120Z: 0.$g11i;ES fiscalconservatsm seemsto have helped shape postcolonialpolicies.In

no.
 MACRQ ECON~

;:.~~~~~~~~~~~~~I
Figure
3.1 Revenues
fromMoneyCreation
I: asa Percentage
ofGDP.-
Examples
fromEastAsiaandOther
SelectedEconomies
(Percent)
4
4

12 - ~ I iMalayhia -a

I . ~~~~~Thafland
:: ..10 IIa- Rep.of Komea -
: -I

1970 1972 1974 1976 1978 198 2982 1984 1986 ±988

2
8
190 22 194 1976 1378f- 1980 198.2- M2fy1984 1986

NReene
ie iw moe c:ato Ji a pecng of G' -sde afined
asroo oaIl
12- - ZaireM
Arge.tnna -

10 - o t o GD-- --
6-uce World^
Ban data. j, s

-- -2_ _ __ _ _ -- _

frode ansfrgincrase goenmn spending.# A | ,

; 1970 1972 1974 1976 1978 1980 1982 is"4 1986

of GDP is ds fi ned as ra2do

Nofte.Rev:nucs from moncy crmcaonas a pucrcntagep of nominal
nge in hirgu-powerdmony monominalGDP.
O
iwcff Weridl
S Bankc
damc

: Thailand,the traclitionof responsible

fiscalpoliciesdatesfrm t:heniine-
: teenthcentury,whiena strosngcurrencyhelpedthffekingdomretainiftsin-
v dependence.In HongKong, coloniial rWfie has insffJated
the government

; from demandsfor-increasedgovemmrent spending.-

- . 0 One result.of low tO moderatei&fltion'ratespardicula-rly
welco'me
tO
:businessis stable real
interest rates.Figure3.2 shows real interest rates in

V .- . -: - : . - - - . . - - ~~~~~~~~~~
::-
 ; 4 S 'MIRACLE;

Korea,Malaysia,and Thailand,comparedwithArgentina,Ghana,and
Mexico.Aswith moneycreation,thecontrast is rcmarktble.In the East
Asiancases,lowinflationand flexiblefinancialpolicieskeptrealinterest
rateswithina narrowrange.For the comparators,the combinationof
nominalinterestratecontrolswithhighand unstableinflationwasdeadly.
wildgyrationsin realinterst ratescreatedsevereuncertaintyfor investorsm

3.2 RealIntemst
Figure fromEastAsia
RatesExamples
andOther Economies
Selected
RealInterestrate (percent)

20~ - ~ ~ I

-70 JI j
i I * C -
Rep.ofKorea
Malandu:
1

SO-80._R.~~~ Malaysia-
. ,~~~1~

1970 1972 1974 197E 1978 1980 296 1984 2988

20- 7 -

40

o4-A-.
-30

"-
-70 ' Ghana
-80 -1 j - !- '- +_ Argentina X

-- . ' -. 400
400
i7[7
!_ _ _] _=

19-70 1972 1974 1976 197 2S80 1i982 1984 1986

Nte. Realintercstratesaredefinedas diedepositrate deflatedby the consumerpuce index.

WorldBankdata.
&-wuree:

. 11-2., --.
 MACROECQ -M

Keeptmg
Extenal
Debtunder
ControL
Of the sevendevelopingHPAES, only
Indonesia,Korea,Malaysia,and Thailand havepublic or publiclyguar-
anteedforeigndebt. The governmentsof the others-Hong Kong,Sin-
gapore,and Taiwan,China-have not borrowedabroad. None of the
four with foreign debt has faced a aisis, in the sense of having to
rescheduledebt, but sharp incrases in debt have led to rapid adjust-
ment. In some economiesduringsome periods-for example,Koreain
.1980-85, Malaysiain 198248, and Indonesiasince 1987--debt-GNP
ratios have been quite high comparedwith other indebted economies
(seetable 3.3). As with fiscaldeficits,however,fivorablefeedbackfrom
other policiesenabledthe HPAE debtors to sustain higherexternaldebt
to GDP han othereconomies.High levelsof exportsmeant that foreign
exchangewasreadilyavailableto servicethe foreigndebt. Similarly,high
growthimpliedthat retrs on borrowedcapitalweresufficientto pay
the interest.
Korea'ssuccessfulhandling of a very high foreign debt illustrates
thesetrends.Beginningin the early1970s,Koreaborrowedheavilyto fi-
nance privatesectorinvestmentand build up foreignexchangereserves.
By 1984Korea'sforeigndebt was fourth largestin the world;by 1985 it
equaledmore than halfits GNP.Yetbecauseof its high export-GNP rauio
and rapid overallgrowth,Koreaneverlost creditwordliness.From 1986
the govermmentpursued an actve debr-reducton policr, drawing on

Table3.3 Intional lndebtedness

;t
4g-." -~S ' ,xen
.- Ra,. ofor deb to
>Akioftvnal

WStFrnom'rgo Pea pt 199 d Pk U' -:9

HPE -- POW -

x0 hdonesia ¢qv478r 690 s66A -455

2
~~MaI~jysxa
~865 47 1384 54

* ,: - *, t * -`3408' 47"- - 390T98-

joCr'.; 177

'A Dc';-. A..4. -,'."-' * ''

'''0
;;f-
*
r4+44;--
;-t' 7:2
0hte;nme•
~~-.-is~,t- -
,
4:
-
r.
-
-'!r-sA 38 .. -.
. 026B0
tx -1762II

"3
 AaNt 1VRAC

burgeoning international reservesgenerated by exports to make pay-

ments aheadof schedule;by 1990the debrtCNPratiowas down-to14
percent (In contrast, when Mexicofaced scvereproblemswith its cred-
itorsin 1982, it had a much /owerdebtto GNP ratio than Koreain 1984
but a much higherdebt to export ratio.)
KeepingtheExchangeRatein Une. The HPAEs avoided the severe appreci-
ation that besetSub-SaharanAfricaand Latin America(seecable3.4). In
contrast to such economies as Bolivia and Ghana, the East Asian
economies did not cling to a given nominal exchangerare (or inade-
quare rare of nominal depreciation)in the faceof continuing inflation
but depreciated when necessary,someumes quite sharply. Fiscalpru-
dence preventedthe cxcessivedemand pressuresthat appreciatedthe real
exchangerate in such economiesas C6te d'Ivoireand Nigeria.
The evolutionof exchangerate regimesin the HPAEShas been broadly
similar.Hong Kong, Malaysia,and Singaporepegged their curencies to
the Britishpound during the Breon Woods period, then floatedthem
in 1973 or 1974.The Taiwan,China, doUlarwaspegged to the U.S.dol-

Table3.4 AverageAppriation Index,1976-85

.- Indx
;t : Pe;rcent -.- aamgerankr>:
. . f. . f(highier
vau means. . ( O: si.an10
most.-
.Ecoonomy
.. mareapp recr .e ap.-p. at0a&O .r.s .

: H Kong * -
*Indonesia. 98 25
Koreo,Rep. of 1 41
Malaysia ~~~88 1
Singapore- 87; I-

Oce - . e-co
;;Slidiend
"nomiei-s

Bolivia 114' , '- ,,

,- ' "' ' ''' 0 ' '' ~~ ~~~~~~0; ,-

... .8 D- . - . 90

4a -- * ..- .2- J--";

 MACRORC

larfrom 1960 to 1973, then appredated twice,and wasfloated in 1979.

Thailand had the longestfixedrate regimc the baht was fixedto the dol-
lar during 1954-84, with a singlesmall devaluationin 1981. In 1984
the baht wasdevaluedand floatedas part of an adjustment program. In-
donesiafixedthe rupiahto the U.S. dollar from 1971 to 1978. Because
inflationwas higher in Indonesia than the United Staes, the rupiah ap-
predated vis-a-visthe dollar,necessitatingmnajordevaluationsin 1978
(51 percent), 1983 (38 percent),and 1986 (45 percent),afterwhich In-
donesiashiftedto a managedfloar.Korea,too, triedto tie its currencyto
i-thedollar but resortedto four major devaluationsbetween 1961and the
start of a managedflnat in 1980.
Most moved from long-term fixed rate regimes, to fixed-but-
: adjustablerate regimeswith occasionalsteep devaluations,to managed
.* -: floating rate regimes.Hong Kong, the single exception,reintroduceda
fixed linked) rate regime in 1983 in the face of fierce speculation
againstthe currencyprompted by politicaluncertainty.Under the man-
aged floating regines that began in the early 1980s, policymakersno
longrerset rates but attempt to influencethem at the margin, generally
to movein parallelrothe U.S.dollar Becausethe United Stateshas been
the region'smajor export market, maintaining a stable and at times
slighdyundervalued exchangerate vis-a-visthe U.S. dollar has assisted
exporters.
The HPAES'successat maintaining stablcexchangerates isapparent in
figure3.3, which contraststhe remarkablestabilityof real exchangerates
since 1970 in Korea,Malaysia,and Thailand with rhe severeexhnge
rate instabilityin Argentina, Peru, and Sri Lanka.Argentinarepeatedly
attempted to use the exchangerate as a nominalanchor againsthigh in-
flation (fr example, in 1973-74 and 1980-81). But failure to keep
other macroeconomicfundamentalsin line led to the collapseof the real
exchangerate and sharp real devaluations.In contrast, the East Asian
economies'pragmatic macroecononic management eambledthem to
avoidswingsof the real exchangerate, evenin the faceof major external
shocks (seebox 3.1).

RespondingQuickly
to Macroeconomic
Shocks
--The HPAEs rapid responseto macroeconomicshockshas been greatly
. dflitated by two characreristics First, by limiting distortions and
tighdtysupervsingbanks, governmentsreduced the spilloverfrom the

"5
 Fiupre33 Exmplesof RealExchangeRatevariabiryy
in EastAsia
andOtherSelectedEconomies
Realexchangerate (percent)

300 1 Rep oKorea-

25D 0-

200

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

0 I ; I I' : I i .

1970 1972 1974 1976 1978 I980 1982 1984 1986 1988

300-- -- Arg ntarM

300- ------- - -- e-i.o

mo 1 t _ 7 I X A Tr. . Z.

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 xchangerce: I980=100 cal depreciaton is down.
Sot t:World Bankdam.

real sector into the financialsector that in other economiesexacerbated

fiscalwoes. Second, flexdblelabor and capital markets enabled the rcal
sector to react quiddy to governmentinitiatives,seting off new growth
cyclestha easedthe recessionaryimpaa of stabilizationmeasures.Thus
the HPAEs recoveredquickly from rnacroeconomnic shocks.Box 3.2 de-
scribestwo more typicaldevelopingeconomies,Cc-e d'Ivoireand Mex-

Wia
 ROECNO:
MACRAC

gik"I!xL~R -I U

. .~~~~~~~~~~~~~~A

J5tczur &q i"-, '-#- c7 ' ?YvsK:r~~

dafiii " g'%k c yx;rvn m; rm'n
* dXc~~~~~6aecord4icsi.&acrh. aia hcdo

o-n'
uq`c1ie.aRAsianiconj
x1B bbiisl'btt`~i~

4
*1 ~~~~~~~~
:t- 10 t "'1"t.>-iJ 0 r*T.

.; K>e,;efiile(nu t io*eeldvomgm Uroole'Eithti 'a,d jt, -

wini-omm CZ>-

- *-;-- tgl g$> >0 b *t> t-;gTl, * -F'

b&AkfrAhe cnsxilngwod4sing

thinaUe.Weinri or is a mn n boxa3& anca frgr ttSF

ath-:

.m4e for caes l- E' Nh, ; + <succdi,

in whidi*&e.mtvt'w>.ttiS.................-i>
athe HPAESinrsonde -,=- j*fd . variety of
-i to.ea
uya .;

-13eeoiOmlecs rheismdy4bun&m ihakd&Rod&M->

-lci *,te>;t,r<,.,
that ,,3ii'-g\;-,..)-sVl,<w;-;='i---
fMaulettaocltaket'imEely to <6_. f $X
action in rs,'l;lponsewr; uhshcsanc-v ,^f-
_ j, 2,;
=tr * g *_

sjeoomamr awer"'ou
£t'F"'vhe '.-s - -wlWtora-- -..
term.asOft2dcbrnzoi.EhcEastArsiaccn mi*>>icshavebeoinaica
wstas unfa-
voabe
.,as cfa-vscin ohocsf dvlo96u-t.aggl -Bld-ow we exa--
ico,din flledto tak toiml ac5tidoninepos to uc shockl
i-tand
s *su
sods:Indonesia,
maCrOe:oneomic
SomxenIobsrvr hva rge ta ore,h, Asia has bee +lucky7
Ea'noooS>ast
Siengapoenloand rather
Thand..

vorale sthose ficing othernadevl,topingzeonomis.Below e exam-4

* n ou ae in whc th HPE respode sucesll $
to aaieyo
macroeconomicshomcks: Kra,Singi;Dapore,andThailand.
Indonesiai,snc
0 %-"7J
2.eex14f-;RS,.-~*S.-4,~-tk;k.te-fSKe-t.w,WNit..hej
 ''-'~~~~~~~~-.'i'-:a~~~~~~n~;' :1

. ~ ~~~~ ~ ~. ~ ~ ~ ~ ~ ~0' - ~ ~ ~ 4

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

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

X y 5tf g t 1 XR2SOZSFU.uas=g ~~~~~

OilPries hi.sIndonesias.
: > w>W~~~~~~~~~ein;iug 1986 Inonsi *fuced
From 1982cto§itls'l
- tMcreaeq-puD^ltMndg- rarpily worsenuing ters oqf traecuedrmriyb decli%ni;ngi
t t~~~~~~~~rcs Strtn in_ 1983,4> jgovernment,
thej! r esponde.di with avremarkabl
.. . Uestlrsrlvs~~omplrehelnsieanF-dvsucesuladjustmenstpro2gDram.Idealuedthe Din-
.~~~~~~~~~~il
in 198 awntd an cur exedtus
1986>= minl by recedln

.~~~~~~~~~Ddnn adarLsIndonesa. OF
aalcs inilntacil90 rom.ut1982 to 196,Iadronosi faed
oDmff
. remmentSemMewolskomeffmeasuremteachieveen 98tillPustated
fosqrom i bthe ollow
>erStalfizaonx DastalpriXeoa.Srng n18,tegvranenTaa-nit lreOspondkled.u with a.eaml
*-. . -~~~~~~~n fo
;Littlean
calculoytaEnViamSoaetion othes orthaocomng). The exces of
"8~~~~~~~~~E
~~~~~~~~~iprso goos adnnactoservices overannon-oltexots f*i:3ls
frormta¢b
SS robinssnaett)eltSmitocapiml-intcnsivepoDjeC OIhened
t eue urn account deficit

.~ ~ ~~ ~ ~~~~~1 ca'lto (fo Li' anc otes foromn) h xeso

- ; of si
~~~~~~~~~imports- F ove 192no
scdvices exportsi facedfi
A96
-- . td
rapidlyw~~~~~~~9orsligtrso lslpiaiyb elnn i
: rosSarigi 93 hegvrmn epoddw8 l
 MACROECO

15 percent of GDPi 44 to 7 percent in 1988.This shiftof 8 percent

OfGDPmeasuresthe extentby whichabsorptionhad to be reduced-the
ects--eof increasedprincipalrepaymentsof long-termdebt, higher in-
terest payments,and reducedincome from oil and gas imports. This
massive resource shift, also associatedwith trade liberalization,was
brought about withoutincreasedinflation.
*-.~ - Adjustment
M was not painless.Growth initiallyfell 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 adjustmentprocess,which actuallyac-
celeratedin 1986. In the later half of the 1980s,perseverancepaid off in
a boom of manufacturedexportsthat pushed the exportgrowth rate to
25 percent a year for the period 1985-91. Since then Indonesia has
chalkedup averageannualgrowthof about 7.1 percent.
Respodimgto M economic CrisisinKorea.In 1979, Koreaencountered
a varietyofproblemsthat threatened to undercut the 1970s'impressive
gro
wthu Risingoil pricesbattered Korea'stermnsof trade, the world re-
cessiondampenedexporcdemand, and high interestrates boosted debt
servicecosts. Koreawas not unique in these troubles,of course; these
werethe samewoesthat led to debt cises in many economnies outsideof
East Asia.
Koreahad plenty of specificproblemsbesides.Realappreciationdur-
ing the 1974-79 fixedexchangerate regine had made exportslesscom-
petitive,the rice crop had filed, and the assassinationof PresidentPark
* ~-Chung-Hee had exacerbatedpolitical uncertainty (Coffins and Park
1989).
Koreahad a few structuraladvantages,however Unlike economies
thatquidyfel into debt crises,Koreawas not running largeandgrow-
ing budget deficits.And while privatesavingshad dropped due to de-
cdining output and incomes, investment remained high. More
important, Korea responded quicldy to its troubles with an aggressive
January 1980 stabilizaton packagebackedby UmFstandby credits.The
governmentended the fixedexchangrerate regime,devaluedthe won by
17 percent, and dghtened monetaryand fiscalpolicy
Things got worse beforethry got better In 1980, output fell 5 per-
Centr,inHfationsoaredto more than 25 percent, and the currentaccount
deficitapproached9 percentof GDP.The strong medicinewaspardy re-
sponsible for the economys worsning symptoms: the devaluation
spurred inflation, while tighter aggregatedemand polcies excerbated

:i9g
 * drop in output. Even so, and despite politicaloutraiesoverKords
~~~~~the
risingforeigndebt,it continued foreignborrowingthroughout the cri-
sis, thus maintaning-high investmentlevels.
'Withintwo years,the medicinehad begun to rake the desiredeffect.
* ~~In1982inflationdroppedto 7 percentand in 1983 to 3.4 percent.The
:!current account defict fei to 2 percent of rot in 1983. Overall, the
gov:edr nes proumptand effectiveresponseto a potential crisis
stresngthenedthe conomy,preparingit for rapidgrowthin the 1980s.
Mjusting to nin
Sho i Thailand. Thaland onlypartaly adlustedto
the first oil shock and in the late 1970s engagedin a mild privateand
publicspendingboom then
Thmcmeh secondoilshockand the risein
worldinterestrates.By 1980-81,thetcornlidatedpublicsectordeficit
was 7 percent of GDP,nearlyhalf of which-wasthe deficitof nonfinan-
dal publicenterprises.
Thecurrentaccountdeficitwasalsoabout7 per-
ctentBecauseforeignbo rro inghad been moderate-the debt-GDP
radtowasonly35 percentin 1982-Thailand wasnot &cinga dbt crn-
sis and continuedto borrow.Evenso, the new governmentthatrook
- : wover in 1980 perceivedthat macroeconomic adjustmentwas needed.
Monetry policyoptionswere lifmitedbythe fixed xchangerateandthe
relativelyopencapitalmarke. The governmenttherfciretookthealter-
nativepath,fiscl contraction,movig gradualybut consistentydudrg
the next severalyearsto cut expendituresand boost revenues
Policymakerssteeplycur deficits of the nonfinancialpublic enter-
prises,then gradualyreducedthe centralgovernmentdefici. As a rt-
suit, the consolidatedgovemrnentdeficitdecined from 8 percentof
GDPin 198142 to 1.6 percentin 1986-87,whenadjustmentwas es-
sentiallycomplete.Meanwhile,steepertaxrates and toughercolection
efTortsboostedcepntalgavernent ax revenuef'rom13 percentof GD?
in 1982 to 16 percentin 1988.The amdjustment processwas facilitated
by a 1984 devaluation.
Thai gradualismwas possiblebecauseforeign borowing had been
mode:ra and the economydid not yet fatcea crsis.Butgrdualismwas
not h-esitancnconservativefiscalpolices were consistentand weresus-
tained into the late 1980samid an exporctand foreigninvestmentboom.
Sin 19878Tailand has becenaccumulatingfoeign excamngere-
;- . : - -- ~~~~ecnomy
serves, and the governm eficlpiiswreosset
hesanqconev has werosu
fan
ent rcgularlyrecordeda fiscalsurplus.
- ThaCocing PolicyFailwesh Snpor. Singaporeescapedthe 1980-81
world recessionwathsecacelyadip in its robust growth.But in 1985the
-economyencountereda sudden and severerecessiongrowthfel hom

120.
 MACKtE~9

in 1984 to 41.6percentin 1985. Reasonsfor thi~sunusual

.83 perce-nt:
episode included governmentencouragementof high wages between
1979 and 1981, which was intended to speed the decline of labor-
intensiveproduction but backfiredby erodingcompetitiveness,and the
apprecation of the Singaporedollar'strade-'weightedexchangerare.
These were eamcerbated by a sharp declinein public investmentL
due to
the simultaneouscompletion of severalmajor projects.Privateinvest-
-: N
ment, whichhad been declininggraduallyfor severalyears,alsodropped
in 1985, apparentlybecauseof flling externaldemnandresultingfrom
Singaprs decining comperitiveness
Recognizingthe problems,the government devisedan integrated
policypackagethat reversedthe high wagepolicyto restoreSingapore's
competitivenes and stimulated domestic demand. The govermecnt
also cut the emnployerscompulsorycontribution to the Central ProvW
dent Fund from 25 percent of wagesto 10 percent, reducedcorporate
taxes,and introduced accelerateddepredation.Income taxeswere cut
andtdevelopmentexpenditurewas boostedby 21Iperrcent.Thesepolicy
repnsesand a fortuitousdepred'ationin the Singaporedollar,due to
the globaldepredation of the U.S. dollar,contributedto a rapid reco-
My-cIn 1986growth recoveredto 2 percent,and by 1987it had reached
9.5 percent, a levelmaintainedthrough 1990.

HowMacreconic StabiriyContnb-- to Gmowth

It cannot be a coincidencethat all of theseseveneconomieshavehad

exceptionallyhigh growthbyworld standards,and all havehad unusual
successmanagingtheir maa.oeconomnics over the long run. Al but In-
. alsocut th
contibutio th CenrMProvi
emploers' tOcmpulsoy
donesiaand Koreahave also been long-termlow-inflationeconomies,
while Indoncsia and Koreafal into the moderatelylow-inflationate-
goesvLowor modate inflationfor lbongperiods providesa fivorable
envirorinmentfor growth.
Thaereare four main reasonswhy high inflationis likelyto be adverse
Forgrowth.and why rhe conservatvepolicaesfollowedin the HPAs are
likelyto have been favorablefor growth. Fsrl economiesthat are not
fuly adjustedto a givenrate ofinflationusuallysufferfromrelativeprice
distortionspcausedbyrinflation Nominal interm rates are often con-
aolled,
and hence real interest rates become negativeand volatle; de-
preciatons of the exchangerate lag behind inflation, so that rel
appreciatonsand exche rate variability
GompeiitiveneSS domestic
rsult.
andsumulated As wc
demand. the
satw,
The grelin-
governmenr
 7---7-
Pen MAL

terestrateand the real exchangeratewereunusuallystablein most

HPAEs,compared withotherdevelopingeconomies, whichmeantthat
thesecrucialrelativepricesweremoree&fctive
at guidingresourceallo-
cation(a topic to whichwe shallreturn in chapter6).
Second,realtax collectionslaginflation,becausecollectionsarebased
on nominalincomesof an earlieryear(the Tanzicffect),andpublic util-
ity pricesare not raisedin line with inflation.For both reasonsthe fiscal
problemis intensifiedby inflation,and publicsavingsmay be reduced.
Publicsavingshavebeenan importantcomponentof the unusuallyhigh
levelsof total savings in the HPAEs,comparedwith other low- and
middle-incomeeconomies(seechapter5).
Third, high inflation is inevitablyunstble. There is uncertainty
aboutfutureratesofinflation,and thisboth reducesthe efficiencyof in-
vestmentand discourage it. If the inflationrate were high and stable,
therewould,in theory,be no problemon thisaccount.But in realitythe
higher the inflation,the morelikelyare measuresto reduceit. Thesein
turn have a contractonary impact on privateinvestmentin the short
run. Finally,highand variableinflationimposessubstantialinstitutional
costsin manyeconomies.Duringperiodsof pricevolatility,scarceman-
agerialresourcesin the economyare drawninto financialas opposedto
real sector management, aswas the casein Latin America in the 1980s.
Howirportant to growthis maaoeconomicstbility?Cross-economy,
econometricstudiesgenerallyfind that higherinflationreducesgrowth
(Fischer1993).But the relationshipis not robustfor small changes;an
economywith a slightlylower-than-average inflationrare for a longer
perod doesnot necessarilyhavea somewhathighergrowthrate.For ex-
ample,Thailand'slong-termgrowth rate is well belowthat of Korea,
eventhoughKoreisaverageinflationrate has beenhigheLFurthermow,
thereare manyeconomies that have,at varioustimes,had lowinflation
rates and low growth.The most importantcaseis that of India.There
are alsocasesof high inflationassociatedwith high growth.The most
importantexampleis Brazilfrom 1968to 1980.Morerecenty,Turke/s
growthratefrom 1981 to 1990averaged5.4 percent,whileits inflation
rate averaged46 percent (ittle and others,forthcoming).Lowto mod-
crateinflationmay be a necessaryconditionfor growth,but it dearly is
not sufficent.
Relativelycautiousfiscaland fbreignborrwing policiesmeant that
seriousdebt criseswere avoided,which reducedthe stopgo pattern of
crisisand responsethat characterizedmanydevelopingcconomiesin the

In
 MACROACCjj_~.

1980s.Asnotedabove,therehavebeenproblems,notablyin Koreaand
Malaysia,but theyweredealt with swiftly.The Indonesianand, even
more, Thai reactionswere more gradualbut also effective.Avoiding
crises and the need for reschedulingmeant that creditworthinesswas
maintained, and it was easierto borrow in the shorc term and to avoid
very deep cuts, especiallyin invesmenL Suddenreductionsin aggregate
demandand in investmentcompelledby debt crisesweremajorcauses
of the sharp declines in growth rates in many of the heavily indebted
economiesof LatinAmericaand elsewhere.
Although macroeconomicstabilityand prompt responsesto macro-
economicshockswere not the whole story of the HPAES' success, these
factors created a basis firomwhich policiesintended to affect the real
economy-the supply side-could be launched in an environmentof
stable realinterest and exchangerates.We now tum to the most broadly
shared of thesesupplyside policy initiatives:creatingan export push.

Creating
anExport
Push

GOVERMENTS OF THE HPAEs HAVE ENCOURAGED

exports by fostering a supportive macroeconomicclimateand
by providing suitable microeconomic incentives. Macro-
economic stability helped exports by easing the liberaization of re-
straints on tade and by facilitating realistic and, in some cases,
undervaluedexchangerates.We discussthese issuesbelow.Few general-
izations can be made about microeconomicincentives,however,since
these economies differed in the degree and selectivity of promotion, and
each economypassedthrough severalstages.Becauseof tis and because
ecport promotion has played such an important role in the East Asian
mirade, we condude this chapter with a survey of the evolution of
export-push polices in all seven developingHPAEs(these arc summa-
rizedin table 3.5). Appendix 3.1 givcs brief policy histories of each
economyin the fiormof timelines.

MacmecononicStabilityFacilitated omre
Open Economies

The successof the HPAESrests pardy on what they have done and
partly on what they have not done. One thing theyhave not done is to

n3
 - IRA~~~~CLUE

Table3.5 Phases in HPAETmade

Regimes

.~Nad&z anlimad 'War and construe- Mket-led develop- Land reformnanid Natural-resouirce-
guidcddevclpmcnc,
do'a,1950-O0 met .150-70 reconstruction, ae prs
194846' ~~~~~~~ ~ ~~~~~~1949-.521900

Impori-substinicing
industrialirzatiLon-,
1953-57

J AF:K~~~~~~~~~~~Eport
promotion
1958-72 -

ZTilting.twr

~ incur
73 1967
tiC .;;J*-Export promotion. Fvoring import sb
tSelectiv6inter:vendion>-2andirnPortssubst-~ Indi'srial. con-solida- sio 197180:
Omlandcommoitydioi4hii:heavy2nd . no stontut7-o5
-boomIn,974-81 cciakuwe growth,1973-080

* ~~Functional
abdlbcraizaon,
AdjstmntEto aex- 1809 .- .': ~~~~~~~Refborm
ad epr
~ ~ ~ ~85~ ~
shodci` ~~~~~~~~~~~~icntvs -1980-
Hihtecnlgn present
Derc~~~u1adon:and
oriU.nrarion,T Adjw~~~mciz-andh- moeniztion,
19iiod-ors
* outward d .. cAIimno,
A_ - - ,11986-9 CMLEMUOn,

present.

impose general import restrictions to redress balance of payments

deficits. Hong Kong, MaIayrsi-a, Singapore, Thailand. and Taiwan,
China, had no cause to-imposesuch restrictions,since their current ac-
count balancesneve-rfaced serious long-term deficits. ndonesa and
Korea. with more troubled current:accounts, might have imposed re-
strictions but dlidnor.
The benefit of avoiding import restrictosi iely understood.
Such measuresstrve enterprisesof imported inputs and are particularly
hard on exporters,who tend to depend heaviy on i.mportedmaterials

124
 MACRO-EC;~~

and capitalgoods. But easingimport restrictionswithout accommiodat-

ing maacrconomic or exchangerate policycan womsencurrent account
problems,as consumersand firms buy imported goods not previously
available.The trade regimesof the eight H-PAEs havedifferedwidely,but
each has gradually liberalizedwithout incurring a serious current ac-
count deficit. Often trade liberalizationhas been part of a policypack-
age chat included devaluation(usuallyto cushion the blow to import
-substitution industries), exchangerate unification, fiscal reform, and
foreign aid or concessionalloanisto offset a tempomarilydeteriorating
current account.
The dose link betweensuccessfuilmacroeconomicpoliciesand trade
liberalizationcan be seen in the experiencesof Indonesia, Korea,and
Taiwan,China. Taiwan,China, took its firstbigstep toward trade liber-
alizationin 1958 with a policy packagechatincluded a 25 percent de-
valuation,a unifiedexchangerate, exportincentives,and the widespread
remnoval of quantitative import restrictions.In Korea,devaluationsand
tradeliberalizationhavegone hand in hand, beginningwith a dtrastkc de-
valuationin 1961 (Collinsand Park 1989). Indonesiabegantrade liber-
alization in the late 1960s, together with the stabilizationefforts chat
marked the start of the Suharto era. The new government abolished
comprehensiveimport licensingand in 1970 united and devaluedthe
exchangerate. After an inward-lookingperiod in the mid.-1970s,trade
liberalizationresumed in the 1980s and has since proceedecdtogether
with an export boom.

Exchange
RatesandExports
As Wehaveseen,severalHiPAEgovernmentrsused exchangerate poli-
cies to off-setthe adverseimpact of trade liberalizationson producersof
import substitutes.A few went beyondthis objctive, however,and used
* deliberatelyundervalued exchangerates to assistexporters.In thesein-
stinces, exchangerate policy and the fiscaland monetary tools to carry
it out becamea part of an overal export-pushstrteg Taiwan, China,
avis
the most notable exampleof this, but Koreaand Indonesiaalsodelib-
ecratelyundrvalued their currenciesto boost exports.We brieflydiscuss
aglthree belowo

The very large current surplusesthat Taiwan,China, ran

in the 1980s, especally from 1984 to 1987 (when the surpluses

125
 CL E

averaged16 percent of GDP,with an etraordinary peak of 20 per-

cent in 1986), resultedfrom governmenteffortsto managethe ex-
change rates What would have happened if the New Taiwan,
China, dollar had been allowedto appreciatemorerapidly?Exports
would havebecomelesscompetitive,reducedexportgrowthwould
probablyhavehad a deflationaryeffect,and this in rum could have
reducedsavings.Alternatively-and more realistically-thc poten-
tial deflationaryeffectwould have been offset by increasedpublic
expenditure,leading to a budget deficir.
* Koreaused exchangerate protection from 1986 to 1989 when it
ran a current account surplus (which peaked at 8 percent of GDP
in 1988). A desire to protect the export industrieswas certainlya
fictor in Korean exdcangerate policy,but tie main concernwas to
reducethe debt ratio and build up reservesto avoid repeatingthe
closebrush with a foreigndebt crisisin 1984-85.
m The Indonesiandevaluationof 1978 can be classificdlaelyas an-
ticipatoryexchangerate protection.No immediatebalanceof pay-
ments problem existed, as the adverseeffect of the previousreal
appreciationof the rupiah on non-oil exports had been offset by
the rise in the quantity and value of oil and natural gas exporcs.
Rather, the aim of devaluationwas to encouragenon-oil exports
and slow import growth. By 1982, when the balanceof payments
had sharplydeteriorated,the wisdomof the 1978 devaluationwas
clear.
One can see a fairlyclear relationshipbetweendevaluationsand ex-
port growth in the 1980s.Taiwan,China's,real exchangerate relativeto
the United Statesdepreciatedsharplyfrom 1980 to 1985 (the period of
generaldollar appreciation),and the result can be dearly scen in an ex-
port boomto the United States.The Koreanreal exchangerate waskept
fairlystble by numerous nominal devaluationsduring the period of the
first great Koreanexport boom, 1963-72. The effectson exportsof real
devaluation from 1982 to 1988 can also be seen very dearly. The
Malaysianreal exchangerate steadilydepreciatedfrom 1987 to 1990,
and this must have been one factor in rapid export growth-.
The effectsof Thailand's real devaluationin 1984-88 on export
growthwerequite dramatic,eventhough therewerealsoexternalfiators
explainingexportgrowth, notably the availabilityof capitl fiom Japan
and Taiwan, China, for developingexport indlustries. From 1986 to
1989 the dollarvalueof exportsrose 12 percent.All three of Indonesims

126
 MAC ROE CO NO M ICr

devaluationshad dear effectson exporcsof manufiactures,

and this was
particularlytrueof the 1986devaluation.From 1986to 1988the vol-
Iumeof exportsof manufacturesrose80 percent(froma Fairlylowbase),
and growthcontinuedright through 1990.
We now turn to our surveyof the evolutionof aporc-pushstrategies
in the sevendevelopingHPAEs.

'KoreaPushes
Exports
andIndustrial
* Korea'sdevelopmenthas passedthroughthreesmgesand is currently
in the midst of a fourth Unlikeseveralof the largereconomiesin East
X Asia,which evolvedfrom protectionist,inward-lookingtrade regimes
toward relativelyopen economies,
Koreadid not have a sufficiendylarge
:: domesticpopulationto contemplatea strategyodtcr thanexport-ledde-
- velopment.Its performancemay be describedas forcedgrowdt,because
it did not stem from exploitationof natural resources,an influxof labor,
flowsof speculativecapital,or the adoptionof new meansof produc-
tion.Rather,growthresultedfrom a systematicprogramof importing
raw materialsand intermediategoodsfor processingand exportwith
addedvalue.
War, P oanstrclion,andLand Refonn (1956O-01.Koreahad a relaively
well-developed infrascructreat the end of WorldWarII, but the parti-
tioning of the peninsulaby U.S. and Sovietforcesseveredeconomic
linksbetweenthe heavilyMidustrial north and agriculturalsouth. The
war tooka heavytoll,taking1.5 millionlivesand destroyingtwo-thirds
of the south'sindustrialcapaciy.With a poor naturalresourcebaseand
'oneofthe world'shighestpopulationdensities,thesouthwasalmosten-
tirelydependenton U.S.aidafterthe war Forallits devastation,thewar
mayhavehelpedto prepareKoreafor an industrialtakeoffbyloosening
a rigidsocialstructure,openingthe wayfor fundamentalchangesin
oudook. Whiledevelopmenreffortsin tie 1950sinduded severalfalse
srans, progresswasmadein reconstruction, includingthe restorationof
transportaton and communication networks. The government also
completeda land reformprogramthat had stalledbeforethe war.
Export 1961-73).
Takeoff UnderPresidentPark,aggressive
promotion
of exports was combined with-classicimport protection at home. Ko-
reanpolicyakers nmainLiained dose control over trade, exchange, and
- 1financialpolicy,as wel as aspeccsof industrial decisionmaking.
In con-
trastto othercontrolled economies,theyusedtheseinstrumentsto pur-

127,
 L~~~~~~A4 41 ,. C, E-

sue the primary objectiveof export growth. The trade regimewasbiased

in fiavorof exports as a whole but essentiallyneutral with respectto dte
composition of exports.
Even so, the first instruments of export promotion were highly dis-
cretionary.Exporters were supported with multiple exchangerates, di-
rect cash payments(seetable 3.6), permissionto retain foreign exchange
earningsto import restrictedcommodities,and permissionto borrowin
foreign currcncies.This system not only avoided hampering exporters
with restrictionson capital and intermediate inputs for their own use,
but it also gave access to he firavombleexchange rates determined by
scarcity rents in heavily protected domestic markets. Even as discre-
tionary incentives were gradually replaced by more automatic instru-
ments, exporters received significant exemptions from import controls.

Table3.6 Effective
Exchange for Imports
Rates andExports,
Republicof Korea
' ' ' '' ' ' S' '''' ^'#ren'ccs due to:
'"~ ' - -~~' '

Ecnge-
p'remiums phs: ':D'treet :.ax ax a n.'
; 4 f- ~d6r tcashX ji. materb--
-,--'-,"-, of2&i<E&YZY.5.
suGbsud - - '.' u a equwa..
inSu-
Yen . wE rceMt) (percent)
-1958 1.79 97 3 ~ '0
1959
.19 0
-1961
1.962 10.7173 23
`19638 7 29 .
1964 114 '~~ ~~29: 21 .5K
19 . .4 -225

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

:.~~~~~~~~~16 113 - -69

; ,- ....J1.,--:-:5

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

xz8
 MACROECON

Tariff exemptions were given to indirect as well as direct exporters, and

generous wastage allowances on imported intermediates allowed some
resale (table 3.6). These enabled exporters to avoid distortions from pro-
tection and in some cases to benefit from the protection of the domestic
marker.
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.
Heavyand Chemical Indostries Drive11973-791. 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 all
its 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, and
machinery-to receivesupport, induding tax incentives, detailed engi-
neering, subsidized public services, and preferential financing. Thcegov-
ernment chose the fir dutee sectors to enhance self-sufficiency in
industrial raw materials, while the latter three were meant to be
groomed 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, projects
tended to be forward-looking--only current technology was imported,
and U.SI-trained. Korean scientists and enginaeerswere 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 already
high, and the exchange rate had apprecated; capacity utiliation in the
:HC: seccor was low, and exports were faltering. The government
switchled course.
FunctionlIncentives 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
 -* L E' . fr:tU.C

:. ~ -%. -Cj P1...L ,,mdopmentBaEcL:

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

,-cmpnc ,,-,-
033

B EW4 mt.'w"gWW $8!stft47

ha&obtide3".*>¼mt'i l ;!l:
.PM.wodsbMwoucra'ovcnsaSo l.F.Sdcparunrnjstwe4*
&lSCca4O :ioo
apancsg-'co
ramadcaelz cn
t S-It?' __~~~~~~-w ,

. * .5
i.-1OI6W4lKLQJon s-Cot r 2t m.~ oi jwtiglrttm&ask*Glitbaej&fl
-EsSw-S^h

-
Moi~~~~~~~~~_S
an a~-
___
a
.Stot F,ritiEgt,,fE,,.,^WK drafited in 97 recogiJ,zed of the>>z=>
thait the comp_9;:lexity economy ws
E rcouldnororZfoeveceedingwmatheurg;ovrnments anagement capEacitie.h plDan's emphsi
indictive planningand a greatr role
. |~~~~~~~~~o *themaret srss4-wtrg-|gs#Qz;btam
s entually

SadiNosindustres,
R$kZsEasl@rteE;s f echolgyand
scupport for h develokmopmcent the promo-

don
competitiont.ti
of in it apprs:oa
poSosoulcylis
ME-Z><dustM-iDsSoxrxSStesrtrialed o fnaLncasetor rneform.The overn-si

oh bndcartiuepn eateinbaleofo
glarlyaact susete mand wlesacdcu in
dustries. ith remnantsoffnancsr liaeanr port-oitepizaie Ko grais
pay-

mfedhaien
mdn*e nt1has
been
197 opngithd thersdeat
incgnwized th reideopei theompexcesse
of theonesomy
of the ex th 1970m
ws and
wasand
industries, sidpdetfnna h devsector.en of t-chnology and the promo-
-1309.y, ssovemmenn1.officms Wwortionhas bendciv parclarlyn andv in brailorsrole sunrt and oveaseveragedlly
ofcompctiton.
; . ~~~~~~~~~~utrieslathedita remants ofpsfinancies are ipart-libtepraiceio progreamis.p
-- ~~~ ~ ~~~~~m *Infoprointervention
ntrvention apolicies79and
for pror sicp9 olcies foruedo
ad forthe
ther failucure
iue tof oest ablishgandkme
edbishresse
: - - ~~~~~~industrident
fiupportfoste dvlpetofrc.lg adtepoo
 MACRO: EcoNP.

of Export
TheEvolition PushinTaiwan,China

Developmentpolicyin Taiwan,China,has consistedof fivestagesin

which the governmenthas implementedcomprehensivebut changing
poliqcpackages.Throughout,low inflationand macroeconomic stabil-
ityhavebeen a foundationfor growth-enhancingpolicyinitiatives,and
sincethe late 1950s-exportgrowthhasalso beena fundamentalgoal.
LandReformandReconstruction
(1949-52). When the Taiwanese author-
ities took over in 1949, one of their first initiatives was an ambitious
land reform program. The program fosteredsocialand politicalstability
and increased agricultural production. Greater agricultural output pro-
videdraw materialsfor exports and earned foreign exchangeto fund im-
ports of machinery,equipment, and industrial raw marerials.These in
turn helpedto make possiblesubsequent cxport-led rapid development
Impat-Substitg Industializaoun(1953-571.During the second stage,
the government attempted to developindustry as the basefor economic
self-sufficiencyThe government invested heavily in infrastructure, ex-
panding transportation and power necworksbuilt by the Japanese;U.S.
aiidwas an important sourceoffiriance,funding 49 percent of public in-
vestment in infiastructure. Excensivequantitative restrictions and high
ta-riffratesshielded domestic consumergoods from foreign competition.
To take advantageof abundant labor, the government subsidizedsome
light industries,particularlytextiles.Consumer goods industries such as
textiles,apparel, wood andlearher products, and bicyclesdevelopedvery
rapidly. By the end of the 1950s, industrial production had doubled.
-'However,the costs ofimporc substitution increasedover time. By stim-
ulating the import of capital and intermediate goods while penalizing
exports, imporc substitution contributed to a growing trade deficit, fi-
nancedlargely by U.S. aid. As the small domestic marketbecamesatu-
rated, overallgrowth declined, fiom 9 percent in the early 1950s to 6.5
percent in the mid-1950s.
Expot Prmotion 11958-72). Anticipating the termination of U.S. aid
2andhence a need to obtain foreign exchange,the govermnent shifted to
a policy of outward orientation and export promotion. Staring in 1958
it adopted a seriesof measuresaimed at promoting exports and foreign
mnvestment. Amultiple exchangerate systemwas replacedwith a unitary
rate, and appreciation was avoided. Tariffs and import controls were
-graduallyreduced, especiallyfor inputs to export. In addition, the Bank
of Taiwan, China, offered low-interestloans to exporters.The govern-

131'
 R~ ~ ~~TWL

ment alsohired the StanfordResearchInstitute to idendifjypromisingin-

dustries for export promotion and development. On the basis of Tai-
wan, Chinas, comparative advantage in low-cost labor and existing
technical capabilities, the institure chose plastics, synthetic fibers, and
electronic components. Other indusrries subsequently promoted in-
duded apparel, consumer eectronics, homc appliances, watches, and
docks. Direct foreign investment(Dn) playeda catalyticrole during this
period and replaced U.S. aid as the main source of foreign capitl. Al-
though DFI was only 6 percent of gross capitalformation in the 1960s,
nearly 80 percent of it went into manufaacuring.More important, DFI
facilitatedtechnologyand skill transfers,leading to much improvement
in quality and the diversificationof industries.
The impact of these measures was dramatic. Exports, which had
grown less than 12 percent annuallyberween 1953-62, grew 28 percert
a year between 1963-72, rising from 123 million to almost 3 billion
dollars. The transition firm import substitution to export promotion
wasthe most important policy changein Taiwan, China's,economicde-
v-elopment.It shifted the economy from a relativelydosed to an open
economic system and exposedit to the forcesof international competi-
tion and technological change.
IndusirialConsoidationand NewExportGrowth(1973-80). As hic 1970s
progressed,intemal and externalchallengesthreatened the contunuation
of export-led growth. The rapid manufacuring increasesof the 1960s
strained transportaton, electric, and communications systems.The
island suflrd profound extemal setbacks. More important, Taiwan,
China's, light manufacuring industries faced new competition from
lower-wageproducers abroad.As foreign investorsrushed to the newly
opened mainland Chinese market; international confidence in Taiwan,
Chin;:s, economydedined. The 1973-74 oil crisis had dramatc leper-
cussionsfor the Taiwaneseeconomy. RealGNP grew only 1.2 percent in
1974and inflation climbed tO 47 percent, while exports declinedin real
terms by about 7 percent.
Beginning in 1973, the govenment chose a more self-relianrdevel-
opment strategy based on industrial consolidation and renewed export
growth. Once again, it mrnedto foreign expertsand commissionedthe
U.S.managementfirm ofArthur D. Litde to find solutons to the eco-
.nomiccrsis. Basedon Taiwan, China's,economic needs and capabili-
ties,the Americans recommendedheavy investuents in infiasrmrure,
industrial upgrading and secondaryimport substitution.A government

132
 MACRO'ECO

plan incorporating the recommendations focused on development of

capital-intensive,heavy, and petrochemicalindustries to increase pro-
duction of raw materialsand intermediatesfor the use of export indus-
tries.Thle governmentalso launched ten major public sector projects,at
a total cost of 8 bilion dollars, to rcvitaize the economy and remove
bottlenecks to economic growth. These included highways, milroads,
airports, and construction of nuclear power plants..
HighTechnology
and Moderunihaon
11981-I. As it entered its fourth
decade, Taiwan, China, confionted a challengingdomestic and inter-
nacional environmenc.The spectaculargrowth of the 1960s and early
1970s sputtered to just below 7 percent in the late 1970sXTaiwan,
ChinaTs,continued inrcgration into the world economy revealedstruc-
tural weaknesses,particularly the financial system'sinability to match
the increasingdemands of induscialization and external trade. Exter-
nally, Taiwan, Chinais, persistent trade surpluses with major trading
partmes led to growing protectionism. In the later half of the 1980s,
Taiwads exports faced an additonal loss of competitivenessdue to the
appreciadon of the Taiwan, China, dollar and rapidly rising wages.
Manufacuing wages rose, undercumrngTaiwan, China's; advantage,
and local firms moved production overseas.As in other first-generation
East Asian NiEs,Taiwan, China, manufacturerswere squeezedbetween
lower-wageNIEsin traditional, labor-intensivemanufacturing on the
one hand, and high-technologyproducts from industrial economies on
the other.
Once again, the governmentmoved to rescrure the economy.After
extensiveconsultationswith domestic and foreign advisers;,the govern-
ment decided to focus on high-technology industries: information,
biotechnology,electro-optics, machineryand precisioninstruents, and
environmentaltechnolo&rindustries.Theshift to ahigh-technologyecon-
omy has necessiated the close coordinaton of industrial, financial, sci-
ence and technology, and human resources policies. In 1984, the
government revised laws to provide tax incentivesfor manufacturers
who allocatea percentageof their revenuesto researchand developmenr
(R&D). Incentiveswere given to industry to diversifyrand improve pro-
duction techniques.The government encouragedthe establishmentof
venture-capital firms and revised university curricula to stengthen sci-
-nce, mathematics,engineering, and computer education. It began to
recruit rechnicalmanpowerfrom abroad by offeringcompetitivesalaries
to forner Taiwan, China, residentsliving overseas.

'33
 ~1R'A CLE

In addition, in 1985 the government launchedfourteen major infra-

structure projects, including expansionof the energy,telecommnunica-
tions, and transportation networksand developmentof water resources
and national parks. The government'sdevelopmentplan for 1991-96
callsfor 330 billion dollarsin public sectorprojects.To support theseef-
fbrcs,the governmenthas adopted an overallstrategy of economic liber-
alization and internationalization, including the lifting of foreigi
exchangecontrols.
While it is too earlyto assessthese policiesfillly,a few observations
are possible.A growing number of small, high-technologyfirms pro-
duce incrcasinglysophisticatedand higher-value-addedproducts. For
most firns, however,the transition to hi-tech industrieshas been diffi-
cult. On one hand, the rapid rate of technologicalchange and rising
protectionism in industrial economies make it increasinglydifficult to
obtain advancedtechnology. On the other, the small-scalestructurc of
industry is not conduciveto the costlyinvestmentsin R&D and skills-
trainingneeded to shift toward high technology.As a result, most Tai-
wan, China, manufacturers are still assembling imported high-tech
components.

tD Manufacturd Exporus
Malaysia'sShift knm Resource-Based

During the firsttwentyyearsafer independence,Malaysiacontinued

the essentiallyfree market trade and industrial policiesof the colonial
govemrnment, aldtough it intervened extensivelyto promote mral devel-
opment and providesocialand physicalinfrastruacure.While the govern-
ment protected import-competng industies, protection wasgenerally
less strong than in other developingeconomies.Government was re-
strainedfiom heavily biasing incentives against agriculture by the eco-
nomic and political importance of theplantation and mining sector In
1957, Malaysi's exportsof tin anctnatural rubber accountedfor a thiid
of GDP(Bruton and others 1993).
ImportSubsitution[1950-701. The objectiveof Malaysia'slimited im-
porc substitution was the same as that in odter developingeconomies:
reducingimportsof consumer goods and increasingprocessingof nat-
ural resourcesto create industial employmentopportunities. The gov-
ernment did not promote individual sectors, and effective rates of
protection were very low, averaging 7 percent, compared with a range of
25 percent to 92 percent in other economies at similar levels
of income.3

134
 MAC ROE CO

Nevertheless,between 1960 and 1980 import substitution combined

with domestic demand expansionaccounted for virtually all of Ma-
laysia'sgrowth of manuFacuringoutput (Salich,Yeah,and Meyanathan
1993).
CombiningExport Promotion andImport Substituion
(197145).Malaysia's
1969 ethnic conflictstriggereda reexaminationof developmencpolicy
The New Economic Policy(NEP) launched in 1971 had many dimen-
sionsintended to promote growth with equity. In trade policy,the gov-
emmnentbegan more acive promotion of natural resource exports,
particularlyrubber, timber, palm oil, and petroleium,and light manu-
factured exports,particularlytextiles, footwear,and garmens During
the 1970s, export-promotioneffortswere designed to complement the
modest import-substitution regime inherited from the 1960s. Major
NEP export incentives induded taxableincome deductions linked to ex-
port performance and domestic input content, tax allowances for
export-related promotional expenses,and accelerated depreciation for
finmsexpormngmore than 20 percent of their output. Credit policies
promored exports through guaranteesand automatic rediscountingof
exporcfinancing at low interest rates.
Export-processingzones, fiee trade zones,and licensed manufactur-
ing warehouses that permitted duty-free import of materials to be
assembledor processedfor export were crucial to the successfillcombi-
nation of import substitution and export promotion. Foreign invest-
ment, particularlyfrom Japan and elsewherein NorcheastAsia,poured
in. By 1980, about 70 percentof manufacturedexports originatedin the
export-processingzones, primarily from foreign-owned firms. As the
economy has developed,the proportion of exports from the zones de-
dined to about 40 percent in 1989.
AdjustmentandLiberalzation (1986).A combination of terms of trade
shocks and fiscal imbalances (describedabove) prompted the govcrn-
ment to move in 1986 away from state-ledindustrialization.Promotion
ofprivateminvestment acrossa broad rangeof sectorswas combinedwith
macroeconomicadjustment and continued effortsto increasemanufac-
tured exports.4 Tax incentivesfor exporterswere increasedand imports
wereliberalized.As a result, the averageeffectiveprotection to induscry
declined 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 changesin export incentivesincreasedthe variation in cf-
fectiveprotecCton acrosssectors.

[135
 -ALCL
I E

Indonesia
TriesIt BothWaysandOptsforExport
Push

Indonesian trade policies have swung from protectionism to open-

ness.The current open phase has gone fiurtherthan any earlier efforts
and obtained the most impressive results.
NationalismaudGuidedDevelopment (1948-66E
Followingindependence
in 1948, economicpolicy was shaped by a strong sense of nationalism
flavoredwith anticolonial and anti-Chinesesentiment. DLespire several
early attempts at liberalization,the policy regime became increasingly
inward oriented and interventionist.A pervasiveand complexregimeof
import and investment licensesfostereda new group of Indonesianim-
porters and traderswho earned substantialrents and becamea powerful
lobby for trade restrictions.Followinga centralizationof politicalpower
in 1958, PresidentSukarno expoundeda populist platform of 'Guided
Democracyand Guided Economy,"and the economy entered a period
of more direct state control of production and trade. Dutch enterprises
were nationalized, and stare enterprises took over all aspects of the
economy-including the recendy emergedimport monopolies.
Growing mismanagement resulted in chaotic economic conditions..
Inlation acceleratedto 1,000 percent by 1965. Exports and foreign ex-
change reservesdwindled, and debt serviceexceededforeign exdcange
earnings.Economic growchstagnated. By 1965per capita income was
15 percenr lower than in 1958. Widespreadsocial unrest and violence
ser the stagefor a change in leadership.
OitwardOriented NewOrder Goverment (1967-731.The new government
of PresidentSuharto moved quiddy to restoremacroeconomicstability
and adopted a moreivorable stancetowarddomesticand foreignprivate
investmentSomenationalizedenterpriseswerereturned t previousown-
ers,and a newForeignInvesamentLawthat provideda thirty-yearguaran-
ree of nonnationalizationwas enacted.The exchangerate was adjusted
throughlargedevaluations.In 1970,the governent establishedthe mcli-
anismfor a market-drivenunifiedexchangerate and abolishedcontrolson
capitalmovements.Sweepingchangeswereintroducedin the tradeand in-
centiveregimes,ildg the abolitionofiport licensingthough litde
change was made to tariffi. As a result, non-oil exports increased at an
annual rate of nearly26 percent from 1967 to 1973, comparedwith an-
nual increasesof barely2 percentduring theprevioussevenyears.
OilandCommodtBoom(1974813. Indonesiabenefitedfrom the sudden
surgesin c-i prices in the 1970s, and the boom in primary commodity

x36
 MACROECOO

prices. The surge in revenue gave the govenment an opportunity to

intensifr development effort but it also posed the problem of how to
protect the competitveness of the non-oil economy from the adverse
consequences of oil-windfiil spending. While Indonesia was relatively
successful in moderating exchange rate appreciation, the appreciation
that did occur increased pressures to protect domestic industry from
imports.
Compared with many oil exporters, Indonesia used its enhanced re-
sources well. Daring the 1970s, about 40 percent of the government
budget went to infrastructure for the economy. The government also
placed 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 progress
during the 1970s in reducing poverty and improving social conditions.
Economic growth averaged dose to 8 percent through the 1970s and
early 1980s, as a result ofstrong expansion of public and private invest-
ment- Non-oil activities reinained buoyant, especialy agriculture and
manufacturing. The government maintained good macrocoonomic
managerment and, following a debt crisis with the state oil firm Pertam-
ma in the miid-1970s, a conservative foreign borrowing straregy, By
1980, 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 became
increasingly state dominated and inward oriented. Fluzh with cash from
oil 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 the
cconomy grew, the govemment tightened regulations on foreign and
domestic private investment Meanwhile, gradual appreciation of the
exchange rate eroded the competitiveness of non-oil exports, and pres-
sure from domestc interests caused the trade regime to become more
protectionist and variable.
AdMjusuenttExtenalShocks(1982-85. The high-cost nature of the in-
dustrial scaor and die end of the oil boom sparked a debate and re-
assessmentof 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,more
inward oriented and subject to government intervention. In addition,
the adjustment resulted in slower growth of output and incomes, re-
duced public and private investment, low rates of capacity utilization,
and the emergence of financial problems among industrial enterprises.
Moreover, although textiles and plywood both emerged as important
exports during this period, overall manufacturing growth slowed in the
face of weak non-oil exports and weak domestic demand.
By 1985, a plethora of decrees had brought a wide range of products
under different forms of import control. Overall, the production regime
favored import substitution rlatidve to exports. As a result, domestic re-
sources were drawn into relatively inefficient capital-intensive activities
supplying the domestic marker- This disadvantaged the more labor-.
intensive, efficient downstream producers and reduced export growth.
DerigflationandOutward 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 efforrsto restore macm-
economic stability, the goverrnent launched a program of broad trade
and regulatory reforms that opened the way for a decisive shlfif o export
Figum3.4 EffectiveRatOs promotion. Exporc-oriented deregulation has gathered momentum
since 1986-8 7, culmninatingin a series of major reforms in trde policy,
...
Import-:X
Impoflb 2sminvestment licensing and transport regulatons.
The first step toward trade reforrn bad actualy been taken in 1985
Tramnston with steep tariff curs. In 1986, pushed forward by extrnal shocks, the
_ l Iprocess 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 accessto im-
(percent)
* Mlaysia 31Indonesia Thabnd porcs (see figure-3.4). Later the government deregulareddomestic and
-cSalieb,
Xcah. and Mcyanadmn foreign private investment, and for the first time permirted the private
-mcr
507zSo&k, Yeah and Mcpnathan
(1993); BbarracharyaandPangx (1992w); sector to invest in power, telecommunications, ports, and roads. It also
Brinmbc (1993). firher deregulatedthe financialsector

138
 'MACRO E 'CON

The economyrespondedrapidlyas manufacturingoutput, exports,

and investmentall increased.Growthacceleratedfrom 1988onward
as a resurgencein domestic demand accompaniedthe continued
growthof non-oil exports.Plywoodand textilesremainedthe largest
manufacturedexport items,but other manu&cturedexportssuch ats
shoes,apparel,and electronicsbeganto growveryrapidly.Foreignin-
vestmentprojectsapprovedby the govcrnmentincreasedtenfold be-
tween 1986 and 1991, to more than $8 billion, and domestic
investmcntroseby a similarproportion,to $30 billion.Perhapsmorc
important,70 percentof the foreigninvesrmentapprovedwasexport
oriented,comparedwith 38 percentin 1986.Since1986, GDPgrowth
has averaged9.3 percent annually,the highestsustainedgrowthrate
sinceindependence.
Indonesiacontinuesto encounteroccasionaldifficulties.
In 1990and
1991,the economyexperienced an importsurgeand deceleradonof ex-
port growthdue to increasedaggregatedemand promptedby a prior
easingof monetarypolicies.This led to a wideningcurrent account
deficit and a surge in foreign borrowing.As with previousmacro-
economiccdistortions,the governmentrespondedby tighteningmone-
tarypolicy and curbing borrowing. Meanwhile, the obvious successof
export-pushpolicies can be expectedto generatemomentum for firTher
export-oriented
reflrm. If this is the case,future swingstowardprotec-
tionism,should they occur,are likelyto be lesspronouncedand less
damagingthan thoseof the past.

Thailand's
Shiftfom Resource-Based
Exports
to Manufacturing
LikcIndonesiaand Malaysia,Thailandhistoricallyexportedprimary
and agriculturalproductsderivedfrom its rich naturalresources.Early
tradepoliciesemphasizedtaxationof the resource-based rent associated
with theproductionof theseexportablecommodities.Tradewasheavily
controlled.Rice exports, Thailand'smajor commodity export, werecon-
trolledby a statemarketingmonopoly,exporttaxeson othercommodi-
tieswereheavy,and a multipleexchangerateregimediscouragedexport
producton (Jitsuchon1991).
Exports
Natural-Reso.rce-Based 11955-70).In 1955 the exchangerate was
unified,and the state marketingmonopolyon rice exportswas abol-
ished.Althoughthesepolicyreformsencouragednatural-resource-based
anrdagriculturalexports,Thailandmaintainedsubstantialimport pro-

'.39
 ffl ,, ]\,, .RWCLE

tection for specificindustries.As late as the mid-1960s,however,overall

levels of effectiveprotection to industry were modest by developing-
economy standards. This was especiallynotable in consumer goods
manufacruring,where nominal tariffs were in the range of 25-30 per-
cent ad valorem.Tariffson machinery and intermediateinputs were in
the range of 15-20 percent.
FavoiingImportSubsfition (1971401. In the 1970sThailamd, following
the import-substitution strategies favored by many other developing
economics, raised tarifli on consumer goods to a rangc of 30-55 per-
cent. Capital and intermediate goods continued to enter at low duty
rates, leading to an increasein the effectiveprotection to valueadded in
import-substitutingindustriesand to dedines in the eff-ective protection
to agricultureand traditional exports (figure3.4). The biasagainstagri-
cultural and export production increased.
Textiles,pharmaceuticals,and automobile assemblywere particularly
favored.At times, vehide importswere banned, which resulredin limit-
lessprotection for domesticmanuficurers But therewasa quid pro quo.
Domestic content requirementswere establishedto promote upstream
suppliersofcomponentsand parts. Thesenontariffbarriersto competing
importsof componentsand parts raisedproducton costsin the assembly
sectorwhile subsidizingthe produce- of inputs. Consumers were in ef-
fect taxedto promote an intermediategoods sector.
R-aionandExport IncentWes(19804.As was noted in our discussionof
macroecononic polic, the second oil shock exposedweaknessesin the
Thai economythat werethe resultof the import-substitutionpoliciesof
the 1970s. In 1981 Thailand's trade poliqc shifted explicitlyin the di-
rection of export promotion. Remainingexport taxeswere reduced, and
as describedabove, the exchange rate was devaluedand subsequently
moved to a managed floaL The government also began reducing pro-
tection of local industries and making tariff lower and more uniform.
The maximnumduty rate was reducedfiom 100 to 60 percent, and the
averagerate was reduced.The tariffreductionswerealrnostimmediately
offser, however,by the government's decision to impose a tariff sur-
chargeto raise revenue for macroeconomicstabization objectves. Ef-
fective prntection remained high into dte mid-1980s at 52 percent of
value added, higher thaL other EastAsian economiessuch as Korea(28
percent) and Malaysia(23 percent).
During theseearly stagesof the export-pushperiod, the govemment
Board of Investment (BOI)played an important role in promoting cx-

140
 M ACROECO?rw1
. .~~~~~~~~~~~~~~~~~~~~~~~~.

port growth. The BOI,which had been establishedin 1960 to supervise

the applicationof investmentincentives,had long been a tool for im-
plementing the import-substimtionstrategy.One of its keyroles was to
distribute (and often increase)the incentivesfor firms operating in pro-
tected domestic market niches through waiversof tariflis,import sur-
charges,and occasionallybans on competingimports.5
The shift to export promotion brought a differentrole to the BOI.A
-major study of the exporc-promotionsystemwas carried out under a
uNDP-World Bank technical assistance project. The study recom-
mended radicalreformsin the wayin which the BOIdid business.It ad-
vocated using moreautomatic,sectorailybased incentives(in contrast to
- the firm-specificdiscretionalincentivesof the past two decades) and
shiftingthe targetof incentivesto labor-intensive,export-oriented,geo-
graphicallydispersedactivities.In 1983 the BOIannounced new promo-
* tional criteria that f&vored,among other objectives,exports and labor
intensity.Box 3.4 describesthe impact of this shift on direct tariff in-
vestment. Other incentivesfor exportersinduded tax exemptionsand
rebates,reductions in electricitytariffs,automarticaccessto credit, mar-
keting assistance,and promotion of tracing companies.The govern-
ment also stramined customs procedures and abolished unnecessaty
regtuationsto expediteexport shipments.6
Automatic and concessionarycredit was another major elementin
the export-push strategy.The Bank of Thailand had taditionally ex-
tended refinancingflaciiities(RFFs),through the commercialbanks, to
key economic sectors.7These RFFswere mostly used to rediscount cx-
*- port billsof largeexportershandling traditionalproducs As part of the
export incenuvespackage,the Bank of Thailand revisedits rediscount
irulesto focus more explicidyon small, nontraditionalexportersand on
other productive activities.
Many of the export-pushincentiveswerespecificallydesignedto off-
set remainingdistortionsfrom Thailand's import-substitutionera. Tax
rebates, duty-freeimports, and export-processingzones correced par-
tiallyfor existingdistortions.In this sense,Thailand'sexperienceis sim-
ilar to that of the northeastem HPABs; the initial surge of export
promotion tookplace in the contextofa domesticmarket that remained
modctey protected from the rigors of foreign competiton. Thailand
has recentlybegun a more broadly based liberalizationof imports. In
-1990tarifs on capitl goodswerereduced,and in 1991largereductions
werc made in tariffson automobilesand computers.

.. -,J - , g f :
 ; -_ ii~~~~LE'

~~ ~ ..g ~. . ~ ~ ~ ~ .
~
-~ ~ .,L~ . ~ ~~ie E-or s

-. ~~~~~~~~~~~~~f wS_ .n
.'',M '.ape'roa9AesS>'~~~~~~~~~~~~~~~~~
'

''.~~~~~~~~~~~~~~~~~ ' -" - -

D X = ~~~i ar =CU roa es

:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ NT
:~
-: g :~~~~~~~~~
f - ~~~~~~~~~~nstet
sm W-96 ne
:--:: ::~~~~~~~~~~~~~~~~~~~~~~~~~i3
!:g :- ;~~~~
- l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~a

- u - ~~~~~
- ~~~
- The-outcome
of thesepol~~~~~~~~~~~~~~iy
alaacB
96*
X - - : - manufacoJres
reprcsented
30.6pelt ofa gowingvolumeofThat ex- -

- pora;.Ladingsecorsinclu ed
d Woin,
~IO W f oj-
: - -.- elrr, ad integrarcd is E Oirect
foei n
ilVtmetpl la ao

.~~~~~~~~~P
. .. A

.- 0 . X - ; - * -
:xS.; 50 -; .: - 0-- - - 0
 MACR0O J~~

v3: Distribution
: 0 -0Table'3.7 of Man.factnriug,
;n.uMtA Thailand C R1O36 _
(Percentageof grossdomesticproduct)
Senor'
,97.
'1986
Hev inutis3.9 42.6: 3.
High-skilllabor-intensiveidusre 9.7 11.1 .13.3
Traditionallight inutis19.2 201 .23.'3

* . Total.- 100 100 . 100

mTab ofTi2n
3 e Banko aflTiland

TheElementsof a Successful ExportPush

These summariesof the evolution of export push revealthe plethora

of policiesconsistentwith exportpromotion. (China'srecentpoliciesfa.-
voring manufactured exportsare outlined in box 3.5.) AssistaLnceto ax-
ports varied over rime and across economiesand induded p..ferential
.financing,promotion stubsidies,tax incentives,subsidized inf1auc-
tLre, and foreign investment incentives.Moreover,at any given time
and in any particular economy,the magnitude of the incentivesvaried
acrossactivities.Some governmentsfavored nontraditional exports,
some favoredparticular types of manufacmured exportshEven so, there
aresignificantcommonalitiesamongtheseponlCis.Each contributed to
one or moreof the four elements of a successfl export pushi accss for
exporters to imports at world prices;accessfor exporters to lon- and
short-term financing-,goverment sistancnin penetrating markets;
and flexibilityin policy implementation.
at WormPricer When imports are restricted or rtaed,
Accessto Iviports
exporters are: disadvantaged vis--vis exporters from other cono e
: psorts-tewrld prinacerin aitn npntrin bonmd
aed
who do not face similarly expensive imports. To combine import pro-
tection with export promotion, these ill effctcs must be mitigated. HPAE
governments, have found myriadeways to grant exporters access to ir-
-portsatworldipories free trade zones, exppot-processing zones, nd

warehouses, luty drawbacks, orexemptions from tiffs. However, the

0 ovrnen:shn foud yia wystogan epotesapssst img
higher thelevel of irmport-substitution protecton, the greatcerthe level
wof
adminisrative cormiptenc needed to cseate countrvailing imcentves
for exporters. Perha because of these difficulties and the dedining

benefits of import-substitution protection in successful export-oriented

economies, each of these economies has gradually liberalized its trade

'-43
 = X _5
-tCL~CE

... .~~~~~~~~~~~~~~~~pt, g. , ar, L .*

G

.. ." ~ ~~ ~ ~ M ..
~
. . ~
~~ ~ ~ ~ ~ ~ <._ ...

.~~~~~~~~~~~~~~~~~~~~~~~~~~~ , MC

- pogamtoenux aces o reir oce a sbizex apries,.FButagion-,.,.

- : 0 we oberve i ipesvartyofhye j
ofcreit ntnvru
0 - - shorr-tan), the degre
oSsubsidizaon9
( !;aMantS; subsi
as
a,
: f 0 dizedraresX,
theselecizviyallQqporT
ve9u ex
5age
, ortatirres

T~~~~~~~~~~~~~~~~~~~ ~ Lgo-. -a
.~~~~~~~~~~~~~~~~~~~~~A- - T_

.. ~~~~~~~~~~~~~
.V. .r, .5
. .-:.: . ~, .:: -~ ;: ~ ~ ~ ~ ~ ~ ~ ~~I
 MACRO.ECRO:E

Mabrket Penebaiom. NearLy all governmentsrecognizedthe difficulty

expoers Fce in crackingfbreignmarkersbut againchosevariousmeans
to encourageexportersto overcomethese hurdles. Somedirectlysubsi-
dized export activity (direct income tax incentives),some subsidized
marker penectration(through exporter associations),some subsidized
small and medium-sizeexporters to offier their difficultiesin marker
penetration, and some promoted the creation of international trading
companies.
fleibfit. Pragnatism and a commitment to resultsmark the HPAE
governments'approachto exportpolicy:Policyflexibilityprovedimpor-
tant becausehitting the right strategyis not easy, fir three reasons.The
right strategydepends on the circumstances.It changesas the economy
changes.And it is not alwaysobvious.Making mistakesis not necessar-
ily bad, as long as the mistakesare reversed.The dassicexampleof pol-
icy flexibility is Koreals reversal of the HCI drive. After a large
commitment to promoting heavyand chemicalindustries4the Korean
government neverthelesschanged course when circumscanceschanged
and the policy threatened exports.But Koreawas not unique in this re-
gard. In each of the lPA.Es, policiesevolvedoverrime and oftenrequired
major revLsions.
In chapter 1 we showedthe impact oftheHAEs' export-pushsategy
on the rapid growth of manufacturedexports. Another outcome of the
export push can be seen in their changing compositionof exports (see
figure 3.5).With the exceptionof Indonesia (whoseexporcsof refined
petroleumproducts are induded under chemicals)all of the HPAEsdra-
maticallyincreasedthe share of exportsin machineryand equipment at
the expenseof such traditional manufiacmresas food products and tex-
tiles. In 1991 the share of rnachineryexports in total manufacturd ex-
ports for the developing HPAEs was about double that for other
developingeconomies.This shift clearlyshowsthe responseof exporters
to changing market opportunities, rising labor force skills,increasing
technoogical capabiliy, and changingcomparativecosts.
What is surprising,however,is the extent to which textiles and ap
parel remainedimportant exportsin the faceof shiftingcomparativead-
vantage.As late as 1991 most of the developingHPAEs had shares of
textdlesin manufacured exports which cxceededthe averagefor other
developingeconomies,suggestingthat thesesectorsretainedsubstantial
cost competitivenessdespite rapid capicaldeepening and rising real
wages(a phenomonon to which we shall return in chapter 6).

:-45
IE~~~~~~~~~~~
 MA CR- O

The successfulEast Asian economieshave had much greater macro-

economicstabilityand much more rapid exportgrowththan most other
developingeconomies.They achievedmacroeconomicstability by ad-
hering to macroeconomicfundamentals-particularly by keeping fiscal
deficitswithin the linmitsof prudent financing-and by rapidlyand ef-
fectively correcting major macroeconomic imbalancesthat emerged.
They achieved rapid export growth by fostering a favorableenviron-
ment through macroeconomicstabilityand by applyingon a trial-and-
error basis a vast array of market interventions and microeconomic
incentives.In contrast to the many lesssuccessfiuleconomiesthat have
clung to filled policies,the HPAEs have pragmaticallychosen and as-
sessed policies accordingto their impact on macroeconomicstability
and export growth. They have adapted policiesas necessary,kept those
that worked, and dropped those that failed or outlivedtheir usefilness
'Why did the EastAsian economieschoosemacroeconomicstability
and exporrsas their policy yardsticks?And why have they been more
succsfil than other economiesin selecting,implementng, and alter-
ing policiesas necessaryto attain thesegoals?We mrnto thesequestions
of politicaleconomyin the next chapter.

3-47
 R'A CLE

Appendix
3.1 Economic
andPloflitcal
rimermes
Indonesia

OpemdanatsubhJJisa':-
M.~ ~ ~~nnhcfup mita
ho & b m n m _ _ _ _ _ _ _ _ _ _

~~~
mawiled ~ ~ ~ ~ ~ ~ ~ ~~Ol dm
Ci Wed ccanoiny.. a.
dkea,mtecuainl Rdninrpaaaim McPtsUniicatone. -fr'nlto.'--totim
atpinduainnmmd Th~Nwndp in. -f
rank gugjsajfraden Wailvs!m cak-
. - r.hghmafi
firms ewzuedwt
HjyCIfDAhtuIKI Sq 01 . -minm

ins-LM 1New
bcip - ntu aio asI&uiamoat iila

c=7- Su&,wont

7. Gicderiy
z.kdqaSns,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
"~"_...

148
 MACRO

i i- - - Rqmn.

.h.c 1 J;.Dfdb
4
it

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

,;-b Fa t h- ' appmv.4mpa; '.- ;elto m-pr

. .rtat& ~~~~~~n , ; ,,.E,--.-.ica
.,p9& ;9?,
:cauua ,had;, . rus,* ;t
cpu! Sore
~~~~~~~~~~~~~-__
t~~~~~ 4(** 4 __ __ _ Applbw
mun-of
* ~~~~
*. aduap ~~~~~~~ Duicamny&mns~ EWuaJn kauga .DcpIaa
-Oc -iiil0
.--. . .:PN

Smwod ai sar bb-l 6hd

SB S lBS. 1982 ' 185 298 2 8 9 0- : 9

'149
 ::::~~~~~~'W --
t L

Malaysia

; E' XFin;Ma/ape ' . RmMaj' P-a

t,. -_
-; -d -d Flu' -|-.:
;' - 1 _o - ! . , ' -.

:Invutcu Guarimeu
., REr;Fads''\
'
-; ,,hiu..m- ' Wea
- 'eT
:. V '-- ''H
.>::-* a' ; ~ ElbI-
rn f -'''t Inra A 'a 'za
. .- -. ;, - -:

-t* --. lnancc .;Me:.i C.nnhzmoc P-o

-d n: --
.~~~~ -it.hl,duc.rf TaiFhvki
...;.i:AeMof nduth
T : .- ; Ai.
.. s-- Dcv
F uIn tr*m
=t. PmThdc F-
. . .UIfllf
- -
§-!
: .,,ar -,'1 -' F95
'., .as -f-l _ak '. CoaiI .-.edpam Atikd 'Z,Aa -- .uc-y:-
19:7'2 -- ,"
*e t ;--T

-, - _.

155 1;:' 1.:; - 2.95 286 26. ;36 .56 2.;6 2.6 _39 27 8h

puawersii.linccihispendcin

- .~~~~¾

:~ ~ ~ ~ aidnfrminMlyCnrn . . -.

150
 MACROE!cONOI'd

f'aasuanu
Jlae 4 mm w?Lntae
flundd ~~~~~~~~~~~~Hwedpma
A H~ d

rA ~ ~ 7mteaa& fmAapaL ~ w?apAL ik.;

Caaveuaan -.-- -~tw

Ccep0
mdon.has 6dm; Pummlga9onc .9 g -95

hakswete,
flatten
CoalIuonAa DFMdapt ate ZntaAa~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 SzittqfSis~~arhz4wteutitym k a cnpi rwmpato

alirsics~~ ~ ~ ~ ~ mqtpmoD - lctor

Fmna , Dndh p mn Praadnfpn)- aun
Avl&mmmac=:-cdtdldnrc RelasionsAct
kbca~hsrt' Ddpsasll
imInanics .lard
emda.A
indusrs
toa rsc
indusif
apusrake f Epl'e q byp(uu
Vzmz * A Cud adlcldorrr

ad~~~~~~~~~~~~~~ asnr
~
tarn - ~ ~ ~ ~ ~ in cir

dvilnnm Z

152~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
 A 'I
p
.1 ijr Ii
1
I:
S.
4
5.
I
N

E. . .. . .

.2

Liii j r
*j.
U Fgfl

iI jS

* H* I
I:

* lB *
K

*
* up
i i.irr
.
I ¶
'
a

Lr
U I"
r I,
N I I
'(U
* -

I i-Ii
JJ'L[ ill
-

I
Id

**
a
I
 CLE

Tainan,China

141gbmaaflntmt
aauvlapdcpnrIndprn
~~~~~~~~Tbhir
liur-Yac;Ema,nonc
i

D mv"nt
Kmr

Pi'lt *' . -cnt t Stwn fa

I of~~~~~~~~~~~~~~~ [WI oa ~ e

LaNd
to
*
LiRnClp
law- ~~~~~~isarml
nd
Smax far
Tnug.IJiar
SWiPCle panacfmcf-
herpm meat
Nto
'CLWO'

Ddnr ~~~~~~~~~~cadlazao
lan lWborm 'InniCal RateL .7. .-

1550
1845- 1952 1953 1965 1571568 1359 290 86 54 185-19 16
* -. .93'164

'54
 MACROE A C.

. ~ ~ 1 .

. ,;s s ;-,?anma,nr,aklbrc . 7 -. - , -RL.---- - -

;[7'<
Rauanachimn ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~f
; pn.ductviycatcr vith f fd;

Urnlimdon
Dewlcpmmr a Foreign

-ofIndustr[al FAmU6k,cn& SAkbmcnt

.d;1972:
197n - &ah
O~Iiutr
=,
- ^ ¼' - - f - S T.-;
Estahlubman 0 :-,- 7 ,, - i-,0 ' &uU'nihmc, ,:
t; - ' -;*< u " laboriaws a 'x

-.d zial .,Ne l.:-:,

Wdncbu -d' c - - :infmu,-c 'J;D';; . , d '5en;-
Ins e D -l Initkial-

- ,.,,'N D ,' T. -'','.or .' s&.a. _______ pu4n *-.' .d .; d''',''

010 . -. ~ WoddEatak
. . . oppm-IIn

. aiwan. Liwnadotpl

..
.. I

m7i~~~~~~~~~~~~~~~~~~~~~~5
0 Nat
'' is, i1ce - - m--a-r"-
f .... -. 1 * . ..
,,,me%'to
 Notes
1. Bradfordis among the first to categorizea combina-
tion of functionaland sclectiveinterventionsthat resultin
the effectiveexchangerare for exports exceedingthat for
imports as an export push. By this definition the South-
eastAsian NiEswould not qtalifl (Bradford1986, 1990).
We we dse term somewhatmore broadlyro indude sus-
tained movementtoward parity of incentivesbetweenex-
port and import substitutes,combinedwith institutional
supporr for exporters.
2. The analysisin this section draws on Easterly and
Vchmidt-Hebbel (Ibrthcoming. The data on consoli-
dated public deficits,as wellas the resr of the data in this
section except where otherwise indicated, are from the
same source. Consolidated public deficits, though less
widely availablethan central government deficits, are a
much more reliable indicator of fiscal management be-
causetheyindude operating deficitsof public enterprises
that have played a critical role in some macroeconomic
cases.
3. Although the mean level of effectiveprotecrion was
low, varianceamong sectorswashigh.
.56
4. The Promotion of InvestmentsAct of 1986 pro-
vided incentivesfor investmentsin agriculture,industry.
and tourism.
5. The BOi's policies of selectiveinterventionto pro-
more firmsand specificsectorsare discussedmore fullyin
chapter 6.
6. Onc impormtn exampleof these effortswas the es-
tablishmentof export-processingzones (Ens) and bonded
warehouses.Drawingon Korea'sand Taiwan,China's, ecx-
perience,Thailand used the EPs to streamlinetemporary
admissionof inputs for exportersin the zones. The zones
also offeredlower-costelecrricity,frcedomof immigration
for managers and technicians, and. common-facility
bonded warehousesto reduce customs and storagecosts
for enterpriseslocatedin the zonc. In general theseincen-
tivesalso favoredlargeand often foreign-ownedfirms.
7. A firmwishingto obtain accessto the schemeissued
a promissorynote, which was discoLntedby its commer-
cial bank and firther rectiscountedby the central bank
both at subsidizedinterest rates.
 R 4

An Institution Basis
forShared Grot
N EACHH-PAEEXCEPTJAPAN,NEW LEADERSPACEDAN URGENT
need to establish.their plitical viability before the economic
takeoff.The Republicof Koreawas threatened by invasionfr-om
the North; Taiwan, China, fr-omChina; and Thailand, from
Viet Nam andt Cambodia. In Indonesia, Malaysia,Singapore,
and Thailand, leaders faced formidablecommnunistthreats. In
* addition,leadersin Indonesia, Korea,and Taiwan,China, having taken
power, neededtto prove their ability to govern.Leadersin Malaysiaand
Singaporehad to contend with ethnic diversityand attendant questions
of political representation. Even in Japan, where the competition was
less immediate, leadershad to earn public confidenceafter the debacle
of World War [I. In all casesthen, leadersdesperatelyneeded to answer
*a basicquestion:why should they lead and not others?
Whateverstrategythe leadersof theHPAEgovernmentsselectedto an-
swer the basicchallengeof legitimnacy, they included apindpoIe ofsbazred
growth.The leaders hoped chat rapid, widely shared improvements in
economicwelfarewould bring legtimacy-
Of course, feiwpolitical leadersanywherewould reject, on principle,
either the desirabilityof gwrowhor that the benefitsof growth should be
shared. 'W1hat distinguished the HPAEe'leadershp was thecextent to
which,they.adoptedtspecificinstitutional mechanismns tailored to these
goals,and that workced.In this chapter we brieflyreviewthe history of
the adoption of an explicitcommitmnentto sharedgrowth. We then dis-
* cuss the main institutional mechanismsadopted to achievethis goal and
how they were areated and mainmainedh

* Wealth-sharing programs designed to indude non-elites in eco-

* growth
~~nomiuc

'157
 4~~~~~~ ~C £
KlA
CLE

* A cadre of economic technocrats insulated from narrow political

pressures
o InsEirutionsand mechanisms to share information and win the
support of businesselites.

Legitimacy
Achieving throughSharedGrowth

T tHE PRINCIPLE OF SHARED GROVfH ASSUMED DIFFERENT

forns in each HPAE It was most explicitin Malaysia.In response
to racialriots in 1969, three dominant parties,each representing
a major ethnic group, widened their coalition to indude other parties
under a nationalfiront.The coalition'sgoal was to share wealthmore eq-
uitaMlywithout stifling growh The coalition'ssolution was the New
EconomicPolicy(NEP), an attempt to improve the lot of the Bumiput-
era, the largestbut poorest ehinic group. The coalition'sown early de-
scriprion of the NEP is quite explicir:
The NEP has as its overriding objective the promotion of national
unnitythrough the two pronged strategyoF.(i) eradicatingpovertyby
raising the income levelsand increasing employrnentopportunities
for all Malaysians, irrespective of race and (ii) accelerating the process
of restrucruringMalaysiansocietyto correcteconomic imbalance,so
as to reduceand eventuallyeliminate the identificationof race with
economic function.... The effortsto attain theseobjectiveswere, in
-turn, to be undertaken in the contextof rapid structural change and
expansionof the economyso as to ensure that no particular group ex-
periencesany loss or feelsany sense of deprivation. [Salleh,Yeah,and
Meyanathan 1993]
The principle of shared growth is also explicit in Indo- .esia,where
President Suharto advocates "economic democracr. For example,in
1990 President Suharto himself strongly endorsed research results of
the Indonesian Economists Association stressing the importance of
sharing the benefits of growth. 'I hope that IsEfsfindings on the sys-
tem of EconomiicDemocracyreceivesas m.uchpublic exposure as pos-
sible and that all levelsof society become involvedin the discussionof
this vital issue. This will enable us to reach national consensus as we
strive forward toward the full implementation of Economic Democ-

: 58 ::- -:
 AN..:INsTn'

racy" (isEa 1990). Economic democracyadvocates relianceon the free

market to stimulate growth. But it also explicitly acknowledgesthe
market system's inadequaciesin distribution and actively seeks to re-
dress them.
In Taiwan, Cbina, commitment to shared growth grew out of the
Taiwanese authorities' searching analysis of their failure in mainland
China The leaders' five-point diagnosis included three equity issues:
agricultural tenants had rebelled against exploitation while the party
continued to identify with the landlords; labor unions had run out of
control; and governmentwas beholden to vested interests(Wade1990).
These realizations,combined with pressure from the United States to
undertake social and economic restructuring,led the Kuomintang to
adopt shared growth as the formula for its rule in Taiwan,China.
Unlike the Taiwaneseauthorities, Korea'spost-World War II leader-
ship lackeda synchronizedviewofgrowth and distribution. Evenso, the
sequence of policies it pursued, and the flattening of wealth resulting
fiom the Korean War effectivelyproduced egalitariangrowth. Pardy by
accident and pardy by trial and error, the leadership graduallyadopted
the principle of shared growth.
In both postwar Japan and post-independence Singapore, growth
with equity was the leaders' primary concern. In Japan, the war's de-
struction of landed wealth actually helped the leadership to implement
its redistributivegoals.In Singaporelack of a significantrural sectorand
the absence of a landlord elite also made pursuing equity goals easier.
But neither set of leaderswas complacent; both adopted effectivecon-
sultativearrangementsbetween labor,government, and business.'
Chapter 1 presents evidencethat the leadersof the HPAEshave been
unusually successfudin achieving shared growrh. Growth-rate and
income-distribution performancemeasuresshow that the HPAEs signifi-
cantly outperformed other low- and middle-income economies.How
did they do this? If adopting the principles were sufficient, certainly
achievementwouid be widespread.The answermust partiallylie in the
insttutions and mechanismsand in how they worked.
To win the support of non-elites, the leadersof the I4PAEsintroduced
mechanismsthat drasticallyincreasedoppormtnituesto share the benefits
of growth. These mechanismsvaried from economyto economy but in-
cluded education (in all the HPAEs); land reform (in Japan, Korea, and
Taiwan, China); support for small and medium-size industries (Hong
Kong,Japan, Korea,and Taiwan, China); and governmentprovisionof

'159
 suchbasicamenitiesas housingand publichealthservices(HongKong
and Singapore).Nearly all HPAEgovernmentswalked a delicate line re-
garding labor by limiting the power of unions and intervening to check
labor radicalism, while at the same time encouraging a cooperative di-
mate in which labor was rewarded for increases in producriviy.
These wealth-sharing measures have differed from the typical redis-
tributdve approach of most developing economies. Instead of granting
direct income transfes or subsidizing specfic commodities (for exam-
pIe, food or fuel), HPAE leaders have favored mechanisms that increase
opportunities for upward mobiity. The frequent result is that indivlidu-
als and families, provided the opportnity and convinced that efforts
will be rewarded, srudy more, work harder, and save more.

Universal
Education
Education is arguably the most important of these wealth-sharing, op-
portunity-creating mechanisms. The provision ofuniversal primary edu-
cation and wide access to secondary and higher education contributed
substantially to opportunities for upward mobility. This mobiity in turn
mirigated the feeling of non-elites that society is unjust and made them
more accepting of the market-orientedpolicies needed to foster growth.
In Korea, for example, the government invested heavily in the expan-
sion of education at the primary and secondary levels soon afier the wan
In rhe mid-I 960s, Park Chung Hee's export-promotion drive boosted
demand for educated labor, providJingjobs for the first crop of graduates
to benefit from the postwar education push.

Equitable
LandHolding
andLandRefonn
Theory and empirical evidencesuggest that widespread ownership of
land not only improves equity but also improves land prodlccrMry
(Berry and Cline 1979). All the lPA.Es with subsEantial agrarian sectors
have widespread land holding, resulting from either taditional owner-
ship patterns (Indonesia and Thailand) or land reform (Japan, Korea,
and Taiwan, China). In Malaysia, corporate-owned plantations have
dominated agriculture since the colonial era. But with a relatively small
dpopulaton and ample land, it has avoided equity problems common-to
other developing economies with highly unequal land distributions.
Hong Kong and Singapore have almost no agricultural seroc.

x6o
 AN

Koreaand Taiwan, China, began land reform under broadly simnilar

circumstances.In both cases,authoritarian governments&cing a com-
munist threat were dependent on the assistanceof the United States,
whose advisersurged them to adopt more egalitarianland holding. In
Taiwan, China, rhe government seized land from the landlords, com-
pensating them with shares in state enterprises.It then sold d-ieland to
the tillersat fitvorablecredit termsand fivorableprices.The government
then helpcd the tillersupgradeproduction for domesticand export mar-
kets. The program worked, economicallyand politically.Land reform
helpedTaiwan,China, achieveone of the world'smost equitableincome
distributions (Kuo 1976). Political stability benefited in two ways.
NewlyIandedfarmers,focused on boostingproduction, had little inter-
est in radical activities.Former landlords, as new shareholdersin state
enterprises,had a vestedinterest in the successof the Taiwaneseauthor-
iies economic program.
In Korea, land reform occurred in two stages.The first, initiated by
U.S. forcesin 1947, distributed the land confiscatedfrom the Japanese
at the end of Wurid War II to the tillersand put a ceiling on rents of
other land. The second, begun in 1950 and completed after the Korean
War, was undertaken by the Korean governmentafter a lengthy debate
in the legislatum The government took over landlord properties, paid
the latter nominal compensation, and distributed the land to 900,000
renants, effectivelyeliminatng tenancy.

Euuterpises
Smalland Medium-Size
AreBeautiful
Just as numeroussmall land holdings improvedequityand efficiency,
the HP'AEsbenefited fiom a profusion of small and medium-size enter-
prises (SMEs). The large number of SMEsgenerally reflected market
forces rather than government intervention. But several of these
economiessupported SMEswith preferentialcreditsand specificsupport
services.Rapid growth of labor-intensivemanufacturing in these firms
absorbed large numbers of workers, reducing unemployment and at-
tracting rural labor As firms shifted to more sophistcated production,
efficiengrrose and workers'real incomes increased.
Support for SMEshas been most explicit and successfulin Taiwan,
China. As shown in table 4.1, SMEscomprise at least 90 percent of en-
-r-erpnsesin each secor. Not surprisingly,the SMEs also dominate the
export sector, producing about 60 percent of the total value of exports

r16i
 - g,sji~~srA 1 N I IACCL E

andMedium-Size
Table4.1 SmaUl inTaiwan,China
Businesses
--.
- ,.-; ;e - -. -; ..-- ,, .d ;-u.,.nt -s Smell -medim--e .
-i - -;f -- - 0E u~0--:--::-Tou
tal-
- --- -. Percentageof
Cue.ory nipX resX Number0 : a -ni vi - rts
^Agriculur, 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
gm a r. . .: 1t.. . 86- 168. 90.32
Constructon , -. 28,419 26,456 93.09.
Commerces: 489,864 - 475,106.96.99
Tmansporr,-srorngc,and ommunimcations 36,897 . 35,818 - 97.08-;
Funancc,.isurancerel estatebusiness services 30,534 28263 92.56: ;
--.:.Comxruunit$,social,andpersonalservic .69,550 69,229.. 99.54
-';
Total 0 818,061-
=- :': -794,834
0 . . -- ,97.16
-- : - ,: :
StmrnnTai China(1992b).

(see tabIe 4.2). Other HPAEshave also encouraged small and medium-
size industries. Japan has direaed enormous financial resources toward
dcveloping small and medium-size enterprises. Public financial insti-
rutions have allocated an average of 10 percent of lending toward
SMES.During the rapid growth period of the 1950s, as much as 30 per-
cent of their total lending for fixed investments went to SMEs.The var-
ious government-supported direcred-credit programs have proven
particularly helpful during times of transition and rapid change (Itoh
and Urama1993). Although there are mixed evaluations of the success
of these programs, one thing is dear. The SMEsector has become an
important cornerstone of Japan's economy. In 1989, SMEsaccounted
for about 52 percent of both manuficturing value added and sales,
and their share of employment in various manufacturing subsectors

Table4.2 TheExportValueof SmaliandMedium-Size

Businesses
in Taiwaan,
China
(millons of U.S. dogrs)

.,.i .,. .:: -V -e~tkr.f

g . Po Auqon.

'M2d .436.89 J 64
-2.'...L.', ;,23525 - 42

674'-'4 3--- -'

i6z~~~~~~~~~~~~~~~~~~~~~~~~~~6
,: ~ OSrTIIJ hn 9b. . .~. _ - -' , .ni& _
ranged from a low of 41 percent in transport machinery to a high of
100 percent in silverware.
Korean developmenthas been largelydriven by the expansionof con-
glomerates,the so-calledchacbols But beginningin the early 1980s,the
SME 2 SMES'share in total manufacturing
sector began to grow rapidly.
employment rose from 37.6 percenr in 1976 to 51.2 percent in 1988;
while at the same time the SME sharein mnanufacturing value added rose
from 23.7 percent to 34.9 percent (Kim and Nugent 1993). Korea es-
tablishedan extensivesupport system for SMEs.And, as in Japan, finan-
cial support systems, such as export financing and credit guaree
programs. have becn d-iemost imporcant.

Housing:
Successfully
TargtingLow-lncome
Households
Two HPAEs,Hong Kong and Singapore, intervened heavily in hous-
img markers to win the support and cooperation of non-elites. By pro-
riding low-cost housing for the majority of residents, both programs
have helped to decrease inequality and minimize social unrest, thus pro-
viding<the long-term stability attractive to investors. Moreover, the mas-
sive construction effort created lobs when both economies faced higb
unmemploymenrtsubsequendy, the wide availability of low-cost housing
for workers helped to hold down wage demands, subsidizing labor-
intensive manufacturing.
In Hong Kong, which has generally followed a laissez-faire approach
to the economy, mass housing programs were a iesponse to a massive in-
flux of refiigees and migrants from China. The rapid increase in Hong
Kong's population from 600,000 in 1945 to 2.4 million in 1950
spawned slum areas, high unemploymenr, and poverry. These led to so-
cial disturbances, which culminated in riots in 1967. To difflasethe ten-
sion and improve living conditions of the general public, in 1972 the
govemment launched a Thst- track public housing program that has
since expanded to include the construction of entre new towns outsidc
the city propet By 1987, more than 40 percent of the population lived
in public housing.
Like Hong Kong, Singapore experienced a dramatic influx of mi-
-grants in the late 1950s and early 1960s and faced similarproblems. Its
expulsion fiom the Malayan Federation made matters worse. The gov-
ermuent responded by creating the Housing Development Board (HDB)
in 1960 to provide public housing to low-income f2milies. As these

-63
 -FIAA CL E

needs weremet, the HDB has turned to developmentof middle-income

public housing and self-sufficientnew towns similar to those in Hong
Kong. Today 80 percent of the population lives in public housing, and
more than 90 percent of the familiesin public housing own their units.
While successin Hong Kong and Singaporehas been aided by small ter-
ritoriesand populations, the experienceof Indonesia suggeststhat pub-
lic housing carefully targeted coward low-income groups may have
benefitsin larger economies as wel (seebox 4.1).

TheLabor Cooperative
Trade-Off: UnionsGeta BiggerSlice
In Japan, Korea,Singapore,and Taiwan,China (and to a lesserextent
Malaysia),governmentsrestrucured the laborsectorto suppressradicalac-

MM. J"F ~ ~ ~ .3>titY$~ t

-. -AiiGwn~~Skj&tiEboy,
Urox
'-V.- \,\4NC PIN# GECONONESf-EU-'E N,:i_OPULt-;

d l RPe%mcn9do
y,W ojcalits c JU,SF
1 new.urbanj
*~.~FLt ~L ,gos aalHrnlaaf.ok
ia.tj. 1 t !A? -, unablcosafford,...

i.>rohow nrir
schaSuetck&^le conca4 ctconianrsbdI'nolidal .
.,,peratlvece; isarbcle uron, lu; d.:-oLo6.
NI- -4 &tr;$
,dii,i potntil
.$QyiTOi'eSpottiid2*¢it& cisr;+f_,';i 'i=rt%v(i
s,the lndoneslan 0E e shCazi -

14PT
'jA'A ThdlW1I&wnroSfi
.2 praflvhezoeinxa e4!zng.aa-WpdcaaunOp;pflennca 4 ::
oaraW-X;
mmp~~~~g
V,
0~~~~~~~~~~~~~' . mmnnns tt
O' q own4paymentt,
.rp4rEonrir. atM
2 ~,
a., .~~i
5L~;
- .. .rcen
4zh4overnnenthshepd,h
b__ rl
.
ooen3 4' ropenu~1auz
,- ;OR¼S..

01 1onapmcwohWidin-ahd.s
l ao0O&n1nttl&n4

1c64
 AN IN4S-TQ -*

tivityin an effirt to ensurepoliticalstability.Governmentsabolishedtrade-

basedlaborunionsand pushedthe creationofcompany-or enterprse-based
unions. Managementand company union representativeswere then re-
quiredto joindyformulateand implementwork-relatedpoliies.
Labor movementsin Indonesia and Thailand, whilenot subjected to
systematicrestructuring, were nonetheless routinely suppressed at the
first sign of radicalism, primarily becauseof the governments' fear of
communism. In Hong Kong, conversely,union radicalism has been
held in check by a steadysupply of new laborfrom China and by the in-
fluence of Beijing, which has encouraged moderation in the unions
under its Hong Kong umbrella organization.
Governments in severalAsian NIEs, particularly in Malaysia and
Singapore, have modeled their systems of labor relations after Japan,
where labor radicalismwas quickly suppressedafter World War It and
labor-management consultations were gradually institutionalized. In
Japan, these consultationshave led workers to expect that adjustments
in their wageswill correspondto trends in the national economydespite
the unions' lack of politicalweight. 'Labor has been compensatedfor its
decreasedpolitical role through wage policies tied to increasesin pro-
ductivity" (Johnson 1982, p. 151).
Like the governmentin posrwarJapan, the governmentin Singapore
sharply lirmitedunion autonomy soon after independencedue to fears of
political instability Lacking natural resources following the separation
from Malaysia,Singaporecourted foreign investment in labor-intensive
manufacuring by suppressing independent unions and assuring in-
vestors industrial peace. To meet workers' rising expectations, the gov-
ernmnent in 1969 instructed the National Trade Union Congress
(NrUC), the officially recognized labor umbrella organizaton, to suggest
new directions for the union movement.
The NTUC,which is in effect a branch of the ruling People's Action
Parry (PAP),made two recorimmendations:govemment, management,
and labor should work together to avoid antagonistic wage negotations;
and workers should organize into cooperatives with a stake in the econ-
omy. To firrher cooperation, the NTUCestablished the trpartite National
Wages Council (Nwc). Each year, the NWCreviews wage and economic
trends before advising the government on wage adjustment guidelines.
Although recommendations are officially nonbinding, the govemment
usualy endorses them wholesale, and business and labor use them as
wage settlement benchmarks. The NWCseffectivenesswas evident during

I65
. , 1, ,t77;tS>ri C LE

the 1985-86 recession,when it persuadedlabor to accept wage cuts of

about 12 percent. The rAPalsoreorganizedthe NTUcitself,transforming
it from a negotiatingagent to a providerof goodsand services.The NTUC
has since establisheda number of worker cooperatives,rangingfrom dhe
FairpriceSupermarketCooperative,with more than 240,000 members,
to the Comfort Taxi Service,with 5,500 members(seebox 4.2).

i-ONQO5tTHEE1R.
FIRCOOPE K
ATVESORAGANIZD*.BYSINGAPORE'S
LW. ,>NtwOACoA't
bf.lrodany, byOf&r( fridhbYii ari
a I I ..... t

f,ruung_. eforp1crncuorVrne
-r aarryrJo 80.ebrrelxelw:pa
6N-ii!'Ode4if&1I te .Tdfoms;. iAti4niz oflisr int

tt-. m,drNeriiBccause
lcenseds
n q,qfgoods41n1.p 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 *

-ac,a.l |cd h
rt
!h
,-s;-'kasb'Vid'e6B'6&f
eranve
'j
;.si iservitlones,
.tovbay4larg&numbers.ofwasxn
;~~~cas.D tmhmr
dnL1c Uii dehWdiilcdt

,s e sout__ro z6b-d_ admzn, tamnve5

*~~~~ g

2 .,i.. Mos D;9iS!si

3'r-7.aWef i . ,5i2&~o ,.Kt,> Zrix
~~~1miUtw Xonda;toso&.faealtrncensei
unDedarUG.
id.@ 2
r,=LTz.eiC nd.sotonuceSE
oarEs;;2ancasc. ,C

;~~~~~~~~~~~~~g . .:.
30 X123 ,T.4 ,5r S 25R"-iTr7:aCPVSi:*' rtt''''';>r*CA -RX 'SgBgWtt
f- I;"
-~ ~~~~~~~~~~~nf.wn . Ei .ii
-~ ~ ~ ~~ ~ ~ ~ ~ ~: ~ -:"~ ~ .~~ - .. - ..- _'

, N

6 A
 :s~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I

Other economieshave forestalleddevelopmentof largeand politically

powerfulunions with less formal arrangements.In Taiwan, China, for
example,closemanagemcnt-laborcontact in small and medium-sizeen-
terprisesreducedworkers'interest in union representation(seetable 4.3).
Moreover, the emphasison labor-intensivemanufacturing generated
rapid increasesin the demand for labor. Consequendy,as figure4.1 il-
lustrates, from 1952 to 1979, inflation-adjustedbasic wages increased
more than sixfold in manufacturing and more than fourfold in agricul-
ture. The total increasein compensation has actually been higher, since
basic wages are supplemented twice annually with profit-sharing
bonuses.

Technocracy
the Economic
Insulating

IMPLEMENTING SHARED GROWTH TO SOLVE THE PROBLEM OF

politicallegitimacyrequiressharing but it also requiresgrowth. We
now discussHPAEinstitutional traitsthat have been critical to achiev-
ing both thesegoals.Foremostamong them is technocraticinsulation-
the ability of economic technocratsto formulate and implementpolicies
in kceepingwith politicaUlyformulatednationalgoalswith a minimum of
lobbyingfor specialfivors fiom politiciansand interest groups.Without
i, technocratsin the high-perforningAsian economieswould have been
unable to introduce and sustain rational economic policies,and some
viralwealth-sharingmechanismswould have been neutralizedsoon after
their inception, as was land reform in the Philippines (seebox 4.3).

Table4.3 LaborUnion
Activit byScaleof Businessin Taiwan,China
(pnenetz

,p- *-3,--es No . -ojo.itX6

-i-:;. Norrenponre
4:i30-49 f 4 .-. 9: 06.M 008
9 . .85.
29S. .; -f67- So28 ,a 015
51:05 .44 . .51 0
, -2 - .t 19.45 . 6.26i_'
:X05 r 37T'..
r'-SourerTaiwan,
clnnChial
Bure of rheTii _Proviial Government,
Annual
tSt ~ &'980-84 :

167
 4.1 IndexofRealWags 1952-79
Figure
o;#{~~~~~~~ ii. I
:.: Percent(IndEx:1952;100)
l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -

I I ' ~~~~~~~~~~Murnsfacturln

300
.:: 0~~~~~~~~~~
Fue41Inv of RelWgsI927

400 - glAtr

200-

p1952 1955 1955 1M1964 1967 1970 1973 1976 1979

SPrcent, Kuo, and Ranis (198 1, p.2).

Direct Mechanisms insulaton

Hong Kong offers the cearest case of insulation, since the colonial
government, which has been stronglycommitted to free marker forces,
needior dependon thepoliticalsupportofprivateintereststomaintain
power But insulaion does not necessarilydiminish as democratic insti-
tutions proliferar. Indeed, the technocracyin authoritariansocieties
-maybe utterlywithout insulation, while those in some democraciesare
heavilyinsulatedfromoutsidepressures.
In Japan, insulation chacacterizesnot only the economictechnocracy
but nearly the entire bureaucracy.Japanesebureaucratsdraft laws in
consultationwith the policycommitteesofthe rulingliberal Democra-
tic Partyand privatesector representatives;the diet (parliament)merely
ratifieswhatthe bureaucracyhasprepared.Becausethe bureaucracyhas
independentpower,it can oftenignorepressurefromindividualactors
in the privatesector. Moreover,rhe bureaucracycan achievepolicy goals
through administrative guidance-nonbinding recommendations.
Guidelinesale enforcedwith generousincentives-suchaslicensesand

ixG
 0 ~~~~~~~~~~~~~~A
N- -INS

. ~ ~~~~~~ ~. . k t.di i h
.u' . .____.~~~~~~~~~~~~~~~~I

-: EP,Pr
::~~~~~~~~~~~~~~~~~~~~~~I:
u G 1;ivestex
::: 4W"-
n6nLY:Ji CVt'i: 9

L c 3 utyt~~~~;g ,psuN'_O ,i dgm

.
:~~~~~~~~~~~~~~~~~~~~O
Ce
Ico,
u~~~~~~~~~t)r,, V ormmg c

0 -wsC
0:00000~~~~~~~~~ICR 3986 :P3p
-~~~~~~~~~~~~' a an rcotan aswep

0.''0 .~~~~~~~~~~~~~~~~~~~~~r

.f ' .wLtS~~~~~~~~~~~~~~~~~~~I
- 3 f : ' -:~~~~~~~~~~~~~~~~~~7

}:g_D
,-. . - D~~

, X; 0~~~~~~~~~~~~~~~~~~~~S
. f , . : .
.t: - := -- f - f . D6d .t- :

D r r ;0~~~c-un0000.;:0-0--0I6.:
 'R'AtCL

foreign exchangeallocations-and the implicit threat to withhold them

from companies that refuseto cooperate.
One of the distinct featuresof the b'ueleary processin all the HPAEs
is the primacy given to the bureaucracy over the legislature in drafting
laws. Though the.legislature must approve the laws, the bureaucracy
studies, analyzes, and drafts the bills. It has considerable control over the
agenda and can use this to minimize political pressure.
IThejapanesebureaucracy is further protecred fiom political pressure by
the Natonal Personnel Authority, an independent body that sets the bu-
reaucray's pay scales and promotdon policies, administers civil service
exams, and makes most appointments. Japan's prime minister names only
his ministers and, except in a few caes, one of the two vice-nunisters in
each ministry; the National PersonnelAuthority is responsiblefor the resrt3
As in Japan, the bureaucracy in Korea has a long tradition of inde-
pendent policymaking. This was boosted in 1961, when army General
Park Chung Hee seized power and reorganized the govemment to pro-
mote rapid development. He createa a powerful Economic Planning
Board that has had broad budgetary authoriry. Acdministraive control
over the banking system has given the bureaucracy strong leverage over
big business, which relies on government-backed financing to fund
rapid expansion.4 The partnership has generaly been uneven, with the
bureaucracy dominating the relationship.
Whe few of the HPAEshave such thoroughly insulated bureaucracies
as Japan and Korea, each boasts at least a small, ideologically consistent
technocratic core that answers directly to the top leaders and therefore
has some independence from the legislature and other sources of politi-
cal pressure. Indonesia and Thailand, where bureaucratic traditions
more dosely resemble thse in other developing economies, are instruc-
tive on this score.

Inflation
Control
and DebtManagement
in Thailand
andlIndonesia
For historical reasons, Thailand and Indonesia have been pro-
roundly concerned with low inflation and debt management. Thailand
is unique in Southeast Asia for being the only economy in that region
not to have been colonized, in fact one of the few in the.developing
world. One reason for this remarkable alchievement was the openness of
the Thai Kingdom to trade with foreigners. Another was the monar-
chy's concern with avoiding indebtedness to foreigners. And a third,

-170-
 A IN.r:a.n.s-ANNm.

and perhaps the most important, was the establishment of a bureau-

cracy in 1892 that enabled the monarchy to centralizepolitical control
over the kingdom'smany regions.Historically,then, the bureaucracyin
Thailand has alwayshad a powerful influence over public administra-
tion and policymaking.But it was-not until 1957, when Gcncral Sarit
took over the reins of government, chat bureaucratic reforms were in-
troduced to address macroeconomic imbalances.Since then, keeping
inflation low and efficientdymanaging debt have been among the gov-
ernment's key concerns.
These same issueshave been extremelyimportant to the Indonesian
governmentsince President Suharto took power in 1967. The previous
Guided Dem'ocracyregime of President Sukano had let goverrnent
spending run out of control, leadingto annual inflation rates of 1,000
percenicand other seriouseconomic problems induding a rapidly dete-
riorating infratructure. Suharto'sNew Order govemment ascendedto
power with an explicit conmitment to fight inflation and develop in-
fratructure, particularlyin the rural areas.This led to passageof a bal-
anced budget law and to financial controlsthat addressedinfrastructure
ddevelopment.
To achieve the rwinobjectivesof low inflation and prudent debt
management, both governments created mechanisms tO insulate their
-: .economic redcnocracies.In Thailand, the government'sBudget Bureau
has tight control of the budget drafting process and has maintained a
stable exchangerate and low inflation. To draft the budget, the Budget
Bureau consultswith the National Economic and Social Development
Board about proposed public investmentsand with the Finance Min-
istryabout expectedrevenues.It then deternines together with the Cen-
ttral Bank how much deficit financingthe economy can toleratewithout
increasing inflation. Having detcrniined aggregateallowable expendi-
ture, it calculateshow much each government agency may spend and
:forwardsa broad outline of proposed expendituresto the cabinet.
The cabinet may propose changesto the outline before forwardingit
to a parliamentary budget committee for evaluation. But since neither
the cabinet nor the committee receivesbudget details, the bureau'spro-
posals usually survivewith few alterations.After approval by the com-
mnittee,the budget outline is submitted to the endiic parliament fior
routine passage.
Throughout, the bureau draws support from budgetary laws that
-'-'lUriiit the governmentdeficit to a small percentageof the year'stotal ex-

'7'
 VMRCLE
,

penditures and that cap the percentage of the budget that can be spent
servcing the foreign public debt (the current cap is 13 percent). Parlia-
mentary rules offer a fiirther unusual guarantee against runaway deficit
spending nonrninisrerial membe of parliament on the budget com-
muitte may propose-only minimial changes and in any event may nor re-
vise the budget upward. The results have been impressive (see figure
4.2). Except for the diruptions caused by the 1972 and 1979 oil shocks,
the annual inflation rate has been dlose to 5 percent. Few. developing
eoinoies can lay dlaim to such an achievemet and even more ad-
winced HPA.Rs such as Hong, Kon-gand Korea have had higher rates. This
demonstrates how a small technocratc core, insulated from politics, can
set a positive tone for an entire economy.
In Indonesia, too, technocrats keep a tight rein on the budget. Under
Suharto, a balanced budget has been the cornerstone of government 6i-
nancial policy: In 1967, the legislature approve-da law limiting expendi-
ares to domestic. revenues plus foreign assistnce. Since then, he
finance ministry has institutionalized a review that requires each min-
istry to justify prposed expenditures on a line-by-line bAsis. In addi-
spend ~ing
nminisr Iemer
of priaet onIh ugtc

Figure4.2 ThailandConsumerPrice Index

(percentagechangefrom previous year)
Percentagchange
30 --

20--

10 -i' k.ii---- 4j
If

1955 1958 1961 1964 1867 1970 1973 1976 1979 1952 1955585 1

&unr Innualtional MoneratyFund da.el

r72
 S~~~~~~~~~~~~~~~~~~~~~~ N ST

dion, parliamentary rules restrict the legislature'sdiscussionof thebud-

get to broad policy issues.Armed with thesc instruments, the finance
ministry has establisheda macroeconomic environment favorable to
growth. Inflation has dropped since the 1960s and has generally re-
mained low and stable (seefigure4.3).

Indirect
lusulation
Mechanisns
SomeHPAEshaveenhancedbureaucraticinsulationindiiecdyand gen-
erally inadvertently.For example,wealth-sharingmechanismsreducethe
potential gain to interestgroups of intervening in policymakingand im-
plementation,sincesomeshareof the firuitsof growthhas alreadybeen as-
suled. In addition, when wealth-sharingmechanisms reducethe size of
oroductionunits, aswithland reformand support for smalland medium-
*. size industries, the marginalcost of organizing and lobbying increases.
'Withthe benefitsof lobbyingdown,and the costsup, interestgroups tend
to be more willingto leavethe policyprocessto the technocrats.
Reozganizationof labor from industrywide unions into company
unions, along with labor federationprovisionof goods and services,has

PnceIndex
Figure4.3 Indoneia Consumer
(perentage changefim previous year)
Percentagechange

- 0 f -'-N-

30 -- rv--
20 -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

0- . ; I I I A | ~~~~~~~~~~~I
1 j I

0~~~~~~~
- ~ ~~~~~~
- f I

-~01970 1972 1974 1976 197t1 "SO 1982 lS84 1986 1.98S 1990

-5400 l- :n eItional Monc4y Fund daa

h . _ _'. -17_
. .
197, 197
0 2.97 197 29: =8 196
;8 819 .98
;our.neranaa Mocx Fun darn00. ,
 similarly reduced the marginal benefit and increased the marginal cost of
collectiveaction. Thus, in contrasr with wokers in many other developing
economies, workers in the HPAEsarc more lihklyto refrain from work stOp-
pages and other disruptions and fiom lobbying the government for man-
dated wage increases. Because employers faced fewer demands fiom labor
they, too, have had less incentive to press demands on the reclnocracy.
Of course a powerfil bureaucracy insulated fiom external pressure
can also be dangerous. Indeed, the description ofjapan's bureaucracy-
an independent force that controls generous incentives and issues vague
guidelines-is nor unlike that of bureaucracies in ocher economies that
engage in corruption and malfeasance on a vast scale. In Japan, itselE
some ministries are perceived to be captured by the interests they regu-
late. These ministries enjoy less power and prestige than those that resist
politicization. 5 Even well-intentioned bureaucrats may become counter-
productive or lose sight of the goals they are supposed to pursue if bu-
reaucratic insulation means that they are not held accountable for their
decisions. How have the HPAEskept insulated bureaucracies honest and
responsive to legitimate private concerns? The next two scctions address
these perennial problems of bureaucracr. honesty and responsiveness.

Buildinga Reputable CivnlService

While insulation of the techmocracymay be necessary,it is hardly suf-

ficient-in the long term. To sustain growth, a bureaucracy must have the
competence to formulatEeffectivepolicies and the integrity coiimple-
ment them fairly.The more policymakers attempt to fine tune the econ-
omy, the greater the need for competence and honesry. Among the
HwAEs,Hong Kong, Japan, Korea, Singapore, and Taiwan, China, have
been successfiLlin building relatively competent and honest bureaucra-
aes. The Southeast Asian NiEs,Indonesia, Malaysia, and Thailand, have
gradually introduced measures to upgrade theirs, with Malaysia in the
forefront of the process, but they still clearly lag behind more industrial-
ized economies.
Despite the stress on culture in much popular writng about the
Asian miracle, the most successflil bureaucracies have not relied on cul-
ture alone. HPAEbureaucracies have employed numero us mechanisms to
increase 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

* Recrauitmentand promotion must be merit based and highly

competuve.
* Total compensation, induding pay, perks, and prestige, must be
competitivewith the private sector.
* Those who makc it to the top should be amply rewardecd

Merit-BasedRecnuibnentandPmmotion. Recruitment in Japan revolves

around highly competitivecmil serviceexaminationsadministered each
year by the National PersonneIAuthority. For the higher-levelexams,
the rate'of successhas been less than 8 percent. Indeed, in 1980 the av-
eragerate of successwas 2.7 percent (Kim 1988). Promotion within the
bureaucracyis basedon a combination of seniorityand a host of perfor-
mnanceindicators that dif&r across the ministries. Becausethe number
of personnel in the bureaucracyis fixed by law, compention for promo-
tion can be intense (seePempel 1984)!'
Korea also relies on exams and performancebut has supplemented
thcsefilkersby,,recruitingheavilyamongacademics with advanceddegrees.
(Box 4A describesthe turnaronundin the Korean civil service in the
1960s.)This has resultedin a prolifcrationof researchinstiutes associated
with ministries. Since die Economic PlamningBoard established the
Korea Development lnstiute in 1971, most other ministries have fol-
lowed SUiL Most institute researchershave a PhD., many from a re-
spectedforeignuniversity(im and Leipziger1993). LikeKorea,Taiwan,
China, recruits heavilyfrom academnia, primarilyto offsetweaknessesin
the civilserviceexamsystem.Academicsare recruitedfrom major univer-
sities,usuallyfor fixedperiods rangingfrom three to sixyears.
In Malaysia,civilserviceexam resultsaresubjectto affirnmative acion
guidelinesmeant to increasethe number of Malaysin gaovement. Be-
causethis reducesthe pool of eligibleapplicants,it probably hinders de-
velopment of the bureaucracy
Incentie-BasedCompensation. In bureaucracies,as in nearly everything
else,you get whar you pay for. Table 4.4 suggs that relativepay is sig-
- nificandy berter in the Four Tigers than in other economies. Relative
pay in Mvialaysiaand Thailand is about the same as the averagefor other
low- and middle-in&me econo-miesbut is brillsignEficantlyhigher than
-in the Philippines,which is widely perceivedto have one of the weakest
bureaucraciesin SoutheastAsia.In general, the more fa-vorablythe total
public sector compensation package comparesto compensation in the
private sector,the better the quality of the bureaucramcy

'175
 .... ~~~~~~~~~~~~~~~~~~~~~~~~~~~
'e.:.
2 ~~~~~~~~~~~~~~~~~,

~ ~~ ~~ ~.~ ~.4 ~
.~~ tyJQ.y~i*S ~ t
L~~~

trta r'A)l4A'

~d. .
Wotsurp'sne4ingapre,uowhidii'widely' peroivd' tohv> he
;1 tNrnsnorecdRBErmwt 1 an oinon
'fie
bCS ontlybaejmlayofacputHministercvdiningapordranws d:o
JapanF.t anfiS Korca alsomke^6 -
a, dcieatem; effor*;t
>Jto<> match the private
sector I Japn,o£teNationaul Pcrsnne icAuthmForityHo annuay sur y
thepulic sectorwforsentry-level,m_id-l n posi_-io dif-

Koreabascsldsaelwrthan salnre.otarein
lare pivaslrbed
copnis-
bu W-pine aflownrs hlp to narrowr th;e ; *iffcen- Tain acon
of salary
plus* bonuss
in thlet
pvteo seaorzan sahlary
plus hatltowne
in
S$22,100to S$7,2
;3iF (aou US$i3,812'l_e9
to US$1,390)
s whl .. a n
Japan ndKreaalo ak dliert effort tomathte s
privaste
seconn Jeapan th@eNationalt Per,.sonnelkAu*thoit annuallyS*,-,{F.4
surves
thu dso
o mpanis toderin :job pay rates$(Kimo1988 p.) In
Koea bs p;sariesarcloer than saares
gin's mtost compentsan uaprihtabreacracyb in largepmays
pr,;iFitvate
cmpaies.x
itasn iburacr
ofisalar plusthly onuss
b7estT basensalayoaflmister
te riae sector amnd alrypls a llowncsin
iSngorerang

-~~ ~ ~ ~ ~~~~~ne in totasrnly compnsatione appear to biel smlerie tOhane in most

176-go' otcmern n pgtbracaqspy t ueurt
 A 1NSTwt:ANWpNS

Table4.4 ICPEstimates
of PerCapitaGDPandRaftioof Publicto PrivateSectorSalaries,1992
(penent)
Seir leve 'Mddee GDPpacai
HPA&; &ononq/regran
,.f.4i. -a), A B~A _B-A ;Ennyluas

';f - i ; *;
apo*r; - n4:b 114. 1150
- 107-'.
7.
. 920
:Korea,Re.. of -. ,':o- 98. 8 8 .3 822 83.6 -710
; Ta aM China ;65.2 *.; 60.3 63.5- 65.8 : 60.0 7,954
0LMa1-aysia 40 33 343 - -5001
5oo
Thailand. 47.G.6-;
0 f; . 0:--2-
i7.50
.: . X -;33.6
. - .-43.0 9 4418.0

0 - .Philipie i277 L2r;4:;.3 25.0 325 62. 232

A-

;Tind-anToa,
4 . ,3 63.5b - _. 63.5 76'9 7 779 8,51 ;0°

SnbSsjz70Afia
~~2 tmgu; - - 3 1P 3 7.1
007 6-g, :---06
60.0
;~aLua 2416 24 25.0- 2 4

Somi- a ;. 35 11.2=, -' - 2 122r-

2. i~ '.

;wCili toth UN4. Ineraoa ComparZionPr e. ,. reth to i- ernational ' ! ;. -

-LEsciae
a. indude alloaneos and boiisetDtarnibrpriarteseor comnpmsdoui ae ased in
asurveyof impranis with500or more;-

lAor smalrilab,
on average compdranontn iconomige
u te o (o

public sector.
In other HPEs public sector base salaries are consistendy lower than
:-their privare sector counterparts (table 4.4).7 W7hile figures were not
availablefor Indonesia, anecdotal evidence suggests tat te private--
: :tpublic wage gap tre is the largestamong HIAEs. A stdent graduating
-from an Idonesian universitycan expectto earn t o to four times as
much in the private sector as in a government job. Nor surprisingly,
c Som
frsle 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,
job:seciny,
cas
c cll -=AE-s.
reare
and other
amn
advantages
H
of public employment will persuade

se y, and other 2civa

-- - oDfpu - plymn : wil: peruad - -:: r~~~~~~~~
 V ItALE

some talented individualsto forgo higher earningsin the privatesector.

The trick for governments is to hit on a combination that will attact
competent individualsto the ci-vilservice.
Prestige is of course partly a function of culture; thus East Asian
cultures that regard dvii servants highy have an advantage in building
bureaucracies.Confurcianthought, with its venerationofscholars and pre-
remainsa pow-
occupationwith written tests,especiallycivi serviceexamns,
erful force in Hong Kong,Japan, Korea,Singapore,and Taiwan, China.
Not surprisingy,thesesocietieshave producedstrongbureaucraces.
But bureaucracy building need not be culturallyconstrained. Merit-
based screening with economic incentives to attract a large pool of
applicantscan produce high status for civilservants in any society.Pro-
vided that the exams are sufficientlycompetitive,individualswho pass
them will be relat-ivelyrare. Thus passin the examnor winning a com-
petitive divi servicepromotion becomessimilar to being inducted into
a hail of fame; it signalsthe public that an indLividualis special. 'While
the value of prestigecannot normallybe quantified, the monetary value
accordedto some Thai royal decorations,describedin box 4.5, suggests
that the value of such pre-stigecan be very high indeed.
job securitycan also offsetslightlylowerpay.In most HPAE bureau-
cracies, dismissa is unlikely unless the bureaucrat commits a serious
mistake. Security of tenure translates into lower income variability,
which in turn providesincentivesfor public-employeesto accept a lower
salary.Furthermore,in many economiesa public employeecan look for-
ward to a retirementpension, a benefit not normallyavailablein the pri-
vate sector exceptin large corporations.
AClearPathtothe Top.A final requirementfor a successfuilbureaucracy
is a well-defined,competitivecareerpath with a substantialprizefor those
who make itto the top. In Japan and some other E*AEs, retrement comes
early,and the rewardsto a successfUlbureaucrataresubstantial,extending
beyond the pay; perks, and prestige to indude a golden parchute. The
Japanese have a specialword for it- amakuckr4which means "descet
from Heaven"'and connotesa lucrativejob in a public or private corpora-.
tion or occasionallyelection to politicaloffice.Similar practicesexis in
Korea,Malaysia,Singapore,and Taiwan,China, in differingdegrees.
Unlikthe much-criticizedrevolving-door,which whisks bureaucrats
in other economides fiom the pubfic sector to the privatesector and back
again, descent from Heaven is a one-way tp. Moreover, rethringbu-
reaucrats in Japan do not choose their sinecuresbut are assignedthem

178
 0~ ~ ~~ ~ ~ ~~~~ ~A IN.ST-W A NijN

,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~
- B The Iu1e.f i NXA
&A, nL~

: IIIGHLYCOMPErITJV& MESBASED iR:ACkAUnENrTO` A BUW-

, g
-auaa

E6~~~at
creaeswuciors

fieai
Ilentrusns%lenlowingimem-w.?-

ftheel -cr o
the
r .
witha'L,
| - r-
3'-
r
ra;cmtomsfo:grslcne ,1
unnv,zes;2ix cra~tcsanzespnt de4Ff*
-~~~~~~~~~~~ruCG-2~eP

roc=on Dbutmntiad-pr
MouShaMW no noiv;Enr Be--'-4

iWZIi&
rbiidMdZl2&• ~ 6n,isui, -.

~
1,4u2 ~ ~ ~ -
i` :2O to

idii vatue beisreweeninstrvies,i

int.scl
except cotxmpedarive
l mana levimnmcnt
e iteia e vy

faos Whl thcaw

re is n denying
whohatv firns welcome etredseniceor bu-

faos Whil there is no deyn tha firs elom rtiedseiob

crrcntueo b strs~eopsilngthe
reuras partyo oaucrac pusfultiesih
h bureZaXi the reatiree ar
mens thatgzovenmersternno 'i,sg ofteni2/
aaftiexceptonal
naueo mnagers \alubl
in theivrown t9ightF,.4^S-t>><is*
h v ueacaymasta the rtirees LiPare
ofe "

i
CiCompetemncBmet~ids
ompetitivniscourgesenvironmesny
M %ehanimsthtidc witein ciamonalso
nbrityicospoDBSimpeteny anbtworyof e ministries,
Ienace
gofclmgi hownesty
guard againrs
Competiimrlon firsq o iffern apst-rticrement jobs
cin etis hae
dscoraesdhoetsice.sgdistar<acsuovryo a corrupt specia for or dis-^
hoet diquaif nyiapplicant,. in.rsocietiEes
c wull4ould havlue publicr

bya acommite
wthatincthei cminitrycThese faturs, combined wihoanesy

179
 i;RA ACLE

service, dismissal brings disgrace on the entire family; hence family

members have an incentiveto discouragecorrupt activities.This mech-
anism appears to be especiallycffectivein Korea, where tradition has
placed a high premium on goveniment service.
Strictly enforced dismissalpolicies also help reduce graft. In Singa-
pore, large or small corruption is pcnalizcd by expulsion from the civil
service and withdrawal of one's pension. In addition, distissed public
employees are unwelcome in the private sector. Under such circum-
stances, the cost of engagingin comrpt acts becomes nearlyprohibitive.
Other economies in the region have governmentanticorruption squads
that focuson enforcementand ptnishment
As Japan and Singapore in the early 1990s illustrate, rich econo-
mies have an advantagein building honest and competent bureaucra-
des that further the creation of wealth. For example, economies with
higler per capita GDPcan afford to pay bureaucrats better and to
fund the high-qualinr education that results in a better pool of appli-
cants. The challenge for developing economies is to concentrate
availabie resources where the can most effectively replicate the
advantages of the industrial economies. This is nor to say that cor-
ruption and bribery do not occur in ithose HPAEswith reputable bu-
reaucracies. These social evils exist in Japan, Korea, and Taiwan,
China, and even in Singapore. The critical point is that they have
occurred less frequently and in smaller magnitudes than in other low-
and middle-income economies. Indeed, most of dteHPAEs scored sig-
nificantly better than other newly industrializing economies in survey
rankings of the prevalence of improper practices and special interest
group lobbying contained in the 1992 World CompetitivenessReport
(World Economic Forum 1992).

Developing
BetterBusiness-Govemment
Relations
Whereas most other low- and middle-income economieshave only
recently turnedto the private sector to stimulate growth, the HPAEshave
done 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 establishingcooperativerelations
with big businessand in some caseslabor.
Legaland ReguLaztoy
Mmlh Envirounet.Compared with other developing
economies,the HPAMEs have been more successfulin creatinga legal and

-80
 - t;- - - ~~~~~
4N.l ~ ~~~~~~AN

regulatoryenvironmentconduciveto privatesectordevelopment.Based
on selectedcriceriaor fourteennewlyindustrializingeconomiesin the
1992 WorlzCompetitiws Report, in whichlowerscoresindicatea better
environment,the HPAES had an averageoverallscore of 32.3 (with a
standarddeviationof 18.4) comparedwith an averageof 56.9 (with
a standarddeviationof 13.3)for the odtereconomies.8
EncouragingSrall Enterpriss Smalland medium-sizeen-
andMedinm-Size
texprises
generally
havedifficultygeridngcreditto financecapitalimprove-
-ments, obtain better technology,and break into markets, particularly
export markets.Hence, a nontrivial SMEsector cannot be established with-
out external assistance. Studies indicate that the northern-tier HPAEs-
Japan, Korea, and Taiwan, China-lhcv been able to develop relaively
effecdve 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 successwith itS support progams(Berry and
Levy 1993); however, the private banking system appears to have been
hdpf3i irndeveloping some sNEs. Both Malaysia and Th h atiV
V
support prograams but not much is known about their effectiveness.

WooingBigBusiness

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 can
be assured of a share of rents. These art, in effect, an institutionalized
form of wealth sharing aimed primarily at winning the support and co-
operation of business elites.

EastAsia'sDeliberationCouncils
Japans efforts to establish such institutions are the most widely rec-
ognized and have been the most thorough. Since the beginning of the
postwar 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 as
pollution or finance; the second is organized according to industry, fr

: 8ii
 exaple-, automobiles or chemicals. Government ministries establish
the councils,which are generally associated with. a specific buLreau
within the establishingministry
Each councl providesa forum for governmentofficial and represten-
tatives from the private sector-business, labor, consumers,academia,
and the press-to discusspolicy and marker trends and gencrallyto ac-
change information. Consensus is encouraged. Figure 4.4 shows the
processof 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 research
group prepares a report for the approprate deliberativecouncil, where
policy 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 MITmpolicy stands little chance of success.9

Figre 4A Indusial WsionFonnuationProcess

Conductsurver r| MMjuniorstaff Hearing

compile
data studygroup - Learned
individuals
- Interested parties
- Overseas erployees
- Local rpresentadves
Prepare draft - oUers

> Research group (Outside lcturers)

(Subcommittee) Otdelcue;

{eedbacM (Briefings.subcomnittees' reports)

.~~~~~~~~~~~~~~
-|Deliberativecouncil

Publec r Ationr
- Public Aions

(Report) ---- -* - Explanatorymngs

- Lectures
-Others

S&urn-.Ono (1992).

: X - . ' . -~~~~~
 In Korea, from the mid-1960s to the early 1980s, goverment and
private sector relationswere also dloseand cooperative,although some
in the p'rivatesector argue that goverment was coostrong handed and
dictatorial.10 Goverment and businessleadersmet oftn and regularly
although less formdlly than in Japan. Government solicitedbusinesses'
views and included them as a criticalpolicy component.
Until the early1980s,the most imporgnt communicationchannels
were monthly rexport-promoton
meetings, at which the president of
Koreahimselfpresidedoverdiscussions berweenthe economicministers
andictopbusinessleaders.Businessmenexpressed theirviewsabout mar-
ets,regulations,and potentialplans.The president:
theninstructedspe-
cific ministers to attend to each important issue;at the nt meetng,
* mnUnisters
dlivered a progressreporton theirassignedtasks.
Thesemeetingsweresupplerm ented bygoverment-initiateddis-of
siongroupsinvolvingcompanymanagersimiddbe-level governmentof-
-ficials, and expertsor scholars.As in Japan, there were basicaluytwo
types: fumnctional
(for example, taxation) and sectoral (usuallyfocused
on a particularindustry).In both, bureaucratsinvitedincerestedpartes
cxperts topresenttheirviews.The focuswason informationgath-
and
- eren bureaucratsmadethe finaldecision.E vernmsonoicitied oin-
-cor poate al viewpointsin policies.
'WithKoreaisgradualdemocratization sincethe raid-1980s,irlations
betweengoverinmetand businesshave becomemore distant and the
meetingslessfprequent.Recently,however,as the conomy hasslowed,
thereis a growingconsensusthat monthlymeetingsshouldresume.
'Similarconsultativemechnisms appearto be evolvingin Malaysia.
From independenceuntil 1970, the governmentfocused on infrastruc-
tueandgenerallyLeftthe privatesectorto fimction freely.From 1972 t'o
1985,as the goverment respondedto the racialdrots-of1969wichthe
NewvEcovrnomic Polncyd
relationsbetweengovenmentand bness
tendedto be contentous,in partbecauseof the requirementthat aslen-
terpriseshave at least 30 percent Malayparticipation.Since 1986
business-government relationshavegradualyimprOved, culminatingin
1991indhe New DevelopmentPolicy-(ND), whichhas strucka more
Ne balance
cn betweengrowthand distributon.For example, uner
cvenalac beCe rwhaddsrbnn o xml, underthe
NEP,all firmswith morethan M$250,000(US$104,167)capitlization
--wererequiredto complywith ethnidcy-based
shareholdingrgulations.
Under the NDP, the cut-Offpoint ha been raised to M$2.5 milion
(US$1.04rmillion).

183
 The NDP specifically seeks to emulate Japan's government-business
consultative mechanisms, a determinat-ion reflected in PrimecMinister
S~~~~~~~~~~~cni taiv Bin
Mahathir stmcFu I
Mohamad's 'Look East Policy." Formal consultations
began in 199 1, when the government invited capt-ainsof industry to dis-
cuss the national budget with finance officials in a much publicized
budget dialogue. In thessame year Mahathir aeaaed the pinnace of the
conulttivstuctre, the Malaysian Business Council, which'i chaired
bycth pruie minister and includes more than sixty members from in-
dustry, labor, and govermennt.
Thailand has been slower to develop consultative mechanisms. In
1983; the govenmmentestablished the National Joint Pubfic and Private
Consulkative Committe (NJPPCC)with the prime munister as chairman
and the head of the plnning agency as searetary general. Othmermem-
bers are top government officials, particularly from the economic agen-
aies,
and nine private sector represenmives dmawn from the Thai
Chamber of Commerce, the Association of Thai Industries, and the
Thai BankersAssociation. Provincial consultative comrnitees have been
established to supplement the NJPPCC (Samudavanija 1992, p. 26).
Views about the effectveness of the committee differ widely.
Public-private consultation is most cxplicit in Singapore. The private
sector paricipates in policymaking in many ways. For example, private
citizens serve as directors on government statutory boards and as mem-
bers ofad hoc government advisoy commirtees. In both capacities, they
review policies and programs and make recommendations for official
considemraion.In addirion, the government regularly invites chambers
of commerce, trade associations, and professional societies to subniit
their views on specific issues (Chia 1992, p- 11).
The Singapore ents efforts to manage dhe economy are most.
fully apparent in the National Wages Council, which indudes rp-resenta-
tives fiom govenmen; business, and labor Becauseof this tripartite struc-
tumr, the Nwc fulfills several coordination funcuons, simultaneously
fiurthering the govenment's guidance of business and of labor, as wel as
business-laborcooperation. Having rejected more typicaldenocratic ifrms
as too adversaial, the PAPhas also etablished a wide variety of citizenasso-
ciatiorLsto provide channels of communicanon with the government
Formal mechanisms for business-government interface are almost en-
tirely lackingin Taiwan, China. Instead, coordination is handled tirough
large public entrrise that provide basic inputs to manu £cturers
and
through enterprise owned by the Khomintang the dominant political

184
 AN NST
.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- ----- - ---

parry.-1 Indeed, public enterprises,broadlydefined, account for a larger

share of GDP in Taiwan,China, than in Koreaor even India, where cen-
rualplanning plays a key role. These public, quasi-public, and party-
owned firmsgivethe governmenta high degreeof control overlarge-scale
sensitiveinvestmentsand a means to influencethe generallysmallerfirms
that oomprisethe privatesector.Over the years,the largerole of the pub-
lic sector has become lessobvious, as some parry and govemment firms
weresold to cooperative private interests. Coordination is also facilitated
by middle- and lower-level officials of the Industrial Development
Board, who make routine, infornal visits to thousands of small and
medium-size enterprises within their jurisdictions.
Like Taiwan, China, Indonesia lacks the formal government-business
coordinating links of other HPABs;this may be pardy explained by the
weaknessof its bureaucracy.Coordination appearsto be handled by in-
fbrmal networkslinking senior officialswith major enterprises.'2 Little
is known about thesenetworks,but theyappear to have produced a high
degree of cooperation in the economy. In contrast to the deliberation
councils,where rulesare more transparent, thesenetworksare more sus-
ceptible to captLre by participants and to corruption.
Deliberation councils natually reflectthe history and culture of the
societyin which they operate Evenso, the experiencein the HPA:Es sug-
geststhat their applicabilityis not limited, as some analystsmay suggest,
to the Conf&ciancultures of Northeast Asia. Becausethe operation of
deliberation councils has been so importanr in winning the support of
business elites, we now look in greater derail at the operation of these
councils in two very differenteconomies,Japan and Malaysia.
Japan'sCouncils: e Foreunnes The concept of deliberation councils
in Japan is most closelyassociatedwith Mm. Because of m1Im's role in
developing and implementing the economys industrial policy, the
public-private consultativeprocess was most dearly manifested in the
councils attached to it.
Government-businesscollaboration has long been a hallmark of the
Japanese economy. The zaibauu-Japan's large corporate holding
groups-were dosely linked with the military-bureaucraticcomplex of
the 1920s and 1930sand were a key element in building the economy's
industrial and military base. At that time, the influence of big business
matched and sometimesexceededthat of the bureaucracy.
The U.S. occupation forces,ideniliing the zabatiu as a keyelement
in the build-up ofJapan7swartime machiney tried to limit their power,

185
 = - . i QSI.YMiI
R-AC L £

in part by conferring more power on the politically independent bu-

reaucracy.Ironically,this not only strengthenedthe bureaucracy-
particularly MITT,which became the center of developmcnt policy
formulation-but also faciliraced
the resurrectionof industrialholding
groups,organizedthis timearoundbanks(theso-calledkid'musystem).
Thesenew groupsbecamcthe main.playersin the government-initiated
consultation procss.
The principal instrument of the bureauicracyin managing this
processhasbeenadministrative guidance,the useof threatsand rewards
to induceprivatcpartiesto followor implementmeasures suggestedby
the bureaucracy.Administrativeguidancedoesnot depend on any law
or regulation.Rather,it employseidter threatsof adverseconsequences
should a private party refuseto conform to a 'suggestion
or recominen-
dation"or, alternatively,rewardsdhatcan be grantedshould the party
concede,I MITEis wellkcnownfor havingemployedthis withgreatskill.
Until the mid-1960s, administrativeguidancewas a powerfil4tool for
Mm bureaucrats.Mm had firmcontrolovertheallocationof foreignex-
changeand overimportationthroughthe ForeignExchangeand Trade
Control Law.But intermationalpressure,combinedwith the need to
maintin an export-basedeconomy,led the governmentto graduallylib-
eralizethe rade and investmentregime.The processbeganin the mid-
1960sand continuedtill 1979,whenthe ForeignExchangeand Control
Lawwasdismandedand the ForeignCapitalIav liberalized.Of course
this meantthatMmTwaslosingkeycomponentsof administraiveguid-
ance. By the late 1970s,guidancehad taken a distinctiely different
form-mergers of industies spicedwith financialsupport from the
government-and coerciondeclined.
Ealaysia'sCouncls: mnulalingJapan and Korea.Malaysia'sexperiencewith
public-private cooperation through ddiberation councils has paricular rel-
evancefor developing econqomies. UnlikeJapan or Korea, Malaysia is a mul-
tiechnicsociety,and unlike Singaporeit has a eladvely largepopulation and
land area.Thus Malaysiais in many waysmore tpical ofa devloping econ-
omy than are te high-perfomiing economies of Nordteast Asia.
When the Malaysian govermment introduced foirmal public-private
cooperation in 1983, it embodied-the concept in the slogan "Malaysia
Inc.,' an overt attempt m emulate Japan Inc., partcularly the dose rela-
rionship between Japanesecministries (especially )and big business.
In Malaysia, officials have applied this concept to their quest for growth
with equiit, a particularly important goal in a multiethnic so6ey-

-I86 :.;
 AN INS'

Before 1983, Malaysia had some consultative panels within min-

istries that dealtwithspecialized issuesand wereconcernedmainlywith
cutting red tape. Many had litde privatesector participation; those that
did, however,produced concreceresults. Since then, more deliberation
councils have been formed,dealing with broader and more complicated
issues (the budget and trade,for instance). One of the most ambitious,
the MalaysianBusinessCouncil, which was formed in 1991, resembles
Koreashighly successfulExport Promotion Council.
Although it is too earlyto know how effectivethesenew councilswill
be, thev would be hamstrung without an efficient and rcpurable civil
servicebureaucracy.The government recognizesthis and is trying to in-
crease the bureaucracy'sefficiency,eliminate corruption, and educate
*. civilservants on the value of cooperatingwith the privatesector.

WhyDeberaton
Councils
Assist
Growth
From an economicstandpoint, deliberatve councilsfacilitate infor-
mation transmission-They enable the bureaucracy to gadter infbrrna-
tion about world markets, technology trends, and the impact of
regulationsdomesticallyand abroad, synthesizethe information into an
action plan, and communicate the plan back to the private sector. Self-
interested behavioris more or less constrained by the repeatednature of
the collaboration.This in part helps establishcredibility-private sector
participants believe cheating and reneging are less likely. Politically,
these councils serve as proto-democratic institutions, providing direct
channels for big business,labor, and academniato the seat of power.Fi-
nally,becausethe rules that governan industry are effectivelyesablished
- rwithin
the council, everymember is assuredthat the rules canolt be al-
tered arbitrarily.With dear rules established,memberscan concentare
on market competiton and not worry about others tring to curry spe-
cial &ivorsfrom the government.In effecr,their share of industryprofits
is determined more through competition than through rent-seeking.In
this sense, a deliberation council can be viewed as a wealth-sharing
mechanism.
: . T . - -. ,

Studies of the successfulAsian economies have fiequently empha-

sizedthe authoritarian nature of their leadership.Leadersof these soCd-
- - are
etiesdescribed
as benevolent
dict;ators
who,incontrast to mostother- 1- -

0 0 X H -:-: 0 - - :. I~~~~~~~~
 - - - ~~~~CL
E

dictators, uscd their authority to musde businessmen,workers,and bu-

reaucratsinto a single-mLinded pursuit of development.This chapter of-
fers a different interpretation. Leaders need the active cooperation of
their peoplc to stimulate and sustain growth. Heavy-handedtreatment
might subjugatethe ruled, but it would also increasethe perceivedrisk
of exprpriation, driving away capital. Lack of capital would slow
growth, in turn driving awayskilled workers and entrepreneurs.Thus,
even authoritarian leaders cannot rely on coercion alone Rather,all
leadersneed a strategy that will induce people to participatewillinglyin
growth. Whether by designor accident, the leadersof the HPAEs have to
various degreesdevisedmechanisms thar effectivelydeliverthe promise
of shared growth.
Although policiesto promote growth are generallywell known, they
tend to be eitherdifficulrto initiate and/or to sustainovertime. Growth
makes those at the top of the economic pyramid uncertainabout the fiu-
ntre of their wealth.It introduces changesthat are likelyto open avenues
for fellow elitesand new competitors to contest control of their e-
nomic fiefdoms Elitegroups therefore generallyfind it in their interest
to try and maintain the status quo unless theyare assuredthat they will
benefit from rapid growth. Furthermore, the initial stages of growth
often requirethat the generalpopulation make sacrifices.Hence, evenif
growh can be initiated it may be interrupted by socialand political un-
res; unless its fruits are shard by the non-elites.
The taskof initiating and susmininggrowth thus presents a very dif-
ficult coordination problem. Lcadersmust persuadeeliregroups roshare
the growth dividend.They must demonstrate to non-elites thar part of
the dividend will indeed accrue to them. Governments must create
acediblemeans to persuade everyonethat their plans are nor all rhetoric.
The governmentsin the HPAEsall establishedmechanisms that opera-
rionalizedte princple of shared growth. The NE in Malaysia,deliber-
ation councils in Japan, public housing in Hong Kong and Singapore,
and small enterprise promotion in Korea and Taiwan, China-all
helped to demonstate the realityof shared growth.
Finally,govemmentsmust foster the conditions in which the bureau-
cracy,or at leasta dedicatedtechnocraticcore,can designand implement
effectivepolicies.To varing degrees,and with a varietyof mechanisms,
the HPAEshave performed these tasksunusually welLIn the next two
chapters, we examine
the policies:and instruments that HrAE govern-
ments used in their effort to rialize the promise of shared growth.

;I88
 Notes
1. For example,the National WagesCouncil in Singa- 8. The criteriaselectedare (a) the transpare-ncyof the
pore has helped coordinatewageriseswith increasingpro- government,(b) the effectivenessof antitrust laws,(c) the
ductivity. In Japan, the establishment of the Japan securityof propertyandtperson, (d) the faimirssin the ad-
Produtctivity Center helpedi institutionalize labor- ministration of justice, (e) the extent of improper prac-
management consutrations. tices in the public sector, and (1) the extent to whiich
lobbying influences government decisionnmaking.The
2. In fact, the smE.sectordeclinedonly when the HCI rankinigsfor eachcriterion arebased on responsesof busi-
push wasinitiated in the mid-1970s. ness executivesstationed in each of the economiesin the
survey to questionspertaining to the criteri-on.A Borda
3. In those cases where the prime minister appoints scorewas computed for each country by summingup its
both vice ministers, he is constrained by custom to ap- rankingsfor eachof the criteria.The overallrank score for
point a career bureaucratfrom the ministryto the post of a group of economiesis the avenageof their Borda scores.
admninistrativevic minister.
9. Industrialassociationshavebeen important partici-
4. The debt-equityratio has beeninordinatelyhigh in panesin the consultativeprocess.In many cases,theseas-
Korea. Official statisticsindicate that this is in the range sociations are empowered by law to represent their
310-80, although the 'true' figuresare said to be lower, sectors.T'hus,the businesscommunity has strong repre-
at about 160-80 (Wade 1990). Nevertheless,this is still sentationin a council.
high comparedwith the ratio in the United Statesand the
United Kcingdom,where the rangeis50-90. 10. Though dhe channels of communicationare still
operational,the democratizationof the polity has led to
5. Okimoto (1989) ranks the power and prestigeof somechillingof relations.The country is currentlyunder-
ministries as the inverse.of their degree of policicization. goingrapidpoliticalchange that has affectedthe business-
* The heavilypoliticizedconstructionministryrankslowest govemment:interface. There is a recognition in many
and the relativelynonpolicicizedMmu the highest. quarters, including business,fir the need to reestablish
good relations but under a new form of collaboration.
* 6. Senioritybearsheavilyon promotion decisions.It is That form is likelyto reflecta more evenbalancebetween
assumedthat, onceone hassuccessfulfly
hurdledthe diEf- the tWOSCectr.
cult entrance exams,she or he is capableof doing (effi-
* ciendlyand eff-ectively)what is askediof her or him. This is I11.It is alsowidelyknown that the relativelyfeiwlarge
not -to say that performance does not count. The pyra- private businesseshavedloselinks to the party.
midlikestructure of the bureaucr-acy guaranteessome c
* greeofcmeiin A neigds ok oehrt 12. In fuct,inte'rviewsrevealedthat even.somemidsizc
ensure that its membersprosper.. Who becomesa bu-
.. firms haveestablishedrelationsof this type.
reau chief,a director-general,or ultimatelythe (adminis-
trative) vice minister is a source of intense competition 13.Johnson (1982) givesan interestingexample:"The
amongthe classesin a ministry"(Johnson 1982, p. 62). city of Musashinoin the suburbs of Tokyo issueda policy
statemnentsayving that housing contractorswhobuilt large
7. Finance-relatedagenciesin Thailand, for examnple, projectshad to cooperatein provicdingor helping to buy
the Bank of Thailand and the BudgetBureau, havetheir land on which to build elemnentary schools.Whena. con-
own personneland recruitmenitprograms.There, salaries tracor ignoredthes guidelines,die city cappedthe water
are said to be about 30 percent higher across the boardt and sewagelines he had built for the project with con-
than in the rest of the public sector. crete. The contractortook die cityto court, but the court
upheld the city" (p. 266).

189
- -w~R S

Strategies for Rapid

Accumulation
CCUMULATION OF PRODUCrIVE ASSETS IS THE
foundation of economic growth. It is therefore noE
surprising that, as shown in chapter 1, the high-
performing Asian economies (HPAEs) accumulated
both physical and human capital much more rapidly
and consistently than other economies and that ac-
cumulation accounts for a large portion of their superior performance.
More interesting is the question of how the HPAEs were able to achieve
these rapid rates of accumulation and, in particular, to what extent
policies were responsible. This chapter examines the HPAEs'superior
performance in three accumulation-related areas: human capital accu-
mulation, financial savings, and investnent. For each area, explana-
nions fill into three broad categories:

- Conditions that were not the direct result-of accumulation policy

: Policy fundamentals: areas in which nearly all governmenrs inter-
vene, but in which HPAEgovernments acted more effectively
i ActivisEpolicies: selective interventions underraken in some HPAEs

to increase accumulation.

Rapid economic growth and the attendant changes in economic

structure not directly related to accumulation policies were important
factors in increasing all dtree forms of accumuladion. For example, the
HPAEs'rapid sLift from high birth rates to low birth rates increased the re-
sources per child potcntially available for education at home and in rhe
dassroorn. At the same time, the rapid deceleration of population growth
increased household saving, in turn creating resources for investment.
:n addition, the HPAES performed well in selectng and implementing
. ..
fuindamental economic policies related to accumulation. Every govern-
gover9
ry
econon
fundam.nml
:~~~~~~~~~~~~~~~~~~~~~~~~~~g
 ment spends on education, but the HPAEs spent their money more
wisely, emphasizing universal primnaryand, later, secondary' education.
Similarly, every governmen tcries to control inflation and secure prop-
erty rights, and most regulate banking institutions to protect savers
against default, all of which encourage savings and investment. The
HPAEsdidt these things better than the rest. Moreover, as the HPAFsde-
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 HPABshave deliberately intervened in marker to
solve specific coordlination problems related to accumulation. These in-
terventions aimecdat boosting savings and investment rates are the most
controversial of the HPAEs accumulation policies. Generally, interven-
tions have been more comrmonand more successful in the three northern
HPAEs-Japan, the Republic of Korea, and Taiwan, China-al of which
have strong govermnnmetl institutions. With some notmble exceptions,
attempts to replicate such efforts among Southeast Asian newly industri-
aiigeconomieishave not been successful.
Atrthe outset we must acknowledge a fundamental methodological
diffickty. Since we know the HnPAFs bad rapid accumulation (and to
some extent were chosen as a group on that basis), we also knoyw,even
before examining their policies, that they wer not inconsistent with
- aantdfut
rapid accumulation.l The
fwikecorg
challenge is to separate
aig policies.that
n netet were neu- h
tral or perhaps negative but not sufficiently negative to hinder accumu-
lation significantly from those policies that actually faLcilitatedrapid
accumulation. We do thtis,noting that our findings are based on antalytic
and empirical judgmecnts rather tha-n derived from statistical models.

Explaining
EastAsia'sHi'gh
Human CapitalFonnation

XCEPTrFOR IN THAE AND, THE QUANTITYOF BASICEDUCATION

provided to boys and girls of school aehas been consistently
iEtiigierin the HPAEs tha in economiieswith 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
 STRATEGIES FOR RAIMDACPIDA

tion and were particularly effcctivein reducing gender gaps at that

level.
What accounts for this extraordinaryperformance?We focus first on
three enabling factors; high income growth, early demographic transi-
dions, and more equal income distributions. Each of these greatly in-
creasedthe resourcesavailablefor education. We then shift focus to two
policy variables: the overall budgemry commitment to education and
the distribution of the education budget. We seek to show that the allo-
* cation of public resourcesto primary and secondary education was the
major determining factor in East Asias successfuleducationalstrategies.

RapidEconomic
Growth
Rapid economic growth is the first factor that accounts fbr East Asia's
extraorlinary quanrity of education. For example, during the decade
1965-75, CDP growth averaged6.7 percent in Malaysiaand 4.1 percent
in Argentina. This implies that over the decade a conscantshare of GDP
allocated to education would have doubledthe resourcesavailablefor
education in Malaysia,while in Argentina they would have increasedby
- less than half Rapid growth also crearesjobs, increasesreal wages,and
raises the rate of return on labor forceskills, thereby increasingthe deo
mand for education (Ing, Anderson, and Wang 1993). Thus, rapid
growth operates on both the supply side of the market for educational
services,by increasing the potendal resourcesavailablefor education,
and on the demand side.
; The risingreal wagesthar area product of rapidgwdicould have a
dampening effect on educational opportunities, if the relativewages of
teachers rise. Fortunately,rapid accumulationof hurnan capitalfacilitates
subsequent educationalexpansionby increasingthe potential supply of
eachers,thereby reducing the premium teacherscommand.' The wages
of educated workers,includingteachers,tend to rise at a slowerrate than
averagewages.Therefore, the impact of risingwages on the cost of edu-
cation and hence on enrollmentsis weakened.2 For example,per pupil
operatng costs at the primary level average roughly 13 percent of per
capita GNP in Indonesia and Malaysiabut more than double that, 29 per-
cent, in Sub-SaharanAfrica. In Indonesiaand Malaysiathe annual earn-
ings of primary teachers are roughly 2.4 times per capita GNP. In
Sub-SaharanAfrica, however,where human capital is relativelyscarce,
that ratio is much higherA-The
pattern is similar at the secondary leveL.

'I93
 --- RAC_LE

DecliningPopubtion
Growth

EastAsiaprecededotherdeveloping regionsin movingthroughthe de-

mographicuansition,and the educationsectorwasa majorbeneficiary.
One important outcomeofa decelerationof the rate of populationgrowth
is a deline in the rateofgrowthof theschool-agepopulation.Duringthe
1980s,the growthrareof the populationage6-11 yearswasverylowin
EastAsia-so low that the absolutenumberof schoolchildren in Korea,
Singapore,and Thailandactuallydeclined-but wasphenomenally high
in Sub-Saharan Africa.Asa result,from1965to 1989theshareof thepop-
ulationup to 14yearsoldin Singaporefellfrom44 to 24 percent,whilein
Kenyait rosefrom47 to 51 percent.Similarcontrastareevidentbetween
otherHPAESandSub-Saharan Africaneconomies(seetable5.1).
When the school-agepopulationis growingrapidly,as in Kenyaor
Pakistan,risingexpendirureson basiceducationareneededjust to keep
enrollmentratesconstant.With decliningor slowlygrowingschool-age
populatons,however,similarincreasesin expenditurescan go for more
schoolingor betterquality.Alternatively,expendituresasa proportionof
GDP can be cut while current standards are maintained (see tableA5.1,
in appendixA5.1).
Table5.1 Sizeand Growthof School-Age
Populabon

.,
of tota pp (rnl

*.FIPA&E2.1:
ong-Ko 402 1: 0
-- f-.- -rcaRPf - -i'o n 143r4o- i iio:' 26¾--r:'0.7. -0,.,3'
p"t.of." ' '4 - _- 37 ' 0.2
d' ' . '.4-4

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

. .

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

1~~~~~~~~~~~~~~~~~~~6 a ' 2 9' . ...-...

T- 47s51_3 4
46 ~~.
igrn.~~~~~~
, ~~~~~~ 9~11 r! 7 t
* I94 0 - . Brazl ; t, 4 35 -
120 ' "1-

'194.
 STRATEGIES FOR. R-A:P-D` I

Accordingly,declining,stagnant, or slowlygrowing school-agepopu-

lations have allowedsubstantial increasesin per pupil expenditure in all
of the HPAEs. Table 5.2 shows the magnitudc of the savingson expendi-
ture on basiceducation in East Asiadue to the decline in the proportion
of the population of schoolage.5 Consider the most striking example:if
the share of the school-agepopulation in 1988 were as large in Koreaas
in Kenya,the Korean govemment would have had to spcnd 5.6 percent

Table5.2 Percentageof CNPAllocated

to Education
Saved
Dueto LowerFertility
Rates

t _f 9 ;\; i X~- dne torowthm;s#e

Expendtre '.on ; rhuol-aepopnltdon #t
basic eduation: --ere tower an.
a prcenage
-a
:Economyp-
- :.-; = :of GNP: . fi!ny' Mcio P X.n
Ho,gKon'
,
1975 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;~~~~~~~~ ; l9-; w ..

1975 13 0~~6 ~ ~~~

04 ~~.
04-
1988-89: 28 . . 20 %8
:14

1980-81 ~~~4:4
11 3. 0 4'0
;1988-8 4.40 16 04 0

* ''J1975 ; .21 08 20

"_J ''i 5 :ijS" < t'' ''05i'! '.' ','

-9 8 268 1 0.3 ' 08

4t,'-,'^ .iNnc' ofcul lan o. [Cur on basic ed an an asapccara2e aof ) '

Fr.
zpercencddference
s S * q ;~~r between ErAsiznan4orhcrccononucs inscihoolWsg popadi)'n2u',_ -+,L{';
on .a;p ied ;oFGN)-
;2,-,;j:u;f t *, , 1;-th.iii
(cxpcn mmeon an cCai
=a f addal
itht would havte been rqe h p gw e m rp
, . s. A 0- C-
4 %Th%isanndImrrofd1Saviui
-d . . - t - - r-
LTvuuut1accoarcstuiooup.½ .

Thris ra ndxue n asic dlihearo,usso(iouy",brn"~

':95
 .~~~~~~~~C C

of GNP on education instead of the 2.8 percent actuallyspent to achiieve

Korea'shighenrollment rates.Thissuggeststhatlowerfcrtilityratessaved
Korea2.8 percentof GNP.Conversely, school-agepopula-
had Pakistan's
tiongrownat the relativelylowKoreanrate, thegovcrnmentof Pakistan
couldhaveincreasedenrollmentratesby as muchas 50 percent.

Equality
inDistributon
ofIncome
The moreequalan economy'sdistributionof income,thehigherpri-
maryand secondaryenrollments tend to be. In a crosa-section of more
than eightyeconomies, significantneg-
thereis a strongand statistically
ativecorrelationbetweenbasiceducationenrollmentratesand the level
of incomeinequality,asmeasuredby the Ginicoefficient(Clarke1992).
As we discussbelow,educationalexpansioncan havean equalizingim-
pact, so the causality in this relationship could run from enrollment
ratcsto the distributionof income.However,thereare a varietyof rea-
sons why low income inequaliry,which we docurmentin chapter 1,
might havebeen a third factorco:ntributingto highenrollmentratesin
East Asia.
Due to greaterincomeequality,the incomeofthe poorest20 percent
of the populationin EastAsianeconomiesis higherthan in economies
with the sameaverageincomebut greaterinequality(seetableA5.4,at
the end of the chapter).Becauseof budgetaryconstraints(and capital
marketimperfectionsthat precludetheir borrowingto investin educa-
non), verypoor householdsare unableto investin theirchildreneven
whenthe returnsarehigh:the pressingneed to subsistcrowdsout high-
return investments. This is lesslikelyto occurin EastAsiabecausethe
incomesof thepoor tend to be furtherabovesubsistencelevel.
Measuresof the income elasticityof the enrollmentraio derived
fiom cross-economy educationexpenditureequationsare0.31 and 0.43
at the primaryand secondarylevels(Schultz1988).This impliesthat in
EastAsiathe positiveeffect on schoolenrollmentsof lowerinequality
canbe quitelarge.Forexample,givenan incomeelasticityof basicedu-
cationenrollmentsof 0.40, if the distributionof incomewasasequalin
Brazilasin Malaysia,enrollmentsamongpoorBrazilianchildrenwould
be roughly33 percent higher.Moreover,the incomeelasticityof de-
mand for basicschoolingamongpoor familiesis likelyto be substan-
tiallyhigherthan averageand mayexceed1.0.In this case,enrollments'
amongpoorBrazilianchildrenwouldbe 80 percenthigher. . . .

-96.- -
 STRATEGIES FORR-R A P

Coordination
Probemsthat Hinder
HumanCapitalFomation

Numerous studiesshow thata person'scarningsare higher the higher

their human capital,as measuredby their education and 'iealthstatus.As
shown in chapter 1, the same is true for nadons; we know that invest-
ments in education contribute significandtyto economicgrowth. But
neither of thesepoints necessarilyjustifiespublic provicionof education.
Indeed, if the privatereturns art as high zs socialreturns on theseinvest-
ments, individualsand householdsare likelyto make adequate (from an
economywidepoint of view) investmentsin human capitalon their own.
However,two kinds of problems generatea gap betweenprivate and so-
cialretumrns,which-if not countered by governmentaction-will result
in fnilies investing less in their children's education than is in society's
interesLFirst, theremay be filures in the capitalmarket and in informa-
tion, discussedabove, that reduceparents ability or interest in investing.
Second, educationalinvestmentshave positiveexternalities,which imply
6
that familiesthar invest in education are not the sole beneficiaries.
The difficulty of borrowing to send children to school affectsthe
poor especiallyCreditors cannot easilystake a fuitureclaim on embod-
ied human capital (as they can for other types of collateral).Evenpoor
&milieswho might be willingto borrow,becauseschoolinghas high pri-
ratcreturns, usually cannot The poor are also likelyto be less awareof
fixturereturnson education-and thereforeinvest less in their children's
schooling than would make sense even from a strictly private point of
view.7 These market faulres in principle suggescmaking loans available
and improving information about future retuns. A simpler, and more
common, alternative is for government to reduce the direct costs of
schooling by making public schooling availableand free.
The second classof marker failure-externalities-has more imme-
diate implicationsfor what type and levelof schooling the government
should subsidizeand for whit groups.There are severalsourcesof exter-
nalities.An educated person may increasenot only her household'sin-
come but also the income of other households, because her ideas and
innovations spill over to others. Similarly,education may have social
benefits not directly compensatedmonetarily,for example, a reducton
in the spreadof contagious diseases.In these and other cases,coordina-
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 LE

Figure5.1 Returns
onInveshnent Policy responses generally consist of making schooling availableand
byLevel(Latest
in Education Year) freeand in some economies compulsory.The differencebetweensocial
Ratreofrtuwm(percent) and private returns from educautonis probably higher at tle primary
:Primary
Pri_
i
1 and secondarylevels than at the university level (see figu 5.1). Many
--- c- ndia1 -- I-i positive spilloverscome from literacyacquired at lowerlevelsof school-
SondarJ Kr while the retu:s from training at the universitylevel are almost
fhing,
-Hi4her I fuly! captured by te higher
i income of university graduates. Vocational
05 ID2may L training also havc high social payoffs, if it improves worker pro-
X 0~~~'5 10 15- 20 25 30 3!i
3 SocIalreturn U Pilvato ductivity not only for the trainee but also for her co-workers. More im-
*; Social retrns N Nvaf mt
portantly, evidence suggests that vocational training is most
Soww Pswchaopoulos(1993). cost-effectiveif trainees have a solid baseof primary and secondaryedu-
cation. All of this argues for universalprimary arid broadly based sec-
ondary education as a means to improve economic efficiency and
income distribution.
Table5.3 PublicExpend4tnreon
Educationas a Percentage
of GNP Policiesthat PromotedHumanCapitalFormation
;E con>msrnyI +.
if 5 .-
-1960 196.i9 ! f Higher shares of national income devoted to education cannot fiuly
explain the larger accumulation ofhuman capitl in the HwAES.In both
oegg
:~5 Kong2.8 ;1960 and 1989, public expenditure on education as a percenrage of GNP
2.0 p.oE
Korc ]46 - was not much higher in EastAsia than elsewhere(seetable 5.3). In 1960
;:,--Smngipoin -2.8 34
?g1ayi '.97~ pV W - the share was 2.2 percent for all developingeconomies,2.4 percent for
T ¾ 23.",
J.'x',tfg>R; '.3]9.2. Sub-Saharan Africa, and 2.5 percent for East Askia During the three
z lndocs 2z3 0.9 decadesthat followed, the governmentsof East Asia markedlyincreased
*~~~~~~~' ovr nts- ogVIg
f! Eas.............t2T-
-J
*;*ZAvctnX ;.sc ,.1 2i.5{ Si ;. ntheof national output they invested in formal educaion, but so
share
--- ; governments
' '_t_AX'l in*'ii;iother developing regions. In 1989 the in Sub
;;rg
sa
1;9 q r SaharanAfrica, 4.1 percent, was higher than the East Asian share, 3.7
-l

PakfisraaV$1'I .-- J 26{-

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 wer initial conditions, for example the colonial legarcy,decisive.
\ 9'n 5|; >Nor

; -6' - J Et 4 41/ V's e Korea did have much hier

W 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 enrollments
over other middle-income economies. A comparison of lndonesia, a
success story, and Pakistan, a laggard, is also illustative. In 1987 In-
;' .- donesia had achieved universal primary cnrollmcnt and a 48 percent
0- - ~ 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 secondarylevel.
-. Q:- t> What proportionof thesegapsis due to initialconditions?At the pri-

.98
--Ef g .- . ... -.. 0 -.
 STRATEGIES FOR RID7ArCCU-M4

mary level,Indonesia increasedits enrollmenr rate by nearly 80 percent-

age points since 1950, while Pakistan managed an increase of only 34
percentagepoints, implyingthat inost of the current gap is explainedby
the pace of increaserather than initial conditions. For secondaryschool-
ing, Pakistan'senrollment rate in 1950 wasactuallyhigher than Indone-
sias; all ofde current gap is explainedby the rates of increaseduring the
past thit-seven years
Prmawy andSecoudary Education.The allocation of public expenditure
between-basic and higher education is the major public policy factor
that accounts fbr East Asias extraordinaryperformancewith regard to
the quantiry of basiceducation provided. The share of public expendi-
ture on educationallocated to basic education has been consistently
higher in EastAsia than clsewherm:Koreaand Venezuelaprovide an ex-
treme examplethat nicelyillustrs the point. Table 5A indicaresthat
in 1985 Venezuela allocated 43 percent of its education budget to
higher education;,by conwzst,in the same year Koreaallocatedonly 10
percencof its budget to higher education. Public expenditureon educa-
tion as a percentage of GNP was actuallyhigher in Venezuela(4.3) than

Table 5.4 Allocationof EducationBudgets,1985

f~~~~Pbi'K>. r;j6Z.et.t
Pub&-, Peretge . of Percenag f
~~~~~~~~~* ; L;;te>ni8s.
s -- g.
k
; r ' ; - on eoncain - '
on'! .- .t
cn. cio*

-* ~- as , ;- educanonsit - -

* t-- ... i --. -- . .. ..

-Indones23:
Mn . ,9;
72.. n '
.0 6
iKoeaRepo 0 *W- ", 10.3, 0f
1vIalay~~~~~ia
79 2-~. - 14.6~ 749
;E~ A:LdcaoonzyGW
Singapore 5.0 o 32,N- Siid,cinollnt
.
30.7: bsc ducanon
4 i:980. .

'F. .4. .. -

-'~~~~~~~I"
t'tL' a6E
7 ; . e-3 .2' a9d4.3itwfl95.46,iiuhatrthe
;E-310
per
- , eniag
-.Taln.t3.ofthe
. eu4oi buidgealarcmbic?Iuconwns_81
2..10 m 9S48; and80,
.8.

-1 '- ~tini985-8
.- .- dc: ng St.adalu`p*
e it -c- Mg(* 'I -- :

P-d -- W5Ofli
: . t-.- .,Lltf -atv.Sosc Qu-.

'99.
 £

in Korea (3.0). After subtracting the share going to higher education,

however,public expenditur availablefor basiceducation as a percent-
age of GNPwas considerablyhigher in Korea (2.5) than in Venezuela
(1.3). Box 5.1 shows how Indonesia'semphasis on primary education,
contrasted with Bolivia'srelative negIlectof primary schooling, is re-
flected in rural educationalopportunities.
The share of public funds allocated to tertiary education in East Asia
has tended to be low, averagingroughly 15 percent during the past three
decades.9 In Latin Americadie share has been roughly 24 percent.'0 In
South Asia, the share is dose to the Latin American leveLThis had been
the case in Sub-Saharan.Africa as well, but in recentyears the share has
dedinedrto East Asian levels.
By givingpriorityto expandingthe primaryand secondarybasesof dhe
educatoa pyramid, East Asian governments have stimulated the de-
mand for highereducation,while relyingto a largeextent on the private
sectorto satisfyrthat demand. Eniall developingregionsthe probabilityof
going to universityis miarkedlyhigher for secondaryschool graduates
from high- thianfrom low-incomefamirnls.Typically,in low-and middle-
income economiesgovernmentsubsidiesof universityeducation are nor
relatedto need, implyingthat they benefit fimilieswi 'hrelativelyhigh in-
comesthat could aflordto pay feescloserto the actual costof schooling.
At the same time, in many economies,Braziland Kenya being no-
table examples, low public funding of secondary education results in
poorly qualified children fr-omlow-income backgrounds being forced
into the private sector or entirelyout of the education system. Because
of the higher conce-ntrationon basic education in East Asia, public
finds for education are more likely to benefit children of low-income
famnilies who might otherwise have difficultyremaining in schooL
VocationalTwining.Human resources and the trainingr to upgrade
them have been important to the HI'Anssuccessfulexport drives,despite
the high degree of labor intensity in their manufactured exports. High-
levelskills are essentialfor such manufacturing-relatedaaiviries as mnan-
agement and entrepreneurship, information -technology, finance,
markceting,accounting, and law.Moreover,adaptive innovationson the
shop floor, which are responsiblefor a major share of productivity in.
manufacruring, demand both higher- and lower-levelskills. But while
vocational taining is widlreonzda imotant such traing is
rarelycost-efficientwhen providedin the school systems.Firmsprefer to
do their own trairnin pardy becausemany skillsare firm-spedific..

-200
 STRATEGIES FO.R. k-Ai

-Wl
UsN
Ed SZ 31
Mud
gg n :p quesm--l-

. ou 0
0

to .,H, citi0ii"bf- propoipomp

Y,
$61
-5-i
PC-= -r. n.
ZOM ulca'don- ud et,' 00
Pff pq*"
90
r w=ntpmbamr- ucauonz,-,F-,,"j,,.
C uc2uoq
diiii2b M-ffiki6
0,-percmrp -,-PFF
2=enta o
Bft a c;-,MC:rcmamclcri,,-,,
CUOn.
arsozaLs= R -'zcu -m
y,'one in
C
O j'an
jopPrO

7 o non,
ZhQ?
1987'
gr
:ax
"ea
.11;2pcm
):PM m 's, ;",O
iham
liEfic,ar-kniv P C
Li
C=O caiiNwduj
&-Co

zm
owcvcr
Lti SCIOLOO

201
 Few studies have been made of training in the HPAESor ocher devel-
oping economies, in part because of the difficulty of defining tamining.
One lage survey of 48,00 0 manufliLccring firms in Taiwan, China, con-
duded chat firm-level trining rised productvity by encouraging the
efficient use of technolog (Aw and Tan 1993). The study also found
chat returns from training were higher in industries with well-educaced
workers in an environment of rapid technological change. Another Tai-
wan, China, study with a smaller survey sample and stricter definition
of training found that only 24 percent of Sins provided trining and
only 7 percent of technicians receivedit. As would be expected, training
was positivelyaffectedby the levelof technologyand negativelyafficted
by labor turnover and high opportunity costs of training (San 1990).
Other studies in the HPAaEshave reached similar conclusions.Enter-
prise and preemployment training have produced social rates of reTur
of 20 percent in Maaysia. In-plant taining for welders in Korean ship-
building had a socal rate of return of 28 percent, higber than in non-
firm. training institutions (Middlecton and others, forthcoming).
On-the-job ntrainig was an tremely systemaincand powerfulingredi-
(onishni 1989).
entbin the rapid growth ofjapanese companieas"
In some instances,govern-mnt effortsto pomote taining have gone
awry According to one study, Koreis 1974 SpecialLaw for Vocational
Training, which required firms to provide six months' training in am
proved schemes, discouraged firm-level training; firms considered the
period too long and opted to pay a fine instead (Kim 1987; van Adams
1989). More narrowgovernmentefforts to forecastoccupationdemand
or providetnained workersto fillanticipated jobs have also generallynot
been successfil.
One exception is Singapore's use of training to promote the inforna-
ton technology sector. Singapore has achieved world leadership in
information-relatedservicesthrugh a concerted program thar involved
educationalinsintitons (specializngin businessand engineeringsoftware
training), training subsidies to schools and office workers, computeriza-
don of the dvil service,and establishmentof TxadeNer,an interatonal
infornation network (Hon 1992). This successillustrates the importnce
of a govenmmenesability to foresee a major trend and coordinate comple-
menary privare invesctnents. At the same dime, businesses must stand
rreadyto take advantage of the comprehensive support that the gvern-
nmentptovides.Other East Asian economieshave had imilar,if lessspec-.
tacular, successesin telecommunications (Mody and others 1992).

zoz
 STRATEGIES FOR RAP

Has training been more effective or widespread among the HPAEs

than other developing economies? The evidence is too limited for a dear
conclusion. However, ir is dear chat retunms from training are aug-
mented by emnomic growth and consequent job creation and are there-
fore higher in the HPAEs.Moreover, the HPAES'extensiveparticipation in
international markets aided the success of their training pmgrams. In-
tremational competition encourages finls to train workers and increases
the number of skill-intensive jobs, thereby ensuring that new skills do
nor erode. Training, then, appears to function like many other aspects of
development in East Asia. It contributes to rapid, susained growth but
does not, in imselfmake such growth possible.

Why East Asia's EnducationalPoicies Wirked

Ln most of the ecorarnies of East Asia, public investments in educa-

tdon were not only larger than elsewhere in absolute rerms-they were
also betten They responded more appropriately to coordination filures
in the market for education. Emphasis on universal, high-quality pri-
maryeducation had importantpayoffs both for economic efficiency and
for equiy. The excess demand for secondary and tertiarv education,
generated by rapid ateainment of universal primary education, was met
largey by a combination of expansion ofa public secondary system with
mentocratic entrance requirements and a self-financed private system.
This stands in stark contrast to many other low- and middle-income
economies, which have stressed public subsidies to university education.

Explaining
EastAsia'sHighSavingsRates

A5
: - S IS THE CASEWITH EAST ASIA'S RAPID HUMAN CAPITAL
formation, the region's high savings rates are in part an outcome
of rapid growth and rapid demographic transitions. Govern-
ment policy also encouraged (and sometmes compelled) increased sav-
ings through a variety of means. The most basic of these was the
maintenance of macroeconomic stabiliry. In addition, the HPAEshave
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

addressan absenceof depositinsurancemarkers),to targetedinterven-

tions (suchas restrictionson consumercreditand forcedsavingsplans).

Growth
and Demographics:TheEffectonSavings
Manypopulareffortsto explainthe EasrAsianmiraclehaveinvoked
culturalfactorsto explainhighsavingsratesand usedthose highsavings
rates to explain high rates of investment and growth. Neither element of
such explanationshas receivedmuch empiricalsupport. Indeed, studies
of the income-savingsrelationshipin a broadcrosssectionofeconomies
indicatethat whileincomeand savingsgrowthare highlycorrelated,in-
comesoften haverisenbefbresavingsratesratherthan after,suggesting
that growth drivessavingsrather than the other way around. Recent
econometricstudiessupportthe idea that rapid incomegrowthboosts
savingsrates as householdsacquireresourcesfister than they increase
consumption(Carroll,Weil, and Summers 1993). Accordingto this
view, East Asia'shigh savingsrates since the 1960s are partily an out-
comeof high growthratesratherthan a cause.
We haveattemptedour own testsof the savings-growth relationship
for the HPAEs (describedin greaterdetailin aDpendix5.1 and its corre-
spondingtableA5.2).Incomegrowthhas beena remarkablygoodpre-
dictorof increasedsavingsrates in Indonesia,Japan, Korea,Thailand,
and Taiwan, China, but savingshas not been a good predictor of
growth.Resultsare mixedfor Hong Kongand Malaysia,and causation
might run eitherway.In Singaporeincomechangeswere not a signifi-
cant fctor in the speccacularrise in savingsrates in the 1970s and
1980s,consistentwiththe viewthat demographicfiactorsand, to a lesser
extent, the policiesof the CentralProvidentFund, a forcedsavingspro-
gramn, determinedthe rise(MonetaryAuthorityof Singapore1991).
Rapid changes in rhe size and age composition of households may
have cont-ributed to risingsavings rates for many of the same reasons
that rapiddemographictransitionsfosteredhuman capitalformation:as
birth ratesfell,the dependencyratio-that is, the ratioof non-woring-
agepeopleto working-agepeople-decreased. Theories of savingsbased
on the life cycle of increasing then decreasing income relativeto con-
sumption predict that societieswith a high proportion of prime-age
workerswillsavemorethan thosewith higherproportionsof youngor
old people(Andoand Modigliani1963;Modigliani1970).Of course,
the benefitoffilling bir ratesevntually will be reversedasthe popu-

204
 STRATEGIES FOR At

lation agesand the old age dependencyratio rises,an effectseen in stud-

ies of European economies and Japan (see Blumenthal 1970; Horioka
1990).
For the HPAES other than Japan, there is litle evidence of the rela-
tionship between demographic changes and household savings. The
Singapore study mentioned above showed that most of the increased
savingsrate fiom 19 to 46 percent betwi.:n 1970 to 1989 was attribut-
able to dedining dependency ratios (MonetaryAuthority of Singapore
1991). In Taiwan, China, where household savingsrates are generally
very high, households headed by prime-ageworkers tend to save more,
possibly reflectingin part preparation for old age. (Bursavingsin older
households are also remarkablyhigh and stable, contrary to the view of
the life cyde savingshypothesis [Deaton and Paxon 1992].) Demo-
,raphic factors may also interact with corpwt"c behaviorto increasesav-
ings. In Japan the tradition of employeesworking beyond retirement age
at reduced wages appears to contribute significantlyto higher savings
rates (Horioka 1990).

PolicyFundamentals
andHighSaings Rates

In addition to benefitingfrom higher growth and demographicshifts

conducive to savings, East Asian governments selected, ofien through
tial and error, a broad spectum of policies rhat encouraged savings.
Two fundamentalsthat facilitated high levelsof savingswere the man-
agement of macroeconomic policy and the making of budgetary
choices.
TheRoleofRealInterest Rates.Foremost amongmacroeconomicfmnda-
mentals was the effort to control inflation discussedin chapter 3. Be-
cause high inflation rates also tend to be volatile, real interest rates are
often negativeand unpredictable, discouragng domestic financial sav-
ings. Moreover,since unanticipared high inflation erxdes the ral value
offinancial assets,volatilityof inflation ,increases
the riskassociatedwitth
holding,them. Conversely,low to moderate inflation, partculady asta-
ble rates, encouragesfinancial savings.Thus the HPAEs' generally supe-
rior record at avoiding long-term fiscaldeficitsand restaiing inflation
gavepeople the-confidenceto save (Neal 1990).
Table 5.5 documents average real interest rats in the uAEscom-
pared with other low- and middle-incomeeconomies.Real interest rares
on depositsin the HPAEswere either positiveor nildly negative.In con-

2.05
 MC LE

Table5.5 Average
RealIntkeresRatesonDeposits,
Selected
Economies

- *. . r~~~~~~~~~~~~~~nwiu&wrd

MJPA&
AIcegtsx ' ~~~~~~1973-2. -tJ . . 3

* Tan.Iiq,~~~~~d 1973-90 AS3

* MmhI& 1 . 177 .2.47
- ', 19,7-91 ~~~~~~~2.43 i
Taiwan C. 94-3 .8 7* '-
ThIml 1177-got 4.4197

-. Naqj.dab ~~~~1976-92 04 .'3

NcpI - - . 117449 ~-3.6 55

M-ils 1791443$3'.4
1935-91 ~~~~~J314 94.9

Un4m7 3976-9' ~~~~ ~~~~~~~~~~LA9

Ism,

EWF -. -. ~~~~~~1976-9 -63m-

Taft,1 ~~~1976-1-5 83
At.tqc . , -~~~~~~~~~~~521 334~~19

Am.- '. 144 -243 * , l

- '*aIa~~~~~~~~17 I . -: :I!-: ;w 9
-. 21033j

uas mAi 12793-q 5

Zsra ore 1Sgonsshw

-5an ecnmc*it1agIeadertso
average ae er 1. 13.1 prciinTASseS an

G2aeret nZaba.Ra intercst ra4e weeas' evlOS e as

2o6~~~~~~~~~~-l 109 3 3
 STRATEGIES FOR 'AlI'

measuredby 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 averageit was 40 in
Latin America, 14 in Sub-SaharanAfrica, and 6 in South Asia. Malaysia
and Singaporehave maintained the most consistendypositiveand stable
real interest rates on deposits, paralleling interest race trends in the
--.. United States.Japan maintained negative,but highly stable, real interest
raceson depositsuntil 1983 and has had verylow btutstablepositivereal
miterestrates since that time. Korea, especiallyIndonesia, and Taiwan,
China, have had much greater volatility of real interest rates than the
:otherHPi-Esbut stl substantially less than in many other low- and
-middle-incomeeconomies.
Several studies have documented the positive association between
real interest rates and the growth of savingsdeposits and broad money
aggregates(Fry 1988; Lanyi and Saracoglu 1983; Gelb 1989). This is
especialy evident for changesfrom highly negativeto positive real rates
of interest. For example, when Taiwan, China, raised real interest rates
on bank deposits from a negative300 percent in 1949 to about 8.5 per-
cent in 1953, the ratio of time and savingsdepositsto money stock rose
from 2 to 34 percent in three years (Chiu 1992). Indonesia and Korea
-achieved similarly dramatic increasesin financial savings after stabiliz-
ing inflation and shifting from negatve to positive real interest rates.
More recendy, Argentina, Chile, Mexico, and Pakistan have done the
same.
TheRoleofPubicSavings One direct way that any government can ad-
dressthe concen that aggregatesavingsare too low is to generate public
sector savings through a combination of tax and expenditure policies.
Although in theory forward-lookingprivate saversshould increasecon-
sumption (and reduce savings)to match any increase in public savings
(becausetheir future obligationsare reduced),in practice householdsare
facedwith liquidity constraints and are not able to act on expectaions.
Thus total savingsmay risewith an increasein public savings(Summers
1985). Empirical evidenceshows that governmentsavingsdoes not fuly
crowd out private savings. Furdtermore, the method of raising public
savings matter.s:on average,increasing public savings via reduced ex-
penditures is more effectivethan raising taxation. For a sample of thir-
teen developing economies, a transitory increase of a dollar in public
savings,miadethrough a cut in expenditures,reducesprivate savingsby
only 1-50 cenrs. If the dollar increase in public saving is achieved
--

207
 UI4MXLCLE

through a current period tax increas, private savingsdedlinesby 48-65

cents (Corbo and Schmidt-Hebbcl 1991).
Most HPAEs have maintained high public savings compared with
other low-and middle-incomececonomiesthrougi. -ombination of tax
policy and expenditure restraint. The extent to waich HPAE govern-
ments have restrainedcurrent expendituresrelativeto other developing
and industrial economies is shown in figure 5.2. which gives average
public consumption as a percentageof GDPbetween 1970 and 1988 for
a sample of 111 economies. Indonesia, Japan, Korea, Singapore, and
Thailand are in the bottom third of the distribution of all economiesin
terms of the share of public consumption, with public consumption
shares below 15 percent of GDP. Malaysiaand Taiwan, China, are ecx-
ceptions, faling in the middle of the distribution. In Malaysiathe high
levelsof public consumption reflectedthe explicit redistribucionalob-
jectivesof the New Economic Policy As a group, the avcrage share of
public consumption in the HPAEsis below both the OECD economies
and all other regional groupingsof low- and middle-incomeeconomies.
In contrast to the situation in most developing economies, in the
HPAEs,rates of public and private savingshave been high and growing.
For example,Singapore'spublic savingsrosefrom 5.5 percent of GDP in
197480 to 18.5 percent in 1981-90, and private savings rose from
22.6 to 24 percent. Malaysia'spublic savingsincreasedfrom an average
3.2 percent of GDP in 1961-80 to 10.3 percent in 1981-90, while its
private savingsincreasedslightly.Th2iland'spublic savingswas high in
the 1980s (87-15percentof GDP), but declined as a proportion of GDP
in the mid-1980s, when the privaresavingsrate rose dramatically (see
table 5.6).
While in some other economies (for example,Pakistan, Zimbabwe,
and Venezuela)public savingswas relativelyhigh and accounted for a
major part of national savings, generally public savings in low- and
middle-income economies was substantially lower than in the HPAEs
and had decdined.For example, public savingsin the Philippines de-
dined from an averageof 10.4 percent of GDP in 1980-83 to 1.4 per-
cent in 1984-87, leadingto much lower national savings rates. Both
public and private savingsdcdined in the mid-l 980s in Argentina and
Chile. Brazilspublic savingsdropped from 8.3 percent of GDP in 1980
to 3.2 percent for the period 1983-85.11Some Sub-Saharan African
economies (for example, Ghana) had negative public savings m the
1980s, offseting increases in private savings.

208
 STRATEGIES FOR RAI3D

Figure5.2 Averag PublicConsumption

asa Percentage
of GDP

-~~~~~~~~~~~~~~~~~~~- I

Taiwan,
ChIna

Singapore

Rep.of Korea

EHPAEs

Oetheeownomies

o,10 20 30 40
Averagepubliccontsumptlon
(percentage
of GDP)

.Soe _,
orldB2nk6 -- -

209
 !&4iIRfACLE

Table5.6 PublicandPrivate
Savings
of HPAEs
andOther Selected DevelopingEconomies
(pereL'ntage of CGP)

u-blesaings ; V

i4o6
J-apa,,*
')0955-70
z | g ' . S 0 17.2-
t^6.2
-19714=0 4.6 '2011

0 ; i . R:|?8
~~191480
6 f.;:: .2;X
312 0f: . 187';
15
-198190. 103 191i
Singapore.
*. .

1981-90'b
-4;---o 1-ft
.-- .3-. -- ->;-i24.0
;:' - I83

1980-B5 . 14.3
* .--j198 8.6. 14
Oird deSpigeconemi

-Argendna~S.Kw'lii -. . ,t''-'.'' s_
-'
_

-- ;--.1980-85* ..... - . 5 I4.2

-19804871,

1980-8. 235
0 :: 985647
: j- 06
1980-84 3.9 2~~~~~~1.6
*'~~~~ 1980-84~~ 1364
*19858 7.9 06
-~~~ 1985~~~~-187 7.
Csti Rica198-8 7. 7

P'ia9i5-8704 161 61",

'Philippines~~~~~~~~~~~~~~~~~~J
,., i.o,9.O
I9
- .4-. ,. 1.
.s '.--'-,0.;;'--. 1690 97- - '.=-'

~~
.2IO: ~ ~ ~ ~ '1503 0
 t I

STPTRATEGIESFORJLAiPPD&

Coorditon Problems
thatHinder
Savings

Savringsare afficted by severaltypes.of coordinationproblemsin-

ducedbyfeaturesof financialmarkets.Below,webrieflydescribeeachof
thesecoordinationproblemsand the policyresponsesof HPAEgovern-
* ---ments+The rangeofthe HPAES' policrresponsesto theses-rual or incip-
marker fhiluresrange fiom standard remediesthat the HPAEsmay
-ient
...haveimplemented better, such as the prudential regulationof banks, to
more acavLst interventions, such as-limitng competition or regulating
tinerestrare spreads.
Abse--.,nf People will only deposit funds with a
Deposit-Insurance
bank or other financialinstitution if they are reasonablyconfidenrthat
'.the insirution will retrn the. funds as promised. In theory, private
-firs.or eVenthe banks themselvescould.sell insuranceto depositors,
in practice,theydo nor (Thesepoints are discussedat greater length in
. 'Stigiz -1993a)T.he reasons are straightforward. Assuming a pae
firm-didLofferdeposit insurance, depositors have no more reason to
tru:st.an insurer than they do the bank: either or both could faiL In-
deed,-afirm that insured severalbanks would risk nearly simultaneous
lossesduring a general economic downturn or-a bank run. The risk of
such lossesm-akesaprivare insurers promise to deliver.even-moresus-
pec than.the banks. Thus, withour government effors to guarantee
that banks will be. able to honor their liabilities,potential depositors
.would rend-tosaveless and keep a higherproportion of theirsavings in
nonfinancial assets..
Governmentshave.a second-motivationfor insuring deposits.-Fman-
.ial.hlures. of banks have a strong negative-excernality;when some
bank fail, depositors lose confidence even in. sound. financial institu-
dons, as the history of bank runs attests-Govemrnments therefore ty to
preventrovertbank 6ailures-, fiequendy acting explicitlyor implicidy as
the-insurer oflascresorr through the central bank.
:Highrnsacti.CostsPrivate banks seek large-depositsand those thar-
are relatvely inexpensive to sevice theg therefore compete for the de-
pits of -corporationsandtaffluenturban-dwellers and.make lite or-no
efforrro attactithe savings-ofworkers, poor people,_andLresidents-of
snulLi owna andshe: countyside In.addition,-banks may-rcfiisesuch
;marg inall po eposits by settng minimumindeposit lewls-
The:pracceaise.the transactons costs forsmall saversancdiscour--
age themfruomn aking,financial saving.

| , t. 0 ;g , . . . . .~ ~ ~ ~ ~ ~ ~~111
 I`- LE

IsolatonandInsufficientSavings.Some economists have also suggested

that individualsacting on their own and their flimilysbehalf may save
lessthan would be desirablefrom the point of view of societyas a whole.
This is becausethey fail to consider the impact of their savingsdecisions
on others. If savings decisions could be coordinated, each individual
would savemore and aggregatesavingswould rise (Dasgupra,Marglin,
and Sen 1972). In such circumstancessocial returns on savingsexceed
private retums, and there may be a role for government efliortsto in-
creasesavings.

Policies
that PromotedSavings

Governmentresponsesto these coordinationproblems arevaried.All

governmentstake stepsto alleviatethe lackof deposit insurance markets
and to promore stble financial institutions. These generally indude
prudential regulation of savingsinstitutions and protection of deposi-
tors from bank defaults. In general, the developing HPAEs have per-
formed these fundamentals better than many other developing
economies.East Asian governmentshave artached great inmportanceto
maintaining saver' confidence in financial institutions. Often their
policieshave gone beyond prudential rgulations to include protectng
banks fiom competition to increasethe financialstrength of banking in-
stitutions.1 2 Sometimes this has been done at the expenseof long-nm
efficiency.
When necessary,all EPAR govenments have bailed out troubled fi-
nancial institutions through financial and managenent assistanceor
mergerswith stronger banks (seetableA5.5, at the end of the chapter).
This isnot to saythat the iEnshave not experiencedbanking crises.In-
deed, some financial instability is probably unavoidable in economies
that are expanding as rapidly as those in East Asia Nonetheless, the
crisesthat have occurred,for exanpie,in the wake of the 1980s' real es-
tate speculation boom in Hong Kongi Japan, Korea, Malaysia, and
Thailand, havegenerallybeen brief and modest, particularlyin compar-
ison with the protracted and disruptivebanking crisesin Bolivia,Chile,
Ghana, the Philippines,and Turkey during the same period.
Regu;llions toProaneBank Solvency.Central banks and departmentsof
'financein the HPAEs appearto have been generallymoresuccessfulin 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,Koreain the early 1980s,

and MalaysiasBank Bumiputera in the 1970s. Generally,however,non-
performingloans have been lessof a problem in the HPAEs than in many
developingeconomics.For example, one study of twelveLatin Ameri-
can economiesfound that nonperformingloans accountedfor nearly 14
percent of total loans in 1987, while the proportion of bad loans sur-
passed 20 percent in Argentina, Honduras, Uruguay, and Venezuela
(Morris.ndothers 1990). In Pakistan,requiredprovisionsfor bad debts
constituted 14 percent of tota pordolios in 1989 (an internal World
Bank report). In Sub-Saharan Africa, many banks have failed, and
25-40 percent of bank portfolios have had to be written off. While
macroeconomicinstabilityand low growthare partly responsiblefor this
poor performance,inadequateprudentialregulationalsocontributed.
The more effectiveprudential regulationof East Asian banks is ap-
parent in the easewith which most of them adopted the international
capital adequacyrequirementsset by the Bank for International Settle-
ments (Bis) to ensurethat banks do not takeon inappropriatelevelsof
risks.Hong Kong,Japan, and Singaporebegan to stengthen prudential
reguation in the 1970s, and Malaysia,Thailand, and Taiwan, China,
oi-llowed
suit in the 198Os.All IPAEshave since adopred the BIScapital
adequacyreo'4 rements. Indonesia,the last to do so,has had some diffi-
culty achieving BIS capital requirement levels because it introduced
proper prudential regulationsrather late.1 3
By conast, in Lain Americaonly Ecuador adheredto BEIS standar4
iwile Colombiawas in the processof adoptng the standards(Morrisand
others 1990). Moreover, some Latin American economies have not
establishedminimum capitalrequirementsfor new banks. In Sub-Saharan
Africa,thenet worthpositionsofbanldngstyermssiarplydeterioratedin the
1980s,and adopting BISstandards wold entail subsantial rcapialization.
Most East Asianeconomiescontinue to rclyon dose contact between
supervisorsand banks to encourageprudence. This routine, often daily
interactonenablesregulators personallyto assessthe riskinessof a bankA
portfblio.This pragmaticapproachis evidentin their attitude toward col-
lateral requir~ments& To reduce the risk of borrowerdefaults,regulators
-encuragedbanks to requiresubstantialcollateral.But becausethe collat-
era. was usually real estate, banks became increasinglyexposed in the
-1980s to the risk of assetprice declines.4 Face with potenial insolven-
-ies due to heavy real estate lending, regulatorsin Hong Kong, Japan,
Thailand, and Taiwan,China, have limitedthe use of real estateas collat-

2.13
 LE

eral and sought to reduce loan concentration by limiting lending to re-

lated parties.Enforcemnenr has been difficult,however,becausecompanies
are doselyheldland dlisdosurerulesare geCnerally weak.Moreoverinm sev-
erl economies,notablyIndonesia,Japan, and Thaiad, banks and firms
rend to have inrerlikng ownership.
Many East Asian financial systems appear to be moving to ara
mixed system of enfbrce-menr,in which the highly structured exposure
rules that characterizeinternational supervision are backed by tradi-
tional moral suasion in which bank regulators use their control of
branch licensing,rediscounts,and other regulationsas leverageto ensure
cooperation.The Bank of Japan, for example, has used constant inter-
action between city bankcsand the supervisoryauthorities, coupledwith
its tight control of branch licensingi to encourage prudent behavior.
Malaysia'sbank supervisorsoperate in a similar flhshion.
CopingwitfDistessedFinancial Iustitutioas.In spite of generaly good
pmudentialregulation, there have been financial crises in Eas Asia
brought about by a mix of macroeconomicshocks,insufficientpruden-
tial regulation,and seculative lending (seetable A5.5, at the endtof the-
chapter). However,goverrnmentsrapid and effTctivcresponsehas been
able to maintain confidenceby acting as implicit insurerwithout exces-
sivefiscalcosts. Rapid growth and spiraling real estate priceshave often
tempted financial instittions into property and sometimesstock spec-
ulation. This has happened repeatedly in Hong Kong (in 1965, 1982,
and 1986), but it has also been a problem in Malaysia,Thailandt,and,
most recently,Japan. By the late 1980s, financial institutions in each of
these economies were facing diffliclties, as nonperformningloans rose
due -to the economaicdownturn.
Korea encounteredsimilardifficultiesbut for somewhatdifkirenrrea-
sons, among them the heavyindebtednessof firmnsthat had been bene-
ficiariesof government-directedcredit. 'When exenldemand fell in
the 1980s, many firms had large excesscapacityand poor -debtservice
capability.Among them were gvrmecnt priority projects in shipping;
overseasconstruction, texties, and machinery.Becausethe firms were
highly leverged, they werefinanciallyvulnerableto changesin demandt
and interest rates.
'When fnnildistres has occurred, HPAEgovernmentshave come
to the rescue_Table A5.5 describesthe originsof and responsesto some
of the significantbanking crisesin the HPAEs.Whil only KoreaandtTab-
Wn, China, have estOablished explit deposit isurance (both in the

2-4
 STRATEGIES FOR RiAPVT

early1980s),they and other HPAEshave for decadesimplicitlyinsured

depositsby steppingin whenevernecessaryto preventbank failures.
In some cases,especiallyin Japan,strongerbankswerepressedto take
overbanks in financialtrouble.When mergershave been impractical,
EastAsiangovernmentshaveservedas lenderof last resort.OnlyHong
Kongand Thailandhave liquidatedinsolventfinancialinstitutions.Yet
iecauseof rougherprudentialsupervisionand better econornicgrowth,
the costsof bank rescuesin East Asiahave been lower than in many
othereconomies.During the 1980sthe estimatedfiscalcost of bailouts
in the HPAEs was 0.5-1 percentof GDP,comparedwith 2.3 percent in
the United States,2.5 percentin Turkey,3 percent in the Philippines,
and 26 percentin Chile.
Nearlyallgovernments regulatethe
Puntein ofBanksfromCompedfioo.
acation of new financialinstitutionsand the entryof foreignfinancial
institutionsto try to ensurethar newcomersare solventand willremain
so. But manyEast Asiangovernmentshavegonefiirrher,limiting com-
petition and protecting existingbanks both fiom foreign banks and
from new domesticcompetitors.With the notableexceptionsof Hong
Kong,Indonesia,and Singaporewhichwelcomethe openingofnew fi-
nancialinsritutionsprovidedthey meet capializationand other regula-
tory requirements,most HPAEs have had rigid restricions on entry.
When expansionhas been pennitted,it has mostlybeen through the li-
censingof new branchesof exsting banks.Despitelimitson enny, rhe
bankingsectorsin the HPAEs areno moreconcentratedthan in otier de-
velopingeconomiesor OECDeconomies(seetableA5.6, at the end of
the chapter).This means that bankingsysremsdifferlitde in structure
among economies,although the onrginsof their relativelyhigh interna-
tionaldegreeof concentrationmaydiffer,as we shalldiscussbelow.
Why were entry restrictionsusedrather than high net worth stan-
dards to encourageprudent behavior?Entry restrictionsgivegovern-
mentsa morediscretio,-. y tool to influencethe behaviorof banksthan
net worth requiremnentr ' tovernmentsmayexercisediscretionin decid-
ing whichbanksshouldexand (throughmore branches)and when.
Bankshave an incentive to respondto govenments' actions lest they
forgothe additionalprofitsfrom asset growth.Net worth requirements
do not providegoveriments this avenue of influence.
Have these restrictionson compentioncontributedto increasedfi-
nancial savings?The answeris ambiguous.Larger, more profitable
bankshavemoreto loseand so a greaterincentiveto maintaintheirrep-
.. . . . . .~~~~~~~~~~~~~~~~~~1
 utations by prudent lendling. Moreover, restricting competition mn-
creasesprofits, and higher profits increasethe financialstrengthiof the
banking system (provided banks do not distribute profits to share-
holders).These effcts would raisesaverconfidenceand hence increase
savings.Yet,in most economnies regulatorshave arguedthat the financial
sectoes tendency toward concentration is so great that govermnments
must interveneto maintain competitionwithin the banldngsector,pre-
venting monopolistic practices that squeezedlients, reduce efficiency
and innovation,and hence, lowersavings.'
Therefore,whethe-rcontrolling competition results in &avorablcor
unfaivorableoutcomesfor savingsdepe-ndson whetherthe increasedsol-
vency comes at the expenseof efficiency.The evidenceamong devel-
oped economiessupports the view that financial systems with fewer,
larger banks outperform systems with many smaler banks (Vitrs
1991). But in many developening econoes financialmarketsare domi-
natal by a few large,veq inefficientbanks whose size is bad less on
economiesof scalethn on restrictionsof bank licensesand othecrlimits
on comp'etinon.
The impact of limited competiton on banking efficiengcvres
acrosseconomiesin East Asia.Japanes commercalbanks, for example,
are regardedas efficient,despite enty restictions, of the pres-
sure of competition among existing banks (Sakaklbaraand others
1982). Thee is no dominant commercialbank in JapanyThe top five
ban are roughly equal mnsize,and they accunt for about 40 percent
of total commercialbank assets(seetable A5.6). In comparisonto large
German banks, Japanese banks tend to have lower lending margins
(spreads),although smal German banks have lower spreadsthan small
Japanese banks. The Japanese banks also have lower lending margins
tha equivalentbankcsin the United States but approximatelyequal
rates of return oncquity (Ueda 1992). Gross nterest miarginsas wellas
net operating costs (as a percentageOf total aSSetS)ofbanks in KoreaI,
Malaysia,and Thailand werein line with oECDeconomiesand substan-
tialy lower than othecrdevelopig economics notably Liberiof Sri
Lanka, and Turkeysand selectedeconomies in Latin America (Hanson
- = ner ~~~~~~~~~~~~~~~~~spo
p cot(asofprcetg
rtn Be
tt t)of
and Rocha 1986). Conversely,studiesof the bankingsystemin Indone-
sia and Koreafind that protecion has led to lessianovautvepracbtces(n-
ternal World Bank reports).
Thus the evidencedoes not condusivelysuprt the efficacyopl
acesto imitentry.CWhereted,encies toward large nancial institutions

z16.
 STRATEGIES FOR I _

are strong, govrnment cfforcs to limit entry will mainly reduce efficiency
and lowerinnovation.This is especially true of Mstrctions in some
economieson foreignbanks,which tend to be the leadersin innovation.
Regilalg Spreads.Exceptfor Hong Kong and Singapore,ali HPAEs
haveat timessimultaneouslyregulateddepositand lendingratesand con-
sequendy the spreadsof financialinstitutions.If spreadsare controlled,
bankswillearn limitedrentsdespiceprotectionfromcompetiion. In gen-
aal, lendingspreadsof the HPABs have been comparableto many OECD
economiesand lowerthan most other devdopingeconomies,indicaung
that controlson intermediation magins were generallyeffcidve.Figure
5.3showsthat avage terestspread of dte iHPAEs (about 3 percent) is
lower than aveg spreadsin other developingregions,whichrangefrom
4.0 to neary 10 percet. Maren in Indonesia,Thailand, and Taiwan,
China,wereabout3-4 pcent in the 1980s.Thesemargi ar compara-
ble to those in the United States (2 percent) and Japan (3 prcenr). In
Koreaand Malaysiathey weregenerally igher at about 4.5 percent (see
tableA5.7,at the end ofthe chapter).Thus, whilerestricingcompetition
greneratedsome rentsfor banks,a combinationofcompettion among in-
cumbents,regulationof spreads,and evenntalliberalizationof entry re-

Fwue 5.3 Spreads in Financiallnternediatioi, 1979-89

P-cntge pentspread

- I I r @ _ l /
14~~~~~~~~~~~~~

- ' --- t -, ! fl---l --- -/- .-

. \. ; . , ~~ ~~~~~~~~~~i I o

- 0- , ' ,-m'-. 1-

0~~ ;(t,~ . K
~ .

2- 1 1-'

1979 1930 1981 1982 1983 1984 1985 1986 1987 2988 1989
a. Indudes countnresin Sub-SaharanAfricaand Nonh AfricL
S:ournranamdonal MoncrccyFunddan.

2;7
 ..- A GL EL

madebanksin East
stricdions Asiarlaivelyefficientcomparedwidtthose
in otherlow-and middle-income economies.
Regulatedspreadsare not alwaysbinding,however,becausebanks
can circumventthem by chargingfeesor requiringcompensatingbal-
ances.Banksin Japan and Korea,for example,chargedcompensating
balances.Thereis someevidencethat thesepracticesweresufficientto
circumventthe regulationsin Japan (Ueda 1992).Intermediationin-
comeofcommercial banksin Koreahasbeenhigherthanthesimpledif-
ferencebetweenlendingand depositrates.In contrast,regulationson
dcpositand lendingraceshavebeen bindingin the state-ownedcom--
mercialbanksof Indonesiaand Taiwan,China. In Thailand,commer-
cial bank lendingratestended to bunch aroundthe ceilingrate, and
interestratespreadsincreasedslighdyafterthe mid-1980sdueto a drop
in depositratesand the relatively
unchangedlendingrateceilings(anin-
ternalWorldBankreport).
CreatngPostalSaiugslnsoboansIn additionto effectiveprudentialsu-
pervisionand regulationof entryand spreads,all of whichencouraged
savingsby ensuringstablebanks,Japan, Korea,Malaysia,Singapore,
and Taiwan,China,established government-run postalsavingssystems
to attract small savers.Postalsavingssystemsoffrecdsmallsaversgreater
security and lower action costs than the private sector and were
thereforeparticularlyeffectivcin attractingto the formalfinancialsector
the savingsof low-incomeand mralhouseholds
Japanestablished firstpostalsavin progamin 1875,with
the region's
the explicitgoalof fosteringsavingsof rual dwelersand peoplewithlow
to moderateincomesin thecitiesand towns.Undlthen,suchpeoplewere
exdudedfromthe financial
effectively ssten, whichlackedruralnetworks
anddiscouraged smalldepositors byrequiringhighminimumbalances or
payingverylowinterestraceson smalldeposits.TheJapanesegovernment
heavily promotedpostalsavingamonglow-income householdsandmade
the interestincomeon smallpostalsavin depositstaxfree.
Similarinstitutionswith thesamesavingsmobilizationgoalswerees-
tablishedin Korea,Malaysia,Singapore,and Taiwan,China. Like
Japan,both Koreaand Taiwan,China,havegrantedtx- exemptstatus
to the interestincomefrompostalsavingsduring longperiods.In Tai-
wan,China,postalsavingsofficesaccountfor aboura thirdof allfinan-
cialinstitutionoffices,and the postalsavingsservicehaslongerbusiness
hoursthanotherfinancialinstitutions(Shea,fbrchconiing). In Malaysia
andSingapore,wherepostalsavingshavealsoaccountedfor a largepro-
,.. - - :.... .:
 STRATEGIES FOR. RAPD

portion of domestic deposits,the governmentsseparated management

of the savingssystemfrom the post officein the early 1970s,when the
proportionof postal savingsin total savingsdedined, cvidendybecause
postal employeeswere not enthusiasticpromotersof savings.Even so,
the savingssystem continued to utilize the post officesas a deposit-
taking branch network.
Postalsavingssystemscan be an effectiveway of mobilizinghouse-
hold savings,provided that governmentstake care thar the cost of ad-
ministerng the systemsdoes nor outstrip the benefits.Geneally postal
savingssystemspiggybackon the mail-deliveryinfrastrucumre, thus min-
ining overheadand fixedcosts. In the earlyyears in Japan, the system
often operated out of the houses of wealthylandownersthat were al-
readyservingas postalbranches.As postalsavingsexpandedand the ad-
ministrative burden rose, posta officals demanded wage increases.
Overall,however,the postal systemscost of acceptingdepositshas been
much lessthan that of the privatebanks (Mukai 1981).
ForedSavins Besidesthe above measures,some HPAEs have tried to
compelsavingsthroughseveralmeasures,includingmandatory pension
schemesand restictons on consumptionand borrowingfor consump-
tion. Becausethese measures consttute more active interventions in
markers,their efficacy,even in East Asia,is open to question. Restric-
tons on consumer choices, induding the basic consumption-savings
decision, have welfare costs. Moreover, similar efforts in other
economieshave been spectacularfailures.Examplesindude the Wide-
spreaddeprivationand massivewaste associatedwith forced savingsin
the now-defunctcommand economies.
Three economiesin East Asia,Japan, Malaysiaand Singapore,have
well-developed,mandatorypensionplans The impact of these plans on
aggregatesavings depends on the degreeto whichthey subsdtuteforvol-
untary savings.Evidenceof the impact of pension funds in Japan and
Singapore,the only two economieswherethe issuehas been studied, is
incondusive.One studyof savingsin Japan found pensionfundshad no
significantimpacton tocalsavings,althoughthey did channelhousehold
savingsinto financialasset therebycontributingto financialdeepening
(Delde 1988). Anoher found a small negativeeffc (Nogui 1985).
This SU that the Japanesepension fund squeezedout a portion of
volunary savings Conversely,there is evidencethat SingaporesCenal
ProvidentFund boostedaggregatesavingsby about4 percentof GDPdur-
ing-the 1970s and 1980s(MonetaryAuthority of Singapore 1991).

.219
 * jILA~~~CLE

Thelackofconsumeraedit to purase housing,constmerdurables,

andotherconsumeritemsmayhaveinducedincreased savings
household
in somcEastAsianeconomies.Bankregulatorsin Japan,Korea,and Tai-
wan, China, rsriaetd credit availablefor consumerdurablespurchases.
Maki(1993)offersevidencethat the rapidincrasein savingsin Japan
afterWorldWarII wasdrivenby theneedto acquireconsumer dumables,
and he showsthesamepatternexistedin Koreaand Taiwan,China.As
household incomesandthedemandfor consumcr dumblesinacased,sav-
ingsas a proportionof incomeroscrapidly.Oncetheexcessdemandfor
consumerdumbles wasmemsavingsratessmbilized andevendeclined.

Savings?
HavePordesIncreased
Howeffective savings?
werepoliciesin theHPAEsaimedar increasing
growthconditionsand goodpolicyfundamentals
Favorable accountfor
Maoeconomicsta-
mostof East Asiassuperiorsavingsperformance.
bilitycombinedwithsolidprudenrialregulationof financialinstitutions
createdan environmentin which savingsweresecureand retainedtheir
value.Limitsto competitionamong banks and the willinpess of gov-
enments to rescuefinancialinstitutionsin troublehad costs-reduced
efficiencyin theformercaseandfiscalcostsin thelatter-but thesewere
moderateandmayhavebeencompensated for bytheincreasein savings
engendered bypublicconfidencein banks.Postalsavings hadthe virtue
of beingbothsecureand accessible instruments for smallsavers.Gov-
ernmentsin Japan,Korea,and Taiwan,China,aswel asMalaysia and
Singapore,operatedtheseinstimtionsefficiently, tappinga nontradi-
tionalsourceof financialsavingsat rdativelylowadminisa tive coss
Havemeasurethatrequired consumersto savemorethantheywould
havevoluntarilyincreased economicwvelfare?Theansweris ambiguous.
Restictionson consumer creditraisedsavingsratestempoarilyinJapan
but at a cost in termsof consumerwdfar Forcedsavingsthrough pen-
sionschemeshavenot beena majorficaorin increasing savingratesin
Eassia, exceptperhapsin Singapore, butpublicsavings arean impor-
tantand complementary sourceof highaggregare Tojusr in-
savings.
trening in households'savingsdeasions,socialreturnson the incread
savingsmust exceedthe privatcoppomruitycos detenined by the rate
ofreturnon investment.Thiswouldbe thecase,forexample,iftotal fic-
componentof
growthwerepositiveand a significant
tor productivity
overallgrowth,eitherasa onsquenceofdynamicscaleeconomies orof

220
 STRATEGIES FOR RAPD
X i.*

cxtnalities associatedwith investment.The social return also must ac-

ceedthe private returnto jusfyr interveningin households'savingsdeci-
sions. In Japan (during its rapid growth period), Korea, and Taiwan,
China, there is cvidencethat economywideproductivitygrowthhas been
positiveand rates of return to capitalhave been good. In Singapore,how-
ever,evidencesuggeststhat aggregae total fctor productivitygrowth is
negativeand rates of return to capitalare low (Young1992). Singapore
may havecompelledits consumersto savetoo much.

Rates
EastAsia'sHighInvestment
Explaining

Q JNE
OF THE MOST REMARKABLEASPECTS OF THE HPAEs IS
their unusuallyhigh rates of privateinvestment.Privateinvest-
ment in the IIPAEsaveragedabout 7 percentagepoints of GDP
higher than other low- and middle-incomeeconomies (seechapter 1).
As with East Asia'shigh rates of human capital formation and savings,
invsment has been fosteredby macroeconomicstabilityand rapid eco-
nomic growth. Growth,savings,and investmentinteracted in a viruous
cirde as high investment initiallyspured growth, which resultedin in-
creasedsavingsto support continued high rates of invesnnent. In addi-
ton, two general conditions have been particularly important to
nvesntmen. Thes-secu property rights and complementarypublic
invesrnenr in infrastructure-are key elements of the market-friendly
institutional environmentdiscussedin chapter 4.
Securityof property rights is so obviousthat it might easilybe over-
looked. Where foreign or domestic private investors fear completeor
partial expropriation of their investments, private investment initative
is naturally reduced. Thus secureproperty rights embodied in the legal
framework were a key to high investment rates. In addition, formaland
informal mechanisms leading to enforcement of contracts were an im-
portant contributing factor.
Data also suggest that the developing HPAE governments have done a
better job providinginfrructure themselvesand, at times, creatingthe
conditions for the private sector tO provide it. Limited econometric evi-
dence suggeststhat these investnents, at least in the casesof Korea'and
-Tiwan, China, generate very high rates of return (an intemal World
:Bankreport). Easterlyand Rebelo (1993) present cross-econonlyesti-

. .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~22
 CLE

matesthat infrastuctureinvestmenthas a high payofffor economic

growth.Theseinvestments havcbccnhighlycomplenenaryto private
investment,
especially manu&cauring
in export-oriented activities.
Table5.7showssomeindicationof infrastructuraldevelopment for
the developingHPAESand selectedlow- and middle-incomeeconomies.
Evenifwe set aside urbanizedHong Kong and Singapore,the HPAEs'per
capita electricity generating capacity is equal to or exceeds that of
economiesat similarlevelsof per capit incomc. Similarly,from 1975 to
1985 the five developing HPAEswith nontrivial mral sectors increased
telephone servicemuch faster than other economies with roughly the
same population per phone in 1975. For example,in Malaysiathe num-
ber of telephones per person roughly quadmrpled,while Brazil,Mauri-

Table5.7 Direct Provisionof infrastructure,SelectedYears

-- ta
;-Wr FB!N
cenw f.
Paetge of w.c4 e odaivr
roads e, padca ; telhone'
Ec; rsonomy^
0-i- 1971- *1986 _.1975 -1985: g
RPAFs
- Hong-Kong-
. 100- 100 160.6-18.3
s ^^. . .....
62.2 77 ... . . 4025 '205
coe.Rep.
-14 of .2 542 332 . 5621-- 2- 5
. $a1 .yuia
' . 85 80 2818 42 12
S ;-in , '= *r,=, ,'-. i 95 8 182.313t3;
3-
Thailand 67 ' 396 --18.2, 174.2 13 6

W 49:C-' ,~~~~~~~~~~~~~~~~~~~~--8
-:351'7 <
- : ; .... : ~~~->-: -:11.6 ,...; 12.4. 16 i95.-- -;309.0, 24
:.......... ¾ 167
-. C'Ivoi-e .0.0 7.9 11' 98.6: '-
Gn . ..528: 79.7
India L 34W ~~48.0- 209 89.5 '352, -20

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

. Si
tnlt 66 1P7o-mlj
FL 7~~~~~~~~~~~~~~~!

=2.2
 STRATEGIES FORRKR.A%W

tius, and Peru had smallergains. Al though increasesin infiastructure

are clearlycorrelatedwith income growth, recenteconometric evidence
suggestscausality may run from infrastructure spending to growth
(Easterlyand Rebelo1993; Canning and Nry 1993).

CoordinationPmrblemsthat Hinder Inrestment

Beyondestablishingivorablegeneralconditionsfor investment,most
HPAE governmentshave employeda varietyof policiesdesignedspecifi-
caly to address coordination problems that may hinder investment.
These problemsare derivedfiom shortcomingsin markets for informa-
tion and in marketsfor sharing risk both types of problems exist in all
economies,but the problemsand the consequencestend to be worse in
developingeconomies.For example,such developing-economy informa-
don problems as poor accounting standaris and a dearth of merchant
banks and other institutions to monitor corporate performancemean
that bond and equity marketsare often weak or entirelyabsen.' 7 Simi-
larly, while market mechanisms for sharing risk are limited in all
economies-firmsin developingand industial economiesoften cannot
buy insurancefor the most seriousrisks they face-the consequencesare
more severein developingeconomiesin whichlower wealthlevelsmake
households and firms more vulnerable.Furthermore, the transforma-
dons inherent in development,such asthe creationofnew firms and new
industriesand the absorptionof new technologies,heighten uncertainty.
Polciesto PmouteInmeshment HPAEresponsesto theseproblemscan be
grouped into three broad categoriesFLrst,as theseeconomieshave ma-
tured,policynmakers havegraduallyincreasedtheiratrteion to the creation
and improvenentof bond and equitymarkets.Hownrer,whilethesemar-
ketsare playingan increasinglyimportant roe they werenot generay a
key factorin mobizing investmentduring the HPAEseconomicmkeoffi.
Second, many Hwss created developmentbanks to ease constraintsin
long-terncapitalmarkets,as wel as specializdinstitutionsto providefi-
nancig to agriculmreand small-and medium-sizeenterprises.These c-
forts-appear,on balance,mthavebeen moresucccssfi4in the HPAEsthan in
otherdevelopingeconomies,for reasonswe shal exploreFmally,HPAE gov-
ernments have encourged investmentwitha wide arrayof other mecha-
nismsdesignedto inacase the atrcdveness of privateinvestment..
The remainder of this chapter describesspecificpolicies in each of
these thiee categories-supporc for bond and equity markers, establish-

223
 ment of decvelopment banks,and mechanismsto increasetheattraative-
ness of privateinvestment-and attempts to evaluatewhethereachwas
effectivein increasinginvestment.Wedevoteparticularattention to the
third category,whiichincludesseveralpoliciesthat involvea high degree
of interventionin markets.At the simplestlevel,many HPALEboosted
corporations'retainedearningsby adopting low corporatetaxes.So'me
alsoreducedthe cost of investmentby adjustingtax,tariff,and exchanige
rate policiesto hold downthe relativepriceof capitalgoods.Severalgov-
ermecnts, particularlyjapan, Korea,and Taiwan,China, triedto reduce
risk for investorsby shoulderingsome of it themselvesor spreading it
acrossother firms.The northern-tierFIPAE-s also stanldout for their ef-
fortsto increasedomesticinvestmentby restrictingcapitalflows,Finally,
severalHPAEs, againparticularlythe northierntier but alsoMalaysia,sub-
sidizedinvestmentfinancingbyslightlysuppressingthe interestrateson
savingsdepo'sitsand corporateborrowing.This last mechanism,which
we term mild financialrepression,playsa k-eyrolein capitalallocaidon,
the subjectof the next chapter.
Bendamd EqLitMarkets, DevelopedaftertheTakeoff.Despitethe phenom-
enalgrowthofAsianco:poatons, bond and equitymarketshaveplayed
a relativelysmall role in Asian finance. As table 5.8 shows,bonds on
averagebetween1970and 1990accountedfor lessthan 10 percentof the
net financingof corporationsfor the fiveHPAEs forwhichdataexist This is
higherthan the shareof bond finanding-inthe oECDeconomies,however
Ironically,pofides that contributed tcosome of the HPAEssuccessin
developingother financialmarket institutionsmay have been inimical
to the earlydevelopmentof a corporatebond market.In Japan, for ex-
ample,the long-ternm creditbanks govermnment-protected monopolyon
the issueof long-ternm debenturesdiscouragedthe developmentof bondi
8
markets.' TeTa governmentcreatedthe samemonopolyfor its pri-
vate long-termcredit banks.In addition,tb"~lack of governmentsecuri-
ties in most HPAEs, a direct outgrowth of their successat minimizing
fiscaldeficits,has deprivedembryonicbond marketsof the most widely
utilize-drisk-freebenchmarkrate. This has impededcorporatebondtism-
sues,sincemarketsfind it difficultto determinesimultaneouslythe risk-
free,rate and the risk premiumassociatedwiiAa specificcorporatebond.
In response,Hong Konghas issuedbonds,re.-n though the government
did not need the financing,to establisha risk-fleebenchmarkfor the
corpomte bond market. Malaysiaand Sina ar now considering
followingsuint.

.22.4
 STRATEGIES FOWIRAP-r"DA&

-table5.8.NetFinancing
Soumes ofNonfinancial
Corporations
in EastAsiaandSelected
OtherDeveloping
Economies

I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

~~'S57D.75 Z'~~~.6 $245 LI14

Malajiula.
I 126-Si $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

jan23.2.&)
!P 1.
LuSti.
v~74;- 23 2 7 8 '
T..wi
b.CaWuW.K
m.a1dua~~~~~~~~~u~~~y.,~~~

_:; immu' a w0.A,i~ -VNIIA RamilkWm B.ubumiawamuiEsu

Wu Iua hl.1z9h2.~
a hitoicAym been small.Makfhzds appeamrs t b

changing- as the Asian-economiecsmature. Asian corporations with

their headquarters in Japan, Korea, Taiwan, China, and othe-r I?PAE
aeincreasinglyissuing.corporate bonds, drawing on the 'strength -of
their worldwvidereputations and accessto international financial mrar-
ks'19 Recognizing-the value -of a healthy bondmarket, Japan has
*eased lalwsthar discriminared agamnstcorporate bonds. Similarly-Thai-
land, a.nd-rec'endy.Indonesia. hav strengthened rating and reg ulatory
agendes, -while-Hong ,Kong has refornictl its regulatory-structures.: .

Malaysiahas introduced acrotebond rating-agency.Thus,_whlile.

'corporatebonds have -notibeen?
imnportant-in Asia'stakeoff,:theyrare
<lkl o play arnincreasing role in its:.financial-fiicure~.

25'
 .~~~~~~~~~~~~~~~I .

The story for equity markets is broadlysimilar to that for bonds: de-
velopment has tended to follow rather than lead the HPABs economic
takeoff.In this, the HPAESare similar to other economies. Indeed, recent
evidenceshows thar equity markets do not play a dominant role in cap-
ital formation in either developing or industrial economies (Stiglitz
1993a). 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 the
1986 merger of four competing exchangesinto 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 has
liberalized restrictions on share issues and allowed foreign-owned joint
ventures to operate secur;'ies firms.
Even so, before the remarkable bull market of the late 1980s, Asian
stock markets played a relativelyminor 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 volatility
of stock prices in response to speculative pressures and capital inflows
may 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-
of3swere 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 provided
by 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 provide
credit to agriculture and small firms.
-Industrial development banks have been substantial. long-term
leders 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 1950sand about half in the early 1960s. Their share in
total lending to industry was small, however,and has dedinedLAt its
peak in 1953, theJDBaccounted for 18 percent ofnewfimndslent. 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 averageof a third of all loans and
guarantees in the 1970s, and the development bank of Taiwan, China,
the Bank ofCoimunications, holds about half of the assetsof the bank-
ing-system. Conversely, Malaysias development financial institutions
accounted for 2.9 percent of the assetsof the financialsysremin 1980s.
Thailand's industrial developmentbank hasonly 1 percent of dteassets
of the financialsystem. Hong Kong has no development bank
Many other developingeconomies have also attempted to remedy
the perceived failurein long-term capital markets. Nearly all have been
unsuccessful in creating development banks. Development banks in
South Asia, Latin America, and SubSaharan Africa are beset with low
repayment rates; a representativesample of eighteen industrial develop-
ment banks in developingeconomies had on averagenearly 50 percent
of the value of their loans in arrears (World Bank 1989c). The most
commonlycited causesof developmentbank flilureshave been political
pressiureto finance bad projectsand the poor incentivesfor and capabil-
ity oFfinancialinstimutionsto 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 evaluateand monitor projects has at least in Japan
created spillovers for the resr of the formal financial.systen (JDB/JERI
1993).23Successfildevelopment 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 projects
that-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 borrowersreceiv-
ing fumdsfrom the U.S. Agency for International Development to put
-upmathing funds and made lending to a nonperformer a criminal of-
fensei imposed on:the loan Officer.

227
 Most EaistAsian governmnents have devisedmeans to contain wilful
politicalinterferencein devlopment banks. Japan, Korea,and Taiwan,
China, have appointed seniorofficialsfrom ministriesoffinance as chair-
men, so they may withstand pressure from other parts of government
Others have controlledthe tyes of lending.Thailand simply disallowed
its developmentbank froamlending to stateenterprises.The high pro&cs-
sionalismand institutional identificat-ionof sraflj which is charactristic
of successfuldevelopmentbanks, is a positive&actoras well.

Policies
tolacrease
theAttractiveness
ofInvetmnt

Besidiesfostering a dimate conducive to investment, encouraging

bond andiequity markets and, in some instances,establishingdevelop-
ment banks, each of the HL'AEsresponded to coordination problems
that hinder investment with a variety of mechanisms to make private
investment more attractive. Like most other governments that have
tried to increaseprivate investment,the HPAEsembodied many of these
pro-investment policies in their tax codes. These codes varied across
economies and over time, both in their specificintentand in the degree
of succe-ss,but have nearly alwaysaimed broadly to encourage invest-'
ment. Becausetax codes are often central to the quait-y of the invest-
ment climate, this section begins with a brief survey of tax policy in
each of t-hewpAEs. We conclude the section, and the chapter,with a dis-
cussion of foiurhighly specificmechanisms usedtin some of the HrWAEs.
to encourageinvestment. low relativeprices for capitalgoods, risk slar-
ing, controls on capital outflows,and mild financial re-pression.
TaxPolicies,to Encorage Investment The HPAEShave employed a widle
variety of tax policies to increase investmnentby raising the retained
earnings of companies. On one extremne,Korecaand Taiwani,China,
used complex tax codes not only to favor investment but also to direct
the pattern of their industrial development. On th: other extreme,
Hong Kong has maintained a simple and largely neutral tax structure
with lo croaetax rates. Other tax policy regimes fll inxbetween..
Singapore uses taes to promote industrialization and to attain social
objectives.Indonesia and Malaysiahave relied heavily on oil revenues
andt so until recently have paid little a-ttention to other aspects of tax
policy. Like Korea and Taiwan, China, Thailand has tried to use tax
policy to provide investment incentives; however,weak administrative
cap.abilites mean such efforts have been distortionsxy rather thin

2oi.
 STRATEGIES FOR RA A

helpful. Box 5.2 further explores these issuesfor Korea, Malaysia, ant
Thailand.
Below we review briefly the investment-relatedaspects of the tax
structr 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 expendituresare exduded from the tax base. New industries re-
ceive a five-year tax holiday, and industry generally is offered accelerated
depreciation of investments. Industries that the government deems
strategic 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 has
continuously fine-tuned itStax 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 low
rate 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 investments
in contrast to the mix of high tax rates and high tax incentives offered in
Korea and Taiwan, China.
* Singapore's tax regime is as interventionist as those in Korea and
Taiwan, China, but focuses special incentives on foreign investors and
on promoting savings. (Contributions to the Central Provident Fund
are tax-exempt for both employers and employees.) Malaysia has the
highest level of taxation. Its tax-GD? ratio averaged 21.45 percent in
1979-88 but fcil steadily during the period because oil revenues
declined without compensating increases in non-oil revenucs, and the
government continued to expand tax incentives to sustain its industri-
ilization drive. The resulting high deficit and emergng constraints on
spending 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, it
embarked 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

- ; obse adeimntd 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 aindii~,lackFTof
b rai capci haecsltti oo m
- pe.n.men.I conrs wihtora s, n Taiwan,Chinaniheri in
-~~~~~~A -ae1nealwrkda
cetie itl, inceninresm hiIn
-~~~~4 apert aedoemr amtl ode h oermn a e

' -.. -~~~~~~~~~~~~~~~~~~~~~~~~~~k

y .. compl .n -:f iuiesnaX :
i:~~~~~-
-2i :xI .~ ~ ~ ~~~~~~~~~~~~~~~~~~11ta-t
raS: _ , . .
ac- . - - .n
~ f.eos~av-~~t-
.. ;.'-V ~ ~ oX-, o1anes--
*~~~~~~~~~~~~~~~~~~~~~~~~er~-n :aea.s o,
t., -,. .,. . - : S; 5.o: , f, , .
,- : f . . - : . : f - - - - - 0; :-f p-:
 STRATEGIES FOIR-

- 17:&tu~t
r, 4e~½~ ,*. 4.

- . Z>- n

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

fr^½Mstoa ts)boni.tacnnve~idrssi. rte mictaves:ant 'lmiberonibo£anM

{g' E aji&ht
w*kw4ti4<_4$SF-> 4nordi#Pnd

.~~~.rDooe>Doeiln at ite: n~
baafpmeERm-@edlafsicrtoqsnutoC*Ddae
~ 0,fl
GDP~Its
- 'zi*:<firtl*8if =lMQl:}enn,>zem ,J -
5
49.W4.,. f1 'tl
e ^D¢ 4 4$i.<v:eiopmencfI rmi4mlulceSeancQgi
jO?9S 3 &rntlolsescllubTncd
PgrOwd ~- S16 !5j7crceXu
alyEboult
-.-.{S
f,!ri< onoImc,3n&213 *^ t.ct sFMi

gu 5brd

* eau:tvrgone:revcnuc. 4 N~S de2Tmcnc *u;lro:et ooteono

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

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

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

Tax poli-
des involve fiscalsacriflice,a trnsfer of income from taxpayer in general to
hii depends
investors. The assessmentofwhether these polices were successf
on similar considerations to those applying to pubc or forced savings, If
the increased investments were absorbed effidiently, as ir appears thcy were
*in mostroftherHrAEswitha2ctvisttax polldies~itrislikelythatrthey 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 Hong
Kon with very low but unifrm. tame on corporate incomes, suggests that:
the major gains accrue from leaving mrtaiedearings with corporationss
ceandpionolvng
fivestment i equipmmen,rather than from attempting to
tfine-mneincentive strucurecsto private investors through the tax system.

7231
 LowRelativePties forCapital Goods.Unlikemostrocher low-and middle-
income economies,the H-iPAEs wereable to hold down the relativeprices
of invertment goods, especiallyequipment, during the 1980s (Bhiar-
tacharya and Page 1993). Figure 5.4 shows the relativeprice of invest-
ment of goods (ratio of the investmentdeflator to the GDPdeflator)for
forty economies for whi'ch consistent real private investment data are
ava-ilable.F-orthe group as a whole, the price of investment goods in-
creasedabout 15 percent fasterthan all otlicr goodsduring the 1980s,in
part because devaluationsincreasedthe domestic price ofinmportedand
import-intensiveinvestment goods.
The HPA.Esarc different in severalways. First, the relativeprice-of'mn-
vestment goods remained lower than in other developingI economiues
:i~~~~~~~~~~~~~~~~~~~~
throughout the global economicexpansionof the 1970s and the adjust-
ment period in the 1980s. Second, the relative price- of investme'nt
goods declined during the early stages of adjustment frm 1980 to
1984-a period when they were rising in most other developing
economies-and
letog od rtoo only began to rise in'1985, when
h necetdfao the HiPAEeconomies
oteGrttso)fo
werewell on the way to recovery.By 1989, relativeprices betweenHPAES
- -foerecnoie fr hih onisenrrel riat iarsten dt0ar
and ocher economies were again aligned. Thus in these economies de-.
NowLMeafe oregopa
104lbe ices fo=aia awoe h middle-ncmconcs
od low-niems
and rieo tenetmn od-n

Figure5.4 InvestmentDeflator/GDP
Defator
- ta~~~~~~~~Suceay
PeTrandPand Mad93).a (1992) hwv h daieprc f net
Index1970 _
120 -

vestme.t godsreaie lower

middler-ievlopin
than . economnies
- ;: th lbleooi
Pout-an
ir~~~~~~~~~~~~ource Madaassa(I 92xaso fe17sand
12±2
: e.peidi2h328sScnd h eaiv rc f neimn

232~~~~~~~~~1
 STRATEGIES FOR RAPI

cdiningreal 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 nominal
investment expenditure is needed to achieve the same volume of physi-
cal investment. For the HPAES and the economies of the non-HPAEsam-
ple, investment efftortand outcomes coincided quite dosely during the
1970s when the relative price of investment goods was essendiallysimi-
lar. During the 1980s, the rising relative price of investment goods in
non-HIPAEs was reflected in an increasing divergence between investment
effort and outcomes. By 1989 the difference was nearly 3 percentage
points of GDP?By contrast, in the HPAES,investment outcomes exceeded
investment effort (for both public and private investment) for the entire
period 197089.25
Tax, marif,and exchange rate policies that kepc the relative price of in-
vestment goods in the HPAEsbelow that for other low- and middle-income
economies undoubtedly contributed to growth. Snce the same volume of
nominal 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 increaseinvestment by lowering the
uncertainty associated with real investment, implicitly or explicitly shar-
ing risks with the private sector. Risk-sharing mechanisms have come in
many forms-recession cartels in Japan, firm and bank workouts in
Korea, financial repression to recapitalize firms in Malaysia. signaling
prorities and policy intentions through the direaed-aedicsystems of
Japan, Korea, and Taiwan, China, and credit guarantees to small and
medium-size enterprises in Korea and Taiwan, Chinat Those that were
effective all exiibited a common feature firms benefiting from shared
risk were monitored for performance. In this way governments miti-
gated the problem that has plagued many public sector attempts to share
risks with rhe private sector in other low- and middle-income
economies: reducing the risk of fihilureto the private sector reduces the
incentive to avoid failure. This can lead to a political dynamic in which
gains are private but losses are socialized.
An importnt feature ofjapanese industrial policywas the creation of
recession carrels and the use of other forms of adjustment assistance in
dedining industries. Japanese andimonopoly law has been consistendy

233
: :: gC -: A CLX
0 _EEFLE V-~~LE

lenientin allowvingthe organization of cartelsunder the administaive

guidance of Mm to ease the exit of firms from dedining industries.
These cartelswere especiallycommon during the rapid growth period
when they were used to ease the adjustment problems of dedining in-
dustries. The cartel allowsfirms to share losses.They may also be eligi-
ble for subsidizedloans to upgrade technologyor restructureoperadons.
In ths way, bankruptcy or massivelayoffs by one enterprise may be
avoided (Ito 1992).
Japaneseadjustment assistancchas been especially prominent in min-
ing, textiles,shipbuilding,and aluminum. In generalit has succecdedin
avoiding the abrupt dislocationsthat often characcerizefirm fiailuresin
market economies, but this orderly decline has been purchased at the
cost of inaeased prices to consumers and lower reurumsto investorsin
firms that would have survived.Moreover,the cartelsprovidean incen-
tive for firms to overinvestin capacityduring good times,since they will
be protected during downns and their adjustnent assistancewill be
related to their existing capacity.
Performancecriteriaplay an important role, even in declining indus-
tries. Interacton among firms and between firms and MITIare a means
to prevent free riding and to monitor performance.Thus the organiza-
tion of an industial adjustnent assistanceprogram in Japan conforms
to the model of contests presented in chapter 2 and suggests that
economies without similarlywell-developedinstitutional capacitymay
wish to avoid the commitments to orderly exit policies.
In some HPAEs,most frequentlyin Japanand Korea,troubled priority
projects have been bailed out by the govemrnment. Often, the financial
cost of the bailout was large. In Japan the govenment took over some
lossesassociatedwith financingdecdiningindustries,such as in coal min-
ing. When govenmuent-supportedheavyand chemnical industry projects
in Korea experiencedsevereexcesscapacity and financial diffidultiesin
the 1980s, the governmenzprovided seventy-eightdistressedfirms with
new, subsidizedloans (totaling abour 16 percent of commercial bank
loans)and rescheduledoutstanding ones. The govrncmentalsoprovided
banks, whose nonperformingloans rose substantiallyand whose profits
dedined (partly becauselending rates werecut), with subsidizedcredit
from the Central Bank and allowedthem entry into attractiveareasof fi-
nancial services.In Indonesia, the govemment bought 35 percent of the
equity of a large ement plantwhen it becamefinanciallytroubledwhile
operating only at 50 percent capacity (Kunio 1988).

234
 STRATEGIES FOR RAPI

Dcvcdopmenrbank lendingmay 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 adjustment
problems in promoted sectors. With government inducement, the In-
dustrial Bank of Japan has led syndications for firms in distress and
forced their restructuring&substituting for liquidation or external
cakeovers.In the 1980s Koreaextended cheap development bank loans
to firms that were nor meeting their debt obligations.
Has bounding risk contributed to growth?Bounding risk by the pri-
vate sector was apparently successfulor at least not a fitilurein Japan,
Korea,and Taiwan, China, but this is almost unique comparedwith the
experience of other economies, developing or industrial. Risk sharing
generates moral hazard problems, which Japan, Korea, and Taiwan,
China, attempted to contain dtroughfostering insitutions to monitor
priority projects. One consequenceof shared risks was shared monitor-
ing by the private and public sectors. Severalinstiutions, induding de-
velopment banks, private financial institutions, firms, and various
ministries in the governmenr,often had stakesin the implementation of
the project. The government's role as monitor of firms was important.
For example, Korea'sfinancial policies promoted high leverage,which
made firms vulnerable to loan rol over decisions of financial insttu-
tions. Since the government was able to influence which loans were
rolled over, it was able to use them as a lever for elicitingcompliance
from firms (Sakong 1993). Similar monitoring was rarelyoresent in the
Southeast Asian HPAEs, with the consequence that attempts to bound
risk there were not successful.
Coeralson CapitalOufflows.Restrictng capital outflows is one of the
more controversialmechanisms employed by the HPAEs to increasedo-
mestic investment.Japan, Korea,and Taiwan, China, all employedsuch
restrictions during the formativeperiods of their rapid growth. In each
case the rationale was straightforward:if people were prohibited from
sending capital abroad, they would saveand invest at home. In addition
the capital controls made repressing interest rates on deposits possible,
since savers were denied higher-yielding assets abroad. Although all
three economies have since liberalized controls on capital flows to vary-
ing degrees,the existenceof restrictionsat a time when these economies
were achievinghigh and risingsavingsrates challengesthe premise that
free and open financialmarkets are alwaysbest for growtk

235
f-a.;;.. 5Y

WVhatimpact, then, do capital export restrictions have on accumulda-

dion?The obvious fuilure of such restrictions in economicesprone to cap-
it-alflight suggests that capital is too fluiidto be retained wvhendomestic
conditions are inimical to savings and investmenc. Therefore the sim-
plest explanation for the seemningsuccess 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 in
eliciting 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 of
sending capital abroad. If so, such restrictons; which have the potential
disadvantage of driving otherwise legitimatc ecoonomnicactivity under-
ground or even provoking capital flight, would rank among the compara-
tively risky interventions that are likely to prove successfujlonly in
economies that have able bureaucracies and have already established the
fundamentals of growth. Moreover, as the trend toward capital account
liberalzation in Japan, Korea, and Taiwan, China, suggests, retrctions
are probably useful orl4 for a lImited dimc: as economniesgrow, becoming
more sophisticated andi more dloselylinked to international markets, re-
strictions on capital exports inhibit efficiencyand must be dismantled.
In aniy event, the success of the more open BPAEsundermines any
suggestion capital flow restrictions are somehow necessary for sav-
that
ings and growti. Hong Kong, Malaysia, andlSingapore have all achieved
high and risingsavingsrates whilepermitidngand-in the caseof the
gional financial hubs, Hong Kong and Singapore-even encouraging
free flowsof capital. Indonesia attempted to control its capital account
from 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 capital
movements,but the economyin peracticeexhibitsa high degee of opn-
nes (an internal World Bank report). Thus, although capital account
restrictionsmay have playeda positivcrole in the earlygrowth of some
HPA:s, conomuesthat have never imposed rthemor have since liberal-
ized appcar to have little to gain by imposing them.
East Asia'slesson about capital account restrnctionsmay in fact be
how to get rid of them rather thanhow to impose them. Both Japan and

1Z36
 STRATEGIES FOR R

Taiwan, China, weresuccessfulin liberalizingtheir capitalaccounts very

graduallythroughout the 1970s and 1980s, adjusting the pace as neces-
sary to maintain macroeconomic stability and minimize interest rate
and exchangerate volarility.Indonesia'scapital marker liberalization is
described in box 5.3. But while gradual deregulation may be best for
economies in which bureaucrats can craft economic policy with rela-
tivelylittle interference,sudden deregulationmay be the only deregula-
tion possible in economies where governments are less insulated and
there is strong resistance to reforn. In such cases policymakersmust
weigh the dangers of sudden liberalizationagainst the risk that gradual
reform may be derailedalrogether.
Moderate RepressionofInteiest Rates.The finalmechanismwe discuss-
the repressionof interest rates on corporate borrowing-is perhaps the
most controversial.Financial repressionis often used to describe the sit-
uation when interest rates are at negativereal levels.A more precise def-
inition, the one we use here, is govermnent intervention to hold interest
rates below market-clearing levels, Of course, in many developing
economies with capitaI account restrictions, managed exchange rates,
and undevelopedcapitaland money markets, it isdifficult to determine
a market-clearinginterest rate. In the absence of such rates, HPAEgov-
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 flight
or significant disintermediarion into the informal markets, where higher
real rates prevaiL
Economic theory and empirical evidence agree that if a government
holds real interest rates on deposits too low for too long people will have
little or no incentive to accumulate financial assets, financat savings will
fall, and econonic growth will be adversely affected. Fmancial repression
usually occurs when inflation runs ahead of adjustrnents in govermment-
regulated nominal intcrest rates. This type of financial repression has
often 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
 CL
AR

e nhgh9 L4i meore,s 4 guhuS

dn rs R k4
/ U2 flV.w% 1"
~oi'nrArf
.a ~4W
d&rgdl
~ $¶ ~ iJ*&V

T heHofNavust
assessme beresoom s codnsitn
se fzinanchial n e

5 fnanr
suggests'rle'op n sfrgsures to Hoevr,83acile
elctdco&nomic naecitlrsea.

time or notherheldrntcrestirtes4below makt-dZaringlcv hl

nnthelsHacShaeveinog used ancmia citen

repressionasgrea and de-nia

liiberatepolic fovrtime tosecurentlrsourcessfto finnc the bdgvlpento

branksanr oevriaca
resourcesntoselectedsaig:dcthAineconomicsscos gre-
pression hascabepenssedion,selectd peraniodsrepusually afterany
extenal
shockteto treAsansemresucles?oadete iigidsra on so
time orfanotherahelfinterestl
raespbelow markeostclaringlEeveSmwhil
achitevighgdn
nonetUyhaels aggregateintmtand n hHancial.
pigrtso find

banksgancoimeodpositinofordsavns diaidltheoAsianleconomiesrowes

23-8
 STRArEGIE RAIDXO
R0
=t~~~~~~~~~~~~~NT 37 5

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

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

X.
2

-; . . nr+ridd
rd4 a4the'interestr iid it andservces'inuluhdprvate
pmasewoduc-s b s-
-uMffiedylwn
.i£is s has mrgina e mchiaseher
9SiytoiBhves
- rpn overrnment
hogusaneo ser,Un
Cld r:Fa 12
ipnvacerafinnci 1
* ie ODW.nm_xc0xnVcbUarQto.%cn
n the cc, .' t ;:l
t
t3.es. *r subidy
-: aatet
* - -ment.by reducUingborrowig oRnted,thes 1 j r poirsii bostp 0 cij
y.$0 d &MCM-.-~~~~~~~~~~~~~~~~~~~~.
iid 2 -

ulbCC&
-:cooiesinmthreeaimpom.nvhent dshedrefeeshiond ubkflt&7Q
wasmo
rElativdy
moderate arldw sh -lir`dexcepions di notcrtoet

in
pen;istndy real.ninterestrat tablea5.5
In Iit as. ae non ieod
- i3rprsso of euryn in a enon h&gbvn
merntnenrghn

. .- ... . - -
A.....

- fiom
.to firsns
househol
inreases the oolorate!profitabljusvof invest
; . --.g a ,J . . .

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

- -FandiadorpressincinthuHAEghs
s iffered froimlatStin ro.swd
o nthernasrxcs i4owe
ousehlds o frms inceaesth crprates proitbiit1o
from iinvest-.trq =--

the euityoffins,
nablig thm toinvet inriskier-projetsnd bo Uo
ings is,-
'more
sufficiently low
financethese
to and t corporate
projec.' Thi sector has a highe
ma hvepadyofe te marginal
eanes

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

idingfirns.anatrnatived toa tequt fiainngesrt(Sdlasiity

of993a)ldsav
Fginancufiadpendyionind theHopAESeseodiardfo
has thathin mostgothe
penois inythre nporthan
th ways.hoFirstao,thedereenofrepressayincrwase
~~ ~
- _ -- X - - ..-

aregav1 oeate ansmnd,wirthaewrsore,t-lied

fcdeptiontrs,tio retcsultid
inopeusistenlydnegtoeirealinterestmes (see tablrae 5.5).tbiSecyofnd,estwe

havteshowninychapktedri3g
rheprsion wasonderctakenoifapenvironmentpro
oFmnancninic
rpestainintye HandEws
nohes unintened
rmta conmseqecofhe
ecoomis W27S.rlrt,
i thee mpotanhe legee f rpresio3Wa
 t - 0 _ L E . 0~~~n-t
,~~~~~~~~~~~~~~~~~~~~~~~~
~ .. 1p -,w

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

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

--lasqueezetL+Si
: - ~unosyxyi}neetrtSadvabnod
~ xb~ "Minterajelnal=knsa T,1
- FQ15g"fim%1,,.,-1B I.;-i g _>9-=
.~~~~~~~~R~p . Xtn<sn
or4-mom
oisblyneFEen^o,rmn
B
eottzSnprXdarmem;
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--k; ii f N...................
{ 4swrs emosd(firsJ.
.lE:reaLtdepoitr atesas iw itres-onldt -. ec: oemihan a6ouvzRcDrcenq-,.t.s
- > ,",snewep.zSerm 68.;e.
&1t'aifer- dw mtnznllts-isl-mmedoucti;.
irg;ont
r ' on+ A* Id_ w a c.en I_ . «
- t '+!F
S.. 4 .%>l- # {-- -tuv--tr>> s rv;s.> t, '-tF H:9 :),.-
U- ] t ! S rv-S , 7 \*.M1 s-
X$Sssrtmus y>dforeacchan^ ~~~~~~contos.ce ponsdw7e.cmpabl Ronn

mxtl!oruenv
sann= 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
.- ~ ~~~ ~~~ ~ ~oovr>navnolX
~ ~ ~ ~ - ~ ~~~~~~~~~r
-icmiemltGsaco 1uuNvdeLAtsts@
cosraire-mti=ur.a: -
f~~~~~~~~~~~~~~~~~~~~~~ 2 _N P
~S- t',S:-
.4
,fF
¢ itrFt..4->><stxts,t<i/gt; ,,>,'gsSw4'St;t
i
i
~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~,-; -tU
~~,r-J;.>a)>iJ--.i v sr-eecv4ec,.ak-o
j>S;<,
=tr2aonnPnrllgno-vnvscLitg.de6tea
6ti-;5f

| l- =J oSaN
,4 t:' _fi< . - ' r St-, , *-; 0 2* s- e r., Xs > ; ; <S:t' n - *t i : -- 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 >*-5s-^st2
Srsge ¢Sg.^=^,
wer s>.w---
¢\*1.tPGK>
fNtf<! -O f mof-
. f.mE~~~~~~~~~~~~~~~~~-m
uS.7;1vt
bzmatH.fRftAS;sf-he;Ho
; S-,R,;wiftW:2.S'sruXs
.{2;;7t>iv1X;742--
r it^.=x
zn-<S'6Stae>-en
"Proodx 4.-*l -. -r-9. Pj
tssXadMeneeRnf>Bcue.i

a.-::'
~~~~~~~~~~~~~~~~~~- ,

A.,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

': . : - . ' e~~~rapifinfancial Av indtwirdl bankt r eguatossquee foth dcepoit. Svrate

| - r - 0 - - stuleding rate sp=,d, ehansth rhatidie that resuts pId

into depositorsla

;.~. . - 0 f' -7-' : .0 : households) wercpaed on to borrw s rorld ations). 8).e' ' .

- 1, ' ' ''

;: .- ' : - - '-. Nonethesltni WSuring tleatlo in raest racsae posauedo efetos'ofmow-

';-- --. 'X S X". - ' . ers, finiandial repression with geneitie. mess dmanct frae :ayic SevAeralc d-':
.-~~~~~~~~~~~~~~~~tde hav arue tha thrd th
o.- ~ ~ ~ ~. ~ ..~ -f0~ ~ ~ ~ ~ ~ ~ ~ ~~~m at reslt is infiin and has

-~ ~~ ~ ~0 ~ ~ ~ ~~eav cosqu-e
. :-- fo grwt (Wrl : Bank 1989: Fr 198. -
- , j z X ,, thsaye, e.msancs.-stngin2da.r pes :on !'paricuarl

-' ~~~
*,~ ~ ~~~~~~~~~~~hs
0,' ' -:f 'earesutin ,''in' ne''4rtv
.- . :0:,-': :.: - e,'-t'o mod--,-
 STRATEGIES FOR P.A P,r.

Evn widhoutfinancialrepression,banks do not allocateloansto the high-

est bidders. Since biddersfor credit offrerpromisesrater than payment,
banks must evaluate the bidders and decide which promises are most
likelyto be kept. Since rationing alreadytakesplace in financialsystems,
the additional impact of moderate financial repressioncan be relatively
small. Indeed, credit rationing may actually be good for growth, if the
mechanismused to selectborrowersgeneratesincentivesto perform. For
-emmple,Japan, Korea,and Taiwan,China, allocatedsubsidizedcredit on
the basisof past performance,usuallywith exportsas the criterionof suc-
cess.Since the valueof subsidizedcapitalwas high, firms perceivedhigh
marginal returns from greater effort and competed accordingly.We ex-
plore this finction of credit allocationin detail in chapter 6.
Did moderate financial repression hinder growth? We address this
question by examining the relationships between income growth, real
interest rates, and inflation using two statisticalmethods and two data
sets:a sampleof nineteen economiesthat includesa subserwith positive
real interest rates, and time-seriesdata for Korea and Taiwan, China.
(Details are in appendix 5.2). In the nineteen-economiessample, infla-
ton rates are far more important in explaininggrowti than real interest
rates. i the subset of economieswith positivereal interest rates, the in-
wrtestrate is statisticallyinsignificantand actually has a negativeeffect
on gowth when inflation is included.
Similar tescscan be made using time-seriesdata for Korea and Tai-
wan, China We use the differenceberweenthe curb rate and the insti-
tutional interest rate as a measure of financial repression.Its coefficient
is statisticallyinsignificant.In Korea, inflation and the curb market rate
(our proxy for the equilibrium interest rare)significantlyexplainincome
growth. In Taiwan, China, none of the variablesis significandy related
totrhe growth rate. While these testsare not condcusive,they support the
view that moderate fnancial repression did not necessarily hinder
growth in Korea and Taiwan, China.

Which were the factors that contributed most to the HPAEs7 extra-
ordinary record of accumulation over the past thirty years?Judgments
are difficult,given our incomplete understanding of the costs and bene-
fits of all the policies used. But on balancethe common elementsacross
the eight emnomies tend to favor the importance of doing the finda-
-mentalswell-encouraging stable and predictablemacroeconomicenvi-
., . = , . .~~~~~~~~~~~~~~~~~4
.~~~~~~~~~~dr '.

ronments,universalprimary (and later secondary)education,sound

and solventfinancialinsctations, secureproperty rights and comple-
mentarypublicinvestmentsin infrastructure,and lowrelativepricesof
investmentgoods. Effortsto improvethe institutionalframeworkfor
capitalmarketdevelopmentcamelater in the processand werenot re-
sponsible fbr takeoff. In some cases well-functioningdevelopment
bankswerea positivebut not a determiningfactor.
Moreselectiveinterventions-forcedsavings,taxpoliciesto promote
(sometimesvery specific)investnents, sharing risk, restrictingcapital
outflow,and repressinginterestrates also appearto have succeededin
some HPAEs, especiallyJapan,Korea,Singapore,and Taiwan,China But
the potential costs of these more selectve interventions if misapplied
can be veryhighin termsof consumerwelfare,and stronginstitutional
capabilityis necessary.Theywouldnot havesucceededwithoutthe im-
portant monitoringand disciplinaryrolesperformedby the banksand
public institurions of these economies. Where other East Asian
economieshave lacked this capability-in Indonesia,Malaysia,and
Thailand-ef%orsat selectiveinterventionsto promoterapidaccumula-
tion havebeen generallyunsuccessful.

Appendix5.1: Granger
Causality
Testsfor
Savings
RatesandGrowthRates

EXISTENCE OF A REIATIONSHIP BETWEEN VARlABLES

proves neither the existenceof causalitynor the direction of in-
fluence.Granger-typetestsareofienusedto test forcausalitybe-
tweentwo variables,in this casethe growthrate of realcD per capita
i) and the grosssaving rate (s).Testsweredone for regressionresult
fir the eight East Asian economiesand the United States. Vectorauto-
regressionswereperformedon equations5.1 and 5.2:

(5.1) g=X g',- + s,+,i

k-I k-

(5.2) s= S*-k + gr-

k-I k-I

242
 STRATEGIES FORItAI I '.

Then F-restswere performed to see whether the laggedsavingsrates are

jointly significant (equation 5.1) and whether the laggedgrowth rates
arejointly significant(equation 5.2). Up to fiveyears of lagged rates are
included. The results show a causal relationshipfrom ODP growth rates
to savingsrates for most East Asian economies (seetable AS.1).
The Granger definition exploits time-seriesrelationships to identify
causality.By this definition, g causess if the prediction of the current
value of s (s) isstrengthenedby usingpast valuesofgt. Thus, gcauses s if
2
1 IA,) < cr (rsA%-gt) where A, = {A*ik<t- 1I denotes the infor-
c2a(s
mation set that includes Ak for all past k up to and including t-1, and a2
(st 1A) is the varianceof s,about the best predictorby A-,.
To deflinecausalityin a time-seriesmodel involvingtwo variables,we
make two simplifjing assumptions. First, that the set A, indudes s, and
g only, and not a third variable.Second, that tansformations 7Tand T
exist, such that ST Tss,and G,= Tg, are a pair of linear, covariance-
smtionarydine series,and that S,and G6preservethe causalityrelation-
ship of s,andg,. Ordinary or seasonaldifferencingisan exampleof such
:ransfonnations.
Causalitycan be defined by using severalrepresentationsof the time-
seriesmodel for the independent and dependent variables.The autore-

TableA5.1 PublicExpenditure
perStudenton Primary
andSecondaryEducation
(US. dolirs)
-~ ______ -. 1975:18 28
,cmy -i , d P -' Seconday Primacy - ;C

-. Hog'Y . 1923.-3 180.5 8.15581 810.2*

'cora,Rep.of-' 90 147 512 32 ' 3898 352.5 ,;j 09699
1 'aaysi=5.
572.5
PT 314.7
~~~~~~~193'3. -8340O
-67i
i49
t0xnanana .176 1881 2 9
t~ e ~ ~~2. I 778.0.' 1r2 a 0j
~~ 'Brazil 76.9 1~~~~~~~~~79.1-'
15.4-22.
4 175 30.9 ~~~~
9 ~~~~~~~~27
62.3 3.
- I _-'3 7 .
India53 713 ~~~~~~~~5038 `6 M†
Mexicot. ~23& 566 1210 312 ' 10. 875' )3

. . ,..,
fl wwevtAunat.wl
c tea, 'UNFUd--. ta. ' '. .

243
 ~~~~~~~
~
*~~~~~~~
TableA5.2 Testsof Bivariate
*2
GrangerCausality
S;
between
theReal
S Per
Capit GOPGrowth RateandtheGmoss Savings
Rate

Si ~~~~Sig
0 ~~~~~~~~~~~~~~~- , 4 .m 1S .Si SBal ;-i -,*-tX

; ~~~~~~~~~~~~ t
jas S -8 MSN NS~~~~S Z g,;t-fI..
. ..
3;.KS. . . g

f ~~~~~~~~~~~~~~~- S, , - ' Kqq,-S--

.. ~~~~~~~ MS'
N'S , -~. -Si

Ng~~~~~~~a
Mdapua1- I Si . S .i-41rs

NS

MS Mg~~~~~~S

.~~~~~~~S . , KS MS ~~~~Si
16 ,~
.~~~~~~~ -, 11 ' 115_ ; ;.

.2 PM-

3 S
la
4 In.i
-S

Si K~~~~~~~~~~Si,

Ti wuyo-u
"S

md &"
~~~ ~
- ' ~ ~: ~ ~ ~ ~ ~2 ' MS

.- ,,- - . . - - :7 - j .

244~~~~~~~~~~
; ~~~~ ~ ~ ~ -~ -;~ ~ ~ ~ luhLgO8 . '' KS gX
 STRATEGIES FO EGERAYIF

gressiverepresentationwas selectedfbr this study. In this representation

: and by Granger'sdefinition,ifg doesnot causes,, laggedvaluesofgcan-
nor add to the predictiveaccuracyof sp,givenpast valuesof s. Using the
autoregressrverepresentations,s. can be regressedon its own past values
and on past valuesofgto restwhetherg, causess, byresting whether the
coefficientsof the laggedvaluesofgare significandydifferentfiom zero
(Granger 1969). The resultsindicatethat past valuesofgare a verygood
- predictor of s, but that the reverseis not true (seetable A5.2).

Appendix 5.2:TechnicalNoteontheRelationship
between interestRatesandGrowth

P; = REVIOUS STUDIES, NOTABLY BY FRY (1984, 1988) AiND GELB

(1989), have tested the relationshipbetweenreal 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 volatileinflationrates (Murdock and Srigiirz1993).
Introducing inflaton to the equation,this correctsfor the rce of inflation
in analyzingwhetherfinancialrepressionhas a negativeeffectron growth:
(5.4) constant fPi,R+ J32 1NF:.+
Third, investment opportuities also aff*ectreal interest races.A &ll in
the real instimutionalrate of interest does not mean a changein financial
repressionif the equilibrium rate falls equivalentlybecauseof lackof in-
vestmenitopportunities. This study, thercfore,introduces better proxies
for the degree of financial repressionin the single economy regressions:
the differencebetween the curb marker rates and the banking system's
interest rates. (These are availablefMrKorea and Taiwan, China, but:not
fMrthe other'high-performingAsian economiies.)
Fourth, OLSestimiatescouldtbe biasedwhen the regressorsare corre-
lated with the error term, as in equation 5.3.If an economyexperiences
a sudden increasein productivity,for example,the demnandfor creditwill
rise,and inflation islikelyto tu'rn out to be lower than expected-Both ef-
fects will result in a higher-than-expectedreal interest rate. Thus, two-
stage least square estimation (2sLS)is used in this study. (In the event,
the results for OLSand 2SLSsuggest thar the condlusionsin this study
would not have been substantiallychangedhad oLs estimatesbeen used.)
Using pooled clint-seriesdata for twenty economnies,this studlyesti-
mated equation 5.3 applying a two-stage least squarcs estimation to
both thLecomplete sampleand to the subsampleofeconomies with pos-
itive average rates of interest. AvenageGNP growth rates were used as
proxies for the growth rate, an1dreal lending races (from the Inter-
national Financial Statistics'database) as proxies for real interest rates.
T'he 25L estimation induded the previousyear'sinterest rate as the in-
strumental variablefor the real interest rate~,a year variableto account
for any sectoral trend, and two measuresof an index of relativeincome
because standard growth models predict that poorer ecoonomieshave
higher equiibriumngrowth rates. All regressionsindlude dummy vari-
ablesfor each economy.
Table A5.3 reports the resultsfor the imact of the real interest rare.
and the inflationrate on the GNPgrowth rare. FoIhCsml of twenty
economniesthe real interest ratechasastatisticallysignificantand positive
impact on growth (equation 5.3).But when inflation is induded (eiqua
donA),the cefficient of the real ineetrare is no longer statistically
significant,while the negativecoefficienton the rate of inflation is. For

246
 STRATEGIES FOR,

TableAS.3 ParameterEstmatesfor Interest RatesandGrowth

, '; 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
<tLLUL
5
LAJIA -000;;
1 0.277
t Subsamplcof&os
o-i;
-. _ AX C5

f tri-ot - -. >tT-.>g
, . .;

Equaiontl ' -0-: '

t i Equaon&
2 Equarioni)..
;'w-
- -. 0.391~~~0.
0.059 - -~-::~;0363u-
t Eqarin3 -- - . A389 x0 0 438a

M6
aeEa
WattTrto-sSt C~1;peratlewd.,' Icr............,;><,.
-~ Ieastst regesons were s
estimated on hmp b of w-
i suhsampl.ofcaniameoLswason
otlie the nm :l' -: -
on - . . ..- * - .

-
swarcw&iBank
dima
2; .;:-;frA. >: < - ~ r..
;:. f!:.y*:...............
,.. ...... , .. - :-

the subsample of economies with positive real interest rates, the coeffi-
cdent of the realinterest rate is not statistially significant cxcept when
inflation is induded. Then the coefficient of the real interest rate is sig.-
nficat and ngative, suggestingthat lower real interest rates may have
had a positive impact on growth. The coefficient on inflation remains
negtive and significanL
'Table A5.3 also reports estimares of a time-series analysis of the rela-
tionship of interest rates, inflaton, and growth in Korea. A third equa-
tion includesthe,realcurb rate as a proxy for the equilibriuminterest
race.For equation5.3, the coefficientofthe real interest rate is positive,
'large, and statisticallyhighly significant. When the realcurb rate is in-
dudced, the real interest rate is no longer significant, but the real curb
rate is. When the infladoti variableis added, the eal curb rate rerains
highly significant,but the real interest rate and the inflation rare do not.
'When the same tests are perfonned on time-series infboration for Tai-
wan, China (not reported in tale A5.3), the esdmatedcoefficientshave
the samesignand similarmanitudes as those for Korea,but the standard
errorsaremuch largerand none of ti coeffiacntsisstatisticallysignificant.

747
 fI. F--ACLE=

. - Table
AIA IcomeDo Rzobution

GNPper shareof. arof: tope

:: - -..
cap i-
ib omm 2 :top20. 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
9AEV-
-Japa f i1~ ; 1969`
4.r 7.9 41.0 5.2
f- *0-XX1979..
' 8.7 ;375 431-
Korea,Rep.of .- 71 1976.- '5.7 453 7.95
4Maaysia"' .0'-. 1973:- 3.5 6.1; --16.03
= R - . .0;198T'0.:4G6 .51.2
' R E11.1 -
tq-
. -gapo 6.5: 5.1. 991982683
6-.-.
Taiwan,China. 1976 95-
9- 35.0 -3.68:
Thailand-
h - 4.4 1975-76 5.6 :49.8- 8.89
* - : , - . .I - 00---

*. f.OzhtAid
PhilPEippines000:
1.i3 g70-71
19e - 03.7 .;53.9 14.6
*-0 f51985
--l985b-
;.- --0
5 -;;
.48.0' : 8.7

Brad!.... ; 3..
3 1972 - 2.0 6 3i3
1983' 2.47 626 26.0.
-. taiRica 1.4
- '1;:-971.
.:. 3;3'.......... 54.8 -
.. ...
;. } :t. ~~~~~~~~~-.;
1986" ~..3 33 -. ......;:--1 -16.61-
8c8--S
54 .51-4- -
-5.6.5

-Mco 2.8 1977 Ž2.9 57.7-- 19.9

p~. 0.2 * 1I72 ;L- -0.
i-9 2.11
-I{ - - ^: --:- - -- . 00 1-98546w 44< - :5139 179
. V&elzuea - '_!. -10 1970
0 -.- T 3.0
.- -:. ;.:-54.0
. .- ....................
{; :
7197' ;76
0~~~~~~~~~~~0
iBoiswana.
---
^ =' -- :i -4- 19856 ---
`2'.: 50 00-
Cbce'Ivi-re 05 198- ;8-T50 527 10 5 5
Gh-.na ',-. -14-. 19s8j9' j rt

dia 1-9 -9 6.7'

'19645 48 7.2
1975-76 7 .: . 49:4 7

: ^ . , 0 -.. . - , : X , - . - n s04.. : 314....

-. ±

a.-~~ ~~ ~ ~ ~ ~~~t'.nerprcznmoc -.. -s.-u

r'|

1 ''-
i '' ' ; . i"' '. q ' '-a~
' r. '-,: bou'e.i"k.c'p.ndz'ut.
'~.D
.o :

--- 248---:- .-
 STRATEGIES FOR RA:PtDiW

TableA5S5Natuire,Causes,and Resolufionof BankingSystemCrises

Naturnfnere. Nature ofbiot
.:'Enfisrj ::s Causes or re :
na
T-TnnHon ..-- Ninedeposi:-aingc- (a) rg exposurc toreal Thegovement rem reg-
-1982-83 . . . eestateiled
lending f-raud ulatoryandaudiuing
system and
- and mismanagement. liquidated troubled DTCs.
* andw
wak prudential
t* regation.

Hihinternational
inhteiiga
g;or&., m or:. 1rg*itrst t- res
1{o4 Kong, Eight banks, includinga maJor large-exposureto real The gormentookoverthe
:1;8-86 :: -international ankwere in. estatelendingand spillover lagebankcandintrduced ne
'fnancial difficultie
. effers fiom 1983crii in cludingtopexecu-
since che banks o-;ccL- -sewn&de
t-v from tceJargeieo rn-
[7stwi .. .' ;..--- ' DTCs
- aSsubsidiaries. - nrial
me. ban.kThose no: taken ovcr
- .... . -- - , : - --;- :-. tr
by thgovernrmenrreaivedcrdt
b.
from tohercom rcialnk
Japan. ACenini Bank repor has esrti (a) Excive Encourage rnergerioFweaker
; -l991
11; ...- mated the sizeof problun loans --- realestatelending,corn-1 banksbwith 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 loansofd
bki systrem:ase,.S. .n and a epdeclinein tO
eofaount.to'be wrt- ' ral esae prices ban-s-
ten off2sabout 1 1-5 percent -nc
Ina-ecjuare - No-~ lon to be
banking systema. transu t
were:.s;r-
ale to P- dalinstidon,and
--- dthecosto, -
r exposuredititugh de&eoftbsie
* .. 4, - - . - ' . Ins to nonbanic among commercia bank.;

- ;. , -a-lureof one deposit-

-k;ng (a) Preoinant causesaThe goverinrrescud -24
7 '1985-88 ~~~cooperatie
ini8cuec ru n speculationi, isletcooperatives and con-~
4. { 88.rins: n`32(if35o) r. : d sldar
In addituo;n4! (o38) bnsstOCks iic npM'.Te eta
and4 (of
= 47) i'Iedfe
naiicecompa ,ic . Bank
-. ,.gee' (b) Deterioration rtemsia-
nieterealsoiinacial dis- l
i OXi iiit of , s b

, r '~~~I ' -' '. :i' ..

-d, .
1' .i44~~~.;'C~~ijrr. M i..,u.. .n7.' ,.- i ae .luo opt (&4kivŽirnnre
-. '-
~ .1C ~
...butOO~~ ~ s,.
i" n.r-'$20 ~
4..h,4.to J8'4r y0'"C-:
.. - trbek~
.
-

SI"' ':ononur - .
OUr1Uust:-&rnpan
_S -
n N Z L'ooeratLVeS ,L' i -J, 'S44. ,,.. ,-4
X . . 4- ,- ' . ~~~ . ~ ~~..
.-... Z9 -

249
 LE

TableASS (coawiaWed)
Economy NantresJiitttn.iI C Aauee fbaib ut - -

ThailandL- Thegovernment'st ofb- (a) Fraudand_speculation - The governmentliqtuidaed 24

-1983-87 out of 50 financecompanies on real,estareand financecompanis and-meriged
w..as
estiinatedtUS$190 mi- exchagerae c--ansac another 9, andthe Centra-:
lion or 0.48percent ofGNP. - ons. Bank cookoverthe other 17 :
- 0f:-; ---ii -
0 .X - t 3 u;9 ; - - ; - 0 45
.an -t soldthmto ncwmvsots
-*-- Five
0 ,commercialbanks () EHighconcenaon of ands tinvestors
accounting for 24 percent of- unscuredinsider - i n o banks)
commercia bank assct were in loans. - It bouiht someshares ofrou-
- financial-difficulties
in -. (c Hit te bledbank ' --

000t9B6-;87
*a 1 - '' .< < ' ->>
g 50t,0 -0"~ " - - inestrae - ksocaeda-Ioarid
to_provid emnergec loans to
rubd -TheRind
would be finlancedby contribui-:
~- . i 0 - 0- - 0- -; -:0~;
-.-Mat-uri-y .6 of tions frdmn
Trobe commercial-banks.-
-m-00-orriom
t
Gl' ;;' :- 'fChide,'
X 7';EntLrc mortgage bankingsys- Matitymismatchesof .Troubled mowwcre'
0-X-1976 - te0 - was
- - insolvent.n - assetsand liabilitiesas liqudat
-' ~~~~~~~~~~~~~interestrates
woscandi
- ~~~~~~~~~~~~inadequate
Capital
* ~~~~~~~~~~~~~requirements-,.
-Chile,- Severn!financialinsdiutons (or -a) Severe.r i . n1981. thegoverrncmntbqu
-*: ' :.':198b-83-
. :-; t: :iabout 60pcrcentofrhefinan-, ,'1981driasi case dat ed:8insticunon&sIn:1983:it
- .0ni'sln
.-.. ;.:al
1porolio) m nterest rts,adX rook ovceS(liquida*ng-3nd
w insolvencii-n1988;ceniral declinei termsof- rccapitzng5).n The:Ceir
-5banlholdingsofbadconcr- tradedueto:ihecol-. Bankalsoreqs--
pe G &iiloutrct coppero ec e rate.-,
cost est dat
-was
US$6bil- schemetomee-±rcgnkabik
lion, or 26.2 percent of GDP.r ( p ' tesd interest rate sub
ltoadsupervisowo er4utpwp o ea
and r2Mpaniisider
0 7 . ~~~~~~~~~C
. .0 !, _,' and speculativelendling raie nc ,.-,, 0, '''t
-. ,';30>-.Maa

'Ghana Thlicnetwot ofthe banking -.Hug foreinecag

-1988 system,-lose
waxnegativsiat andahppor-, -- ,
*c&rof restructuriii as 7 'rionirononiperfrrnng-
* US-$300 millionornealy 6 2ans
o

1hr ng (a) Collapse1in hepic;of -Government- ' 5

~
heavylosses.~ ~ ~oi.
ipuin ~A enig '~tew.
wie oOamoans. resieur
n,' capta
Encourage drpe:and o

L.... ...... I. ; . -. . . . .-

_'50
 STRATEGIES FOR RAPI

Economy ~~Nature
offlwwal Naftre ofLalikt

-1-;
Eco ov .dt-ress ; cause: or rescue
.~*-PhiUippinas,---
-- :. Amongsmaller finanacalinsti-- (a).Inrcrascsin oil prices . Insolventisttutions weas
i- 5;- :1981-87 wituuons,comprising7.6 percent - nd internationalinter- liquidatedand depositorspaid
o- deposit money bank assets,w esi rats, coupledwith off by ie deposit insurance
-.
182iilai In a
In,a~~~~~~~~ddizio deterioration'ian cc
he corporatio.
erment intcrvened
' in thc two tenmsoftradc.
largestp bli f - -banks.
Th 2 publicbankswere.
private banks, accountingF .Ilb) High concentration oforestu-c-tr: in 1986, and their.
more than two thirdsi of the. lending to relate par-, adporios wcr put und raa
asset.s .-.-
ofde|itnolney ban. tienes
and inadequate : epate agencyunder over
-- ;;:-.By-1986,
centl b:.lcassis--- piud al rigon mMnaeent. Five private
t;meat

e.- :;: -- -- :-.tance to financialnstirudons and--:-

suevso, . bak reaie under.........
direct..
j .--i4,
r- c tral-banlk
!ene'X supervision;:
~~6-.~
am9-:outto 19.1 bmiionpesos
; . E .!. y:.Sor 3 percent-ofGDP.i
i;Turkey.- :- Afinancialcrsis
A- eruptedwith: ~Inadequate prudential Thegovernmentrecapitalized- '-

- .'.198245-8
5 -' '-:',. the collapsc ofsal brokers,= superision.---- . banuks an'd isistedonthe-ir
the--a
5. .-.- 8- -adfeba
weres dara :- ~ - - op--
opucrurt,
erriaires

re+ xc=ured. .- * on its prudentialregulaons.

'

.>,Uwe Strs . 'Of appodirel 4,500 .- (a)Predominat &wewa Simia approachtora oth
.21982-87. bank, 175 liqudated.
n in
meried1,575
'andfi eriarporen~ spcltv
-:ofanks'aserand rea esat S&Lgoemn-e
FedraDeoi Isrc-).
A b1~! upcr themers'nd.rnew'ana2ei-
;N ; E >- -'-troubl in 1987.tec en 1982 - loans:;; - nce :Corporationl riquidated~
- and1988,e ou 424banks banks(a d od
-'1. ~~~~~-:.M'
fiwxra.ouc'acdequateat a pru x!srctn&-

sp Berien 1 Sincel9i epe ~ ( ~ ~ ~ ~~"rceve

D988, about
X i57s9-89;
;
1,5O6
..f :-'.:(ofES,147)
al crisis
S&Jwre
uptea wid-
2 lœ ~~
98Ireiducedrhev aluc -;ur;;IeC&porons.ised
-nT :i -
.-

4> i. ~~ ~~ . 1989
~w over60
_-~0-- (one1~fift oif~ mcesineetrob
'to poo7r-q,ueprtc~
alr ecer- er liaoso- -- A. r. noe
-!;4
-. i S&Ls) were solvent
-- '.Toed paid t depostors,. ...e0 S L a acquiredw

Q f, . . . _
~~~~~
(~~~~~~iarioi1987.sB
,~ i:
KiCwewn198
_lrg ,b=., 6
(19)lindz (96;Seg(18A9)Sii&ae
CO'.rbn
.
.

- '
'D?. '. ''
.. 4..~.,
of .VV~'p -
~~~~
aennas supervision;
inic vi let costs'OLcLa.ofA.Lupec,rtsoFGN
t 1~~~~~~~~~~~~~~~~~~~~
'I otspJ !ei2 ge ioanC.Q-
additionlg' up

~il~jii15o
, ~S' ' _ . 4A
' 'f*_ _ _ ' : __ . _ ,_ .A., -:'

- ibi-i,sl9 ~US-$13 bilion r22 (b Dergultion ofS ,s, -t rb -7 >'thertJ5Ss

of

,, .4--
0 . , ' ' S&._ .- to le_h
' -
 r ~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ &rp _.e .' ,. Tnw .;g
TableA5.6Comparison
ofBankingSystem
ConcetainRto ka ses
* g f ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ mi

-tuSI L E -.

Hmlshnet ~~~193 75 77.7 .2

Indaish 28-54 59
lapi '
690.10
2 40 37 1.-
KcrsRep.J 41 50 '51
_d qil n 15 94 2: 56 61 DI OP
: . 0,0.1959
.00 60 6 7. 0i-.,

: >; ~~~~~~~~~~~~~hil.,nd
. 359f8 9 56 66 72 t,
Pidais ~~~~19116-U - 4 O.26
hipln 3986-U 40 CLOW)

0 j t<N-x-~~~~~~~~~~Dhlds)
CA"ac '.19156-111-
- tips-Y;an - . 74w (UGh U.f:t
20..

M e 90
0.50
52

6CM.

X=atbr i ; g ! .9!i
l O s-u . 5. 0040

2989 ' ~50 6 41 LU

.Camdz, 1989 , 5~~~4 69, 431a
Pane
*. . - 1359U .43 ~ - as ~ i
- 993 .r

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

~~4enua1abJc
.. , :2
. ~~~ ~~.C. ~ ~ ~ ~ ~ ~ ~ ~ a a . tlkHcdnMiI main&Inedbna~iuIl.,w
Sahi . ,.M.Ifbbrh mAh6hIfl6&lkj d k=ni htl

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

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

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

; ~~~~ ~ ~ ~ ~ ~.91r ~ ~ ~ ~ Iy . ** .7
 -. - - - FOR0 RA,_s
STRAT~~~~~~~~~~~~~~~~~~EGIES
~~~~~~~~~~*
F ST1_s;RAT
:!i

Banks,SelectedHPAErs
-.TableA5.7 InterestRateSpreadsof Commercial
tS P > la.; s;] s -
r _ps ;.^r
s;=
@
S
>r.z<-i
=-~~~~~~~~~~~~~~~~~~~~~~~~~~~~..
........

i ;J1980
-3, 21. ..
*',i-i#t..
;-;
';,'1;';
'.........
l'-'
'....................

r19909' I--1 S

Silg9O8|e to -- 0S@03e2
-7.0
0 1' r _L _Na
,s; . r ."5

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

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

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

:/_~~~~~~~~~2 e -1aitltwo 4 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,&-

.- ,-~~~~~~~~~~ tiluo,,.ii.rn=ii

-::~~geed .te Z - 4 . .- -.
-r7- :-
 . -IR_AL
E5 EE

TableA5.8 Outimes
of MajorPolicy
Implem on inancing
Institutions

-Aals a -~~~~~~~
(Af it e - . -. -. . . < . ;. - . . - .0,--;0Dl
U..e

I - .. nnii7.5pantna..d

Bath( -PtN009 ' - . (

0rWSM7mIlm, M3 manC4 %nr .
- (GO~~~~~mt) - -- nd.Acwith bath.
bafowa Pksci.fwpa&M.y-
M ;f0Q00Q.-- ; .4n ~tirl;.u;
LnU1S UmmlUll 23
De
- - , f- terstssX
n1ef MO Fw* %

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

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

Ewe uneur Das. '- - - hAS?-

bik we-' 5, p=^ ' nrlmam4ea - Iaaswmi-aawha-uda
h-

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

(US$3LEADe38)
tss9,~lnsgTe 05,
C lat
l ,x ,
, ,
- 351 ,Wtau
H ,
.- - iicrs) inpesmu, -
mmdmmBan
Mas.s-~

r6'w-S> 'f, 'MWkm~

; ,. ra rtu7 'tn -m'.
'' -to .

t DcbuaP.ace- (US$54 i-:)k haEraw$ 0;.. . . ..................

; 15*41=

D,adepeet. BakefMaLquh 821sata .huu - ! hp t beta-1usdm

b*r pre:,: - M s'.-
r"T"'ad ~ W" ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~mylwdncg

iu

* (383fpm
Saa,Dwqscfa f6ai,na Ouse aena,iacpehfehtd cMa&aashqm na

.' "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~w

(U M
F. . .. .
; -. 1
-'ps, :w aucef,u
1-
DZI~hhy
.!

_~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i.anuctl&nl ...

254
-. - .. - ;
, . . .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

a .e
 ST RATE GI ES FORRWD a

notAum 56Na'.".-

bnal.pnwant nkdkdw -17MpcLIpa:Dad.xemowwt mnouvkt

Phllpphim(inw) . UISI WUDNW Caa,g--st aayTciamr

salmawn4 younfixwmldtus

Hai, 3 Ealcmwmaaima
25dluftolankga.safaw,

* -*. Ml~~~~~~~~~~~~~naCr) -. . ~~~~~~~~~~~~~~~~~~~~~~agmriaa

. ftfiwm.Fwre io
fhlsi5p).Amam1ka(vn$ . - . . taandeatnamI
. &Wtuat.,. L U.Jm..
'ra'. Cia...
lwx.~ ...

.. I
flakdC.ma,wdcttaaa. -. - Psunatatd - .. ~~~~~Damu-lag
rumtht ptwmsnt. M.&Sns-toharm spn

Indian] Rime.
Capstans: Jlbadand
~~~~~-
21.1 raiKonbAixtaIwna
~(USSLIS ils.)
Ova.sbmmPt-gbaii.ouwbg - -isllh- s.leantam line. -
hAlfpnaee fa pnaar'AaSlak - fiwrdxsA aee nmt&
nxd.dbamd mAman

Back fhcAgcdswc
(MAO ~45JW
KWon 54 pm-- Drpsu : PiLbSzaa
(W1i5tPSmaxlonwund- aghhinAxn m

0(W)

G Na'orSbn .easmatn
Haaxid2781 m snaenntw ix ai t 13tmx aotnaan d 9 =it dndaif.d,136 (n

bnvarFan&nsn lelmnaina Rantm (199011. nc.Ceainullash ripen Exap nImpaMllk a Japa bmtlwakandblnkLTuixwaLnz
h Eaptut-ap Bank f,V bases
mixs
dhcDdJchi KangacDank.lad. Malayt, teat] banksa.iaa.Dn
Japa Cntlnmmod
Kot.spaeqtaa(nip am TDank 52ldr.
fllnAadRps
w mi lee
33Twhfpa.Ca
N.re hl~C"lbakum
ai u.uo
T
Iqi
a
Bthtapnbahndoknet

Novtes
1. See chapter 6 below for a mor'ederaileddiscussion 3. Of courseGNP per capitais much higher in EastAsia
of the relationship between educational expansion, the than in Sub-SabaranAfrica.Becauseof the bias that re-
* growth of labor demand, and the educationalstructure of sults from using exchangerates rather than measuresof
wages. purchasingpower parity, we do nor attempt to weigh the
effect on costsof higher GNP per capita against the effect
* 2. From aross-economny school expenditureequations of lower teachers wagesrelativeto GNP~per capita.
esinared by Schultz (1988), the elasticityof enrollments
with respect to the relativeprice of teachers is -0.80 and 4. the relationshipbetweenthe abundance ofthumnan
--0.70 at -theprimary and secondarylevels. capital and reachierpay is not firm. In Korea,~where
 CRACL
E

human capital is most abundant, the earningsof primary 1O;Given the smaller size of the basiceducation age
teachersare roughly 5 times per capitaGNP.However,be- cohort in East Asia than in Latin Amcrica (see bdow in
cause pupil-rcacherratios are higher in Korea, operating chetext), this difFerenceunderestimatesthe gap between
costsasa perccntageofper capitaGNPare, at 16.5percent, the nworegionsin the strength of dte public sectors com-
not much higher than in Indonesia and Malaysia,sug- mitment to basiceducation.
gestingan important role for policy both in determining
teachers' vage and in controlling costs. Sec Tan and 11. Howevcr,since the mid-1980s, public savingsin
Mingat (1992), pp. 28-34. Argentina, Brazil, Chile, Colombia, Costa Rica (since
1983), and Mcxico (in 1987) have incrcasedin response
5. Implicit in the calculationsis the assumption that to fiscaladjusrments,leadingnationalsavingsto rise (an
expenditureper pupil would remain the same even if the intcrnal World Bank report).
growth of the school-agepopulation was faster.
12. In Japan, by managingexcesscompetition, indus-
6. There may also be problemson the supplyside-in- trial policymay havealso assistedir the stabilityof finan-
adequatecapital marketsor entrepreneuri deficiences- cialinstiutions. Japan's recessioncartels,fir example,may
that limit the private provision of educational services, have avoided large and systematic bankruptcies,which
Psacharopoulos,Jimenez, and Tan (1986) present esti- surely helpedthe solvencyof financialinstitutions.
mnatesof private and social retrns from education.See
Psacharopoulos(1993)for morerecenrestimares.Behrmn 13. The danger of casing intcrest ratesand entry con-
(1993) citesthe manycritcisms of hese estates, indud- trois without simultancously strengthening prudential
ingtheir failurcto acoountforabiityand forqualiy ofCed- regulationsand capital adequacy requirementshas been
ucation. Neither he nor others, however, question the evidenrin Indonesia, where low capital requirementsfor
existenceof some gap betweenprivateand socialreturns. entry and poor enforcementof prudential regulationsled
to mounting concern about insolvenrbanks.
7. In fact, it is difficultto distinguishcmpiricallyb-i
tween the e1*lrEsof nor being able to borrowand the ef- 14. In severaleconomies, the risk was aggravatedby
fectsof poor inbrmnationon the demand for schooling- the banks' tendency to speculatein securities.TableA5.4
shows that the major source of financial disruptions in
8. Governmentexpenditureon education,exprssed as Hong Kong,Malaysia,Thailand, and, recently,Japan was
a percentageof GrNP,was used as an explanatoryvariable real estateand securitiesspeculation.
in a cross-economy regreSSionin which expected years of
schoolingof the school-agecohort (essentiallyan aggre- 15. Stiglirz(1993a) notes that in economiesin which
gateof enrollment rates) is the independentvariable.For governments do nor take an active pro-competitive
a sample of fifteenAsian and Latin Americaneconomies, stance,there are a limited number of firmsin the banking
the expenditure variablewas insignificant.See Tan and sector.Furthermore, evenwhen there are a large number
Mingat (1992). of banls, e&ctive competirion (if viewedfrom the per-
spectiveof the receiverof the financialservice)is likelyto
9. In Koreaand Taiwan, China, the share of public ex- be limited,This is becauseinformationpertinent to a sin-
penduimre on educationallocatedto higher education has gle borrower is specialized,and each of tie major cate-
increasedduring the lastdecadeor so for two reasons.On gories of loans represents a diffcrent product markerk
the one hand; universal and near universal enrollmrient There are differencesin the kind of infbrmation that is
rateshavebeen achievedat the primaryand siondary lev- needed across product categoriesand few economies of
- , respeccively.On the other hand, the increasehas been scope across the categories. Patterns of specialization
consistenrwith the shift in the strucrure of production among financialinstiturionsconfirm this view.
- and exports to more technologicallysophisticated and
skill-intensiveproducts, and die consequent increasein 16. Data on quality are extremelylimited,but avail-
the demand for engineersand other skilledworkers. able measurements suggest that infiastructure in the

-256

HPAEs may be of better quality than in other developing
economies.Table 5.7 shows the percentageof roads that
-arepaved in a largeselectionof low- and middle-income
economies.
17. There is a large literature not only documenting
the relativeunimportance of equirymarketsas a sourceof
financebut also explainingwhy this isso, both in termsof
the evidenceconcerningthe negativeeffectof share issucs
on share price and in theoretical terms. See Stiglicz
(1993a).
18. Long-term credit banks, but not city banks,issued
coupon debenturcsof arnaturityof fiveyears,and this has
been centrmlto the segmentation of short- and long-term
finance (Packer1992).
19. Korea's bond market has been expanding more
than other East Asian economies,except for Japan. But
most of these bonds arc government guaranteed. Firms
have begun to issue bonds, because governmenc has
begun to control the debt-equiy ratios of largecorpora-
tions (an internal World Bank report). Many of the cor-
porate bond issues may be simply a form of bank
financing that carriesa government guarantec.
20. This is known as the price discoveryfinction of
stock markets.
21. More substantial long-term lending to industry
has been carriedout by other developmentbanks (for ex-
ample, Eximbank, the Small BusinessFinance Corpora-
non, and private long-term credit banks).
22. While the experince offEastAsianeconomicswith
developmentbankshas been generallypositive,there also
have been failed development banks. In Japan (Recon-
struction FinanceBank in 1952) and Thailand(Industrial
Bank in 1959), insolventdevelopmentbankswere closed.
Development banks lending to srnall and medium-size
-.
STRATEGIES FOR R_
enterprisesin Indonesiaand Malaysiahave reported diffi-
culies becauseof a growingvolume of arrears.
23. For example,in the 1950s and 1960s, the ratio of
write-offsand bad loans to rotal assets of JDB was lower
than private commercialbanks ODaIJERI 1993). The De-
velopment Bank of Singapore has evolvcd inro Singa-
pore's largest private commercial bank, largely on the
basis of its excellent financial performance (Gordon
1983). Yaron(1992)suggeststhat traditional financialin-
dicators may not be the best indicator of peformance of
developmentbanks,since they havebeen establishedwith
developmentobjectives.Hc mcasuresinstead the degree
of depcndence of developmcnt banks on subsidiesfrom
the government. With some exceptions (for example,
Malaysia'sBank Bumiputera,Indonesia's Bapindo), the
HPAEs'developmentbanks are self-reliant.
24. There is remarkable umiformity in the pattern of
investmenroutcomesexceedinginvestmenteffortsamong
the fiveHPAEs in the sample.Malaysiais perhaps the least
representativcin the sensethat during the 1970s(and up
to 1982) the nominal investmentshare cxceededthe real
investmentshare. All five economies, however, had real
investmenr shares that exceeded the nominal share be-
tween 1982and 1985. Korea continued to show a declin-
ing relative price of investment goods throughout the
1980s, while Thailand had a major increase in relative
capital goodspricesin 1988-89.
25. De Long and Summers (1991) obtain a similarre-
sult for the relativeprice of equipment (to the GDP defla-
tor). They predict the realequipment price for the sample
of sixty-twoeconomiesas a function of relative1980 GDP
per worker. Of the six HPAEs in-their sample (Singapore
and Taiwan, China, are excluded), Indonesia and
Malaysia lie dose to but above their predicted values;
Hong Kong. Japan, Korea, ancdThailand are all signifi-
cantly belowtheir predictedvalues.Real equipmentprices
are unusuallylow in theseeconomies.
. .~~~~~~~~~~~~~~~~~~5
0; a E R 6

Using Resources Efficiendy:

Relyiingon Marketsand Exports
E
T IS APPARENT THAT THE HIGH-PERFORMING ASIAN
economies were unusually successhflin strategies to achieve
rapid accumulation. But high levelsof physicaland human cap-
ital formation are not a guarantee of economic success,as the
modem economic history of the former Soviet Union and East-
em Europe sadly attests. Resources, once accumulated, need to
be allocated to high-yielding actirities. This chapter first considers the
strategies, explicit and implicir, that the HPAEs used to achieve efficient
allocation inte areas: the labor market, the capital market; and inter-
national trade. It then turns to industrial policy-deliberate, government-
sponsored interventions to alter industrial structurewhich was an
important feature of public policy in some HpAES,and asks whether
these efforts have resulted in superior economic performance. Finally,
the chapter considers the role ofmanufacured exports in promoting the
demand for skilled labor and the acquisition and mastery of best-
practice technology.

ExplainingEastAsia'sEfficientResourceUse

T
0 t|HE HIGH-PERFORMING ASIANECONOMES HAVEEFFICIENTLY
absorbed unprecedented rates of growth of physical and human
capitai for thirty years. As we showed in chapter 1, all of the
HPAEshad total factor productivity growdt that exceeded that of more
than 70 percent of developing economies between 1960 and 1989, and
some-japan, the Republic of Korea, Hong Kong, Malaysia. and Tai-
-wan, China-had among the highest rates of TFPgrowth in the world

259
 - 0 - _EE [+ ~C
LE

This means that fictors, once accumulated in the HPAES, contributed

much more to output growth than would be predicted on the basis of
the average performance of a broad sample of economies, both develop-
ing and industrial.
-How have the East Asian economies succeeded in using resources ef-
ficiently? Good fundamentalsagain tell much of the story. Price distor-
dions, while present, were limited, and markets were allowed to work
HPAE governments have relied primarily on market mechanisms to guide
allocative decisions in both labor and capital markets. Labor markets in
the HPAEs have been remarkably flee of the interventions that restrict
labor mobility and create small, high-wage elites in other developing re-
gions. Whle capital markets have been frequently controlled, resric-
tions on deposit and lending raceshave generally distorted markets less
than in most other low- and middle-income economies (as we demon-
strated in chapter 5). Government efforts to direct credit have generally
been undertaken within the framework of moderate interest rare subsi-
dies, and financal institutions have been given the final decision on the
creditworthiness of investment projects. This has meant that loans sup-
plied through directed credit have been monitored and generally have
been repaid. The relative prices of capital and labor have broadly re-
flccred their relative scarcities.
The HPAEs have also used the international economy more effectvely
than many other low- and middle-income economies. Each of the
HPAEs,except Hong Kong, went through an early phase of protection of
impo rt-subsituting industries. But these policies were later modified by
reductions in import controls and tariffs, combined with strong incen-
tives to export (see chapter 3). This opened up much of the modem
economy to international compeiition and introduced world prices as
performance standards, not only for exports but also for tie intermedi-
atc and capital goods used in export production. As a result, domestic
prices for traded goods in the HPAEsare more doseJy aligned with inter-
national prices than in many other developing economies.
Apart from these findamnentals, however, most IIPAEsseletvely in-
tervened in financial markets and used a combination of policies to pro-
mote thc development of specific industries. Most HPAEShalve directed
credit to priority sectors and sometimes to specific firms. Where these
programs have been effectivein -allocating resources to high-yielding
investrnents-for example, automatic credit for exporters-irt was be-
cause allocation rules were performance-based, and borrowers were ac-

:26
 USING REESOURCiS

tivelymonitored; in short, effectivecontestswere created. These condi-

-tions have been more-common in the three northern HPAEs that have
strong governmental institutions.
* 0 Indonesia, Japan, Korea, Malaysia, Singapore,and Taiwan, China,
have tried to use targeted industrial.policy to promote the development
of specificknowledge-and capital-intensiveindustries.We test the rela-
tionship between industrial policy and productivity-basedcatching up
and conclude that industrial policy has generally nor been successfulin
changing industrial structure or raising productivity, What, then, has
contributed to the iPAEs' apparendy superior performance in adopting
and mastering international best-practicetechnologies?We argue that
the combination of compedtive discipline and well-firictioning factor
markets with a pro-exportorientation-the export-push strategy-
employed by all eight of these econormieswas responsiblefor their supe-
rior productivity performance.
Manufacturet exports acceleratedthe acquisition and mastery of in-
ternational best-practice technologiesin highly imperfsc international
technology markets. Hih and rising levels of htiman capital in the
HPAES,especialy the superior educationd atminment and cognitive skl
of the industrial labor force, helped to ensure that these new processes
and equipment were used and adapted effectively.Thus export orienta-
tion and high human capitl formed a virtuous circle;exports raisedthe
returns from education,and educationiraisedthe retuns from exporting.

the Market LaborMarketsin EastAsia

UsUsng

5 xT rILE HUMAN CAPITAL ACCUMULATION MAY BE A NECES-

condition for sustained rapid growth of output and
wages,it certainly is not a suffiaent condition. The Philip-
pines, the republicsof the former SovietUnion, and Sri Lanka arepromi-
; nent examples of economiesthat have had high school-enrollmentrates
but low rates of growth of per capita income and of wages. Utilizing
humnancapital in activitiesthat yield high returns on the prior invest-
ment in education and training is as unportantto growdhas the accu-
mulation of human capitaL
We have fiequently looked to coordination turilures in markets to
understand why public policy might be used,to help markets function

-- - - ; 0 ~~~~~~~~~~~
-
,'',' e!'kCLE

better.In the case of labor markets,however,coordinationfailuresare

limited, and information, although not perfectly symmetric between
workersand employers,is betterthan in capitalmarkets.'Extenalities,
aswe showedin the previouschapter,primarilyalTce investmentdeci-
sionsin humancapital,not interactionsin the labormarker.Thus, there
is a presumptiontiat, ifleftto themselves,labormarketswouldperform
theircoordinating(allocative)rolequitewell.
Twoconditionsmust be fulfilledfor a growingsupply of educated
laborto be utilizedin high-retumactivities.First, rheremust be rapid
growthoflabordemandrelativeto supplyand, in particular,of demand
forskilledlabor.Second,thelabormarketmustperformwell.It must be
efficient, flexible4and responsiveto changingconditionsso asto ensure
thatworkersare employed in jobsin whichtheirskillsaremost produc-
tively utilized.If thesetwoconditionsarenot filfilled,thereis a riskthat
rapideducatonalexpansionmay resultin labor marketproblemsdtat
erodethe benefitsof human capitmlaccumulationand havea seriously
adverseimpact on the growth of output and wages.
In ourassessmentof whyhuman capitalin FastAsiaappearsto have
beenefficiendyutilizedand whychangesin labormarketoutcomeshave
beenso fuvorable,
we focuson twofacors:the dynamicsoflaborsupply
and demandand labormarketperformance.We alsoexaminethe inter-
actionsbetweenthesetwo factors.Then we considerwhetherthereare
any common denominators in thaepolicies of East Asian governments
that have contributed to the efficientutilizationof human capital.

LaborSupply
andDemand
in the HPAEs
Shiftsin theaggregatesupplyand demandfor laborin EastAsiacom-
paredwith otherdevelopingregionshaveledto morerapidlyrisingreal
wages(seefigure6.1).East Asia'smorerapid rate of increaseof wages,
followingan initialperiodduring whichwagesdid nor increase,.is the
reslt ofslowergrowthof supplyand morerapidgrowthof demandfor
labor.2 Fora givenrate of growthoflabordemand,the slowerthe rateof
increaseof labor supply,the bigger the increasein wageslikelyto result
from competitivepressuresin the labor market. One measureof the
dampeninginfluenceoflaborforcegrowthon the rareof increaseofreal
wageswasobtainedfor a crosssectionof twenty-twoeconomiesin the
1970s;a 1 percentincreasein laborforcegrowthresultedin a 1 percent
declinein realwagegrowth(Gregoryand Lal 1977).
z6z
 USING RESO£5 RC

Figure6.1 Increase
hi RealEarnings

Japan f ; ;- [
Rep.of Komea __ l
Tawan,nChina 7= -Z I0 .

HongKong|
Singapore .
Indonesia } - - -

ThailandI - }

andCertheanF-
LatinAmneuca mi 1 - 1SrO.:-
d980-90
SubSahIranAbica -'A- It
- U
t I
-t 0 1 2 3 4 5 6 7 8 9 10 11
IncreaseInrealeamings
(p'e.nt)

Nwc ldIdocfor Taiw2n,Chin= 1979 = 100. Other Asiancconomicsarr Bangladcsh,India.

Pakistan,and the Philppines.
SouwreWorld BankdacviTaiwan, China, National Stistcs.

Labor Force Growth.During their period of industrializarion in the nine-

teenth century, the populations of the currendy high-income economies
grew, on average, ar an annual race of only 0.8 percent.3 Today, the pop-
ulation of Sub-Saharan Africa is growing at roughly four times ca rate;
the populations of Latin America and South Asia are growing at romughly
three times that race. Only in East Asia have populaton growth rares, and
the share of the population less than fourteen ycars old, declined to levels
approaciing those that prevailed in the high-income cconomies.4
We have already seen how the early denographic transition in East
Asia markedly reduced the rate of growth of the school-age population
to levels well below those in other developing regions, thereby easing the
financial burden of maintaining education enrollment rates. The early
demographic transition also reduced, with a lag, the rate of growth of
new entrants into the East Asian labor force. The annual rate of labor
force growth during the 1980s was 2.6 percent in Sub-Saharan Africa
and Latin America and 2.2 percent in Soudt Asia. In East Asia the rate
was 1.8 percent (see table 6.1). The labor force growth rare in Singapore
is only 0.8 percent, about the same as that experienced by the high-
income economies in the nineteenth century. In Sub-Saharan Africa the
rate of labor force growth continues to increase. In Est Asia, despite in-
creases in the particpation rates of women, labor force growt rates have
been declining much faster than in Latin America or South Asia.

263
 RACL
E

Table6.1 LaborForceGrowth
Rates
Ecenomy,vfonegjn. J--.90-85 1985-2000
H; , s;- 2.5 1.8
Indonesia: . -~-; ; - -2.4 - 2.2
t'rKc,Rep.of. 2.7 1.-9
Malaysia - 2.9 2:6
..Sinore 1.9 0.8
Thailand -; 2.5- .- 0-
1.7
thAsria 2.2 2.2
L4 in Am6icai bhea-n 2.8 2.6
ST--S
m2ranAfr.
. : 2.3 2.6:.
S areWorldBank(1987b).

Growth ofLabor Demand. Not only has the growth of labor supply been
slowerin East Asia than in other developingregions, but the growth of
labor demand has been faster. For the period 1960-90, the rates of
growth of wage employment in manu&cruring construction, and ser-
vices have tended to be substantiallyhigher in East Asia than in Sub-
Saharan Africa, Laiin America, or South Asia. This is not because
gowth of employment has been unusuallyslow esewher Widt the ex-
ception of the 1980s, during which output and employmentstgnared
because of the macroeconomicadjustments made necessaryby debt-
serviceproblems, the pace of industrial sector employment growth in
Sub-SaharanAfrica, Latin Arnerica,and South Asiawas more than dou-
ble the pace in the currentlyhigh-income economies during the perod
1880-1900 (see table A6.4, at the end of the chapter). Rather, the in-
terregional differcnce reflects the unusually rapid growth of employ-
ment in East Asia.
Even as it has been growing rapidly, labor demand in East Asia has
become increasingly skill-intensive.As a share of wage employment,
white collar and technical employment increased steadily during the.
1970s and the 1980s;in Korea,for example,it rosefrom 29 percent in
1980 co36 percent in 1990, and in Taiwan, China, from 32 to 40 per-
cent during the same period (seetable 6.2). The pace of change in the
occupational structure of employment is lower in other developing re-
gions. The change in the occupational composition of labor demand in
the HPAEs reflected increases in the abundance of educated labor, and
consequent declines in its relative price, as well as changes in compara-

264
 USING RESO-UE

Table6.2 Professional
andTechnical,
Administrative
andManagerial,
Clerical,andSalesOccupations
as a Percentage
of TotalEmployment
e i H Kora,JRp of Shngapore Tariano Chbina
1980 29.0: - 42 31.8
.1981; - 292 30.8 -42.8 - 32.7
0 S-; 198200:
f - . 30.6: 0 : .-- 0 .
*;: -;-44.2 33.1
1983 32.1 - 45.6 33.1
-1984 32 - 46.5 33.7
1 l985*. :-- 34.3ta;; - - - |--.46.5 34.2
1986. 34.3.3 379 46.2- - 34.3;
- 1987 2 34.1 - 47.0 35.1 :
-1 98 8:. . . : 34.6 - .-X 46.1 37.2
,1l989~0 0 - 35A4~> V- -? - -t-46.8 38.5
;1990_
36.2 . 60.0 X 39.8
*£/ r'-_Notvanilablc;
-
employmentforHoneKongindudesLinmployed
aToml whohavepreviously
bald
KIidudes serviceworke -X
:Somiwfe::ields
(1993).

tive advantage. East Asian exporters shifted into more technologically

sophisticated, skill-intensive goods as rapidly rising wages of unskilled
labor eroded international competitiveness in labor-intensive manufac-
tured goods.
What accounts for the extraordinary pace at which labor demand,
a-nd in particular the demand for skilld labor, has expanded in East
Asia? Since the demand for labor is derived from the demand for output,
the explanation must be sought, in the first instance, in those factors
that contributed to East Asia'srapid output growth: high rates of savings
and investment and high rates of return on investment B at it is also im-
portant to consider the extent to which the path ofgrowth of output has
been more labor demanding in East Asia than in other developing re-
gions. As we shall see, the dynamism of agriculture and export-push
stratcgies contributed to rapid increases in the demand for labor and to
rising skill-intensity of labor demand.

Letting Markets
WoricPublicPolicyin the LaborMardt

The term "labor market?' generally refers to the complex of interac-

tions between employees and employers. The labor market thus bears
the responsibility for melding the changing needs of the economys pro-

265
 iR_ACLCE

ductiveapparatuswith the changingskillsand preferencesof laborforce

participants(Berryand Sabot 1981).In the courseof economicdevel-
opment,workers-most of whom are followingeducational,training,
and vocationalpaths differentfrom precedinggenerations-tend to
shift toward more dynamicsectors,from self-employmenr to wage-
employment,and towardhigher-skilled 5
occupations.An efficient,flex-
ible, and responsivelabor market allowsrapid matchingof skills to
needs, improvingresourceallocationacrossfirms and industriesand
therebycontributingto growth. Whilein most developingeconomies,
labormarketshavetendedto unctionratherwell,wherethey havenot,
the costshaveat timesbeen high.7
In East Asia, more than elsewhere,governmentsresistedthe tempta-
tion to intervene in the labor market to counter outcomesunpalatable
in the shortrun or to particulargroups(forexample,wagesbelowwhere
workersbelievedtheyoughtto be, educatedunemployment,or the em-
ployment of educated workersin lower-leveloccupations).A relatively
high levelof efficiencyin the allocationof laborwasachievedby allow-
ing wagesand employmentto be determinedlargelyby the interaction
of thosesupplyingand those demandinglaborservices,ratherthan by
governmentlegislation,publicsectorleadership,or unionpressure

DidEastAsianLaborMarkets
HowWePl Perform?
In EastAsia,wageswerepulled up by increasesin the demand for
labor,whereaselsewherethere was a greatertendencyfor wagesto be
pushed up artificially.Earningsgrowthwas determinedmore by the
growthof the economyas a whole hn bygrowthin anyparticularsec-
tor (Fields1992).In economieswith highlysegmentedlabormarkets,
the oppositeis the case.
EastAsia's
FlexibleLaborMarkets.Bynot allowingthe price of laborin
somesectorsto risewellabovewhat workerscouldearnelsewherein the
economy,most HiPAEs avoidedthe creationof a high-wagelaborelite.
For example,thereis evidencefor Taiwan,China,of a remarkablyinte-
grated labor market; there is no significantcorrelationbetweenthe
growthrateof earningsand the growthofoutputwithinsectors.Manu-
&acuringwagesfor unskilledlaborare only20 percenthigherthan agri-
Lulturalwages.By contrast, in Colombiaand Jamaica,where labor
markersare highlysegmented,workerswith the sameievel-ofskillearn
nonagriculturalwagesthat exceedagricultmral wagesby 150 percent

266
 USING -R

(Fields 1992). Still larger intersecroralwage gaps have becn observedin

East Sub-Saharan Africa (Ranis 1992). Another noable feature of the
-structure of wages in East Asia, as illustrated by Korea and Taiwan,
China, is the modest size of the gap betweenskilledand unskilledwages
in the nonagriculturil sector (see figure 6.2; also see Ranis 1992). The
small gap did not result from minimum wagelegislationpushing up un-
skilledwages.Rather the growthof demandfor unskilledlabor, in com-
bination with marked increasesin the abundance of educated workers,
compressedthe occupationalstructure of wages.
One benefit of the compressedwage structure in the HPAEswas that
it reduced the incentive for educated workers to conduct a lcngthy
search for a relativelyhigh-wagejob, and thereby remain unemployed,
rather than fill a job slot at a loweroccupational level.Conversely,a seg-
mented labor market may provide an incentive to workers in low-in-
come employment, and to the unemployed,to lobbythe governmentto
provide more high-wagejobs than justified by the derived demand for
labor.8 Thus, the share of the public sector in the increasein total wage
employment in recent decadesis a crude indicator of the magnitude of
public sector surplus labor, and hence labor market segmentation.9 As
table 6.3 shows, in all the Sub-SaharanAfrcan, Latin American, and
South Asian economies for which data were available,employment in
the public sector grew more rapidly than wage employment in the pri-
vatesector. The median share of the public sector in the increasein total
employment was between71 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 increasein total wage employment in those Et
Asian economiesfor which data were availablewas lessthan half the me-
di-anshare for other developingeconomies.The limited growth ofpub-
lic employment in these East Asian economies is evidence that public
surplus employment has been kept in check and stronger evidencethat
the scope Sorlobbyingfor make-workjobs has been limited. The need
to maintain international compeitiveness helped to limit growth of
V public employment becauseimporters who purchased intermediate in-
puts from public enterprises lobbied for low prices. Moreover,govern-
mentsstrivingfor macroeconomicstabilitystricdylimited the sizeof the
public enterprise deficitsthat they would finance.
- In-Malaysiathe share of the public sectorin the increasein tota wage
employment is as high as in other developingregions. The increasein

267
 6.2 Industrial
1Figum WageGaps
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~t
' . ' ' Wage a Taiwan,Ch._a
of 1955 NTS)
Meanmonthlyearings (thouK;mds

.. ~~~~~~~~~~~~~~ .... L T-
:~2~ - :

to-I A -l rea walp

O. IF= F4

; -- ~~~~~~0.5- - -

. I| ' ]I I . U UnakflledlIndustulal
realwags4
- °°- 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 odustrialreal 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

Table6.3 Cmwthof WageEmnployme.t

inthePublic
andPrivate
Sectors,
Selected
DeveloPing
Economies

__ __
__ __ __ _ __ __ __ p b tsector In
&onorny/~on
Peno Public rkwre Toal inrase p/otl
wageemplyment
LnAmericaazndC"bantbe:
* :
- ~~~19733 1.4 00. 0.3 92
CostaRica ' 1~~973-43. :6 , 28353
Panama *':, :1963-82 . ~7.3 ,.,18 2. 45
Peru ~~~~1970-44 * 6.61- 061.4
Trinidad` :1970-84 4 12 ,195

Venezuela ;1967-82 ~~~~~51 3Ž4 '. 37 27

Middle ase and Notnfipc
Egjp , . . 1~~966-76f 23-. . 103
n*nj.I tA -c

Gh~~~~ina ~~196048-734 -5.93-4.

,Keny 196381 64.2. . 67
Tanzania ~~~196246. 6.1' --3.8' 1.619
LZambia -1966-080 7.2 -r0943

K India .~~~~~~1960-80 42213.2 7

SnI nka ~~~197183" 8.0 09393

? Taiwan,China . ' -196.5-85 23- 665501

~~Thailand .: ~~~~ 1963-83 ~ 6. .. 7 33,
I' .~~~~~~~~~~~~~~~~~~1

Unwciglired.mC"ean"
bl'emlyetwsntarsostolbyninud by2.la0or mar2- 6

publennat0. ahr,pbicaresons
employmentEwswns tolbyn
nue bynlaborcma-
ployment opporrtunitiesfor the historicallydisadvantagedethnic Malay
Majority.The Bumriputeraprogram can be seen as an aggressiveaflirma-
tiye action policythat resulted in considerablepublic surplus labor (Free-
man 1992). In contrast to many gvrmnts in similar circumstances,
however,when the macroeconozmccostsof public employmentbegan 'to
mount
theMalasian government, manifesting the adaptability that
sewedgovernmnents so well -throughoutthe region,changeddirection.-
Why WereEas AsianLabor MarktsMoreNiedble? Rapidgrowthwasake
-fcror in making wagesmore flexiblein the HPAEs. Workersaregenerally
more willingto accept flexibilityof wagesaround a rapidly rising trend
bc&ausedownward adjuLstmentimplies a:slowerrate oinrease rte
than an absolu:tedeclinein reai earnuings,(Shiahand Mathur -1992).This

2.69
 is manifested in wage-setting practi ces that tie compenation to enterprise
performnance,another example of how the extraordinary hibor marker dy-
nainics in Eas Asia contributed to superior labor marker perforrnance.
But some HPAEgoverniA.."ts also intervened in labor markets, pri-
manily for political reasons, to suppress the activities of industry- or
economywide unions and to ensure that wage bargains were set at the
enterprise level. We have suggested chat the rapid growth of wages con-
tributed to the explanation of why, in contrast to odchr regions, wages
were not maintained above market-dclaring levels in E-astAsia. An alter-
native, or perhaps complementary explanation is that governments, in
collaboration with employer repressed laibor markets.
Wage-seuingpractices One manifestation of the link between rapid
growth and wage flexibility is the apparent willigness of workers, espe-
diaily in the northeastern economies of Asia, to accept wage structures
that include a substantial element of-performance-based bonuses. This'
practice links compensation to enterprise, rather than wholly individua
perfbrm2nce. It is comnmon for current cash compensation to have two
comnponents-a base wage and a bonus (Shah and Mathur 1992). To
the extent th at "special cash payments" are more responsive to changing
economic circumstances than are base wages, thiis arrangement could
help account for the greater flexibility of wages in East Asia than else-
where. During the period 1958-83 in Japan, econometric evide-nce,
shows that rthe ratio of bonus payments to base wage-svaries positively
and significantly with such indicators of aggregate economic conditions
as cnrporate profits. By contrast, there was no significant relationship
between base pay and corporate profits (Freeman and Weitzman 1987).
Nevertheless, the impact of changes in bonuses on coral compensa-
tion appears to be marginal. Whlile bonus payments as a proportion of
total current compensation (15 percent or more in Japan and Korea) or
of corporate pois(40 percent or more) rend to be large, the elasticity
of bonus payments with respect to profits is small. In Korea, most bonus
payments have become contractually linked to the base wage, and in
Japan the implicit link of bonuses to base wages appears to be strong
(Shah and Mathur 1992).
How represseduor H-PAEla bor ma rkeex?Repression in the labor mar-
ket can rake two forms: dlirec restraint on wage levels or growth (wage
represion) or restraint on organized labor (labor repression) with indi-
rect restaint on, wages. Wage repression was dlearly nor a necessary
ingedintin -the East Asian recipe for economic sucesThgorn

7-7 --
 USING RESO

merntsof Hong Kong, Japan, Korea,and Taiwan, China,-did not repress

wages,ant yet in these economicsoutput, exports,and ultimatelywages
all increased at rates well above the international norm. These
economies were adept at adjusting to changing comparativeadvantage
as labor abundance gaveway to labor scarciy. At the other extreme, the
govcrnmnentof Singaporerepressedwages, at times quite severely,and
the consequent costs of labor shortagesappear to ha'vebeen high, lead-
iing the government to abandon. the policy (Fields 1992; Freeman
1992). Several HPAE governments have reprcssed labor organizations.
Korea, Singaporc, and Taiwan, China, suppressed indepcndent
unions.11 Malaysia has independent unions but highly restrictivc labor
laws. Hong Kong and Japan do not intervene in labor relations. H-ow-
ever, ir Japan most unions are company based.
In Korea and Taiwan, China, real wages increased at roughly the
same pace as real GDP. Nevertheless labor relations in Korea have been
deteriorating in the 1990s, so that labor relations problems serious
enough to threaten economic growth have arisen, partly in reaction to
repression of unions (Freemanl 1992). In Singapore during 1972-79,
rreal wages increased at less than one-quarrer the rate of increase of real
GDP.This divergence was the consequence of the governmenes aggres-
sive, and quite open, intervention on the side of wage repression in the
market for 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 a
tighter grip on wage increases.... If we donor tquicldy and willingly
change to low gear on the wages front, we shall fisrher discourage in-
vestrnent and aggravate the unemployment problem" (reprinted in
Fields 1992).
* . EIfwagesfall well below market-dcearing levels and employers are pre-
vented by the government from using wargecompetition to bid up the
price of labor, chronic shortages of labor are likely to be the result. This
is what occurred in Singapore, with the result that the rate of growth of
output, as well as the rate of improvement of living standards, declined..
In 1979 the government recognized the costliness of this corsraint and
* changed course: a policy of "wag,ecorrection" was adopted with the ac-
plicit aim of alleviating labor shortages and shifting the economy away
from labor-intensive and toward capital- and skill-intensive industries.
Although the high wage policy was abandoned in 1981, in the 1980s
..the grovth rate of wages accelerated:and matched the growth rate of
SGDP(Fields 1992).-

271
 *CL

Hong Kong repressesneither wagesnor. unions. With respecr to the

labor market, the govemment of Hong Kong is among the least inter-
ventionist in the world. Despite labor laws that are favorableto union
formation and action, union membership in private sector fims is low;
unions engage in relativelylitde collectivebargaining and are generally
considered to be "weak." Perhaps becausewages were rising rapidly in
Hong Kong, workers did not fcel the need for collectiveaction. Hong
Kong demonstrates that East Asian economies need not aggressivelyre-
press unions to have a weak union movement, suggesting,for example,
that in Korcathere may have been frw benefits to offset the politicaland
economic costs imposed by labor repression(Freeman 1992).
Japan'sexpenencewith labor unions is another counterexampleto the
beliefthat East Asiansuccessrequiredgoverrnmentrepressionof labor.The
Japaneseexamplemay be the more tellingof the two becausein Japan, in
contrastto Hong Kong,union membershipishigh and unions are, at least
potentialy, quite strong.Yetunions did not extracteconomic benefitsfor
their membersat the expenseof econornicgrowth(Freeman1992).

FowDidEfficientLaborMarketsCoutnIbute
to Growth?

The benefitsof maintainingwagesat or marginallybelow,rather han

well above,marker-cleawing levelswere numerous and substantial.13 FiSt,
becauseinflatedwageswere avoided,the profits of firns and, most likely,
aggregatesavigs rates were higher.As a consequence,retained earing
accountecdfiora higher proportion of investinent finance than orherwise
would have been the case,redung reliance ornunderdevelopedcapital
markets. The probable net effects,therefore,were higher levelsof invest-
ment, greatercompeitiveness in internationalmarkets,and fasterrates of
growth of output, employment, and ultdmatelyeanings. 14

Superior labor market performancein East Asia thus contributed to

sustainingthe extrordinary dynamicsof labor demand we documented
above.Moreover, by not allowingthe price of labor to become inflated
relativeto the price of capital,wagerestraint encouragedthe use of more
labor-intensivetechnology.Ar a means of avoiding excessivecapital in-
tensity,this was particularlyimportant, given that government interven-
tions tended to lower the price of capitaL
Efficient use of skilled labor also contributed to the high levelsof
publicsector savingsin most of the HPAEs.If the govenment capitulates
to demandsfor make-workjobs the subsidiesrequired to finance the te-

.272
 USING RESOURO

sulitinggap between the wage and the marginal product of labor are
likelyto divert scarcesavingsfiom productiveinvestments,loweringthe
rate of capitalaccumulationand stcadilyerodingthe grwth potential of
the economy.15 For a time, Malaysiawas an East Asian illustration of
this point, but there are manrymore examplesoutside the regionwhere
governmentshave been slowerto rake correctiveaction.
If the government provides make-workjobs, squandering the pro-
: ducdivepotential of human capitalmay be another consequence.In the
: worst case,which unfortunately is quite common, workers in whom a
substantialeducationalinvestment has been made are paid the prevail-
ing (relatively)high public secrorwage to do jobs in which their mar-
ginal productivity is zero or even negauve.1 6 In this way creation of
makei-workjobs impedesthe processof hunan capitaldeepening that is
cen-tralto education'scontibution to economicgrowth. Had the labor
market been allowed to functon without intervention, these same
workcrswould have entered lower-payingoccupations,and the consc-
quent mcreasein the educationallevelin those occupationswould have
contributed to an increasein labor producdviry.
Becausewageswere responsiveto changesin the demand for labor,ad-
Justent to externalmaaoeconornicshocks,suchas those inducedby the
oil crisesof the 1970s, was ofien quickerand lesspainful in the HPAEs
(Mazwrndar,forthcoming). Ifthepriceoflabordoesnotadjustquick-lyand
smoodily to macroeconomic shocks or to secular changes in the macro-
economic environment, then either the quantity of labor must adjust or
the ratesof output growth or inflation are liely to. In relaton to fluctua-
uionsin aggregatedemand, employment stabilitytends to be much greater
in East Asia than in other rgions, so that East Asian econonies genelly
adjust to downturns without laying offworkers (Shah and Marhur 1992).

Assisting
theMarket FnancialMarkets
andAllocaioDn

-T-L MARKETSSPERFORM
THEIR ALLOCATIVEFUNCTION IN
-ways:they aggregate savings, allocate funds to competing
a_J uneat,and monitor performanc Without financial
intermediation, firms would have to rely on retained earnings to fi-
nance investment, so marginal retums to investment would diverge

2173
 . '., .. J:RACLE

Markedly,especiallyin the short run. Thus the policies descrilbedin

chapter 5 that contributed to the deepening of financial sectors also
contributed to better capital allocation. In addition to private capital
markets governmentsin East Asia,like governmecnts in many developing
economies, have attempted to allocate credit directly to high-priority
activities. Elsewhere,directed-credit programs have been catastrophic
failures.In some HPA-Es-parricu.larlyJapan, Korea, and Taiwan,
China-these programs have caused less damage to capital allocation
and may have been beneficial.
Our examination of capital market performance begins by docu-
menting the highirare of financial deepening in the H-PAEs. We then
describethe nature and magnitude of directed-creditprograms and as-
sess their performance. We show that, where these programs appear to
have had the most beneficial results, they supplemented fundamen-
tally sound capita market policie-swith improved selection and mon-
itoring. Interest rate subsidieswere small, loans were repaid, and strict
performance criteria were used to ration fumds.These preconditions
for efThctivedirected credit were present in only a few of the HPAEs.
Where they were absent, directed-credtitprograms largelyfailed.

RecordofGrowing
EastAsia'ls FinancialDepth
The growth of financial assets in East Asia has been ve-ryrapid-
faisrerthan the growth in aggregate savings-leading t usata
grwth in the financial depth of East Asian economies. Figure 6.3
comparesEast Asia'sratio of M2 (rime plus demand deposits) to GDP,
which is often used to measure the depthi of the financial sector, with
the M2IGDPratio for a group of low- and miiddle-incomeeconomies
and the OECDeconom-ies.The relativelygreater financial depthiin East
Asia far surpassesmost developingeconomies and comparesweli even
with the OECDeconomnies.. A broader meaure of financial depth, the
ratio of financial liabilities(M3) to GDP?,
shows a similar trend.
Korea is unusual in that its ratio of M2 to GDP?is low compared
with many developing economies at similar levels of per capita m-
come, and theseratioshaveinacrasedmoreslowlythan Korea'sincome
growth would predicc. The largest incremental change was evident
after the inte-restrate adjustments in 1965. However,M2 fails to cap-
ture the presence of a vast nonbank financial sector. The rapid growth
of income in Koreasince the 1970s has been accompaniediby large in-

:74
 USING RESOI5

Figure
6.3 Financial
Dept in IPAEs,OtherDeveloping
Economies,
andSelected
OECDEconomies

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

LaiMaidde
Etropeand East
HAmaEs I 190:7

Sub-Saharan
Afrks

OECD
economiesI_________________

0 20 40 60 so 100
M2/GDP (percent)

Not latin America includes Argendina Barbados. Bolivia, Brazil. Ch_c. Columbia, Costa
Trinidad and Tobago,Uruguay.and
Rict, Ecuador,El Salvador,Cuatemala,jamaica. Panamna,
Venezuela. Earopc and Middle East indudes Algeria, Egypt, Greece. Morocco, Portugal. Syria.
Tunisia, and Yugoslavia.Sub-SaharanAfricaincludesCameroon,Congo. C6te dt'lvoire,dhe
Gambia, Gharn, Kenya. Madagascar, Madriania, Nigeria. Senegal, Sudan, Tan.a.ia, Togo,
Uganda. and Zirbabwe. OECDeconomies include Atriza, Belgium, Canada, Denmar
Finland,lGermany,Ireland,Netherlands,Norway,Spain,Swecden and Switzerland.
NoweM2IGDP= [currency + demand deposits,+time deposits,+ savingsdeposits(except
largecertificatesof dcposit)jas a percentageofgross domesdtcproduct.
SawrrrIntemnatonalMonetaryFund data.

.Mg-~~~~~~ CP

creases in die assets of' nonbank financial instirutions (King and

Levine 1993).
Fnancial assets in industrial and developing economies, incling
East Asian economies,are held largelyin depositswith banks and other
financial institutions and to a very limitedtextent in government or pi
vate securities (World Bank 1989c). The public policies described in
chapter 5 that have-emphasized the development of banks and other
savings institutions may partly explain this composition, but there also
are market explanations.Equity and bondtmarkets developslowly,espe.-
daiy in emerging markets. In East Asia, the combinaton of rpid
growthof financial assets in the banking system and rising savings dut-
igthe past two decades has meant' that bankcshave played an maceas-
inly i an role in capital allocation.

2.75
 AL
C EACt

FinancialMarkets,Coordinaion Failures,and the

Roleof Government

The lackof perfect information that characterizesall markets is most

seriousin capita markets, where acquiringinformation to ficilitate bet-
ter selectionand monitoring of projectsis a key function. Imnperfecrin-
formation has severalconsequencesfor allocativeefflciency.As we noted
in chapter 5, bond and equity markets are slow to develop. This places
increasedemphasison the role of banks in selecringborrowersand mon-
itoring their performance. But credit markets are not like auction mar-
kets or markets for ordinary commodities. Because banks cannot
discern perfecdy the probability of default, they cannot simply aUlocate
credit to the highest bidders-those who promiseto pay the highest in-
terest rates. Rdther, they must chooseamong agents willing to pay the
going rate those that seem most likely to repay the loan. This allocativc
process is, in effect, a form of rationing. Moreover, it gives banks some
control over entrepreneurs, who must maintain their good credit stand-
mgin order to renew loans.
Financal institutions evolvein responseto these market limitations.
For example,bank-dient arrangementsrespond to incomplete in$onma-
don through relationship banking. Loan contacts are characterizedby
more than lust interest rates. The namre and amount of collateral be-
come key. Yet despite efficiency-enhancingmarker responses, capital
markerSremain imperfct. Market power by banks may raise inrermedi-
ation margins, resulting in monopoly rents for the banking sector
Banks also faceincentives to misallocateresources;underfunded banks
rend to finance excessivelyrisky projes. Social returns on investment
may differ markedly from the private remrns to investors, and social
perceptionsof the riskinessof investmentsmay differsubstantiallyfrom
pnvate assessments.
For the abovereasons,most governmentsare reluctant to rely entirely
on private markets to allocarecapital. Twotypes of government involve-
ment are thereforecommon: first, efforts to reducemisallocation of re-
sourcesby financial institutions; and second, eflorrsto control banking
spreadsro limit monopoly rents. While these measuresare by no means
unique to East Asia, the HPAEs have been unusually active in exercising
them. In addition, most HPAs have gone a step further and attempted..
to direct credit to specific sectors, industries, and even firms. We discuss
this more complex and controversial topic in a separate section below.

-76
 U S I N G R E SOUlCM

Portfolio requirements,
ReducingBadImuestmentChoices. eitherto avoid
excessivelending in some areasor to increaselending in others, direcdy
influence the allocationof credit by banks. In chapter 5 we argued that
East Asian govemmcnts bave emphasized the safety of banks through
both regulationsthat monitor and enhance the solvencyof banlksand
measures that avoid financial disruptions. These have positive effects on
resource allocation. Solvent banks have fewer incentives than under-
funded ones to take excessive risks in lending. Solvrentbanks can also
lend more.
Portfolio requirements to enhance prudent lending-such as limits
on lending to real estate projects- -also may guide credit away from
some activities to others with higher-perceived social returns. In general,
restrictions on lending have favored producers. For example, East Asian
governments have often restricted lending to real estate and other spec-
ulative activities. These governments are concerned nor only that such
lending increases the riskiness of banks' portfolios but also that it diverts
lending away from manufacturing. Thailand, for example, does not
specifj how much of their portfolio banks should lend to industry, but
it provides lending guidelines that favor lending to industry and agri-
culture and limit lending to other activities, induding real ,-re. and
securities.
Most BAEs were assisied in their efforts to influence lending deci-
sions because they directly held a high proportion of savings deposits.
Table 6.4 shows that in Indonesia, Korea, until recendy, and Taiwan,
China, very large shares of total deposits of the financial system were
held in publicly owned commercial banks. In addition dte governments
ofjapan, Malaysia, Singapore, and Taiwan, China, have discretion over
a large (and growing) prop ,rtion of total deposits iri postal savings
and/or provident and insurance fi^nds. Some governments have also
been able to complement those resources with budgetary allocations or
borrowing fiom abroad.
Japan and Singapore have the most well-defined programs fior allo-
cating s;-vingsheld by public finanicial institutions. The financial inst-
tutions created under the Fiscal Investment and Loan Program (FuLs)of
Japan allocate finds originating from the governments postal savings,
national welfare deposits, and national pension deposits. The FILaab-
sorbed 20-40 percent of household savings during the rapid growth pe-
riod. The Development Fund of Singapore, whose resources originate
primarily from the provident funds, lends direcdy to the governrmcnes

277
 Table6.4 PropotinofTotalDeposits
withGovernment
Financial
Instutionsin EastAsia
(percentage oftotaldeposits)
c3 - - - - .:mmcw Wi.h
IM or Wi.h iteI P . .3-
pus ed singspwvidearinsturance rid aemn
m
Economy/year
S
insa*adem
~~- ^
fiends -
s~Pecielixzdi
banks' Thud
XjiOQ~~~~~~et-r
.,- E -. ; -, :0; 7M 1 1
1971-76 N' NA
>ing-sy - N-. ;- : . NA8.79.9a 7
,91960
81-8 5 NA NA. 7 1. 7 1. 7
-6 5F-A ' 61.3 461.3

- -10-. . 108, 4. NA. 151

1971-75' 1981-85 ~~~~~~ Noa ~

NA61 ~~2.280.7 855

_1971-75-
197 44308½ N' 4
14t3} '.0 NA. 22-.3;-':r.
_
1976'-80 36 24.7 02. .~~~~~~~~~~8
..

. -. - .19 0
. ~~ ~ ~ 22,N.37 .

--- I 0-~~~~I

198145:~ ~ ~ ~ efiieg dcigtevlrl9o

57rsueealoainb netel
~
1906-89 ~ ~ 15 itede 30cora4 prdn beavo an3nhne 1ovnq9 e

.~~~~~~~~~2
. .2 .

stradbtutoryboardsinand goernment
: -0 - coprateos. - --~~~~~~~~~~I
; ;. - - - =.: -ndesprad Kreauete
Mhalaysa and Thailand hyagven guided theisaig.Tate-ownedcomerieal
banks ofinanen priority projmbiiets. cilsngsb rmougi

duvencthoeef fcmptto banks and

mn inceas thecen
averageins.-
n~ ~~ ~ ~~~~sra bewe borwn.n edn aes ihsrasrdc h

t78
2- -:
 USING. RESOI1

- With pvernsment or Vit? pensoi 4 J pbi

** ~~~~~~postalsavings
~provident awntrance commercdialad

*SingPapre
1971-75 2.1- 21.6 -23.7

1976409. 30.0 39.4

19814~5 -1.0- 34.6 456'
1986-90 ~~~~~12.7 34.9- 47.6

-. 1971-75- 10.8 A67978.7

~~~~~-13.81
1976-80 NA. 63.6 77.4
t18-5 - 92N.53.5 727
-:1986-90 19.4 NA 52.4 71.8-
.Thailand
-1971-75 .12.7 N.1.1 -13.8
1976-80 10. NA. . 13.7
* -1981-85 8.2 NA1. - -

*1986:190 7.6 NA- 2.2 . . .

is fbr 1978-80.
a.Th fguxre P6rpotd
c Te figure-for-Taiwan, China, are bed on the depositsof all COMmercialbanksand'accneunl vrsiae eas ny1
f he16 commnrdalbans&aregov&nmncnr
- owned.Nvtlds,xgocnnrwndbanks acon forthe mjrshare of!commer-,
cilbankdeposits.
SomorrlndoesizIndonesian
Financial-Sratiirk(Vadlouis
issues).Japan:jDDIFJI (1993); ooi Staijsnal(aiu er)
Kore= fronmndcWSraisdtYew-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& QnarurlyBulletin(vrosise)

Except for Hong Kong and Singapore,all HPAESat Varioustimes have

simultaneously reguated deposit and lending rates and consequently
the spreads of financia institutions. As a result, rents derived from im-
perfrct competition have been modest. Our evidence led us to condude
-in chapter 5 that, in comparison with other developing economies,
bankcing system spreads were relat-ivelylow in East Asia. and have de-
clined as entry restrictionsand interest races have been liberalized (see
table 5.7). Given the rapidi growth of dhe financial sysems in the HrAs
and their limited intermediation margins, we conclude that regulators
-have done a good job of balancing the need to limit competiton in.
order to enhance bank solvency with the need to limit imperfect ornm-

279
 ~~~~~~~A
LE-

petition renrs. HPAEregulatorsdid not alwaysget it right, however.It is

notable that, in Indonesiaand Thailand, regulatedintermediation mar-
ginswere substantiallyhigher than those afterliberalizarionof entiy and
interest rates.

UsingDirectedCredit

Like many governments,those in East Asiahave gone beyond the in-

direcr guidancedescribedabove to target credit direcdy at priority acriv-
ities.The categories to which they directed crduit have differed litde
from other developingeconomies.However, HBPAEs, particularlyJapan
and Korea, have uscd unusual mechanismsthat have helped to increase
loan repayment rates and overallsuccessof directed credit. They have
limited the size of credirsubsidies,and theyhave applied stringent stan-
dards to the selectionand imlplementationof projecrs.
CriteriaforSeectingDreeed-Cree. Tagts. There are three broad types of
directed-credit interventions. First, governments can direa credit to
specificfirms or ind ries.Second, they can direct credit on the basis of
broad functional criteria, such as promotion of exports or small and.
medium-sizeenterprises.T ird.they can direct credit to accomplishso-
cial objectives,such as mass housing or redistributon of assets among
ethnic groups.HimAE governmentshave engagedin all three.
Ta?etingfirms or indutdes. Japan cuinng its postwar reconstruction
and Koreaduring the 1970sdirectedsubstantialcredit to specificsectors
and firms, mostly in heavy and chemical industries (Has) (see table
6.5). Japan'sFILPaccountedfor about a third of newequipment lending
in the 1950s. Between 1973 and 1981, Korea'spolicy loans were about
60 percent of the total loans of its deposit money banks. Most ofJapan's
priority lending targets through the early 1960swere industries associ-
ated with large optimum scalesand increasingreurns to scale; so were
Korea'sduring the HCi period. Indonesia and Malaysia,conversely,had
disappointing experienceswith credit interventions targeted to specific
industries in the 1970sand early 1980s and abandoned the schemesin
favor of more functionally directed credit Hong Kong and Thailand
have not been active in using credit instruments to push selected
industnes.
In contrast with many other develop'ingeconomnies,where directed
credit often went to public and quasi-publicenterprises,direcred credit
in the BPAEs has generallygone to the private sector.Consider the cases

o80
 USING RtESOURGWE,j

Table6.5 Pmoportion
of Loans
Accounted
forbyGovernment's
PolicyLoans,
Japan
andRepublic
of Korea

I96'':.iV Japan
-22. ,
Republic~~~~~~~~6
fKa:reat
4965.
09 . 1 lons 'Pa/C
.SE5.9uS: P k'onsas a' o -: a
Pol. sare f ne'.7-- sa or

.. ~~~~~~~~~*
;,..e i

1955 J3~~~.2. 32. .. .-

~~,1970 ~~~~~~9.8 -13.8 47.41 "11.~~~~~~~~~~~J771:

.,W ~ :
............ ' t; .. ', '' I' ................ w: .,1' 7, oF
1975 W 106' :1.75.390
*~~~ PoS loan indude loans0t
udrJ some kind oftexii ;rrna ;&I prg;m tht, lon eted&emnali term
1980 1 141 ,~~~w'.u½ 176,
.D,>,,0inuerestzat
,;; eandaada8llt or suppor;tend
b'' ch Cenra Bn's auomr
8:12'.~~~~
'nicws&tdnglar 59.78"
h au ofal lan 19.4
and d-

;Wr;-;rfX(
Srnc:JpaCiDqE 19).0 i-r
uS-.rn;n-oth: (192. 8- -
:, ' - :;rl)w

Table6. loanspertainwi Lows Adeounder byiavemmentad Imn'Policgram(FIlP) Outstanding

Rloancre f2 finanKial

c-unts -ob -',ey b- ',

* fi:'ir
privasector madeby &- '::'; of ICor&de-'o'
j'Ban ', cd _ ,_,_
iwo _

va7&bni ~2U'Foreig
rh6~~~~~~~~~dveometinttuin
'rmd.nn
o icudkeyal-lansfo foa- i n seb,-dt banI d a d loa : ;te a

of Trkeyand
exio. Lans y te bakng systemto public enterprises
in Turkey'in 1975were 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 developmentbank
provided about 75 percent of its industrialloans to fbur state-owneden-.
terprises.UJhfortunatelyv,t-hesepublic enterprisesoften did not perform
---
a n econmic c193)ep
well economicallyor fina-ncially. oInS-goes 1 9t* Malaysia,and
East Asia, Indonesia,
Singaporehavedirectedcredit to state enterprises,but the proporions of
total acrdit were not p tversistently
Ehigh
the parastatalstended to perfbrm
better financially,and thaeintLerest
ratesubsidieswerenor large (exceptin
7
Indonesia).1 1Koreawas an exception; its lending to the public steel
*plant, Posco, was substantia but is widelyconsidereda succes story.
Ta.rgeting eorerne and smalland medium-sizeenteTiprises. The broad-
est functional-targetingof credit in the HPAEshas been to exporte-rs.All
Easturkeanecnmisexcept-f creit Hong
o steKong
-- Sngaore and Sigpem
avedircte
nterriss, ut hehavesubsidized
popotios o
export credits, often through the centralbank redis onSt te (see
table 6.6). Exportersfrom all industries'have had access.Smalland
medium-size-firms,whichoftenhad to relyon curbmarkets ibund ax-
portfinancingone of theirfewreadysourcesof fnral finaneow.Box6.1
onpagd284)
rAa, Maltznus
 .AC SKAN MIRACLE

of the ExportCreditSchemes
Table6.6 Characteristics in EastAsia
Hong Kong OfQersexport credit insurancewith a slight subsidy through
the Hong Kong Export Credit Insuance Corporation
(HKECIQ.

Japan Between 1946 and 1972, the Bank ofjapan (BO,Iredis-

countcd axport bills at low.ratesof interest (I to 2 percentage
points lower than that applicable to general trade bills).
During this period, the HOydirected 10 percent of its loans to
export finance (which financed 50 percent of toal exporr-
related bankIcnding).
Kara, Rep. of Offers concessional rediscounting of prreshipment credit
based on Ictters ofcredit (tcs). The differential betwecn loan
rates of this cxport credit and general loan rates was between
3 and 20 percentage po;nts until 1982, when this difFerence
was eliminated. In 1984 the dififrcntia was restored ar 1.5
percentage points.
The KoreanExporr-Inport Bank provides longer-term post-
shipment finance.
Provides export credit insurance, but this has been used spar-
ingly (by about 3 percent of exports in 1980).

Malaysia Since 197i7, the Export Credit Refinancing (ECR)faciliry has

rediscountd export bills fbr postshipmcnt financing
Since 1979, the ECRrefinances peshipmenr export acedits at
subsidized rates for cligibleexports, based on lectersof crediL
A domestic leter of credir was also introduced to enible the
financing of indircct expores.
Singapore In 1975 the government established a rediscount facility for
exports andre-exports.
The Export Credit Insurance Corporation provides insurance
for exports by Singapore.
Taiwan,Cbina Offersfirms short-tcrm preshipment loans based on LCs.
After products are shipped, firms are cligible for loans based
on LCs.documents agait acceptance of payment, and ship-
ment documenrs. In addition, the Central Bank provides a
special discount rate to designated domestic banks for export
loans, usually at 1 percentage point lower than the usual dis-
count rate The oeporrloan rate was about3 to 5 prcentarV,-
points below the minimumn intercsr rate for secured loas
between 1970 and 1980, and it dropped to I tO 2 percentage
point after 1980.
Thaand The Bank ofThailand rediscoun:s at subsidized rates 50 per-
cent ofthe value of Icnding to exports by comrmeraal banks.
Commercial banks lend on the basis of LCs.

28z
 USING RESOURCS

-Cref'itWork.s-
Box61 Directed -
r as aSupplemen,
,Bed -

MANY ECONOMIES HAVEmID TO USEDIRECD CREDIT TOE-:

smalland mediuni-sizeen erriss;,
- counge fewhavesucceeded.The
-contrast betwcen an unsuccessfiullandulumately unsusinuiable pro-
i.. in K
,.. -a .., =;.tE,.,;-n'-
;
gram in J-Idonesiaand a successhflprogram in Koreatshowswhy.
* .- In Indoncsiawhich itdl 'th_early.198 s hada heavilyreprcssed
f--'
financialsystemsimilartothatof' niany,developingeconomies,heav-,
'..ysubsidizedcredit dixectcdto enterpriseir. pnvdmnent-d&ignared
- categonesoften went awry.Cr&dizallocautions wre divcn-bypoliticalf
=aswellas conomicconsiderations.Repaymentrateswereas iow as 53:
percent imone program supporredby'the World BanL Meanwhile,.
porentialy aedrcwdrth
..
medum-ze ; firmsi that lacked
d R - ' ;,m a um' V
the politcal
-,P. q.R. d-,--,!j,:-
-:influenceofireenterprises and the speci.alcret access cnjoyedby
.. smallfs werefirzen-'ourtofh finanial systeCM..--
the govenimentbegan.o liberal-'
PartyPto re=edj thesewproblemts
"'-izz he fiianial systmmin te -early1980s, permitringa rpid increase
in thenuiiiber of'banls, 'iralizin intert -d. dimann -
cdirccted-credit programs.Aa result,medium-:sizefirms'access to iii'
.nance hasevidentlyimproved,althoughte situation mayhavewors-
died somwar for vey small firms lack-ur established
l-that
:commercialtack record.
.; In- contrast,
: . .= -Korea's
.. -. ~h;--.;.S ectcaei
Me- -M., t-progrmn
- er- -or small.. nd
medium-size. enterprises has beencompaiblefromthe outsetwlm'-a- a
disci&pnec-miarket-responsiveFminal:isysen.B ks; whlchflave
''..had ialarge degr¶eof autonomyin:choossingborrowers,,havepref-r-ed.
frinwith
'' i 'estiblisheduack'record:of 122firms su2rvd, only20
:.,drew.onKflnand inaik&trcsouraesduring rhei.start-Tuhje Kiorean.- -.,
program raiieed hrms withournor acesothe banking sector To-
keent&rpdsesgoi MepapMrv idcdfoSorldng,aswdl s-
5
iutvesm nt finance,\xth workingcirzal far eexp,r&avafable
on par-
.:cuailawly -libenltem&,Repayment.rates ii Kbrea hare beeAin.-excess-..
npn;rnr . .. -. '. . .- :

Perhaps most important-todmesuccess&4I{eKoreanPprogram,Jn,:.

-. teresr.iates
subsidieshaveben modest- wih thoseTin
wtcompason
.:I.
n*edonesiaand other developzg-e'onoiiest h'hdy;epies-ed fi-
rt
nancal>systems
ysr ,.,{,c lx,
Despite
e the
- KoreanD
perveitnrern
gov.emnenr's
_',pi,' -a..-.,;.Fs.,,. ,.!],

r,0I. 2n
ti7;nfiacamar dleedcr =n iumsz
dr:f4o'r.srn4all its;'-

_.. -4h 0-- - - r _ * - '

Prlss- 2:rCm SU- c.

 a *II.MIRACLE

effortsforsmalland medium-size
contrastsdirected-credir enterprisesin
Indonesia,wherethey fiiled and werediscontinued,and Korea,where
they have been moresuccessful.As we arguedin chapter2, the sizeof
the subsidywasprobablylessrelevantto firms,especiallyto smalland
medium-sizeexporting firms, than was the automatic access.Allocation
wasstricdybasedon exportperformance,and commercialbanksmoni-
toredperformance,sincetheyonly collecredrepaymentson realizedex-
port sales; governments, through customs, had the means to verilfr
performance.This createda simplecontestfor exportfinance.
There areother morelimitedexamplesof functionaldirectedcredir.
Indonesia,Korea,Malaysia,Tlailand, and Taiwan, China, have re-
quired banksto extenda proportionof their loansspecificallyto small
and mnediun-scale enterprises.Smalland medium-sizeenterpriselend-
ing throughspecializeddevelopmentbanksis a growingproportionof
Japan'sdirectedcredit There are also attemptsat facilitaringcrecliac-
cess of technology-inrensive producers,especiallythrough Taiwan,
China's,Bankof Communications.
Targeting
socialobjecaieL.SomeIuAEs haveuseddirected-creditpro-
gramsto achievesocialobjectives. As we noted in chapter5, Japanhas
directed credit to facilitarethe adjustment of decliningindustries.
Malaysiaand Indonesiahavedirectedcreditto promoteincomeredis-
tributionamongethnicgroups.EvenHong Kongand Singapore,which
have fewother directed-creditprograms,providesubsidizedcredit for
mass housing.
KeepingSubsidies
toDirectedCmditSmalL
Compared nithmost other de-
velopingeconomies,the subsidiesfor directed-creditprogramsin most
HPAEs weresmall.In part this was due to the way in which directed-
creditprogramswerefinanced.The creditprogramsof Japanand Tai-
wan, China, werefinancedby postal deposits,whose repaymentset a
limit on subsidies.Singaporealsoensuredthat the DevelopmentFund
financedby the pension funds wasself-financing.It was also the case
thar,with theexceptionof the HMcperiodin Korea(1973-81),the more
prudent macroeconomicand interestrare policiesof all HPAEs meant
that realinterest rates to borrowersof directed credit were generallypos-
itive, if not at market-clearinglevels.'3
Measuring effectivesubsidiesto borrowersacross economies is diffi-
cult,since it requiresestimates of equilibrium interest rates for eachtype
of credit in each economy. Table 6.7 contrasts real interest rates on di-
rected credit for the HPAES with a sample of other low- and middle-

7,84
 USING RESOURCES*'EflI

Table6.7 RealInterestRatesonDirectedCredit,
SelectedHPAEsandOiler DevelopingEconomies
(percent)
Economy Direcedit Nonireted credt
FIPAESI
Indonesia, 1981-83, liquidity credits -1.7-4.0
Japan, 1951-60 0.5-3.0 3.1-4.6
Korea,Rep. of, 197040, industry -2.7 2.9
cxporrs -6.7 2.9
Taiwan,China, 1980-89, industry 1.9-3.9 4.6
1984-85, exports- 1.5 4.6
QOterdevelopingeconomies
Brazil, 1987 -23.5
Colombia, 1981-87, industry 1.5 13.5
India, 1992 - -2.5-4.0 7.0
Mexico, 198748 -24.0 6.0
Turkey, 1981, indusay -4.0-15.0 13.9
1980-89, cxports -14.0 13.0
,,- _No: mIvilable.:
.iurerJDB/JER (1993); World Bank d2t:L

income economies. Directed-credit programs in the HPAEs had higher

real interest rates than in other developing region&sGiven capital's high
international mobility, we conclude that interest rate subsidies in the
HPAEs were lower than in other regions.' 9
The HPAEsgenerally have smaller shares of their credit systems subject
to government direction than other developing economics. For exam-
pie, in the early 1980s, roughly tiree-fourths of loans from Turk-ey'sfi-
nancial system were made at government directive. In Pakistan directed
credit as a proportion of nongovemment credit was 54-68 percent in
the late 1980s and I990s (an internal World Bank report). In Brazil,
about 80 percent of the total average stock of credit was directed in
1986-87. Colombia's directed credit to indusay in 1984-87 was on av-
erage 56 percent of total credir. By contrast, in EastAsia, Korea's policy
loans in the HCIperiod were more than 50 percent of total credit, and
Japan's never exceeded 15 percent of total credit. Indonesia briefly had
48 percent of its total loans in priority liquidity credits in 1982, but it
reduced these rapidly following the liberalizations of 1983 and 1991.
Moreover, dircted credit is becoming less importanL WhiLe earlier
efforts of govremments in East Asia concentrated on developing the

285
 ' S!WNMIRACLE

banking systemand guiding its lending decisions,more recent policies

have been directed at deregularingfinancialmarkets:Japan started in the
1970s and the others in the 1980s.The fearuresof deregulationare sim-
1 directed credit is reduced,
ilar: interest rates are gradually deregulated
and entry restricrionsare slowlyeased (seetable A6.5, at the end of the
chapter). The majority of HPAEs deregulated their interest rates during
the 1980s, jointly with efforts to liberalizetheir economiesand their fi-
nancial markets. The deregulation of entry has been much more lim-
ited, however, except in Indonesia. This process in the East Asian
economies has been gradual and was meant to minimizedisruption in
the manufacturingsector.
Institutional
Mechanisms andMonitomng.
forSelection In contrast romost
other economies that have rried but fiiled to use directed credit, Japan
and Korea appear to have strong institutional capacity for project de-
sign, appraisal,and monitoring. Becauseof this, direccedcredit has gen-
erally gone to projects that are creditworthy and viable. More
importan-and in sharp contrast to many developing economies-
funds have usually been utilized for the purpose for which they were al-
located. The result has been a high level of loan repayments and a
correspondinglylow levelof loan losses.
In Japan, officials working within government sectoral guidelines
operated much like loan officers in private commercial banks, scruti-
nizing projecr proposalsand the firms involved to ascertain their abil-
iry to repay. Once a loan was approved, close cooperation between
development and commercial banks ensured continuous monitoring of
the performance of the borrower and enabled development banks to
take early action if loan repayment was in arrears.Coupled with Japan's
overall economic success, effltive pre-appraisal and monitoring re-
suited in very low loan losses.The Japan Development Bank, the main
government lender,experiencedwrite-offsof a mere 0.09 percent of av-
erage loans outstanding in 1951-55 and just 0.01 percent in 1956-65,
lower than comnmercialand trust banks that focused on short-termn
lending and had a more diversified portfolio. The JDB's superior per-
formance may be partly explainedby its greater reliance on collateral
security, particularly equipment purchased with the loans. Fear of los-
ing collateralized equipment gave borrowers a strong incentive to repay
their loans.
In Korea, the government, banks, and industrial firms worked to-
gether and shared responsibility for the success of directed-credit pro-

286
 USING RESOURCESO

jects. Various institutional arrangements such as monthly axport pro-

motion meetings and monthly briefingson ecoonomic trends, discussed
in chapter 4, supporred this close consultation (Cho and Hellmann
1993). Enforcement of government goals--primarily the imperativeto
export-was achieved by rewarding successful firms with continued
credit support while reducing credit to less successfulfirms. Such con-
tinual and pragmatic reviewand adjustment of policy implementation
reduced, but did not alwayseliminate, the moral-hazard costs of gov-
ernment intervention.
Other HPAEs fiallalonga spectrum betweenTaiwan, China, whichhas
utilized mechanismssimilarto Japan and Koreaand has had similarsuc-
cess,arndIndonesia,which wasplagued by the problems discussedbelow
and has largelydiscontinueddirected credit.
In contrast with Northeast Asia, credit allocation decisionsin many
developing economies were often motivated by political and noneco-
nomic considerations. Projects were designed with conflicting objec-
tives and lacked proper evaluation. Moreover, the very large rents that
could be obtained from subsidizedcredit werea strong incentiveto cor-
rupt practices. In smaHereconomies,credit allocation to specificsectors
became credit to individual firms, because of the limited number of
companies involved. Consequendy, projects were subject to govern-
ment-monopoly negotiationsand were prone to poor appraisaland dis-
bursements without proper documentation. Credit allocation to large
and politically powerful firms or state-owned enterpriseswere charac-
terized by weak monitoring and follow-up on debc recovery,producing
high loan losses and eventally large-scale failure of directed-credit
programs.

HowDidFinancialMarketPoliciesContribute
to Growth?

Capicalm;arketpolicy fundamentals are undoubtedly the major ac-

tor explainingthe contribution that capital marketshave made to the 6f-
ficient use of resources in the HPAEs. Rapid deepening of financial
markets, combined with prudent restrictions on lending by banks for
real estate and stock speculation, are probably the most important con-
tributions of East Asia's financial policies to growth. Recent cross-
economy evidenceshows that the averagelevelof financialdevelopment
is strongly correlatedwith the averagerate of per capita incomegrowth,
both before and after contuollingfor initial conditions (King and Levine

287
 3iR A C LqR"A`ThMIRACLE

1993). Moreover, the level of financial development imporEandypre-

dictseconomic growth.2 0
What about attempts by government to direct the allocation of
credit? Here tie record is mixed. NWe conclude that credit directed to
exporters was a positive factor in acceleratinggrowth. All East Asian
economiesprovided automatc, generous access to credit (or credit
guarantees) to exporters with varying degrees of subsidies. Japan,
Korea, and Taiwan, China, went a step further they used exports as a
criterion to determine whether priority firms, regardlessof the market
for the actual priority investmenr, continued to receive credit.21Other
more specific directed-credit programs had both positive and negative
results. In general, the economies with strong civil servicesand highly
professional public financial institurions-Japan, Korea, Singapore,
and Taiwan, China-apparendy had some success in implementing
credit policiesto promote priority industriesand firms. Key to this suc-
cesswas the use of economic performrancecriteria and the monitonng
of firms by banks and bureaucrats.
T hese traits were evident in an extensive study of Japan!spolicy-
based lending in the immediate postvar period (DB/JERI 1993). The
study found thar interest rate subsidieswere no more than 2-3 per-
centagepoints and that effecive monitoring kept loan lossesvery low.
Moreover, direct,. :redit was a minor aspect of the overall financial
system, accounting for less than 10 percent of total loans. The study
identifiedthree characteristicsofJapan's policy-basedlending: reliance
on the private sector; the dose correspondencebetween national eco-
nomic plans and directed credit; and respect for the managerial inde-
pendence of the government financial institutions responsible for
channeling funds to the targeted sectors by selecting individual firms
accordingto purely economic criteria.
By comparison, Korea's use of directed credit during its economic
takeoff in the 1960s and 1970s was more extensive,and interest rate
subsidieswere larger. As in Japan, the governmnentin Korea consulted
dosely with businessesand banks, attempted to monitor perfbrmance
dosely, and relied on a competitivebusinessenvironment to providean
objective measure of performance.Accordingly,policy-based lending
probably contributed to the rapid expansionof Korearnindusy, partic-
ularlymanuFacturedexports (Cho and Kim 1993). Most of the directed
lending to priority industries in Korea translated into investments in
equipment. Despite Korea'ssevereadjustment 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 utilizationof the HCI investments,
which were in physicalcapital, as demand increased.
However,the cost of extensiveand highly subsidizeddirected credits
was substantial-and likelyto increaseas development proceeded. Di-
rected credits burdened banks widt nonperforming loans, interfered
with the financialsystem'sefficient operation, hindered the growth of fi-
nancial savings, and sometimes threatened macroeconomic stability.
Koreawas able to weather thesedifficultiespartly becauseits closed cap-
ital market at the time enabled the governmentto suppressinterest rates
on deposits, thereby funding the subsidiesfor loans to corporations. In
addition, relatively easy credit in international capital markets and
Korea'sstrong relationship with two key lenders,Japan and the United
States, helped to ensure the availabilityof foreign capital, which was ei-
ther fimneled into directed credit or went to bail out troubled enter-
prises unable to repaysubsidizedloans. Today,boch the greater openness
of capital markets in most developing economies and the heightened
competition in international capital markets suggest tiat financial re-
pression and directed credits are now less viable options for developing
economies. 22

To argue tlat directed-credit programs in general reached their tar-

get groups and that loans were effectively monitored does not establish
that they contributed positively to growth. To do this, microeconomic
evidence is needed to establish two facts: government lending pro-
moted investment and did not simply displace investment from other
sources, and the social returns on the incremental investment ex-
ceeded private returns, justifying the targeting of specific sectors. Such
evidence is rare-even for the northern-tier HPAEs-but two recent
studies attempt to address these questions for a sample of firms in
Japan.Z3
A careful microeconometric study of Japan's industrial credit pro-
grams in general machinery, electrical madcinery, and precision instru-
ments during the 1980s-sectors selected because of their high
potential for externalities due to innovation and learning-reaches the
following conclusions:

* Government lending had a large, positive, and statistically signifi-

cant impact on both gross and net investment; the three-year cu-
mulative effect of a one-dollar increase in debt financing from

289
 MRACSN MIRACE

government sourceswas more than two dollars of additional gross

investmentsin fLxedcapitalassers.
* Government lenderswithdrew credit once firms had earned access
to privatecredit marketsand, in turn, extendedcredit to new bor-
rowers rather than preexistingborrowers,consistentwith (though
certainly niot concdusivelyestablishing) the daim that credit was
being allocated according to objective lending standards, rather
than on thc basisof political considerations.
* The characteristicsof firms receivingincreasinggovernmentcredir
indicatc that assistancewas targered toward firms most likely to
grow,invest, and generate rechnologicalexternalities;firms receiv-
ing credit had researchand developmentinvestmentshares in out-
put approximately twice those of finms not receiving increasing
government financing.

The authors conclude: "Possibly because of the institutional structu

of these lending programs, it appears that economic logic rather than
political capture motivates policy-based lending decisions. Govemment
credit is withdrawn quicldy from seasoned firms, is targeted toward
growing, R&D intensive firms, and produces substancial increases in
investment and access to private credit. The fact that these results are
measured with data from the 1980s leads one to expect even larger ef-
fects for earlier periods? (Calomiris and Himmelberg 1993).
Limited evidence fiom the 1960s tends to confirm Calomiris and
Himmelberg's conjectare. Horiuchi and Sui (1992) examine the role of
the Japan Development Bank's directed crcdit in promoting private in-
vestment during the 1960s, when JDB was the major policy-based insti-
tution lending to medium-size and large firms. Their sample consisted
of 477 medium-size firms listed on the Tokyo Stock Exchange in 1965.
Among their results are:

* JDB lending resulted in a net increase in new investmenL

* JDB loans were accompanied by incremental lending from private
banks, suggestng that JDB's loans did not crowd out private bank
lending.
* JDBs loans to firms preceded the firms' increased borrowing fiom
private banks and improved the access to credit of firmnsthat were
not artached to private "main banks" (often firms affiliated with a
keiretsu bank).

290
 USING RESOURCESj =

Thus, directed credit appears to have acted as a signalingand insur-

ance mechanism. By selecting among borrowers, the JDB encouraged
complementarylendingfrom private financialinstitutions with a conse-
quent net increase in investment. Moreover,policy-based lending was
most successfiulwlhenit was directed at nontraditional borrowers.The
JDB's loans stimulatedgreater lending by private banks to firms not affil-
iated with main banks but did not have the same impact on firms
affiliatedwith main banks.
Another study of credit programs in Japan and Korea based on an
analysisof the objectivesand outcomes of the programs adds an addi-
tional element, noting the positive role of coordination achieved
through govrmment lending programs. During the 1950s the bulk of
directed credit in Japan was given to "basicindustries"-electricity, iron
and steel, and coal mining-that were characterizedby increasing re-
turns to scaleand that produced basicinputs for other parts of the econ-
omy. Once these basic industries were developed,Japan subsequendy
promoted industries whose expected spilover effectson the economy
were large, for example, machine toolS (JDBIJERI 1993). During the
1970s, Koreachanneled most ofics directedcredit for fixedinvestments
in manufacruring to the HCI industries, among them such producer
goods industries as steel and machinery.Using a frameworknot dissim-
ilar to the contest frameworkoudined in chapter 2, the study reachesa
similar conclusion: the coordinating role of government in Japan and
Koreahas diminished overtime, and hence the effectivenessand desir-
ability of using credit programs to improve coordination among private
firms are diminishing as well (Cho and Hellmann 1993).
But many other directed-credit programs, even in the HPAEs,did not
mect their objectives.In fact, the track record of most directed-credit
programs in Indonesia, Malaysia,and Thailand was so poor that their
governments, demonstrating the pragmatic flexibilitytihat we have ar-
gued characterizeseconomic policymakingthroughout the HPAEs, aban-
doned many of them. Limited differentialsbetween directed credit and
market interest rates and smaller volumes of directed credit relativeto
the size of the financial system meant that financial market disrortions
in the HPAEs, even from unsuccessfulattempts to direct credit,were not
generallylarge. In addition, the focus of HPAE governmentson 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

Usingthe International
Market:
TradeandIndustrialPolicy
W HILE THE LINK BETWEEN INTERNATIONAL TRADE AND
economic growth is widely accepted,the precisenature of
the relationshipis controversial.Earlywriting on trade pol-
icy and growth stressed the benefitsof neutral incentivesbetween pro-
duction for the domestic market and production for export. With
neutral incentives, it was argued, resourceswould flow to sectors in
which the economy was most internatonally competitive (Corden
1971), and gains in technical efficiency(or "X-efficiency")would result
from a more competitive environment (Nishimizu and Page 1991).
Studies in the late 1960s and early 1970s convincinglydemonstrated
the extreme bias in incentivestructures against exports and agriculture
in most developingecononmiesand the resulting high costs of import-
substituting industrialization.4 Recent theoretical work has argued
that, where market power, economies of scale,learning, or externalities
are significant,departures from neutral incentiveregimes (lowand uni-
form rariffs)may improve economic performance (Krugman 1986).
This literature, however,unlike the earliercarefiullydoctumentedstudies
of the costs of protection, providesvery litde evidence conceming the
importance of these factorsor the potential benefits to grwwh of depar-
tures from trade neutrality25
Proponents of trade neutrality and interventionboth dee the records
of the high-performingAsian economies as evidence supporting their
views.As we noted in chapter 1, international trade is important in all
the HPABsand is the factor most consistentlycorrelatedwith their suc-
cess.Balassa(1991), Krueger (1993), Hughes (1992), and others argue
that opennessto international trade, based on largelyneutral incentives,
was the critical factor in East Asia'srapid growth. Conversely,advocates
of trade interventions, while acknowledgingthe importance of trade,
note that incentivesderiving from quantitative restrictionson imports,
tariffs, and subsidieswere not neutral among sectors (or firms) during
their periods of rapid growth. They argue that the HPAEgovemments
successfully intervened to change comparative advantge (Amsden
1989;Wade 1990;Singh 1992).
Industrial policy interventions, which often use trade policy instru-
ments, are motivatedby the belief that shifting industrial structures to-
ward newer and more modern sectors increasesthe opportunities for

292.
 USING RESOURCESF

capturingdynamicscaleeconomiesthat resultfrom learning.During

theirheavyand chemicalindustriesprograms,Japanand Koreawerethe
most active HPAEs in promoting individual industriesand sectors.Sing-
apore and Taiwan,China, have also activelyprovidedincentivesfor
tednological upgrading.Malaysiahad an HCIprogrm reminiscentof
Japan'sand Korea's,while Indonesiahas attemptedto leapfrogfrom
labor-intensive
manufacturingto high-technology
industriessuchasair-
craft and electronics.
How havetradeand industrialpoliciesaffecredgrowthin EastAsia?
To answerthis questionwe first examinethoseconditionsthat might
justif governmenteffortsto promotespecificsectors.We then describe
the trade and industrialpolicy regimesin the HPAEsand compare their
domesticrelativepriceswith internationalpricesto determinethe extent
of interventions. Finally,we evaluaterwo approachesto increasingpro-
ductiviry.We find, on the one hand, that governmenteffortsto promote
specificindustriesgenerallydid not increaseeconomywideproductivity.
On the other hand, the evidenceshows that broad governmentsupport
for exports was a highly efFectivewayof enhancingabsorption of inter-
national best-practicetechnology,thus boosting productivity and out-
put growth.

MarketFailures,Trade,and Industrial
Policies
Why should governments inreriere with the level playing field cre-
ated by the international market? Our ditcussionof coordination fail-
ures in chapter 2 touched on many of the reasons,but it may be useful
to reviewsome of the arguments specificallyas they apply to trade and
5 Among the many reasonsfor discrepanciesbetween
industrial policy26
socal and privatereturns cited earlier,four stand out:
X Interdendent investments and economiesoficale. Increasingreturns
fom scaleand capitalmarket imperfectionsmay mean that investments
that could be internationallycompetitive at optimal scaleswill not be
undertaken. This is especiallytrue with large, interdependent projects
for which optimum scale depends on simultaneous investnent in up-
stream and downstreamindustries.The larger the indivisibilitiesand re-
turns from scale,dhemorelikely that private initiativeswill be absent.
a Straegic negotiations.In negotiatons with other economiesand for-
eign companies, governmentscan alter the nature of the marker envi-
ronment by intervention. The outcome of any bargaining problem

293
 CL r
RACR"MM!IRA

dependson the strength of competition on both sides. By coordinating

thc actions of buycrsof technologyand trying to increasecompetition
amnongsellers, governments can appropriate more of the surplus associ-
ated withlthe transfer of technologythan they otherwise could (Stiglitz
1993a).
e Pecuinay externalities. Pecuniaryexternal economiesarise if, as the
size of a competitive industry increases,the long-run supply curve faills
(Marshallianreal externalities).Such gains in productivity are attribut-
able to economies of scope in the use of specializedequipment and
greater specializationof individual skills. When economics are small,
currcnt prices may not convey adequate information about prospective
lower costs of production through largerplant size (Scitovsky1953;
Chenery 1959).27 Externalitiescan also arise from the interaction be-
tween suppliersand buyersabout the design or production of a product
leading to a better or cheaper good than is availableinternationally.In
this case,the sourceof the externalityis the nontradabilityof some types
of inputs or knowledge-otherwise the improved method or product
could be obtained from internationalsuppliers.
* Learning.Externalitiesrelated to learning have traditionally been
identified as important sources of mark-et fai]ures in developing
economies.When firms gain knowledgeof production from other firms
without incurring cosrs, real externalities are present.2 8 Becauseof in-
completeappropriability of knowledge,individual firms may spend less
on obtaining production knowledge than is socially optimal.3 9 Exter-
nalitiesdue to learningmay alsobe conferredon other firms by the first
entrant. These indude the demonstration that the sector is physically
and economicallyfeasibleand the leakageof information on technology
and marketing (Pack and Westphal 1986; Rob 1990).
Economists initially responded to these market failures with argu-
ments supporting the protection of so-calledinfant industries. Protec-
don wasseen as necessaryso that firms could gain the experienceneeded
to lower costs and become viable.30 Industrial policy advocatestakethe
aTgumenta step fiuther. Becausethey associatelearning with capital-
and knowledge-intensiveindustries, they advocate a rapid shift in
industrial structure toward these activities,evenif they are not intema-
tionally competitive at present levels of scale, knowledge, and fiactor
prices. In this view, the short-mn aflocativecosts of establishingintema-
tionally uncompetitive industries will be outweighed by the long-run
benefits of rapid productivity change in the promoted and linked sec-

294
 USING RESOURCE RCE

tors. But as tlc many inf.int industries that have never grown up amply
demonstrate, prorection does not enisurethat the promised learning and
economies ofscale actually materialize.

TradePoliciesin the HPAEs

Most IHPAESbegan industrialization with a protectionist oricntation

and ihave%radually moved toward increasinglyfreetrade. Along the way
(as we showecdin chapter 3), they often tappcd some of the efficiency-
generatingbenefits of international competition through mixed trade
regimes:they granted exporters duty-free imports oFcapital and intcr-
mediate goods while continuing to protect consumer goods. Export
prices were set in the international market and were often substantially
less than current marginal or averagecosts.31 Losseson axpart produc-
tion offset profits in the protected market, while competition in the in-
ternational market ensured that the firm would not suffer from loss of
cost discipline. More recently,all the HPIAEs have reduced their protec-
dionof import-substituting industries.
Belowwe describe the evolvingpatterns of protection in the HPAEs,
with the exceptions of Hong Kong and Singapore, which adopted es-
sentially free trade stances early in their development.
Japan. Japan pursued an early import-substituting industrialization
strategy,similar in many dimensions to those of Argentina, India, and
ocher less successfideconomies. As late as 1968, effectiverates of pro-
tection (ERPs) in Japan were still quite high and exhibited the cascaded
pattern from raw materials (low) to consumer products (high) that is
rypical of most developing economies (see table 6.8). Unlike many
import-substituting economies, however, there was surprisingly high
protection of machinery (final producer goods), confirming other re-
search indicating that the Japanese atuthorities engaged in an intensive
effort to develop this sector. Eiu levels in the machinery sector were re-
duced during the 1970s, only after it was evident from export perfor-
mance that the sector had become internationally competitive (Itoh and
Kiyono 1988). Quite high levels of protection were afforded sectors
such as iron and steel and nonferrous metals as late as 1970. Protection
in capital-intensive sectors such as pulp and paper and chemicals also re-
mained high, to say nothing of the remarkably high levels in textiles.32
Korea. There is considerable evidence, summarized in Pack and
Westphal (1986), that Korea selectively protected sectors that the gov-

295
 A RACLE

in Japan
Table6.8 EffectiveRatesof Protection
(percent)
By type afgoork
1963 19 8 _ _-NRP

Type -A . -P. NARP ERP 1 1968

Rawmatcrials 3.1 0.8 3.9 0.9 -2.3 -3.0
Producergoods -13.7 29.6 15.2 22.3 15.9 7.1
Intermediate producergrods 12.3 28.0 14.1 21.7 15.7 7.6
Finalproducergoods 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 14.4
Tattiles 54.3 282 .18.6
Spinning 27.1 12.5 I.0
Weaving -44.6 30.5 15.5
Products 72.8 32.8 22.4
Woodproducts 14.0 25.6 16.1
Pperandpulp 9.7- 18.0 11.0
Publishing -16.7 1.0 . -0.9
Learherand rubberproducts . 30.9 21.8 12.3
Chernicals 33.4 17.7 8.8
Perroleumand coil products 19.5 14.5 . 7.1
Nonmeallic minerals 22.2 15.7 8.1
Iron andsterl 30.1 30.0. 17.1
Nonrfrous metals 30.4 34.1 22.1
Metalproducts 13.8 -19.9 9.9
Machinery 36.7 20.0 7.7.
Genemal machinery 23.0 14.5 8.7
Electricalmachinery 30.9 16.5 -5.4-
Transportequipment- 61-5 31.0 9.2-
Precisioninstruments - 34.9 22.9 IG.4
a. NRP=Nonial RateofProtcdon.
b.-ERP=EfficdvcRarcofProtecion.
c. ERPbasedon simpleaverages
of riffraes--
Sovw Fortop panel,Itoh mndKiyono(1988);fiorbottompancl,'Shouda
(1982).

erinent hoped to promote. Protection consisted of both tarifis and

nontariff barriers.Table 6.9 presentsa summary measurethat combines
the effect of nominal tariifs and non:ariff barriers.33 The figures span
the period 1966-85, and thus indude the earlyefforcsat industrialpro-
motion, the HCI drive,and the subsequent liberlization. The 1966 fig-
ures show a relativelyprotectionist stance that becomessomewhatmorc

2.96
 USING RESOURC SE E .

Table6.9 OverallDegreesof ImportLiberalization

byMajorIndustry
forthe
Republicof Korea,Selected
Years(ConsoEdating BothQRsandTariffs)
(percent)
Mgjor indunrLy £966 1970 1975 1980 1983 1985
I. PrintaryIndustry 5.5
56420 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 lcathcrproducts 34.7 .36.6 43.8 74.0 74.3 87.8
IV. Woodand its produas 24.2 64.0 69.1 76.5 82.7 92.4
V. Paper,paperproducts,andprinting
and publishing 39.5 54.7 54.5 78.4 88.8 90.8
VI. Chernicals, perroleurm,coal,rubber,
and plasticproducts 47.9 57.1 58.4 65.9 75.2 80.0
VII. Nonmetallicmineralproducts 41i. 77.0 76.8 89.1 89.2 91.7
VIII.Basicmetalproducts 44.7 73.3 74.8 86.2 90.1 92.0
IDC Meralprotucts, machinery, and
eqwupment 41.0 59.3 55.4 63.6 69.7 77.8
X. Othermanufiacturing 31.8 39.5 38.7 58.2 65.4 76.1
Allmanufacturing(II-X) 37.5 47.7 52.6 66.8 71.3 79.8
Lighrindustry(II-VX) 33.7 38.2 43.7 62.3 63.8 76.8
Heavyand chemicalindustry(VI-DC) 44.5 62.5 61.2 70.5 76.5 81.7
All-industryavcrage(I-X) 39.6 50.8 52.3 65.6 70.4 78.5
Near:This tablegivcs
thedegreesof importliberalizationbymajor indusuy theseare obcainedby averngig scrmal damweightedby
productonforrespcriveyer
pricevalucofdom:sric
thecurrent
Sowcan:Sakong
(1993).

liberal by 1975. Nevertheless,even by 1983, when Korea'ssuccesshad

become an establishedfact, most sectors were still protected by some
combination of tariff-sand nontariff barriers.While Koreautilizeda va-
riety of instruments, especiallyexport targetsand rebates,to ensure that
exportersfaccd internationalprices for their tradeableinputs, there was
considerableprotection of goodssold on the domestic markec.
Taiwan, China.Wade (1990) provides a carefuldescription of the ex-
tent of intervention by authorities of Taiwan, China, in product mar-
kets. The pattern of protection is not dissimilarfrom that of Korea.As
late as 1972, a significantpercentageof items were subjeccto nontariff
barriers, and two-thirds of potential imports were subject to nondinal
tariffs in excessof 30 percent. As late as 1980, more than 40 percent of
imports receivednominal protection in excessof 31 percent
Indonesia,Malaysia,andThailand.As we describedin chapter3, Indone-
sia, Malaysia, and Thailand all had import-substitution regimes that
while modest by international standards, nevertheless favored produc-
tion of manufhacturedgoods for the domestic marker at the expense of

2.97
 MURN!iL IRACLE

agriculture and exports. Malaysia was notable for low, if variable, pro-
tection of import substitutes, while protection levels in Indonesia and
Thailand were higher. All three economies began export-push trade
strategies during their periods of protection of the domestic market.
Tables 6.10, 6.11, and 6.12 provide recent data on the structure of ef-
fective protection for these three economies. Effective prorection rates
are declining but in Indonesia and Thailand they remain sufficiently
high to result in some anti-export bias. The free trade regime for ex-
porters in these economies partially offsets the structure of protection.

of EastAsian DomesticPricesto IntemationalPrices

The Closeness

Notwidtstanding the protection thar exists in all the HPhEs except

Hong Kong and Singapore, domestic prices in these economies are
closer to international prices than in other developing regions. Two bod-
ies of evidence lead us to this conclusion. First, nominal tariff rates ad-

Table6.10 Selected EffeciveRatesof Protecticn,

Malaysia,Selected Years
(percent)
Sector 1969 1979 1987
Foodprnducts 83 88
Bev:rages 20 38 -22
Tobaco 125 - -26
Textiles - 58 15
Apparel 400 45 6
Wood products 33 38 82
Furniure 40 84 43
Paper and paper products 140 66 29
Printing and publishing - 26 -9
Industrial chemicals 230 32 12
1Peroleumrefineries - 22 5
Rubber products 60 129 14
Plasticproducts 265 312 163
Iron and sted 84 63 289
Fabricatedmetal products 40 26. 30
Nonelectricalmachinery 1,600 89 19
Electricalmachinery 130 4 12
Transport equipmenr 185 59 65
-Not available.
Note Theeff&cive
mceof prorectionis thepercentgebywhichvalueaddedat
domestic
pnicesexcereds
valucaddcdatworldprics.
SoweSalleh,Yeah.andMeymnathan(1993).-

298
 USING RESOURCE-SC

Table6.11 SelectedEffecveRatesof Protection,

Indonesia, Years
Selected
percent)
Seaor 1975 1987 1990
Food, beverages,andtobacco 336.2 122 124
Textiles,cloth,and foonvear 231.8 102 35
Wood products -12 25 33
Paperandpaperproducts 873 31 20
Printingand publishing - -
Industrialchemicals 28.4 14 13
Petroleumrefineries - -1 -I
Rubberand plasticproducts 426 57 48
Ironaandstcel 18.2 13 10
Fabricatedmeaulproducts - - -
*Nonelectrical
machinery - -
Electricalmachinery - -
Transport equipment
Non-oil : -1 1
-Cement 63.6 60.2 53.6
Manu&crmnrngTotal
(eclduding
oilsector) 74.1 68 59
-Nor ailablo
Nan The cflicrfie rareofprotectionis thepecenngeby wcb valueaddedat
domesticpricesexceeds valuCaddedarworldprices
Pngestu (1993).
Sowr.cBhmacharyaaand

jusred for the presence of nontariff barriersare lower in the HPAES than
in most other developingeconomies.Second, comparisonsof real GNP
across a broad range of economies indicate that domestic relative prices
fortradablegoods in theH-iAEsare doser to internationalprices than in
other regions.
Table 6.13 presents one of the few systematicattempts to compare
nominal taiff rates for a broad range of developingeconomies (Erzan
and others 1989).34The authors have also atrempted to adjust for the
eflc of quantitativerestrictionson levelsofprotecion. (We have added
comparable data for Taiwan, China, and have computed the average
nominal tariffi for Korea, Malaysia,Thailand, and Taiwan, China.35 )
Nominal tariffs in the HPAEs are lower than for any other grouping of
developingeconomies except the island economies of the Caribbean
and the oil states of West Asia. The differencebetween Latin America
(albeit before its recent trade liberalizations)
and the HPAEs is striking.
Thus while these economiesfavored production of import substitutes,
they did so less than most other developingeconomies.

299
 WN MIRACLE

Thailand,SelectedYears
Table6.12 SelectedEffectiveRates of Protection,
(percent)
Sept Marc/i Oct Nov. April
Secor 1971 1974 1978 1981 1983 13984 1984 1985
Averages
Allseccors 87.2 18.6 70.2 - - - - -
andtobacco
Exdudingfood,beverages, 44.2 45.9 90.3 - - - - -
Manfacmuring - - - 7.4 67.4 65.3 57.0 66.3
Textiklprmducts - - - 248.5 144A 127.8 108.0 118.4
Leacherppmducs - - - 151.3 199.7 199.0 167.0 152.7
Woodproducts - - 60.7 53.7 55.0 49.7 62.0
Paperand pulp - - - 39.5 43.0 43.0 39.0 53.5
Chemicalproducts - - - 54.3 52.9 51.9 46.6 44.5
Rubberproducts - - - 48.7 46.0 46.0 39.3 42.0
Othernonmetalproducts - - - 99.8 102.7 102.3 89.8 108.5
Metl products - - 633 62.0 61.9 53.9 70.9
Machinery - - - 14.1 23.6 23.6 21.0 29.3
Consumergoodsandmotorvehicles - - - 34.9 33A 33.5 30.6 45.6
- Nor available.
Noz The effectivemrre bywhichvalueaddedat domesticpricesexces valueaddedatworldprices
ofprmtectonis the percentage
SwxrcBrimble(1993).

Table6.13 AverageTariffsandPara-tariffs,
byRegion1985

Tranip Ceatral Soudt North Saharan West Otler DeveAlpg All

para-ta f Caribban America America Africa Africa Asia Asa PEAt' aglow
Tarffi
Economies
unweighlted 16 23 34 29 32 7 36 19 26
Economies
-import-weightedb 17 24 38 30 35 4 22 17 24
Tariffplus
pars-tar:ffi
Economics
. unweighted" 18. 65 46 36 34 9 42 24 34
Economics
import-weightedb 17 66 51 39 36 5 25 20 30
- a Simpleaverages
acrossproductsand economies.-
*b.Simpleaveages cwssproductsacrosscountries,avcegs weightedby otal impor.
c IndudesKdrca,Malaysia,andTaivan,China.
d. For Taiwmn.China, 1981 dat:.
now=
Erzanandothers(1989).

300
 USING RESOURCEStURECF

This is borne out by comparisonsof internationaland domestic

prices.Figure6.4 showsan indexof outwardorientationbasedon in-
ternationalcomparisonsof pricelevelsand pricevariabilityforthe HPAEs
comparedwith other regionalgroupings(Dollar 1990). The HPAEs as a
group are moreoutwardoriented than other regions;their relativeprices
are doser to and more consistentlyrelated to internationalprices. Of
course,there is substantialvariationamong the HPAEs. As would be ax-
pecred, Hong Kong, Malaysia,and Singapore rank in the top nvo
decilesof the index, consistentwith evidenceon their stmcruresof ef-
fectiveprotection.Somewhatsurprisingly,Thailand, with higher effec-
tive protection,alsois rankedin the top decile,whileIndonesiais in the
top third. Japan, Korea,and Taiwan,China, rank in the fifth and sixth
deciles,below such developing-economycompararorsas Brazil,India,
Mexico,Pakistan,and Venezuela.This is consistentvith rhe evidence
that the three northern HPAEsintervenedfar more frequendyand sys-
tematicallyin their economiesthan the southeastem HPAEs (Dollar
1992).While anylarge,multi-economyeffortat real price comparisons
is subject to methodologicaland empirical criticism, the results are
broadlyindicativeand consistentwithother evidence:EastAsiasrelative
pricesof traded goodswerecloseron averageto internationalpricesthan
other developingareas.

Opennessto Forin Technolo

An important fctor in East Asia's successfulproductivity-based

catchingup was opennessto foreignideasand technologyGovernments
encouraged improvemrentsin technologicalperformance by keeping

Figure6.4 IndexofOutwanr
Orientaion

HPAEs

OECD
economies
SouthAsia I
LatinAmeadca
andCadbbman

MldcileEast I
SabSaharan
Africa

0 1 2 3 4
Svnnre:Dollar(1990).

301
 .SIAN'-MIRACLE

severalchannelsof internationaltechnologytransferopen at all times,

even thoughsome, such as directforeigninvesunent(DFi), werere-
strictedor dosed for varyingperiods.In contrastto mosteconomies
withimporc-substituting strategies,evenwhcnprotec-
industrialization
dionwaspracticedwith respectto the domesticmarker,the searchfor
and absorptionof foreigntechnology wasencouraged iWhile Japanand
Koreaset obstadesto DPI,theywerehospitableto licensing,though
evenheretheJapaneseattemptedto ensurethat theydid not incurex-
cesscosts.Singaporewasexceptionally welcorming to directforeignin-
vesunent,and a majortask of tie economicplanningagencywas to
locateappropriateforeigninvestors.
Malaysiahasaggressively soughtex-
port-orientedDFI, particularlyfromJapan(see box6.2). Noneof the
HPAEs washostileto the establishmentof localbuyingofficesby inter-
nadionalpurchasers,an importantsourceof productionand marketng
knowledge.This selectivelypennissiveattitude towardthe acquisition

Box
:;W.6.:--,
I stEn
6.2.ri-Forign .
BuigsExport
, -
to Mablysia-
Technology : a
.INmTHE 1970SAND EARLY 1980S DIRECr FOREIGN toward eecrcal- and electronic;priductsC
chemical2
Was weloMe in Malaysia.
investment Gonenbbe rduc melproucs, and
policy-encourd foregn investorye the eec pelm. Moreoftheinvtents were-dedicated:-
:
was.inhibitedby- i a of i ing t- exorsiv before.
partcpaton incorporateowner. Then,
;-Dlocal -. e The-benefits toMaysia of DF havibekensub-
-receson of 1985-86 largefiscalantrdedefi stnal par genrang reign echange
anddecliniginvesmt broun
htanaur for
spuh a er, somecitics -argue
ploymenr.-.Howe that
.,pC
. . . . . for. -xpos
. : . PO.. , . o d Z .- - i , eloc
. = no-my
V . ... . .. I.
ax andrladdomesticquity strast,not
t--- cenives atogas i neighig&
bparigpaiong potentialinvesto,
requzrcnmcis-or cpits ofDETwoimportntreasonsfirtatare,
tecleogy and-e-portan& employment thec young :#ny g of the investments LIn
~~-.,,CO~31 Wr-.-an MC MU O
opporrumnes. eaiescaz&cl ofmangainet cac
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Malayuiidd
The newflcdilihad a draiaic confo padiyIskilld h-qualiysupplierate-
lbor,and

U US$300millionayearin198385,jmpe
averagd foreig inveset pj thmselves
cts -h

to US$2 bbi981-. t "became rnbuied to blng. no-hoywin iec y

--
''. Som-L.@ -- ! - f-mc
tii&&ibJanii d-
n>198a qU2rerFEre,T og
moredynaic. 1985tn tirie mrcnsmiocntrm
nare frriiJapzi>fllo~1..arwah,Chi&anc &crrinMlysase9io~ than$0 ilo
y when thegreatest soureof-D.El
hd'bcen_:W i;L u a tsi e -k
gaporet nanfrlly tiicafti&and
afraed bfisb&ffrtedrd hertcnooyr kceepu

302
 USING RESOUPRCEts;EPfl

of knowledge of international best practice was a reflection of the view

that the world market for goods and services provided an opportunity,
not a threat.
In contrast, many other developing economies that tended to em-
phasize the dangers ofopcning to world markets were equally suspicious
of open policies with respect to knowledge acquisition. Suspicion of ex-
ternal trade was often reflectedin a mistrust of DFI and licensing. The
absence of exports reinforced the suspicion. EvenwhiereDPI was permit-
ted in inward-oriented economnies,it was not viewed as providing access
to international best practL e but rathieras a source of additcional domes-
tic production. Thus some economies in Latin America that were hios-
tile to licensing nevertheless allowed DFIin 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~ ~
- - ~
er ~~ ., -- .o
actmis
M2)Sa. , eprs W - i-h& t'u n

semiconductorsto their assembyatvtes hr s toMlyi. The eiets aysihabulupn

also no doubt that Dfl has,creaceda lot of empiloy electronics.is.ConsidemrblcNIxdrTomue AG,
mea-often.fbrpoor. rua aay. women-=to, e- .wiEc recen'dyestablished.aUS$3,''iniIl souibfware:z-
tune of 85,000 inelectronics alone in the late 1980s cetr ocreate4applications6r,UN-aeworkism
MAiyaysi-has found :that,-ainetments mature- sin>a trce oMl yiabth&r'arts
-- adluand ndrginl'ikae aefrtifed ml-. sotaeegnes&roe-fifth of Germany4snre
nainl hat' are fe' huh fa "foodoiose" *.calledii its Mlyineprst_hpstu;ci;.
stay, and the~poiiplovers ~created-ysuccesfu asmlline iinArzo'na.'
vntures encouragemore.investment.Fo xapc Mlasa spanproduce:a:-rang-fpout-
*-~
Si cth 90,DI1 nd aysi '& rom coiorTy<~,xado6Lassett-pyepl;;ar ioatD5asters
cx
word'sithirdargest,--producer f. semic-bondiu"cto&compdze"eiiheras. Mcomom istat scarte&
i--, Tmrld'
I .n.flt.2 li- Ma

-garnering almostbillion.
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:

Unfiterie-s). Birocincmais a&okp os oief die ornporflezfbr:itsnew clri

to _seinc5ndu&rois--vestmentsin dectroAicshave plntffaiya,jn th&tat&-1980 Matus `ia's

r~~eincohduccbrPa4s.A
~uppljilocal i%5Pi~&nr4' rei:ywor1d=
m rke Sr-roomtair,-'-
Motbrolaind
~llo(r&
Hiachilii~e
tSei~a&has ~conchcioner n op.(-1991)

303
 MiR A C LN.MIACE

given the small internal market. Korea and Taiwan, China, with larger
domesticmarkets, created a similarsituation by encourging DFI in pro-
duction for acport and in those sectors where substantial technology
transfer could be anticipated. In contrast, the large import substitution
economies, such as Argentina, Brazil, and Mexico lured investment of
foreign firms by offering access to protected domestic markets.

Poicies
Industrial

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, exceptX llug Kong, have employed industrial policies as
defined above. Japan and Korea had che most systematic set of policies
to alter industrial structure. Efforts in Taiwan, China, were less system-
atic but were nonetheless widespreacL Industrial policy in Singapore was
more functionally directed at the rapid upgrading of technology by di-
rect foreign investors, regardless of type of output. Indonesia, Malaysia,
and Thailand have all used industrial policies but much less systemau-
caly than the northeastern HPAEs.
Industrial
GrowthandProducivit Change. HPAE industrial growth pattems
diffier from the patterns in most other low- and middle-income
economies in the relative size and growth rates of two important indus-
trial subsectors: metal products, dectronics, and machinery; and texies
and garments. Table 6.14 shows the share of value added in nine Inter-
natonal Standard Industrial Classification ('sic) subsectors as a percent-
age of the tota value added of manufacturing. Among the HPAEs, metal
products, electronics, and machinery (isic subsector 38, or MPM) have
grown unusually fast The sector's share of manufacturing value added
doubled in Japan and Singapore, nearly tripled in Indonesia and Korea,
and quadrupled in Malaysia. More surprising than the importance of
growth in MPM, which provides vital inputs to numerous other manu-
facturing subsectors, is the contnued importance of tctxiles and gar-
ments even as the rapidly developing Asian economies shifted fiom
labor- to capital-intensive production.
Appendix 6.1 describes our method for determining the projected
relative importance of specific industrial subsectors in cross-economy
comparisons. As can be seen in table 6.15, both capital-intensive seccors

304
 USING RESOURS O

ValueAdded
Table6.14 CurrentPriceValueAddedasa Shareof Manufacturing

-m TM"& w-d _ - X.- - pA-

' ;dr
Eismjf hwo~~~~~~'" ad ad...d "ad u ~a
HPA&

3973 0.05 0-. ' .I gm 0.12 Ut 0.01 0. 33 gmJZ

3573 L15
.4 0333 "I 0.36 00) DA3 0.0 DA. IW
3511 0.-16 0.35 all 0.01 0.35 0.05 an0 0.34 .03 0.59

3053 IL"9 0.36 WI 0.10 0.36i 035 0.12 ON2 0.02 IL"
MP a" am oi3 - e5 . -0.6 0.4 s os 1 0.46 e a

-Y196 .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 02
0- 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 - '-56 4.36 0

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.~~ ~ ~~~~~
OA05 ~~28
AM3 01 0.37 0.33I 41.1 0.5 amo 3.0

(mpm)and labor-intcnsivesectors(textles) play a greter-than-predicted

role in manufacturing in most of the HPAEs (excepr Indonesia and
Malaysia).Besidesthe unexpectd prmninenceoftextiles and garmencs,
the aaalysi yields another surprise. Chemicals and nonmetallic miner-
als, two sectors often assciated with indusErial modernizadon, ameno
-largein the HPAES by interatdonal s o.ndards.
Detailedsectoralgrowth ratesof total factorproductivty areavailable
for Japan, Korea, and Taiwan, China. There are now sufficiendy long
time-series dat to conclude that in these economies, T* growth has ac-
counted for a substantial raction of the growth of constnt price value
added in manuFacturing. Table 6.16 shows a varety of-esmates oflong-

305?
 AiRAC MIRACLE

term rFP growth rates. Given the length of time of the observations,it
seems unlikely that the measured growth rates of TFP could be attribut-
able to cyclical phenomena or growing capacity utilization of initial
large investments.3 6
Sbtres of SelectivePrmnonion.Here we describe the strategies of selec-
tive promotion that have evolved in the HPAEsas 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

Table6.15 Actual/Predicted Shareof GOPOriginatingin ManufacturingSectors

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 ter
0nay ftrn

1973 e12 lo
10-3 .89 0.53 0,7 e.14 0.37 2.6 0o35 1 2s
9ISBt 0.43 IMF95 036 065 Dh67 0.14 0.17 11.6 0.3 1.2

1973 1.63 [.7 0h30 O3

IAO 0.20 0.00 0.44 0." Oda
15SC 0.75' 03' 1.42 0.6 0.54 095 Z.62 a.ll 0.01 0.57
_,~~~~~~~~~~~~~~~7 :-5
19 on 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. Rxord
1563 0.99 17 2.0 1.04 0.94 0:73 0.52 2.0 0.20 1.04
19 D35
0. 2:4 o.n1 0.64 0.56 02 1.13 276 0.32 1.26

155 0.J5 025 1.55 O." G." 0o.e5 @0A 1.03 1.43 0D4
19B1 - 0.94 0e5 33 0.7 C."9 0.70
Jo57 '.44 0.07 OS7

1973 0.31 2.14 2.40 0.49 L09 0.75 1.53 1236 0.13 IAI
t9S9 0@36 11.32 0.31 0.47 t.7 0o27 oa 50 0.34 to3J

1968 0.!0 121 4.35 al 0.3 1.9 02 3.16 0.09 0.95

19: 6 .34 3.33 1.85 071 0.65 135 033 1.32 022 3.s

SeM -

1I56 0.50 6.00 0.54 233 033 0.37 1so

3,74 0.32 1.10
19S6 0.56 - 2.39 (L 1.02 0.39 0.35 0.63 L51 D.09 052

Lai . AM
'listS_;a - '-

- 1
3963 0.91 I10 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 _32
l

- £. ?dk±cJ uhiew 90711C aku*t-l.t..p..

-306
 US ING RESOURSCOU

Table6.16 Long-Tenn
TFPGrowthRatesbySector

TFPgrowtldrare
Taiwan,
Japan, C-/ina,
Sector Kirea 1960-79 1966-6
Food- 7.30 -1.76 2.0'
Bcverages 7.90 0.0
Tobacco 13.40 -
Apparel - 1.98 10.5
Tcxctiles 10.70 0.47 7.6
Lcather 12.60 1.03 _
Shoes - 1.03
Wood 9.40 2.81 o3b
Furniture 12.10 1.74
Paper 8.20 1.44 2.3c
Printing 10.70 -0.18
Chemicals 13.10 3.36 3.3
Pecroleum -0.30 _3 55d O.Ob
Rubber 11.40 1.02 63b
Nonmetallicminerals 2.80 - 2.4
Basicmetals - - 7.2
Ironand-steel 3.70 134
Metalproducts 7.60 3.41 4A
Nonelecricalmachinery .00 - 230 67'
Elecncalmachinezy 10.70 537
Electrical
equipment - - 7.1
Transportequiptnent 11.20 432 2.7
Prccision instrumencts - - 11.0
Plastic products - o.92f: 0.0
Oilhermanufacturing 7.50 -1.76
Average 8.8 1.2 4.6
-Not availablc.
a. Foodmndbevrages.
b. Rubber,petroleum,andwoodproducts.
c Paperandpaperproducts.
d. peuroluemref and coaL-
c. All machincry.
f Plasdc..
SowrePack(1993b).

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 (Effectiveratesof protection in these industriescan be seen

in table 6.8.)
Japanese industrial policy until the early 1970s had severalstrands
(Yamamura1986; Yamnamura and Yasuba 1985). The government di-
rected credit to large firms and protected the domestic market to help
the firms realizestatic and dynamic economies of scale.mmTt allocated
foreign exchangeto favoredlargefirmsand tried to ensurethat technol-
ogy payments to foreign firms were neither excessivenor duplicated,
often forcing licenseesto share their information with other Japanese
firms.Relativelyhigh levelsof protection were used as part of a strategy
As late as 1978, the averageof effec-
to help firrns to export successfully.
tive rates of protection across the manufacturing sector was 22 per-
cent? 9 Such rates may have been responsiblefor the very low level of
manufactured imports, an issuewidelydiscussedin the scholarlylitera-
ture (Balassaand Noland 1988).
Korea.Korea'spolicieswere similarto Japan'swith respectto selected
aredit,protection, and limitation of entry into specificsectors(Padcand
Westphal 1986). However,the Koreangovernmentpromotedindividual
firms more often to rectifyperceivedentrepreneurialand skill deficien-
des, usingexport performanceto determinewhether firmsdeservedcon-
tinued promotion. Other policies to encourage industrial growth and
exportsinduded making directand indirectinputs to exportsavailableat
world prices. In Korea,the selectivelypromoted sectorswere the heavy
and chemicalindustries:iron and sreel,metal products,machineiy,elec-
tronics,and industrialchemicals.The motivationfbr theseappearsto have
been both strategic-to increasedefensecapability-and economic-to
shift to capital-and technology-intensivesectorsin anticipationof a loss
of competitive advantagein labor-intensivesectors. 4t The costs of the

Ha: drive are still not filly known, but theywere high (seebox 6.3). Al-
though there are no recent esrimatesof Korean rates of effectiveprotec-
tion, there is considerableanecdotal information suggestingthat the
governmentaffordedthesesectorsrelativelyhigh protection.
Taiwan,GCina.The intervention of Taiwan, China, in manufactur-
ing has been similarto Japan'sand Korea's,though less important quan-
tiratively.Wade (1990) has documented tariffs,quantitativerestrictions,
and selectve credit policies, maintaining that the successof Taiwan,
China, was at leastpartly attributable to an intensivegovernmenteffort
to direct the economy'ssectoralevolution. This conflictswith the stan-
dard neoclassicalview that its developmentwas primarilyattributableto

308
 USING RESOURCES.E

low protection, tde availability of inputs to exporters at international

prices, a conservative macroeconomic policy reflected in low inflation,
and competitive factor markets (Little 1979).
Unlike the situation ill Japan and Korca, in Taiwani,China, it is dif-
ficult to discern a pattern of economic incentives. Wade (1990) notes
that a guiding principle in the 1960s may have been the existence of

. ........~~~~..
~

lidtervention
Box63.31TeHCIDrive:Costsof
THE-KOREAN C;OVERMENT'S INTERVENTION sidered the lower bound, however. Those industries
.policy,most notably ihe HCIpromotion, has often whose credit was squeezedwere,frced to borrow at
. been
evaluateCd
from thc.perspectiveof the successor- curb markerrate. The rCI
reaLeffectivetsubsidies
fior
..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 -betoestirmate the fisal/financial cost asso- strugglingheavy machiney; shipbuilding, overseas
r CaCed wit intervenCiin. Govcrnment intervention in-:X-- construcrioh, and shipping indusries in' th mid-
.- ',.cureddiecthefosts thform ofsubsidies to strateic 1980s, nonperfbrming lo of conmrcial banks
sectors through-policy loans andl tax-exemptions, es- accumud rapidly and accordny bank prof-
¶ecialfr during the..1973-79 periodof.HCI .promo- i;abi.y seriously detoed During 198-7, fr
tidon. Intervention McQf
nflosPA
rred idirici c
in:iec
- t exaanplIe, the share of nonperfmii lonsI to
-rmn .of accuimiated nonperorming loans assets readied
eInost 10 percent. Between -1985 and:
resulting port:bli difficties'ofcommeicial banks.: 1988, 78 coqorations were "rationa DrC
ThepornmoUooftea scows spporteda
b. abroaod ran.oft fscalae,dFCa.ncirwa.i'srumets. ti process,
iwntie-o. of principal aonc amounted rn
iy~-oroid-='rangds deovted- totheCninstrent. Far tr*illonwn,-orabour-I.prcentofcNxin 1985.
.-Z Goiernment ids t.devotedal
dtoe-tc
dHC.,isector Farmaging-was
d-am tihe frter issunceofsubsi-
amounwird roperceatord2tc rationaiefis adt p
-romotrionpmo.n aton, ations to Cen Bakdiscoun-rswhich hash:urtbthe
Upfipjwo100pr.fr:hi ea ad 50 .. balaces . ; .0
percht-irfor the` ew yas -provi eK overnmeab r d rf
;'-rate aboutatwo-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 percenmeofrtot
won s o(3t sixt gb'p
<collecned)>a.~~~~~~~~~1ost
'of fi ~ ~ ~ 5qt;ra;n it has taken,mre tan-a
;-XstedT ha..spredomnant -. . + .idede
toawith
.sdircefi- ind .th costs
i n
?-.-,.&asupoirH U-wa:pr&etl pOliclon
1 . form of accuulated.nonpdomn'on an
dirtedtokey-industes.
snp n977r 'squed ba pr N C45
93t
percent tioflthe
wfahe:oiidomeT&
rhplicinin&i&~~~~P ib kaMng governt
oubida
tKore aonce med
n6perceinnt
domesais,ei- pl
.nanci I r iii dHCIi uppsorCofInsfpoic
theiic to lo stlact
,.sectr: I.l HdVseccoI'in. more than40 . of tot
1977-a wroe ati5& 75io & separate.; Da&k- ad to.z adle.t tc
piceit 4 tGNP'(6kIiby aitg an nerest< troughicbuaet atyetto 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
evidencesummarized by Wade demonstrates considerableintervention
in many sectors, there are no studies of effectiverates of protection or
subsidy after cheearly 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 investmentsundertaken by the gov-
ernment to providethe basisfor a rapid shift toward higher-technology
sectors.
Malaysia.In the 1980s Malaysia experimented with a heavy and
chemical industries drive similar in focus to those of Japan and Korea
but dependent on public investment.The Look East policy adopted in
1981was, apart from its socioculturalaspects,an explicitattempt to em-
ulate the heavyindustrializationeffortsof Japan and Korea,which Nvere
regardedas successes.The governmentcreatedthe Heavy Industies Cor-
poration of Malaysia(HICOM) to be the vehide for the heavyindustrial-
izaton push. HICOM targeted a number of large-scale,capital-intensive
projects for development including iron and steel, nonferrous metals.
machineryand equipment, paper and paper products, and petrochel...
ails. Whilepublic investmentswere intended to be catalyticin each of
these activities, die government activelypromoted joint ventures with
private (usuallyforeign)investors.By 1988 HICOMhad set up nine com-
panies, employinga totai of 4,350 workers,involved in steel, cement,
motor vehicle,and motorcycleengine manufacruring.The highest pro-
file of these investmentswas the PROTON car project, a joint venture
with Mirsubishidesignedto produce 100,000 units per year.
The ambitious targets of the HICOM program ran into the maao-
econornic constraints described in chapter 3-declining terms of trade
and a deteriorating fiscal position. Prompt responses to the deteriorating
macroeconomic situation induded reductions in public investment in the
HiCOmprogram. Poor management and low profitability also forced the
treasury to bai out loss-making enterprises, a practice it could not sustain
under the fiscal austerity of the adjustment program. With the shift to
policies of privatization of public enterprises and private sector develon-
ment in the late 1980s, the HICOM period had ended (see box 6.4).
Indonesia and Thailand. Neither Indonesia nor Thailand has made
the systematic efforts to change industrial structure characteristic of

310
 US ING RESOURCESCE

'-'Box6.4Ro-' gnizingeaMistke, 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 - ductiveand innovative.Nor can the macroeconomic
.rapidly,expandedthe umberof stat efirms, pary to bencfits be overlooked.in the end, tihecderal gDi'-
-,;;givehMlay'sopportuniuesand.partlyoutofa convic- -emuent anticipatesthata-tleast15 percentofitsour-
i,noniat the govea entcould s de&elopmentr. snd -in debeen trred to privafizedi
MLSifuch of th& vgrowt i the early 1980swas due to firms.
'HICOM,. the Heavy and, Industrial Corporation of Malaysia'sslow but steady divestiture and' the
Malaysia,which esiablishednine subsidiaries'in such, b enits it realized by 'adaping flexibly to thcir-
=areasassteel,automobils, cement, ant paper. cumsrances of-each--cas provide anI examplefor
.- , :fkis, especiallEy
State in HCOM, performed less. . niany
m oter deveopig cnos, wh privatiza-
,'wl thanexpeced nd in 1984 :thedefici of the- 'tion' is an offical goal, yet much less has been
p'ublic enterpnsestached 3.7 p ercentofGNP.Asris-.- achieved A derailed analysisof the compauies that
,, Dipg fiA:il cEdfictraa&la
mngs of trade threat- -w:er privatized'showsconsistentgains in soial we0
-ened -macroeconomic stability sthe govemenc fare. As two cases belowillte, changes m incen-
begat pdvaize in 1983.iOfm orerhan 800 state tives and a broader role for priate sector interests
fiMrs existing Mi th mid9Os, about a hundred had, have 'frequently been more imporrant.-lthan the
been by'WO.Maire s conmecial change in. oners pe .
.manuring oper s,butsome were quite big,. M A&r?Systems Berhad (MM).- -Wlhi48 .-
incl=du,ding
an--airline; a shipping' company,, and a percent of ASwas sold to the publicin 1985,ther
telecc}tnmuincatiotis fim, wa no gc in an t or de 'jur autl{orit'"
* there
have beeinglitches.-For instance, t public Eve so, managment gnd de t contrl over
offeingshave been heavily esscbed,:suggt investrment decisions and eventulywon permission'
; the:government
in that upde.*ied.i' sha'res
s per- to. raise domesc- fares. Whie the farie har.
6
Imps l sing as much as US$150 million i, t sl of consumers, less, rationing -o -tickets ra c
ur big companies.A-lso,morc atenon to promor- w r Moreoe, ile a b
, b- mg¶5mpedrion
u h wr',p
monopolisriasitu?aribnsvol r; "O';L the early
'in 18sir
Mocvr-1ic becamec
ALA a~ profitablta-paying
l'i2bu ',Rb-,cn
:. even,
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 f Kr was sold to de-izep -
Sect or r-femain
; X sL.,ibni, ,, d . --e.;
;; --. . T - dlS.
. - . to.,t the
;, pia

hysia.
s -meant thoughte ublic secor retained control, the
more.than ounigh sales. Ith. indmcude'l ashosr"of
' ;.rm that took over, many repres'nnt'' Bumi
-p '-;
:schems d h c.thh'
ontro srests, -' ireincentvet to ear Pr6fit.nK
tovrpro
~~rodxicri
tiea
tvi&de
cotrings
6& eII&'
r
'ubl-PiIcsc~
phec s a luc
ui c r es, nopbz-
t H-tover *as- theaiSaxontrall PRO- -- 1,ing container. traicinand out ofdie caf reion
nil
ON,taornil company inWCOM. E
ienad Even
ref.ier o . 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); -

kspoldecrk&4he~~
mor' toit~z wa tha ~ Slkii(199); od Bn (.1988)"'-

3"I
 -C _AIRACLE

Japan, Korea,Malaysia,and Taiwan, China. Thailand's Boardof Invest-

ment (BO})has attempred to promote variousactivitiesviewedas having
the potential for technologicallearning, but, as we pointed out in chap-
ter 3, it is difficult to discern any sysrematiceffort to change industrial
structure from the pattem of promotion of the BOI. Indonesia has used
public investment in an effort to move roward high technologyindus-
tries. Indonesia'stechnologydevelopment efforts representan cxtreme
case of state-supported technologicalupgrading, but the public sector
researchfacilitiesand strategicindustries have fev linkswith the private
industrial sector. One experimentwith leapfrogginginto high technol-
ogy industries-Indonesia's attempt to create an aircraft industry-is
describedin box 6.5.

DidIndushialPolicyIncreaseProductiv?

Have atrempts [o alter industrialstructure helped accelerateproduc-

tivity change?We attempt to answerrhis question by addressingrwo re-
lated issuesin 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 relativefactor prices?If changesin dhe seaCoral
composition of output are largely rnarket-conforming, industrial
policiesmust have failedirhat least one of their objectives:to guide
industrial development along paths thar it would not take if it
wereguided by market forces.
* Second, what were the rates of productivity changein industry in
the HPAEs, and what were their sectoral patterns? If rates of pro-
ctucivity change in industry are low overall or in promoted
sectors,prima facie, industrialpolicydid not meet its productivity-
enhancing objective.
In both cases,our answersleadus to conclude that industrialpolicies
were largelyineffecnive.
Indusbial
PolcyOnlyMarginaly
Altered
Induri Structure.The cross-econ-
omy comparisonsof industrial structure raise quesrionsabout the effi-
cacyof government effortsto promote or discouragespecificsectors. In
Korea, for example, despite the government'sextensiveefforts to speed
the privatesector'sshift from labor-intensiveto capitml-and technology-
intensive industries, the relativelylabor-intensivetetiles and garments

312
 USING RESOUCES

sectorwasnearlythreetimesbiggerthanintemationalnormsprediaed
in 1988, a substantialincreaserelativeto internationalnorms from
1968.Duringthe sameperiod,Koreamerelymaintainedthe interna-
tionalnormin chemicals,a heaviLy promotedsecror,whileocherheav-
ily promoted sectors, basic metals and metal products and machinery,
achievedonly modestimprovements.Similarsurprisesare evidentin
Singapore,
which,likeKorea,hasa governmentthat aggressively
inter-

Box6.5 Indonesia'sTurbulent
Leapinto HighTechnology
* SECr1oRAL-.TARGETING TO ACHIEVE RAPID PRODUCTIN .
changeis a rislybusiness-paTilarlywhenpubliclyfundedfirs
arepemiedmtoily onprotecred domestcmardratherdian bing
subJece to international
compeunon.Thedangesoofsucha coursc
*areevidentin thedifficulties
encountered
byIdonesianai.ramfpro-
- ducerPM. Industri Pes;wat Terbag Nusantaa(iv), whichhas
absorbed
$1 billionin goverment.
fiud since its: esitablishient in
1979 but has yet to becomeinternatonally compmtiveor geniiney-
.profitble
Intheory, IPN is exectcd to usethe domesticmarkeras a spring-
board toiinternation2lsales.Te strategyis to shift-froiproducing'
aircraftunderlicnse, tOco-designand co
-.- cton, tohsoledesigna
-and production.In practice, vns.pcpdo meticxmruueiis tooe:-
.smalltogenciate the economiesofscalenecessayfor efflciehtaircraft.
pro<pductio&Moreovcr,the company hasihad lidelexposureto the
:-worid mker-.Of the 230 aircraft(30 hlic:opteri, 100fixl:-wing,r
-. that IPiNhasmanufacruredsincerits inaiion, -Cl

.beensold to Indonesianairlini<the miiaiyrhe pmoli and th&

stame-ruoi in
Foreignsalesareuniikelyto imptrove.Whileotheraircaaftaimanu-
&u--r&sJcihaive been
zibusy,foginhnological ianc&,to%Iinkrhe
-timefromdesign
compleonoplucon, i is- iesil
alone. Its majorpwd@ forthisdecadan dendeily
igXto~ gDIt
designed fifyaurboproppi .nor.
rs fr el u.
-1995,f ieryearseimdesignicoxpiieon Byh
..for slmalarar f&alr~adywedwill havexseral hern en
rr cn ill affztpnC
t -ranrsallpncing their uairdfrCaggressively.
arompyr srbu195'hle, m ostsrc:,o4i-
:catethat IPTN-was.nor pfiptirabWlewhen
capitaLcoAstsarleen into ac-
COUUL Noduogin xhompP2a,nyspinCrtormancsnce
rhenindic-te3
thgatthesztuandiibs iniproveI:. . >-

313
 :AN
MIRACLE

vened through a variety of mechanisms to promote capital- and

technology-intensiveindustries. Here, the importance of textiles relative
to the predicted norm has increasedeven more sharply.The textile sec-
tor went from double its predicted size in 1973 to eleventimes in 1989.
Over the same period, metal products and machinery declined from
rwelveto fivctimes its predicted size.
Appendix 6.1 describes a simple rest of the impact of industrial
policy on the basis of the factor proportions theory of comparativc
advantage (Hecksher-Ohlin-Samuelson). We assume that given the
relative labor abundance in all of the HPAEs early in their industrial-
ization, the share of labor-intensive manufactures in toral manufac-
turing should rise as they begin exporting. We compare statistically
the pattern of growth of the industrial sector in each economy with
its pattern of factor intensities. We argue that if industrial structure is
derermined primarily by market forces, low wagesand low capital in-
tensity at the beginning of the period of active industrial policy will
predict the pattern of industrial growth. If industrial policy to pro-
mote capital- and knowledge-intensivesectors was important in de-
termining the sectoral pattern of growth, however,this should not be
the case. Indeed, the opposite should be true; high wages (a measure
of human capital intensity) and high capital intensity at the end of
the promotion period should predict changes in industrial structure.
We have performed thesetests on each of the Four Tigers plus Japan
for periods relevant to their industrial promotion efforts, using two-
digir isic classificationsof industrial sectors (seetableA6.2 in appendix
6.1). Our efforr to differendare berween a comparative-advantage-
based evolution of industrial structure versus one characterizedby sig-
nificant intervention is notable mainly for a number of negativeresults.
In particular, in Korea-the economy for which significant interven-
tion is best documented-during 1973480 the most rapid growdt in
seac 'I shares of value added occurred in lower-wageor lower value-
addeC_per worker sectors. In Korea at the two-digit level, sectoral
growth was broadly market confbrming in terms of traditional factor
intensities.The resultsfor Japan are incondusive. In Hong Kong there
is no relationship between our explanatory variablesand patterns of
production. In Singapore, for the period 1980-89, output grew more
rapidly in more capital- and knowledge-intensivesectors, supporting
the view that the Singaporeanauthorities successfhllyintervened to en-
courage increasingly capital-intensive development. But, given the

3"4
 USING RESOURCES F

rapidgrowthin the capital-laborratioin Singapore,this resultalsocon-

formsto factorproportionstheorypredictions.
These findings do not imply that governments were not attempt-
ing to influence industrialstructure. They undoubtedlywere. But
they suggestthat, despitegovernmenrintentions,the manufacturing
sector seemsto have evolvedroughlyin accordwith neoclassicalex-
pectations;industrialgrowthwaslargelymarketconforming.
Ratesand Patternsof Product Changein Industry.
There is both good
newsand bad newsfor advocatesof industrialpolicyin the productivity
performanceof East Asian industry. The good news is that, on average,
rates of productivity change in industry in Japan (before 1973), Korea,
and Taiwan, China, which are the only economies for which we have
derailed sectoral estimates of TFP growth, were high by international
standards;productivity-based catching up was tak-ingplace (Page 1991).
The bad news is that, in general, productivity change has not been
higher in promoted sectors. Japan may be an exception. Between 1960
and 1979 chemicals and the metalworking machinery complex haveun-
usually good TFP performance Jorgenson, Kuroda, and Nishimizu
1987). Japads industrial srructure differs from international norms in
these sectors and exhibits quite high values of the share of value added
in total manufacturing. These industries are those that observers usuaDy
point to as having received significant government support, induding
efforts to stimulate producrivity growti.
A number of calculations of Tr have been carried out for Korea for a
variety of periods (Dollar and Sokoloff 1990; Lim 1991). From these
studies a number of patterns can be identified that are broadly consis-
tent with one another. Most striking are the high values ofTFrPchange in
most sectors by intemrational stndards (Nishimizu and Page 1991). Al-
though the Korean government selectively promoted chemicals and iron
and steel (included in basic metals), the large growth in the share of iron
and steel was accompanied by quite low TFP performance between 1966
and 1985; textiles and clothing conversely, had very high rates of TFp
growth. The promored chemical sector, whose relative size was decreas-
ing, was characterized by considerably higher-than-average TFPgrowth
during this same period.
The government in Taiwan, China, did not attemnptto 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 threesectorsthat exhibitedthe greatestexpansionin the shareof

valueadded-apparel, electricalequipment,and metalproducs-are
allcharacterizedby below-average wages,capital-laborrarios,and labor
product.Moreover,there is no statisticalrelationshipbetweenwageor
capitalinrensityand productivitychangeat the sectorallevel.In fact,the
highestsectoralratesofTFPchangearerecordedin textilesand apparel.
RecentsectoralTFPgrowthraresfor industryin Malaysiafrom 1973
to 1989showasimilarpatternofgreatvariabilicyin cheTr growdtrates
in promoredsectors,from high in nonelectrical machineryand paper
and paperproducrsto lowin iron and sreeland transportequipment.
Textilesand apparel,a nonpromotedsector,areamongthe TFP leaders
(Maisom1992).In short,thereis no apparentrelationshipbetweenthe
sectorspromotedunderthe HICOM driveand rapidproductivitygrowth.
Overall,the evidencethat industrialpolicysystematically promoted
secrorswithhighproductivity changeisweak InJapanthereissomesup-
port for theassertionthatTFPgrowthwashigherin selectedseccors, while
in Koreaand Taiwan,China,activitiesthatwerenot promoted(forcx-
ample,textilcs)had TFPperformanceas impressive as those that were.
Moreover, arremptsto determinewhedterhighratesof TFpgrowthcom-
binedwith rapidgrowthof promotedsectorscan plausiblyexplainthe
veryhighoverallratesof TFPchangein manufacuingyieldmosdyneg-
ativeresults(seeappendi 6.1).The mainreasonsfor manufactrings
successinJapan,Korea,andTaiwan,China,layin the highgeneralrates
ofT"Pgrowth,includingthosein labor-intensive, nonpromotedsectors

Increased
Exports
HowManufactured Productivy

OES THIS MEAN THAT PUBUICPOLICYHAD NO ROLEIN THE

We be-
rapidratesofTFP changefoundformanyof the HPAEs?
lieve not. In chapter I we describedhow rapid UP change in
developingeconomies can be a result of the ability to move quicldy
doser to international best practices.We termed this "productivity-
basedcatching up." We believethat rapid growth of cxportsi a resultof
the export-pushpoliciesof the HwAEs, combinedwith the superiorper-
formanceof rheseeconomiesin creatring and allocatinghuman capital,
provided the means bv which they attained high rates of productvity-
based catchingup and TFPgrowth.

3I6
 USING RESOURCESC

How did the unusually high rates of growth of exports and human
capital contribute to the productivity performanceof the East Asian
econonies? Most explanationsof the link betweenTFP growth and ex-
ports emphasizesuch static factors as economiesof scale and capacity
utilization.While thesemay account for an initial surge of productivity
soon after the start of an export push, they are insufficient to explain
continuing high TFr growth rates. Rather, the relationshipbetween ex-
ports and productivity growth may arise from exports' role in helping
economies adopt and master international best-practice technologies.
High levelsof labor forcecognitiveskills permit better firm-leveladop-
tion, adaptation, and masteryof technology.Thus, exports and human
capital interact to provide a particularly rapid phase of productivity-
based catching up.
Of course, it is possiblethat the move to a higher production finc-
tion occurred beforethe growth in exports-that TFrgrowth causedex-
port growth rather than the reverse.41 But even if expors began on the
basis of productivirychangedue to such domesticeffortsas plant reor-
ganization, the cumulative magnitude of productivity growth over
nany yearsis most unlikelyto have been a resultof purely domesticef-
forts.42 It stretchescredibilityto suggestthat the largecumulativeeffcts
of TFPgrowth in Japan, Korea, and Taiwan, China, could have been
achieved by the plant floor innovationsproposed as important sources
of productivity growth at lower TFP growth rates (Packand Page 1993).
Were that the case,it would be difficult to explainwhy these gains far
outstrip productivityincreasesin the industrialeconomies during their
own rapid-growthperiods, when TFP growthwas attributable largelyto
domestic factors.Clearly,then, an increasedability to tap world tech-
nology has been an important benefit of exports.4 3

Exports'Rolein ImpedectKnowledgeMarkets
Why do exportsfacilitatethe move towardinternationaIb&,:-practice
technologies?The knowledgethat pernits this shift is availableonly in
quire imperfectmarkets. Often the marketsdo not exist-some knowl-
edge is simply not sold, because its owners fear that licensingor direcc
foreign investment will eventually Ieakthe knowledgeto future com-
petitors. Evenwheremarkets exist, they are likelyto be characterizedby
bilateral monopoly,so thata varietyof difficultiesconfirontfinns trying
to purhase such knowledge(Arrow 1969).

317
 4MIfA CL E

In markecswith imperfectknowledge,mechanismsthat help an econ-

omy or firm obtain technologycan conferan cxternality-apecuniaryone
if knowledgeisobtainedat a lower cost,a realextemalityif the knowledge
turasferis more efficient.For example,the transfermay be more beneficial
if knowledgeis obtainedas a resultof the self-interestof purchasersof ex-
ports in the OECDeconomies,rather dtan through arm's-lengthpurchase.
Exporting helps to overcomesome imperfectionsin the market for
knowledgeand permits its acquisition through a varietyof mechanisms.
These include:
mPurchase of new eqiupment.The purchase of new equipment is a
straightforwardmethod of obtaining new technologyinsofaras it is em-
bedded in equipment The two main impacts of exports in this case are
in providing the competitivepressure for firms to make such purchases
(if the equipment is used to produce exports),and, more generally,in
earning the foreign exchangeto finance the purchase of machiner)y
a Direct#irieztinvestmentThe firms that generatemuch ofthe world's
new technology are reluctant to part with it. They perceive that the best
use of tieir newknowledgewith respectto developingeconomnies often is
through exporrsof producrsto them. In East Asia,however,most of the
incoming DFIhas been intended for production of exportsrather than as
a strategyfor domesticsales.An economy'srecentexport performancefre-
quendy signalswhether it is a desirablelocation for export-orientedDPI.
Economieswith rapid export growth are preferredin part becausesuch
growth often reflects good macroeconomic management Moreover,
econornieswith significantexport growth have generallyprovidedmore
infrastrucure useful to exportes than internallyoriented economies44
DFImakes crucial production and marketingknowledgeavailableto de-
veloping economies.It permits them to begin manufacmringalong the
world'sbest-practiceproducton fiLnctionby substitutingforeignphysical
and human capitalfor absentlocalfactorsfi5 As locallaborlearnsfrom the
presenceof best-practiceknowledgeand equipment, knowledgerends to
become diffiLsedthrough labor mobility and informal contacts among
managers.Even when the investingfirn attempts to hinder such diffu-
sion, signicant knowledge ranfers are inevitable.6
* Technology
licensing.licensing exsting technologies,both rights to
proprietaryequipment and derailsabout production pocesses,offersde-
velopingeconomiessubstantialopportunitiesfor improvingdteirlevelsof
best practice.During the 1950s and 1960s,Japan benefiredconsiderably
fiom licensing (see, for example, Nagaoka 1989). The net gains were

3I8
 USING RESO URCC

large,as the licenseswere obtainedat relativelylow cost comparedwith

the domestic sarch expensesavoided.However,there issomeevidence,
anda growingsubjectivesense,that arz's-lengthlicensingis decreasingas
an option for dosingtechnologygaps.Technologydevelopers,the licen-
sorsin the OECD economies,havebecomewaryof helpingpotentialcom-
petitors, even if contractspredude exports to ochereconomiesfor the
duration of the license.Particularlyin R&D-intensivesectorssuch as
chemicals,machinery,and electronics,firmsare increasinglyunwillingto
licensetechnology-, they believeroyaltiesprovidean inadequatereturnfor
actionsthat mayimpairtheirownlong-termcompetiuveness. This creates
an irnperfector nonexistentmarketfor criticaltechnologytransfers.
Exporting economieshave an advantagein coping with this situa-
tion. Licensingfirms prefer cross-licensingagreementsin which they
obtain accessto the licensee's own technologyor to its manufacturing
skills.Given asymmetriesin knowledg. the best test of the potential
parner's abilities is its performancein export markets. Moreover,the
fact that a firn has exportsand the requisitetechnicaland commercial
skillsto producethem improvesthe bargainingposition of the licensee
Salesof licensescontain a significantelementof rent (Pack 1993c),so
the cost of licenseswillbe lowerwherethe sellerperceivesthe strengths
of the purchaser as signaled by exports. Finally,where a mark-erdoes
existfor technologylicenses,the foreignexchangeto pay for feesis more
ikely to be availablein economiesexperiencingrapidgrowthin exports.
* Trasfir of nonpmprietary technology.
The freeor inecpensivetrans-
fer of nonproprietry knowledgeis easierif an economybeginsits in-
dustrialization effort in labor-intensivesectors using relativelyold
machine designsand production technology.When exports are based
on comparativeadvantage,preciselythese conditions are met. Hong
Kong,Korea,Singapore,and Taiwan,China, all beganrapid industrial-
ization in labor-intensiveand low-technologymanufacturing. The
equipmentwas readilypurchased,and the production knowledgewas
availale at low cost from engineeringpublications,trade literature,and
independent consultants. 4 7 The simpler equipment was conduciveto

liocalproductivity-enhancingimprovements,often on the part of blue

collarworkers (Ranis 1973). Moreover,in labor-intensiveindustriesic
waseasierto acquire and absorbinformationfrom customerswithour a
largestable of educatedengineers.
* Infibmwonfrom custorme.In the presenceof imperfectlytraded
information, knowledge providedby purchasersof an economjs cx-

319
 -ALRACLE

portscan be quiteimportant.This hasbeenshownbyWesrphal,Rhee,

and Pursell(1985)in Koreaand independendycorroboratedin Taiwan,
China,bya numberof researchers(see,for examnple, Gee 1989).Buyers
want low-cost,better-qualityproductsfrommajorsuppliers.To obtain
this, they transmittacit and occasionallyproprietaryknowledgefrom
suppliers. Such knoNvledgetransfers
their other, often OECD-economy,
are morecommonin simple productionsectorssuch as dothing and
fbotwearand with older technologiesthat are eitherin the publicdo-
mainor arenot closelyguarded.
a Knowledgefromreturning naiondls. Drawn partly by the high wages
madepossibleby exports,many residentsof Koreaand Taiwan,China,
trainedabroad,particularlyin newsectorssuchaselectronicsand com-
puting, have returned home to work. Many retuming nationals have
receivededucationin OECD economiesand then workledfor OECD-
economyfirms. 48 Their return hasprovidedsignificanttransferof besr-
practicemnethods. For example,foreign-educated nationalsaccountfor
al/the postgraduatesemployedin the industryof Taiwan,China (Pack
1993a).This source of knowledgeof internationalbest practicesbe-
comesmoreimportantaschangingfactorpricesdictarea shift to more
capital-and technology-intensivesectorsin whichhigher-levelskillsare
neededto unlockknowledgethat maybe embodiedin patents,licenses,
or the useofspecializednontradedequipment.
In both KoreaandTaiwan,China,a considerable
* Domesticrseanrck.
proportionof R&D has beendevotedto improvingexportsand reduc-
ing productioncosrs.There is no evidencethat the gap betweenR&D
and commercial needsfrequentlyseenin the import-substituting
econo-
mieshasbeena problem.

HowExportsand Human
CapitalIncreaseProductivity
technology
Accessto intemationalbest-practice and rapidfiomnarion of
humancapimlsupplement and reinforceoneanothecIt isdoubfl that tie
HPAEscouldchave
madeasprductv useofforeignknowledgeand imported
capitalwithout highlysilled domesticengieers and workers.Convey,
without foreign knowledge it is vey untlikelythat total fiLctorprductivity
gowth wouldhavebeenaslamrt IntensveefFortsby ghly skilledmanagers
and techniciansin individualplants in inward-oienredLain American
economiesto improvetheproductiviLy of eitig capitalsmckwith internal
innovationsdid not generate high- productivitygrowIY49

3Z0
 USING RESOURCES :EFI:d3r

The HPAEs' rapidexportgrowthhasoftengeneratedpositiveinterac-

tionbetweenhumancapital,physicalcapital,and knowledge. The ex-
ternalitiesgeneratedby manufacturedexportsin the high-performing
Asianeconomiesin the formof cheaperand moreeffectiveknowledge
transferswouldhaveundoubtedlybeenlessproductivehad therebeen
fewerskilledworkersto facilitatetheir absorption,while the HPAEs'
rapidincreasein educationlevelsreflectedin part risingprivaterowards
for greater education made possibleby exports and export income.5 0
Thus laborforceskills,flexiblemarketsfor labor,lowdomesticdisror-
tions, and exportincentivesall interactedto promote high ratesof tech-
nologicalupgradingand productivitychange.
We can use the cross-economyregressionframeworkdevelopedin
chapter 1 to look at the impact of trade and indtLstrialpolicies on
growth (seetable6.17). We introducetwo variablesinto the basiccross-
economyregression:the index of openness(as describedabove,the de-
gree to which domestic prices conform to international prices) and
measuresof export performance.51 Openness to the world economy
capturesnot only effidencr gains induced by the need to remainglob-
ally competitivebut alsothe likelyallocativebenefitsof having relative
pricesdose to internationalprices.
Introducing these two trade-relatedvariablessubstantiallyincreases
the explanatorypower of the cross-economyregression.The effect of
opennessis positive.Economieswith limited relativeprice distortions
grewmore rapidly. 5 2 We chooseas our measureof export performance

two indicators-the averageshare of manufacturedexportsin total ex-

ports and the share of manufacruredexportsin GDP.5 3 Manufactured
export performanceis stronglycorrelated(at the 1 percent level)with
high rates of per capita income growth. When the share of manufac-
tured exportsin totl exportsis introduced togetherwith the openness
index, only the latter is significant.Conversely,when the share of man-
ufactured exports in GDP is used, the opennessindex remains positive
and significant.One possibleinterpretationof theseresultsis that a high
concentrationof manufacturedexporcsrelativeto total exports, rather
thanopenness,contributesrelativelymore to productity change in a
cross-economyframework,which would be consistentwith our reason-
ing concerningexport 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-
gressionsandadd someinsightsinto possibleinteractonsbetweenexports

321
 -MIRAC LE

Table6.17 Output,Growth, andInvestment

(dependent variable: rate of growth of real GDPper capita, 1960-85)
NrinnJerefobneations: 92 86 69 97 79
Intercept -0.4237 -0.0124 -0.2324 -0.0055 -0.3943 -
(0.1650) (0.0083) (0.1550) (0.0085) (0.1748)

GDP relativeto 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)

Secondaryenrollmcnt, 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)

AverageinvesrmenrlGDr.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)

Averagemanufacturedexpotslrotal 0.0002 0.0003

exports, 1960-85 (0.0001) (0.0001)

Averac manufacturedexportslGDP. 0.0011 oi

0.001a
1965-85 (0-0003) (0.0004)

AdjustedR2 0.3947 0.4912 - 0.5217. 0.3530 03938

RMSE 0.0159 0.0124 0.0132 0.0147 0.0155

Statisicailysignifieanrax
the 0.01leveL
- Smotisdallysignificantarthe 0.05 level
NotwCoefficientis top number.Standarderroris bottomnumberin parnthees.
Suree.World Bank staffestimtnes-

and human capital.We attempt to explainvariationsacrosseconomiesin

TFP growth rates in terms of relative income, educational attainment (as
measured by the average stock of education per person), openness, and
our measures of manuFactured export perforrnance (see table 6.18). The
education stock variable, while positive, does not explain variatons in TiFP
growth among economies. This is appropriate, since we have measured
TFPgrowth net of human capitals contribution. Openness is consistenTly

322.
 USING RESOUIRtCE£ PT

associatedwith superior TFP performance,controllingfor other variables.

Both indicatorsof export performanceare alsoconsistentlyand positively
correlatedwithhiglherrates of T PPgrowth.
We alsofind some evidenceof.- positiveinteractionbetween the share
of manufacturedexportsin total exports and in national income and the
stock of education. The coefficientof the interaction term between these
two variablesis positivebut not significantat conventionallevels,and dte
export share variablebecomes insignificant.When we consider the con-
tribution of the variablestaken together to explainingthe variationin TIP
growth rates,however,it is statisticallypositive.We conclude chatexport

Table6.18 Determinants Growth,1960489

of TotalFactorProductivity
(dependent variable: rate ofgrowtlh of real GDPper 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)

*GDPrelatve ro US., 1960 -4.8047 -5.5757 ** -2.3509 -2.1562

(1.9771) (2.0637) (2.3225) (2.3008)

Educationalattainmcnt, 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: Educarionalattainment 1960 times 0.0032

manufacturedexports/totalexports, 1960-85 - (0.0026)

Average manufaccured expors/GDP, 1965-85 . . J.0686

(0.0966)
- .. . . m: 0.02 0.0284....................,
Interaction term: Educationalattainment 1960 times 0.0625 . 0.0284
manufactured cexpors(GDP,-1965-85 (0.0269) (0.0201)

AdjursedR2 0.6333 0.6376 0.4507 0.4628

** Smzisically
significantat the 0.01level.
SmziricaHllyiignificantat
the0.05 leveL.
;- XNoeCCdicmt istop number Standard error is bottom numbernpathes.-
SoureWorldBakstaffesdma.---

323
 - r&IAN"MIRACLE

performanceand education interactpositivelr;higher levelsof education

raise the contribution ot manufacruredexport concentration to TFP
growth.5 5 This is consistencwith our hypothesesthat manufacturedex-
port orientation and high labor force skills interact to faclitate the ac-
quisition and masteryof technologywith attendant spillovers.
The evidence from our cross-economyestimates is supported by a
number of recentmicroeconomicstudiesthat attempt to test the link be-
tweenexportsand productivitygrowth. Packand Page(1993)present ev-
idence from Korea and Taiwan, China, that ar the sectoral level rapid
export growth is correlatedwith the pattern of productivitychange; ax-
porting sectorshave higherseoral ratesof rF growth.Wei (1993)uses
city data from China and finds a statisticallysignificantrelationshipbe-
tween exporrgrowthand productivitygrowdi.Perhapsmost compelling,
Aw and Hwang (1993), using firm microeconomicdata from Taiwan,
China, find a statistica[lysignificantrelationship between productivity
differencesamong manufacturingfirms and export orientation.

Contibutedto Growth?
HowHaveTradeandIndustralPolicies

The early admirers of HPAEtrade policy dearly overstatedthe neu-

trality of incentivesbetweendomesticand foreignsalesand understated
the variation across sectors. But the revisionistswho see in the HPAEs'
successevidencethat highly targetedindustrial policy worked overlook
the fac that the calculationson which borh reliedwer- static,that firms
based their decisionsnot only on current levelsof protection and factor
pricesbut on the certain knowledgethat thy would need to competein
the fuiture.Effectiverates of protection may shape short-term tactics but
the long-term strategiesof currendy protected firms are more likely ro
reflecttheir expectationsof the future.
Thus, the emphasis on exportswas decisivefor the entire manufac-
turing sector;promotion of individualsectorswas less important, since
all bur the very dull knew that their turn to export would come sooner
than they would have preferred.Given the widespreadnational under-
standing establishedby government statements and actons in Japan,
Korea,andt aiwan, China, that exportingwasthestandard by whichall
economic actity would be judged, even firms benefitingfrom higher-
than-averagerares of protection in the domesticmarkcetunderstoodthat
in the near future they would be forced to compete in world markets.
Sustainedreduction in import protecton sent a similar messageto pro-

3274
 USING RESOURCES

ducersin the SoutheastAsiannewlyindustrializingeconomies.Govern-

ments were credibly committed to export competition. Exports were
important becausethey ensured that, giventhe HPAEs'high human cap-
ital base, productivitygrowthwould be facilitatedby the improvedabil-
ity to tap international knowledge.
*..

What are the m2in faors that contributed to the HPAEs' superior al-
locationof physicaland human capital to high-yieldinginvestmentsand
their ability to catch up technologically?Mainly,the answerslie in fim-
damentally sound, market-oriented policies. Labor markets were al-
lowed to work Financial markets, although subject to more selective
interventions to allocatecredit, generallyhad low distortions and lim-
ited subsidiescomparedwith other developingeconomies.Import sub-
stitudiorn,although an earlyobjective of public policy in all the HPAS
exceptHong Kong, was quicklyaccompanied by the promotion of ex-
ports and duty-freeadmissionof imports for exporters.The result was
limited differencesbetween international relativeprices and domestic
relative prices in the HPAEs. Market forces and competitive pressures
guided resourcesinto activitiesthat were consistentwith comparative
advantageand, in the case of labor-intensiveexports, laid the founda-
don for learning international best practice and subsequent industrial
upgrading.
Does this mean that selectiveinterventions played no role in East
Asia'ssuperior growth? Our conclusion is that selectiveinterventions
were neither as imporant as their advocatessuggestnor as irrelevantas
their critics contend. All eight economiessucceededin establishingau-
tomaticaccessto creditfiorexporters;this simplecontest wasan impor-
tant component of their export-push strategies. Other directed-credit
programswere lesssuccessfulWhere rheyappear to have resultedin al-
locationof credit to high-yieldingactiviries,mainlyin Japan, Korea,and
Taiwan, China, it was becauseof carefulszieening and monitoring of
projectsand an orientationtoward the privatesector.Where they failed,
governments limited the damage by offering more limited subsidies
than the credit programs in other developingeconomies. Finally,we
have conduded thac industrial policy,with the possibleexception of
Japan, did not alter industrial structure or patterns of productivity
change. The most successful selectiveintervention in the HPAEs-the
commitment to manufactured exports-was also the most general.

3Z5
 x&EXkMIRACLE

Flexiblemarkets, low price distortions,high levelsof humancapital,and

industrial and trade policies that tilted incentivesin favor of exports-
all combined to allow rapid growti of manufactured exports and,
through it, rapid technologicalcatchingup.

Appendix6.1: Testing
theImpactof Industrial
PolicyonProductivity
Change

G OVERNMENTPOLICIESTHAT CORRECT MARKET FAILURES

should give rise to growing total facror productivity in the en-
tire industrial sector,though the resulting pattern ofrFrpgrowth
across subsectors is not easily predictable. Knowledge-basedimperfec-
tions are likely to yield TFP gains, primarily in the subsecaors in which
the intervention occurs. Subsidization of knowledge acquisition in tex-
tile production will mainly benefit other textile firmnsas knowledge seeps
out to them or trainecdengineers go to other textile firms. Some knowl-
edge in textiles may benefit dosely related sectors such as dothing
through production of ftbrics that are better suited to local conditions,
that is, are not obtainable on the world marker.
T-he secroral locus of gains is important for several reasons. First, em-
pirical efforts to relate government interventon to productivity growth
are desirable as aweak restofthe efficts of intervention. Productivitygains
in rhe entire industrial sector or the entire economy attributable to other
factors might mistakenly be ascribedto the intervention, even if it had ac-
tually reduced productivity growth. Undoubtedly, the productivity gains
ftom correcting various market Miluresmay diffse to many sectors, bur
most should have a major impact in the initial sector of intervention.
Empirically, many of the intersectoml interactions will be measured as
intrasectoral at the two-digit level. Improved production engineering in
spinning may improve TFPin weaving. Both, however, are induded in the
Ewo-digitisic textiles branch. Equal diffusion to all sectors seems implau-
sible. While there are individual cxdmples in economic history (Rosenberg
1976), there is no evidence that this is the general prototype. Moreover,
recent research on the pattem of spillovers of R&D in industrial
economies demonstrates that the major beneficiaries are dosely related
sectors, often sectors that would be identical within a two-digit classifica-
tion Uaffe 1986). The nonsubstantiated claim that the spread of external-

326
 USING RESOURRCES

itiesis both largeand diffiuseprovidesan all too easyrefugefor those who

argue for the benefitsof selectiveinterventionand then offer as evidence
effectsthat are plausiblyattributable to the effortsof marketagencs.

HowWeCompared
the RelativeImportance
of Industrial
Subsectors

To comparethe tansformation of the sectoralstructure of manufac-

turing in Asia with the transformation projected from international
norms, we utilizedthe followingequation:

log(VA/GDP)= b0 + b1log(GDrPIOP)-
+ b2(PoP)

where VA,is value added originating in sector i in economyj, GDPis

gross domestic product, and vcl; is total population (Chenery 1960).
The equation attempts to capnure dte influence of demand elasticities
and dte evolution of supply. POPfactors in scale economies, although in
a world of free trade this may not matter.5 6 The equation is purely de-
scriptive and has no normnativecontent Indeed, a given pattern might
indicate that all economies follow some initial leader and undertake
policies to emulate its sectoral evolution. Despite its flaws, the equation
provides benchmarks to determine whether the relative importance of
industrial subsectors in the HPAEsdiffered significandy from the weight-
ings observed in other economies.
For purposes of comparison, we utilized equations estimated by
Syrquin and Chenery (1989) to derive the predicted sectoral shares of
value added in GDP. Table 6.15 presents the results for eight branches of
manufacmring, showing the ratio of actual, VI,to predicted, vp, shares
of gross domestic product for various sectors. Thus a value of 1 repre-
sents confbrmance t the international norm, values less than 1 indicate
that the sector is smaller than predicted, and values greater than I indi-
cate that the sector is larger than predicted.
In general, the results conform to our expectations of these econo-
mies. Three economies known as manufacturing powerhouses have
larger overall manufacturing sectors than international norms based on
the international norms for economies with similar incomes would pre-
dict Hong Kong (1.26), Korea (1.26), and Singapore (1.38). But there
are also several surprises; these are the focus of our narrative. Given the
rough nature of the cross-economy analysis, we have not attempted to
test the statistical significance of deviations fom the norm.

327
 43 -yASrAN
MIRACLE

Efects of IndustrialPoricyon AggregateProductivityGrowth

AggregateTFRin any period can be decomposed by weighting each

secror'slevel of total fitcror productivity, A,0 by the sector's share in
value added, v.r The growth of TFPwill then depend on changesin A,,
and changes in v2..Algebraically,this relationcan be written as:

(6.1) ALogA = `,v,tlogA4,- via-1 logAUr.l

Equation 6.1 gives the growth in A due to the increasein productivity

of existingsectors, logAi.> logAi r- or the growti in th.evalue-added
share of thesesectors, vi,>z-, 1 whose productivity isgrowing.57As will
be seen belowvin table A6. 1, small differencesin die rate of growth of
sectoral TFP,A*i,, lead to large cumulativedifferencesin Ai,
As the text of chapter 6 points out that on the basis of international
comparisonstwo sectorswere generallyoverrepresentedin rhe industrial
sectors of the HPAEs:metal products and machinery,and textilesand ap
parel. Only the former sector was promorecdin HPAEs that used indus-
trial policy.We define two types of sectors, those fiavoredby government
policy, f and those subject to benign neglect, h. The implicit model of
those who believethat industrial poLicyhas had an important effect is
that changes in the deployment of reso' rces were nor based on conven-
tional competitive advantagebut that the governmentcreated competi-
tiveness in new sectors. Sectoral growth was stimulated in capital-or
technology-intensivesectors rather than unskilledlabor-intensive sec-
tors that were more "natural" fur economies such as Koreaor Taiwan,
China, in the 1960s. Moreover,it is implicitly assumed but not empha-

TableA6.1 EffectofSectoral on
Composition
Manufacturing-WideGrowth
ofTFP
TFPgrowh, :TFPSrow*,,
Economy actua vae-ad weighn -adIjal wegkh .
Korea,Rep. oE 1966-85 6.7 6.1
Japan, 1960-79 2.3 1.9
NeArWeightod by value-addedshae datwould haveprvaied ifthe meal products
and macduneysectorhad conformneddt redictpd by txcquatuion estimatedby
Cheneryand Syrquin (1989).-
SurarzPack (1993b).Basedon Kuroda.Jorgenson,and Nishimizu(1985) fbr TFP-
esimaresforJapan.

328
 USING RESOURCESOEUF$

sized that rFPgrowth was greaterin the selectivelypromoted sectorsas a

result of static and dynamic scaleeconomies. 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. Sectorswere begun or grew more rapidly in terms of factor com-
mitment than would have occurredwithout intervention.In terms
of edquation6.1, Vf1 increasesmore than woulc havebeen "normal"
from its initial level,vf Thus, ir is necessaryto demonstratethat
the sectors to which resourcesflowedwere not those that would
have grown on a simplecomparativeadvantagebasis.If, however,
the sectorsthat grew1, 1 > 'M' are characterizedby low wages,or
low capita intensityas measuredby capital-laborratios,this is con-
sistent with growth accordingto comparativeadvantage.
b. Favoredsectorshad higher levelsof TFp,Af> A,,,l. Evenif (a) can
be demonstrated, it is necessaryto show that the governmentcor-
rectly forecastfitture growth of TFPin sectorsor thatit createddif-
ferential rates of TR) growthresultingin * >A*t
c. There was a quantitativelyimportant introduction of new, mod-
ern sectors, v,, which exhibitedhigher levelsof TFPthan existing
ones. In terms of (a), the sectoral structure of output is radically
changed by the growth of modern sectors that are introduced at
high levels of TFP, and this raises average manufacturing-wide
productivity.
Becausethe HPAEsweresuchsuccessfhlexporters,predictingthe impaca
of selectveinterventionson the sectoralpattem of productiondependsto
a considerableextent on one's preferredtheory of internationaltrade-
Ricardian, Hecksher-Ohlin-Samuelson,neotechnology,and so on. To
demonstratethe impactof selectveintervention,it is necessaryto posit a
counterfictualmodel.This is obviouslyexceptionallydifficult,and to be
rigorousrequiresa generalequilibriummodel that in turn has difficultiesof
its own, such as con-crdyspecifyingsectorl production fimctons and
learningpatterns.Again, the "natural"evolutionof an economyalso de-
pends on one's preferredtheory of internationaltrade- Nevertheless,a
weakertest can be applied. If the evolutionof the secaoratstucmre of in-
dustry in the HPAEscan be attributedto comparativeadvantage,it is diffi-
cult to support the viewexpost of a strong impact of industrialpolicy.
Implicit here is that the "naturatl evolutionamong industrialsectors
can be predicted by the factor endowmentstheory of comparativead-

329
 ;M-IRIARACLE

vantage,or Hecksher-Ohlin-Samuelson(Hos).Although large numbers

of addirionalexplanationsof the patternsof cradehave been put forward
in the last two decades,many of them are actemptsro explainintrasector
trade and specializationarnong products that broadly require the same
Factorproportions. HOS theory yieldspredictionsabout the factor con-
tent of exportsand imports, not directlyabout Ehe structure of rotal pro-
duction. However,given the extraordinarygrowthof exports in the Four
Tigersand rheir sizein the GDP, the changingstrucaureof production in
theseeconomieswas decisivelyaffcted by trade patterns. This is lesstrue
of the largereconomies,Indonesia,Japan, Malaysia,and Thailand.

the Pattern of Sectoral

Predicting Growth
The newerliteraturesettingout the meritsofinterventionhas not rested
whetherthe sectoralevolutionof successfilAsian economieshas differed
fiom what would have occurredhad therebeen neutralpolicies.Neverthe-
less,showingthat many instrumenms wereselectivelyemployedis not the
sameas provingdat they had a significantquantitativeimpact on the sec-
toral structureof production. Even'whereeffectiverates of protectiondif-
feramongsectors,the impact on sectoralstructureisnot easilypredictablc.
In generalequilibrium,the movementof resourceshas no monotonicrela-
tion with ERPs (Blackand Taylor1974;Bhagwatiand Srinivasan1979).
For each economy,four simple regressionsare 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, relativeto value added in all manuficturing. The independenr
variables,xP,are the wage per workerat the beginning, zv5,or end of dte
period, we,or value added per workerat the beginning or end of the pe-
riod, vZabor va. La competitive labor markers, the wage per worker
shouldmeasurethe skill intensityof workersin each sector.Valueadded
per worker should reflectboth skill and capiralintensity and is a proxy
for technologicalcomplexity.Although valueadded per worker and the
wagerate are likelyto be correlared,both are used, as there is a possibil-
ity of divergence. Governments also may have employed one or the
other as indicators in targeting sectors.Wageper worker may be trans-
lated as "goodjobs,' while value added per worker is a natural measure
of "high productivity."

330
 USING RESOURCES 'E!jl

Consider the movement from inward orientation to more neutral in-

centives.The static Hecksher-Ohlin model predictsthat for low-income
economies sectorsthat exhibit low physicaland human capital-laborra-
tios by international standards are the natural candidates for growth.
Hence, the prediction is that dvu/zdwband dvl.Idvab c 0. While many
newer theories of trade, including neotechnologyand those emphasiz-
ing increasing returns subordinate differenccsin factor endowments.
they are more likelyto describe intra-industry patterns of tradc among
industrial economiesthan the evolutionof the sectoralstructure of pro-
duction of developingeconomies whoseproduction structure islikelyto
reflect differencesin factor proportions.
If interventionhas been of great importance in promoting the expan-
sion of capital- or knowledge-intensivesectorsand, eventually,affecting
tradepatterns, the data should show dvjldwb>0 and/or that do1.Jdva 6>
0. An alternatetest of the impact ofselectivepromotion is to testwhether
the sectors that expandedexhibited high capitalintensityor wagesat the
end of the period. A result fvoring the interventionistview would be
ehi-iIdw or dzIlIdva > 0; the sectors that grew most exhibited a high.
capitalor high-wageintensity at the end of the period.
We have performed these tests on each of the Four Tigers plusJapan,
wsing two-digit lSIC classificationsof industrial sectors. Table A6.2
shows the sign of the regressioncoefficientand its significancefor each
period for each economy. The years chosen reflecta distillation of rhe
existing literature for each economy on significant intervals during
which intervention was practiced. It is not possiblein all casesto utilize
the ideal interval. In Japan, for example, data from before 1953 vary
considerablyin definition from those after 1953.
These are clearlyvery simple tests of the determinants of structural
changes in production but capturc the spirit of the opposingviews. The
statistical results must be employedcarefully.Even if it is slhownthat
there was a significanrnegativecorrelation between 9i and w1 , support-
ing the HOS view,some of the individualsectorsthat grewmost may still
have been high-wagesectors. Thus, we will examine some individual
sectors as well as report the statistical results.
Korea.For most periods, the signs of the wage and value-addedvari-
ables in Koreaare negative.For the period 1973-80, the one character-
ized by promotion of the heavy and chemical industries, the negative
coefficientsof 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

Explaining
TableA6.2 Signsof Regressions
Changein Value-AddedShareof Sectors
Economy Period Wu WE VB VE
IHengKong 1973-80 - + - +
1980-88 + + + +
1973-88 - + - +

Japan 1953-63
1963-73 + 4- + +
1973-80 - + - +
1980-89 + + + +

Korea,
Rep.of 1968-73 - +
1973-80 - - _ :
1980-S8 - -
1973-88 - --- -

Singiporre 1969-73 - - - +
1980-89 + + + -;
1973-89 + -+- + +

Taiwan, China 1966-86 - - -

+ and - signs re sign of cDdenciLent

- Signficncatthe0.10IeveL
' Significnurartie 0.05 klveL
Significantat tie 0.01 leveL
NbOwtWSmZa Wage per emPIoyeeac the b%eing (end) of heperiod. VB = Value
addedper workr at the bginning (end)of the period.
SouxerPack (1993bl)

period. This implies that at the two-digit level the relative size of more
labor-intensive seccorsincreased during each period On the face of it this
result is surprising. Quite apart from policies of selective promotion, this
was a period in which both labor skills and the capital stock were grow-
ing. One would have expected, according to the Rybcyznski theorem,
that the labor-intensive sectors would have declined in importance.
Part of the explanation for the relative growth of labor-intensive sec-
tors despite the promotion of ostensibly capital-intensive suzbsectorsis
that the major branches that grew most rapidly in these years, and in-
deed over the entire period 1968-88, were machinery, electrical ma-
chinery, and metals-all sectors exhibiting somewhat less-than-average
labor productivity at both the beginning and end of the period- Con-
versely,some of the declining sectors induding iron and steel, industrial

332
 USING RESOURCES:E'FPtC ,

chemicals,and paper and paper producrs, were sectors with greater-

than-averageu and va1, A more completeexplanationof the contnu-
ing growthof labor-intensivesectorsisthat they experiencedmore rapid
growthin TFr, and the cost-reducingeffectof productivitychangestim-
ulated expansionof output by more than the output growth induced in
capital-intensivesectors by the decreasing relative cost of capital and
skills due to accumulationand subsidies.
In sum, the regressionsfor Korea indicate that: (a) the more labor-
intensivesectorsmaintained their relativeposition or grew, (b) many of
the promoted sectors themselveswere not capital intensiveat the tWo-
digit level;and (c) subsidiesto the cost of capital may nor have been suf-
ficient to overcome the ongoing differentialsin TFP growth among
sectors.Moving away from the anecdotesabout individualchaebol,the
quantitative importance of governmentintervention to alter the struc-
ture of production is not confinned at the sectorallevel.
Japan. The regressionresultsshown forJapan in table A6.2 have no
significant coefficients.At leasta simple versionof HOS does not work,
probably nor surprisinglyasJapan by the earliest year considereain the
table, 1953,was sufficiendyadvax±ced so that intra-industryrather than
interindustry trade would have become an important determinant of
the sectoral producton structure.Scale economiesor a neotechnology
explanationof production and trade patterns are more useful at this
stage of development. Going back to 1945, which the data do not per-
mit, would not change this picture much insofir as this earlier period
was one of reconstuction, in some casesof firly advancedsectorsthat
had brkenquite important before and during World War IL A simple
HOS story would be interesting to test for the period from the Meiji
restorationonwards.Once an economyis in a world of technology-and
scale-economy-baseddevelopment, the requisite conditions exist to
allowalmostany assertionto be made about the determinantsof the sec-
toral evolution of production.
HongKongandSingapore.The cwo island economies revealdiffcrent
patterns of development. In Hong Kong there is no relationship be-
tween our explanatory variables and patterns of production. In this
sense it is similar to Japan but derived fiom different causes.The slow
growth of capitalstock per workerin Hong Kon& partly reflectinglarge
immigration, miliated in favor of a slowerchange 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 Singaporefor the period 1980-89, outpur grew more rapidly in

those sectors that at the beginning and end of the period exhibited
grearer value added per worker, supporting the view that the Singa.-
porean authoritiesintervened to encourageincreasinglycapital-intensive
development.Given the rapid growth in the capital-laborratio in Singa-
pore, this result also conforms to the Rybcynskiprcdictions.
Taiwan, China.Like the results for Korea, the signs of the wage and
value-added variablesfor Taiwan, China, are negative, although insig-
nificant.Again, this implies that the relativesizeof more labor-intensive
sectorsincreasedduring each period, albeit weakly,at the two-digit level.
Growvthwas market conforming; government intervention did litde to
alter the structure of production at the sectoral level.

The Evoluton
of Sectors:MarketConformning
or Not?

The precedingeflorr to cdifferentiatebesweena comparative-advantage-

based evolution of industrial structure versus one resulting fiom inter-
vention is notable mainly for a number of negativeresults.In particular
for Korea,during, 1973-80, the years of greatestintervention, the most
rapid growth in sectoralshares of value-addedoccurred in lower wage or
lower value-added per worker sectors. Despite government intentions,
the manufacturingsector may have evolvedroughly in accord with neo-
classicalexpectations. It is also possiblethat the government undertook
measuressuch as the provision of social overhead capitaland technical
servicessuch as the Korean Institute for Scienceand Technology,which
facilitatedthe 'neodassical' tansformation.
Singapore, in contrast to Korea, confirms the casual empiricism of
many observers,as well as nurnerous accounts of rhe goals of govern-
menr policy that industrial growth emphasizedcapital- and knowledge-
intensivesectors. The sectorsthat grew most exhibitedhigher wagesand
labor productivity than the laggards.There is no evidence of a factor-
proportions-based expansion. In contrast, the evolution of sectors in
Japan and Hong Kong does not permit a simple view of its determi-
nants. This is nor surprising for Japan given that HOS is designed to pre-
dict the determinants of trade patterns rather than production, and
trade constitutes a relatvely small share ofJapanese GDP. In tie postwar
period, certainly after 1953 when our data begin, Japanese trade par-
terns are more likelyto have been dominated by technologicaldevelop-
ment and scale economies. The failure of Hong Kong to conform to

334
 USING RESOURCfSts$r.

neoclassical
comparativeadvantageis at firstglancesurprising,given hie
enormousimportanceof its trade. Therewas, however,relativelylitde
variationin the dependentvariable,the changein the sectoralsharesof
valueadded.

Patternsof SectoralProducfivit
Growth
This sectionconsiderstwo issues,namely,the patternofsectoralpro-
ductivity growth and the impact of the changing composition of pro-
ductionon the manufacmuring-wide rateof growthof productivity.
The
issuehere is whethergiventhe observedvaluesof sectoralTFr growth,
A*i,the unusualevolutionof the sectoralpatternof productionin some
of the economieshad a significantquantitativeimpacton overallTFP
growth.
Going back ro equation6.1, the analysisfocuseson whether the
changingstrucure of vj, giventhe pattern.of A*j, contributedto more
rapid growth. The value of A* is taken to be exogenous.Of the
economiesconsideredhere,detailedsectoralgrowthratesofTFPare only
availablefor Japanand Korea(surprisingly, sectoralTFPgrowthratesare
nor availablefor Hong Kongor Singapore),but theseare two of the
three economiesin which activistindustrialpolicieswere employed
mostconsistendyto achieveproductivity-based cardtingup.
Korea.Calculationsof A*ihavebeen carriedout for a varictyof peri-
odsby Dollarand Sokoloff(1990)and Lim (1991).What is mnost strik-
ingis the high absolutevaluesof.At i in mostsectors.Fromthesestudies
a numberof patternscan be identifiedthat are broadlyconsistentwith
one another.Thereis no simplecorrelationbetweenpromotedsectors
andTFPgrowth.Wasselectivepromotionimportantin the sensethat re-
sourceswere shiftedtowardsectorsin which TFP growthwas (exoge-
nously)high, increasingA*for the entireindustrialsector?To answer
this we first calculatethe manufacturing-wide growthof A*using the
observedvalue-addedsharesat the end of the period1966to 1985,uti-
lizingthe A> valuescalculatedby Lin. We then recalculatethe value-
added weightsassumingthat the sectorsin isic 38 had been at their
predictedvalue,VP,on the basisof intemationalnormsasa shareof sec-
torwidevalue-addedand,reassigningthe residualvalue added to all
other sectors equally. Thus, instead of accounting for 38 percent of
manufacturingvalueadded,sector38 accountsfor 14 percentin this
59
calculation.

335
 SIRANL MIRACLE

The result is shown in row 1 in table A61. The actual secrorwide

growth rate, A* was 6.7 percent, the recalculatedone 6.1 during a period
in which the growth rate of manufacturing value added was 17.5 per-
cent per annum. Even if the "excess"growth of the MPM sector may have
been attributable to selectiveintervention, its rare ofm'FPgrowth was not
sufficientlyabove that of other sectors to make a large contribution to
overallsectoralTFE growth. The major reasonfor Korca'smanufacturing
success lay in high individual values of A;. for most secrors in most
periods.
Japan.As in the case of Korea, a considerableliterature argues that
Japan's selective industrial policy was critical for its industrial success.
Nevertheless,many Japaneseanalysts are skeptical of its effects,as are
many analystsfrom other economies.60The existingliteramureindicates
that many of the criticalacts of the Japanesegovernment with rcspectro
industrial policy occurred in the ten years followingWorld War II. In-
deed, a verv good case can be made that many of the major policiesoc-
curred in the period from 1868 to 1920, but theseare more difficultto
analyzewith systematicquantitative evidence rather than anecdotally.61
Chemicals and the metalworking machinery complex have unusually
good performance. These industries are those thar observers usually
point to as having receivedsignificantgovernmenr support, induding
efforts to stimulate productivitygrowth.
A calculation, similar to that for Korea, to determine the impact on
A*for the entire manufacturing sector of the unusual size of sector 38
can be done for Japan. Table A6.1 shows rhe actual growth of A* for
1960-79 and the estimated growth, had the sectors constituting 38
been at the international norm, 24 percent of manufacturing value
added rather than 41 percent in 1979. The value of A' would have de-
dined from 2.3 to 1.9, a relativelysmall decreasegiven the growth rate
of manufacturing value added of 8.7 percent per annum. If selective
promotion is interpreted to mean that only a few sectors rather than all
are aided,Japan'sgrowth rate would have been only slighdy slowerwith-
out an industrial policy.It is possiblethat industrial policy significantly
increasedthe values of Ai in sectors other than those promoted due to
one or more externalities.While this may be the case, it is likely that
most externalities occur within individual sectors or in closelyrelated
ones. Given the many branchesconstituting the metal products and ma-
chinery sector, the externalitiesshoulldhaviebeen revealedin its own TFP
value.

336
 USING RESOURCES. -FFYCZR

Appendix
6.2: Testsof the Relationship
between
TFPChangeandTradePolicies
T HE ECONOMY-SPECIFICTFP GROWTH RATESCALCULATEDIN
chapter 1 allowus to make an alternativetest of the relationship
between tradeand industrial policy and growth. We specifya
regressionof the followingform:

(6.3) TFPG= f(RGDfI PRIM,SEC,OPEN, rXPTh)

where rFPGis the estimatedrate of total Ecror productivitygrowthwith

physicalcapital,humancapital,and laborasinputs; RGDP is GDP relative
to the United Statesin 1960;PRIM and SECare the primaryand secondary
school enrollmentrates; OE'ENis the index of relariveprice conforrmity;
and EXPTis our measureof manufacturedexportperformance.
TableA6.3 summarizesour attempts to tcst rhc relationshipbetween
these variablesand rates of TFP change derived from our production
fiunction estimates.The regressionscontrolfor other sourcesof variation
in TFP by introducing a -variablethat has been shown in previouswork
to be significandy correlated widt TFP growdt-relative income level-

TableA6S3Regrssionof TFPGrowth,1960-89
(&tpenkt variable:totalfactorproductivitygrowth, 19610-89)
Na&refdsa . - 67 67 51 St 69 53 5i

- inrn *.634 *419 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

iu.
wwPu7 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) 54),
M0 (0.50

M)uud IV O 0R5579J CO50 .604 0UA Q4457 CAW67 - e70 03673

KNG,'tr. Mba. Sma_rd effrits

N -- >-- ff- : o.dtr

::m - - . n-mrn
. - - . - : - . : - -- -

337
 Fr~~X MI.RACLE

as a control(DeLongand Summers1991;Mankiw,Romer,and Weil

1992; Pack and Page 1993). The coefficienton relativeincome
(RGDP60)is generallyinterpretedas summarizingthe productivity(TFP)
gainsrealizedasa consequence ofmovingfromlowerto highertechno-
logicallevels.This catching up is the componentof TFP,change made
possibleby the diflerencebetweenaveragepracticesand best practices.
As we nored in discussingthe cross-economyregressions,ir rime-series
analysesof aggregatecoralFactorproductivitygrowth,estimatesof TFP
consistof two components-intraseccoralTFP growthand the impactof
factor reallocationamong sectors.Thus, the shift variableRGDP60 may
capture the reallocationeffeccof structuralchangeon TFP in additionto
productivity-basedcatching up.
The use of schoolenrollmentrates is an attempt to test for externali-
ties in human capicalformation. In our growth-accountingfiurnework,
the contriburionof human capitalto output should be reflectedin the
elastcity of output with respectto human capital.We have argued that
primary education in particular can have growth-enhancingexternali-
ties through the better ability to use and mastertechnology.These ex-
ternalitiesmaynot be capturedby rhe human capitalvariableand would
then be included in the estimatesof TFP change.We attempt a similar
rest for secondary education.
Openness to the world economyis another imporcantcan.dlidate for
explainingrapid TFPgrowth. The theoreticalcasefor this viewrests not
only on allocativeefficiencybur also onlextemalitiesassoiated with
tradingactivitiesand on "X-efficiencyfgainsfrom creatinga morecom-
petitive environment for domestic industry. Severalprevious studies
havefound such effects,induding Harrison (1991),Dollar (1991),and
Thomas and Wang (1992). We test one definiton of openness,Dollar's
(1990) measureof the correspondencebetween domesticand interna-
tional relatrive
prices.
TableA6.3 presentsour restsof the relationshipberweenexportsand
productivitychange.We chooseas our measureof export performnce
two indicators-the averageshare of manufacturedexportsin toWtle-
ports and the share of manufacturedexportsin GDP. 62 When the share

of manufa-tured exports to cotal exportsis introduced into the basic

specification,it ispositiveand highly significant(at the 0.01 level).The
share of manufictured exportsto GDP is significantat the 0.05 levelin
the basic spedfication; however,when the degreeof opennessor the
HPAEdummy is introduced,manufacruredexporcsdrop out.

338
 USING RESOURC

TableA6A AverageAnnualRatesof Growthof the LaborForcebySector

Gnnupasdperkvl Atcunm Indury Serikes Thoe

Industrialeconomie?
1880-1990 0.4 2.1 2.1 1.2
DDcvropingeconomicsb
1960-70 1.1 3.8 3.9 2.0
a. Developing economicsincludethecountiesofAricawiththe exceptionof Soudh
Africr LatinAmeica;Asiawith die exceptionof China,the DenocmaicRepublicof
KoDrea,Japn; Mongoli:.andViet Nan; and Oceaniawiththe exceptionof Australia
and NewZealand.
b. Industrialeconomiesindude Australia,Cnada. Japan.NewvZeland. SouthAfuic,
the UnitedScrcs, and countriesinXYVesrrmn
Europe.
- iwrc Squire(1981).

TableASS FinancialLiberalizaionin the East Asian Economies

IN Dl.-prsn .fflaw-dd
a~~~~J m - )Isrssien -In IVA4ime-wgnla 7

Jfrg fly 6kTn

19* sop innrvmewidn Tudiduifr r&d
hum 6W awi 9IS-Tbeerrcrk /ort..rDatEd mknatu
39Q, he irans. fer
5
-i nenuis dtpma .iln - undo th eoniumNMmforfrwva.x. Yeq ib cadmftgeablkxd liauibsnbe fRn HlU5et to Hls 1in litA
dimisonJibe achuop wlln6 n&kr inarusul bsdznmgpam. 19P3I:1iedfeecI odue t
murafltk andscowinnisldu&
A-dairt$EI byvkita&-s Aldwien-cking kn.utim we aesnu -r pin Hanl og a y,'% eiW-Balimcomuimnu cirye me-
bawwnjtunkt ?tlautwrr ' rihadjrnnnat - o inte fini ecano.
St bawithdie sin rdIasEqking,ts G et
nrmaBen Us tanRg Uqunc d ITden imoatnspipsad qBiz.Wl.tobmU mnndma%imaams nid
by do m. neei,i-g bush sad dwpntncdadmifillsiA ,;Mstnie pidhoe hT, dti-nirnma AiedmdnpduWed
inyharlw. iumadiaitsdwaWnim i-flOAbbidIwbbnS annl.
ISMNWIh'NZd kshn*qjuio.n
rwniy ILkand Uno" ikerdflq'ek.it amaiS otnoMpUifMttI Ad f& ksduapnd se
timenateAmhek tug Kn TuingC.o i(DtC4md
Qinkiua hEn.r. 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 W mn k
!tmoba..It I.L . bkh . IZRninddasSihcF'un'au-rmk dsq,. Eri.b dewlukInde= ionm ,a d
eymm.totx is ,kni-
nsw&ayrndeumciwa vith dumt. .syaniydsngd* FunnUmbw 195* Etsd fdsetl9dwSuikmadl Fume
swjydryc-nnpa atn Usn RadA Reuian LiCe Oft C:Gti atm
- ltea CRn
Onindelesun. iehuuansiiYwEmk aa
ben Theen,ttiian
ints Lnb.aid Dlr At danethire Anie num. Cod eFda Edung cmetd puidestpeindn
w-binnlnge1yforbamla 1a lmd
enom (B -fitovr 4apter ut 6ih-k-m
25d5f-imdtrnim. r in pin hi &g.

;nrk"du IMtlnsnemutermsicdsmanitnl IX6-9& Fad trsrininen pe of :CntwmmI

baokcahmn .3hiselnkid s
anchahe bends iuqm.
depCSfbUIadruttnlub dinm. thituNialneutinia tim b innS dthw ntto
,i 1tu- enag tcru
-- : luta~w
Lrmnamo(iutehab. uw ledsnso sakvi - bcnriitss&s.-. ' awstdasd -nqtsin'nionvi_en-
'wedaspheolaiu. Cmavitccilnp S aujxMEn;MCmmi 'w u SW:Smtuib aetbl"i a] 0 line in na. inplnst nalFuiwtnats--
ahh&C ,, -baksiiar.-i.
u capiaIcstwmrlnsfiS. ' tn a eneaominmhemsy tpedenn
ainmn dhbsufl4hiWntoteoI-
dw p- l-dia 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 widfir4ig AMS,EstdghAdNWwr-ducmimmi-T
INnamDU finnkfj,rlnsnntkst5ankmn
(HI) mki&
--- trdnefidlkiusndtelkisteel b6iIss Fd6ER.PwdioInwili. O )etanassp aqu dRtel puMladan
- m'- idadwiegnn nmppoadi ctadseopintrtCD4) pplic ,dbmk&mxli muse hd pwrdfBtk.-
'-- ~ - , Lnalnbemi- -hn -- ,l,ssi dzctc N.o f nedmw
w-i eye Laureleng -
:: - R - - - (^~~~~~~~ ~~ ~~~~~dcwys
tna..sTsdahgaingaina plscigsghe,

p:Xd
im. _ 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 -fnanq tflnwu&gku .f?e.emw ruin JImeqIad ,fiecu M."

_y IVRf= mya
andns tse

Jq_ Lading maawarmneguled bn lleal- flMtAlhyecd cmey.C 6lr ~n . bfldJUb-& EarnWerd elek
- The MinyfFlhlna witi iwn bue is

Idewldeindroddidiol minks.. wIdch by IPRI wee ulpalin anall 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 e
SecuidmitedadEhnd 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
hsr
lotion ofin rm.m Feed bawur IJ uu hu
TniT by Ereiga.
wesneJ flfaRnictiuon.sinurinral iItStnMU.ddl pnd udea son
m
munibehinmrbskalkudIl b mr- hakbd eiclLd n.UhULi Mea dia ndwepirle iaJ
hmaClg7talwdmailunlmeuhaa lWtSinI9 NWiedlum boan.. /misarkar(pulickmmdAutiuL inada
hig CWaet maeberamtnls(I 91 btw6n(Seg)wem endvn aul I9eat*Gw nubeaeeet a 191: EatbiMlu dthellWd Dq* In
Ml (bean.
'ngleanEcesnlldCa BAten nrlca andt u .n Coonamtidenaeda.ceqnr prcedes
eF a'el
amarke ra(nS).1 e- Inf: Batik Iti
l esgi ro
te ng ISW&T mmuyiuco nwaed tN cw lon& I...caaaar.L

queamalAa4d dqeie rgetrorminga sudiarlsnlf iea incs fi.rm Feuip atitia SrngpnErd Itnh beanit anenimpat herniae law
t9t Rea - nemyskLst lhew maianheniphntheTLy.Snk iftiUenUIStrd.
ncmlad. Ilntve rum e ExdaKJ.wafuned wme
m I ag u lanird
r ed

- s thn tint JrpaI nraipal. -tan. Iwkueai p4mt

Sesehreem
(CC)mkask mmneiL
.trdin thteUs Sequlds
Edia nse ered. Dan usMes
rf nd intne arn
almond tv psiidpau inh m
Ca,ndial hajneaaMrt
AtktSmckindfis ma wading
rfTey ru Minr
Ms-Tkofiac
5ketmaut. un.mj
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 lRni6ill tt mt s-
ttacauac rmcatlain1973. 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 cheOQeeOf 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. eagn s up OTCtae
Ieefarkinu i
amkte rits UlflnS a bnY IWSd Xlirlilnab
19-* Minfimumed mem&nn but fwaegunla (enenere r`fieigs lt M*bUn ieUd ahea my an- cpened e nsi ta pe-
h ecago
won incnhAwL Isa btd.ia iama Bn IS is I te . PIa
po4u I bgewen uhefhbne i lee na
mee.tr nae niuluak hding 64 in 199li. Deei. p incndtudi- banldlagnrld nm.nkn-aae la shiauia
n enr CD.
uWhi.Ll3abnUnseA if cngrefAmuiei6c%ukhleaditccank% )mPAwdl kgwidaehID cace?
JW mhlhiscynrf inae(MOljwapeniassal iinpne
-r msigid.sunfreinrdepode tr.nbuineband erpsdamenprc- lamemat lnaancnmpudics ate pn&i6l g;dd rma pe.tionnsftWBk
tna.2 Noa atw bankas ma any &W scaaelckshanSipuaeauadmaqqwd
-~ ~~ tnt Mm bekake .ukflnti-
b he.deeetw (NR loaig
=
n
Dil insmialen (NOR) imnm niem
eiq
qchncAhelFlsmeded
s
ewiInen FIksaJ9c.
JNIPouaiaac
a
licthcc
fmlap
hey.mkannrioait
ipe inpdoeiii
atcljeeitk arW
rnd
a p.*hmfia
-

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 vnXd

r-oel
wu: end vmleuneetaStcdad maini. IA9 rlhid iFa_balloerd M eAs
- m gd t .a CD, :r.'o-Sf.th,

340 _:,_

340 . :. - . .-
 USING RESOURCESSO

aM.opta 1971:hmnaei aianhrfind&pdewicf Mi.hudantf194hmoaku ~iapig fl%-KmJus"s

Li cIS&E if LRid hJAiog&t-eonpickalyddsdn
nmmzho flmyw,g. hiajlty 6cua haiLbmab Howemi.dndc w nab&hd
NW p &hf. Idga Jtaincfbhulu prcaci, of
ffkb loUtes aam.csa11icMFumnl bmil kw Inured aainsO 50, &Ldqwia p-cer. wma- dqiWmnc anvocrUe
i maumpnamu
9 andaural of
Inkik&p.sitmuinr am pr uawiru theyhawha. napmdto insni rdubinmnbedq buil
(End. ~~~~~~~bn~nchh
w6iiu ne&lernn rmaota
nmnduat jMs--aoVmei.fUanaic 6i.& mnd0TC

nhh.rCmbsponlosasd&4 1374-KNew Anaidelimuamsn ngmitoinhclupot..49 prenm 139W8h A kaC.1'dohia Wiath

gasiddnmenogle
57&MA wnianuadu
Laekbaing CD%
.aac ouhAseleiA mnd
hr&r aua0cmdrm n Ewa mu' KManddhcfamdanX AnbudgesZudiddspli
sit
alibe,tindaweL
drpca aaqumcninamn oa sriau dfitpakiLm rordineindng Ktndialaw
Flame.ma
h1 Bwm.l melube. 19~~~~~IN-%O,SshSab,rdan
cff-,bca 1398 mALina%ar
Finadi lasituhm.Ar-~cwjw
plfthklAngSram manroe
tifan 1 UL
biWngto CLCruStbaku6, poad rtheakingAadZ5l'ManlFonmanceGimpma
Act
p wdn f SoL:i..tm,.lacu
.1M. lc&g guidolina
-)flCrmmrulbmk-alminhunced Bar iuqMaCCaKrhia. andactig awtdiinmsu.darda&afiaaCd,hainnd it
LadIng1atehrodcednmanlhe. 1:&ibdla c ue,pWcctc .aitardind rnamisffi.aeWupcidg anLinac
199-.(snt ISZ211lutternilaprnn
at liamfln car- h Sfiogper.l 'n..-
2dIcrjta
/>sindl -taupary.
3 aflhlsc so pwpawapi.Mcnd
il

phmuLSac
cadlgeuzie
- .
aiiua
~~~~~~~~~~~17r0
asa hraiaglirnbna frran ro sliad6 INtn SitdiEuaqrlam )ni cceaDTlc±sdbi

t^ii0ilbding aWeCD* - dhr~~~~~.M aJeimonqefi and -gi vm&g-;ia:-r- tnO -w

Sk~~ar ii7kCamplcc~fi-Nrad libaclin- IWI
9Aanmoderelmcnminoal IAf ai. DAr Miadot-&lai 15M.kEorkWhh MqntuAki& iyaVSaypm
uam.Bankouitn 7 aa J&odeurmia. ofbanWigumwawh bklbal p-x J"ifSupwpurS&Exhap or AlonigwmAt
canal bau'ngAocticm. mope-bnou-

kg fitcennuccd*mow dqaodtand aintuudrlpmz 9laUhlmI

I= dwSIngar pveI. maicsncfainea dut baki mow mimutWit.
1
IcadhguotaJimam Rial aftoe Ads=ODi. AdLs I6IxMA a a!,1niidloocuiy E aFeat BabigAauet oidha.apiiada
aiuhnerma,tin.d-
etehaFownand watplevalldratma- nEabat boideg maskts. Hi limema MqencdcwlS1VpmrCG&1dEn Bu ala r1Mnms 10 .liqddkrruaz..h*j ba-
rlod by 197K in shyumgarn to.aADwiplbikano dbze.pupcdatnimnmsd eulamts
adautraiomtea.opm.perJlomauidlnid

--_iona
tSlgt had..rbankiang lacun
lnnard i Mlaa1fi
and Stock a c..mnnpecl

banlh Fontoeadat,i oaptiax flflSWtFocf dooic buk par- lanti in apitban orfbeeW hniari c L
mir"uiomefluainidey*riwai
pIMatler4l oatior
Jfmgn t inlocalboaokomps.Poretbasik
madlbomgarvhow, nuk(hd
tacnve tehdi,19 poite sula,

re... aim. 1950(Lirbclsd C at.

Ckit AttaInhit 1 ban mw IQiEahlsbehrx T.iwao19t D.SsxhiS ad Exitnu b Crwnuiona
194 lAI,Jnaamsaimds.adini- fmnddradcpitbd Suit.tud.a ucpbcnm.pnvreaepital ni.arktpeaaks.
snladugawa mnd:auccl. d a e ow t mccoy IWi tfmbl&d aedqmiIa rpnce copses .
as. da
B ncc.dbaglachain I*oaow r.vlmrrial
aJbn pad- nTh.alc RappmhscMakanr lflnlSboib rh tlni ucccmopatuy Zn6 -
lylSCosa&lrf hanktulnuadM permttdartoiWtid.scutlq6a1aw mlet11CP. CDaamnW.4Atwit IN*-JflA.."dclSriabTrmanatdn I.A. t

napinlene maewi*inanmnl .ppmnL Madaadnsmcuic- awknntctnwinal. inauacsmm. acpdwMaaenrmandb.tnauborahmIthan ainp.

bank-adaalnimndinapt. Rrmbrd aiomfiurfoaig bade.malm)ined 1956 Craed .W.&bovebadaig Raiwd banbinglausnD puohlbktcrdiemn Iiewma

delnt an Piamcwrugahndimun long-mmlendin hwodam. N Cepnui c.rtahn kdaiuAwld maas aada

hidinpia - naaapmnicaanal
InacknEacsof ~~~~~mn&,uno
ltmcl napdcnacuaad miumntimarrff

*drpltionsadlarcog.it
iSa
- meien nof oiunp anl
ioadligrolumbrod. mW

73&11J Jwifletnnrymecti6pkbocWitnca CdllhOGtoMldkrhgaenIf1 113qsskd.p ITTmanrortoa ltpleamatedo

ate.
,nau.rin. bmI ficoInuSpnnz hid - ew .dn startd buioandoin 197tkdua. gto
f/m"an purcm fiadand a ansby~
iatyiuedbyuatybw lNtPcekd 1562habonglwaIn - pw"-Ph. - FanceCsNp,

kb;-itpebid. rofihltDam bat11strada EnltaqccftluiIp trdnStds ticrmuAdpmdu

*lirinn reTolwet uclikehagnharm.aidndunl.

; r-
caabl Lsa
. J^*LdC> /Is*hrbeglarff@6f:-_IT
Rtrctcioa;fr c
-n
/fftSBpS:r3
ti.dindopMMimdaanigf
I.Ab awidr uatrumap E uubcd d.c trer;Anbc
II"urewse 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
nd iea acsaaahi-
-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
WDmillion bOIL am padh clicducaon InF
ISthSaundsandSabaa 0am.d cmasal wnit

- ,9,d.i4 ,,em ledwiard --

:~ . - -*,-*- . : .
 5jE
eMRSCINMRALE

Notes
1. See Greenwaldand Stiglimz(1988) on the effecrof 7. The costs do not derivesolelyfrom the misalloca-
coordinationfiailuresin labor markets. dionof labor impliedby gaps in marginalproductamong
workcrswith thc same human capitalendowments.The
2. Altematively,conscantrealwagescouldbe due to a re- costs of such static misallocations rarelyexceed 1-2 per-
duaion in surpluslabor(Lewis1954; Ranisand Fei 1961). cent of GNP. Dynamiccosts can run much higherwhen,
for example.employmentin enterprisesis excessive(im-
3. A populationwilldoublein 87 yearswhen it isgrowv- plyinga wedgebetweenthe wageand the marginalprod-
ing at an annual rateof 0.8 percennwhen it is growingat uct of labor) and the subsidiesrequired to keep firms
3.2 percent a populationdoubles in only 22.5 years.See afloat divert scarcesavingsfrom productiveinvestments.
Squire (1981) for a historicalcomparisonof population Sec Gelb, Knight, and Sabor(1991).
and laborforcegrowthin developingcountriesin the cur-
renly high-incomecountriesin the nineteenthcentury. 8. A labor market is segmentedwhen some workers
reccivehigher wages than others with the sarmelevel of
4. For the period 1989-2000 population isesrimated human capital simply by virtue of their sector of
to growat an annual rate of 0.8 percent in Hong Kong, employment.
03 in Japan, 0.9 in Korea, 1.2 in Singapore,and 1.4 in
Thailand.Indonesiaand Malaysiaare forecastto growat 9. Clearly,a high share of the public sector in wage
higherratesof 1.6and 2.3 percent,respectively.The share employmentgrowth is not definitiveevidenceof a rela-
of the populationlessthan fourteenyeawold was31 per- tivelyhigh share of surpluslabor in the "modemsector.'
cent, circa 1900, in the currentlyhigh-incomecountries. The publicsectorcould be findingproductivejobs for all
In 1989 it was30 percentin EastAsiaand 36 percent in its employees.
LatinAmerica;in 1986it was46 percent in Sub-Saharan
Africa. 10. In more than one-third of the economies,private
sctor employrnentactuallydecreased,implying that the
5. Increasesin the rechnologicalcomplexityof non- public sector accounted for all the observedincreasein
agriculturalproductionand in the sizeof establishments total wageemployment.
bring greater divisionof labor and, hence, increasesin
skilldifferentiationat thesame time that greateravailabil- 11. While there is no doubt about the roleof govern-
ity of educationalopportunitiesincreasesthe supply of meat in suppressingunions in Koreaand Singapore,in
cducatedworkers. Taiwan, China, there is some questionwhethergovern-
ment action, for example, prohibitingthe ChineseFeder-
6. Labor market efficiencyis assessedby comparing ation of Labor from engagingin collectiveaction, was
marginalproduczivitiesin losingand receivingactivities. decisive.Market forces,and in particularthe sizedistribu-
Evidenceof changesin laborallocationsuggeststhat indi- tion of firms,alsohelp explainwhyso fe%w workersin Tai-
vidualschooseamongalremativeoccupations,industries, wan, China, are union members.See Fields (1992) and
and geogaphic areasas though they had a fine regardfor chapter 4.
marginalcosts and benefits.Ihough behaviorconsistenrt
with rationaldecisionmakingon the part of individualsis 12. SeeFields(1992),who noresthat until 1972unions
a necessarycondition,ir isnot sufficientto ensureoptimal were freeto bargainoverwages,whichwerelargelydeter-
allocationof labor. It is also necessaryfor workersto be mined by marketforces,and Freeman(1992),who, refr-
price-takersand to be freeto enter any industryor occu- ring to the post-1972 labor market poliqc regime, has
pation for which theyare qualified.Restrictionson free- asserted,"Singaporcis possiblythe most interventionist
dom of cntry arise from such sources as nonopsony governmentin thecapitalistworld."It is irterestingto note
power, discimination, government interventions,and that the sameinstrument-the NationalWagesCouncil,a
workers'associations. deliberationcouncildiscussedin chapter4-was useddur-

342

ing the period of wage reprcssion,during the accelerated
growthof realwagesin the 1980s,and nowto maintainreal
wagegrowth in line wid productivityincreases.
13. Lookingbeyondthegroup ofEastAsianeconomies
on which we havefocused,1aborrepressionis norgenemlly
associatedwith economicgrowth.Freman (1992) notes
that 'most dictatorshipsthat supprcssunions are incom-
petent in economic affaiirs-vide Burma in East Asia,"
though he recognizesthat this generalizationis not an ad-
equate dismissalof the argument that in some economic
circumstances,such as those prevailingin the "miracle"
economies,repressionmight be good fir development.
14. See Ranis (1993)for documentationfor Koreaand
Taiwan,China.
15. Gelb, Knight, and Sabot (1991) show how in re-
sponse to unemployment the creation of public sector
jobs with zero marginalproduct an overlitde more than
a decade reduce a moderatelygrowingeconomy to eco-
nomic stagnation whilehavinglitde or no impact on the
number unemployed.
16. The governmentsof Cote d'lvoire, Egypt, Mali,
and Sri Lanka all haveat varioustimes explicidyacted as
"employerof last resort,"particularlyfor universitygrad-
uates. Guarantees of jobs to workers who cannor find
"Suitable"employmentelsewheregenrally resultedin the
mushroomingof public sectorpayrolls.See Gelb, Knight,
and Sabot (1991).
17. When Indonesiaattempted to direct credit to in-
dustryin the late 1970s,ir wasimmediatelychanneledal-
most wholly to two public enterprises. Indonesia
withdrew this directedcreditschemein 1986.
18. Malaysia,for excanpie,lent substantialfiactionsof
credit to public enterprisesbetween 1975 and 1989; the
fderal govemment lent direcdy to public financialand
nonfinancialcorporations.Its lending reachedthe equiv-
alent of 22 percentof bankingassetsin 1975, dedining to
16 percent in 1980 and 9.8 percent in 1989. These loans
carried real interest rates, which were mildly negativeto
slighdy positive, comparedwith the substantialnegative
real racesof many other economies(Caprio,Atiyas, and
Hansen forthcoming).
USING RESOURCBS$T:
19. Other evidence on the extent of subsidiescan be
obtained from the reduction in financialburdens to en-
terprises that borrow from subsidized lending pro-
grams. In Japan, the cost reductions have been large in
sca transport (more than 20 percent); 5-10 percent in
electrical power, transporc machinery, and the coal
mining industrry but very modest in other industries.
Ogura and Yoshino(1988) estimate the financial bene-
fits arising from the reduction in interest rates of
Japan's directed-creditprograms was lcssthan 2 percenc
of total invcstments in manufacturing between 1961
and 1973. In Korea, some investment projects and ex-
porters receivedsubstantial interest rate subsidiesdur-
ing the HaI drive of the 1970s, but these largesubsidies
havesince declined.
20. King and Levine(1993) also show that high cen-
rralbank sharesin rotal credirare bad foirgrowth.Hence,
the macroeconomicrestraint disLcussed in chapter 3 per-
mitred rapidexpansionof private crediL
21. When exports were not used as an explicitalloca-
ion criterion, Japan often used inremational prices as a
stndard of good performance. Its industrial rationaliza-
dion programs used as criteria cost reductions that were
based on the costsof inTernaionalcompetitors.This was
enforcedthrough the tireat of internationalcompetition
as imports were graduallyliberalized(for example,ma-
chine tools and steel).
22. The meritsof well-managedand fbcuseddirected-
credit programs have long been advocatedbv Japanese
govemment officials.For example,a study by the Over-
seas EconomicCooperation Fund has argued that gov-
ernment involvement in directing crcdit is warranted
when there is a significantdiscrepancybetween private
and socialbenefits,when the investmenrrisk of particular
projectsis too high, and when informationproblemsdis-
courage lending to small and medium-sizefirms (OECF
1991). Use of policy-basedlending rather than other
forms of industrial assistance(for example,lower taxes,
grants, and so on) is premised on the argument that the
main constraincfacing new or expandingenterprisesis
their accessto extemal finance at reasonableterms and
conditions. Directed-credit programs involving small
subsidiesovercomethis consuaint, but Japaneseofficials
havestressedthat to avoid the misuseof funds and abuse
343
 LERAGLE

of credit programs,strong emphasismust be placed on menr, sayingthar if a firm reallywere to be profitablein

the maintenanceof macroeconomicstabilityto minimize the long run, it wouldpay for the firm to encounterlosses
distortionsin incentivesand on effrive monitoring to today.But this counteragument runs into problems:it is
ensure the timelyrepaymentof loans. basedon the premisethat capital marketsare perfect.

23. Thesestudies are preliminaryoutputs of a World 31. Exportpricescan be set in the international(un-
Bank researchprogram on the Effeccivencss of Credit protected)markersat long-run marginalcost. Is is possi-
Policiesin EastAsia,which is intendedto provideempir- ble to show, under certainsimplifyingasswnpdons(see
ical testsof the effectivenessof directed-creditprograms. Spence(1981), dhatoptimal production of the firm en-
tails setting the marginalvalue of output cqual to the
24. Among the best known of thesearc Balassaand long-runmarginalcost.
others (1971), Little, Scitovsk,yand Scotr (1970), and
Kruegerand Bhagwai (1973). WorldDevtdopmentReport 32. In the early and mid-1970s, nominal tariff rtes
1991 providesa good summarvof the argumentsfor the and effectve protectior.began ro dedine (Iroh and Kiy-
linkagebetweenneutral incentivesand growth. ono 1988;Komiya1990).Althoughthere isdisagreement
over the extent of the reductionsin effectiveprotection,
25. For an interestingattempt to test one of these there is agreementthat effectiveratesof protectionwere
propositions-extenmalitiesaccruing to export-in a substantialbeforethe early1970s.
gcneralequilibriumframework,see de Melo and Robin-
son (1992). 33. While effectiverates of prorecrionor subsidycal-
culationswouldprovidea better measureof thequantita-
26. Among the earliestformal discussionof market tive impact of protection, the figures shown furnish a
failuresin developmenrareScdovsky(1953)and Chenery notional indicator,consistentover time, of the extent of
(1959).A thorough surveyof many of the issuesis given reductionof nominaltariff ratesand noncariffbarricrs.
by Corden (1974). More recent discussionsincludePack
and Westphal(1986) and Itoh and others (1988).World 34. It would be more desirableto compare efnective
Bank (1992c)providesa recentsurveyof the issuesand an ratesof protection,sinceit is value-addedpricesthat mat-
application to the evaluation of World Bank industrial ter for resourceallocation,but the most recentsystematic
developmentprojects. cross-economystudies based on comparabledefinitions
udilizeevidencefor only one year in the 1960sand early
27. Economic growth will eventuaHly preclude the 1970s Thereare fewrecentestimatesto providecompara-
need for policiesto capmre the producrivitygains from blefigures.
economiesof eitherscopeor scale.
35. Singaporehas been exduded since its zero duty
28. For a reviewof many of these, see Stewartand rateswillbias the estimatedownwart.
Ghani (1992).
36. Pack (1988) expresseddoubt abour the robustness
29. In general, real externalitiesprovidea validargu- of the evidencefor Koreaand Taiwan,China. When che
ment for trade policy interventionsonly if they allow paper was writren in 1985, the TP estmares for both
goods to be produced at less than the imported c.£f economieswere basedon data that began in the mid- to
(costs,insurance,and freight)price.This is nor, however, late 1960sand enidedin the late 1970s,a sufficiendyshort
sufficientto justifjrintervention.A sociallysuccessfulin- period so that increasingcapacityutilizationor one-rime
tervention depends on whether the present discounted benefitsfromscaleeconomiescouldhave accountedfor a
value (Pnv)of fumtreproducersurplusexceedsthe PDVof substantialshare of unexplainedgrowth. One study of
the cost of subsidies. Korea (Kim and Kwon 1977) was indeed convincingin
demonstratingthat correc measuresof capacityutiliza-
30. More recentdiscussionshavediscountedthis argu- tion eliminatedthe residualfor much of the period in

344

question.Additionalobservationsare now available,and
improved estimaresof the capital stock make it seem
likely that the measuredTFP accurately reflectsunex-
plainedgrowth.
37. On these issues,a particularlygood discussionis
givenby Itoh and others (1991),part 1.
38. For details,see the various chapters in Komiya,
Okuno-Fujiwara,and Suzumura(1988).
39. Itoh and Kiyono(1988).Theseestimaresarebased
on nominal tariff rates,not implicitprice comparisons.
Given the presenceof quantitativerestrictions,they are
likelyto be minimums.
40. For extensive discussions, see World Bank
(1987a).
41. In Koreaand Taiwan, China, the initial spurt in
exports was in labor-intensiveproducts as predictedby
the static theoryof comparativeadvantage.These,how-
ever,arealsothe sectorsdtar Dollarand Sokolofffindex-
hibited greater TFPgrowth between 1963 and 1979 in
Korca.Wang (1990)finds a similarpattem for TaiNvan,
China, for 1966-86. Exportsmayhaveresultedfrom this
initial spurt in productivity. But analysisof Taiwan,
China's, early export growth by Scott (1979) and the
analysisof the Koreantrade incentivesystembyWestphal
(1978) suggestthat factorproportions,supplementedby
a smallpro-exportbias, plausiblyaccount for the earliest
exportgrowth.
42. A recentstudyoFTFPgrowthin Chile,forexample,
finds that followingthe 1970s' liberalization,accelerated
TFP change due to organizationalchange and product
qualityupgradinglastedat most sixyears(Perryand Her-
rara,forthcoming).
43. Undoubtedly,domesticsldllswereimportantin al-
lowing the technologyobtained from abroad to be uti-
lizedproductively.
44. On the roleof infi-astructure
in Taiwan,China,see
Ranis (1979).
45. DFIwasnot a major sourceof invesrmentgrowdt
USING RESOURCES C
in Japan, Korca,or Taiwan, China, but it has been very
importanrin otherHPAEs.Evenin Taiwan,China,DFIac-
counted for 15 percent of employment growth in
1975-84 and demonstratedthe viabilityof new exports.
Moreover,muchof the investmentwasin relativelysmall
firms in which employeesvvereexposedto new equip-
ment and the relevant production cnginecring.Since
labor turnover in these foreign-invesredfirms was sub-
stantial,the total imnpact
of DFIon effeaivelyintroducing
newtechnologywvas quite high (Pack1993b).
46. For efforts to assessthe empiricalimportance of
spilloversfrom DFI,see Blomnstrom (1989) and Haddad
and Harrison(forthcoming).
47. SeeRheeand Westphal(1977)fbrevidenceon the
use of older technologyto achieveexportsof textilesin
Koreaduring the 1960sand 1970sand Ranis (1979)for
Taiwan,China.
48. On Taiwan, China, see, for exanple, WallStret
JounmalJune 1, 1990, and Liu (1987). On Korea, see
Westphal,Rhee,and Pursell(1981).
49. Katz,cd, (1987) reportthe resultsof thesestudies.
The firms investigatedmade intenseefFortsto improvc
productivitydespitethe protectionaffordedby the im-
port-substitutionregime,an exmple of the too oftenfor-
gotten point that protecton has both income and
substitutionefects on theeffortexpendedto reducecoscs.
In thefrms considered,the substituion effec dearlyout-
weighed the income effect. Nevcrtheless,calculations
based on most of the studies suggestvery low rates of
growthof capitalproductivity,an exceptionbeingthe re-
sults for one steelplant analyzedby Dahlnian and Fon-
seca (1987). rTPestimatesare nor easilyextractedfrom
the data reportedfir most of the srudiesin Karz,ed.
50. The impormnceof interactionsmay explain the
phenomenon noted by Behrman (1987) that between
1960and 1981,the averagelevelofschoolinginall devel-
oping economnies increasedby two-thirds, yet in most
economiesthis had littleeffectin stimulatinggrowth.Ed-
ucarionhas a smallereffectwherecomplementazychanges
do not occur,thus, high educationmaybe necessarybur
is hardlysufficient.
345
 lR A
CIRACLE
51. Previousstudies using this approach are Dc Long
and Summers ( 991), Mankiw, Romer,and Weil (1992),
and Pack and Page (1993).
52. Severalpreviousstudies have fbund such effcas,
including Harrison (1991), Dollar (1992), and Thomas
and Wang (1992). We use as the opennessvariableDol-
lar's measure of the correspondencebetween domestic
and international relative prices (1990). This measure
usesSummersand Heston's internationalcomparisonsof
pricelevelsto developan indexof 'outward orientation"
for ninety-fivedevelopingeconomies(1988).
53. Alternativespecificationsusing the rate of grotvth
of tomalexports were also employed with similar but less
robust rcsults.Dara on the rate of growth of manufactured
exports is not availablefora largesampleof economies.
54. This conclusion is reinforced by the arguments
that low distortions relative tO international prices can
arisefrom havinga significantshare of theeconomy oper-
aring at export prices;in that case,the share of manufac-
rured exports in GDPmay be morestrongly collinearwith
the openness index,
55. We ran two joint E-tescson the regressionsinduding
theinteractve term.The Ftests wereboth consistentwith a
high degree of multicollinearitybetween the interaction
term and the exportterm. The Ftest rejectsthe null hy-
pothesisthat, takenjoindy,the coefficienrsonthethreevari-
ables (education,manufacwured exporrsftotalexports,and
the interactiveterm)are norsignificantlydifferentfromzero
(taken together, the three are significantlydifferent from
zero at the 0.01 level).Likewise,the Ftest rejectsthe hy-
pothesisthat, takenjointly,the interactivetermand the ex-
346
port variable are not significandydifferentfrom zero(tamen
together,the twoamrsignificantlydiffrent fromzeroat the
0.01 level).Where rhereisa high degreeofmulricoilinearigy
berweenthe variables,the coefficientson Lhe interaction
term and the export variable,despirebeing separatelyin-
significant,shouldstillbe ueaed as besrpoinresimares.
56. The discussion of scale economies has recently
beenrevivedby Murphy, Shleifer,and Vishny(1989).
57. An alternativeto cquation 6.1 is: AO= iVj A.,
which assumesthat sectoralsharesare constant.
58. Another interpretationwouldbe that there wasan
initial disequilibrium in marginal productiviEiesamnong
secors and that indusrrial policy acceleratedthe move-
ment of factors ro higher marginalproductivitysecrors.
Writhrespea to labor, the issueis moor in the Four Tigers
as they ypicallyhad rdativelycomperitivelabor markets
59. Put differendy, it has been shown earlier that
Korea'svPl4 ratio for mPp was 2.76. If this had been 1,
the actualshare of valueadded wouldhavebeen 13rather
than 36 percent; thar is, the share would havebeen con-
stant ar the 1968 level.If factorshad been allocatedto all
the remaining sctors equally, the sectorwide averge
valueforA*would havebeen 0.061 rather than 0.067.
60. See Itoh and others (1991), Komiyaand others
(1988),Saxonhouse(1983), and Trezise(1983).
61. For a very usefulaccount of many of the measures
in the earlierperiod, see Lockwood(1954).
62. See note 53 for an explanation.
 q!TER 7

Policies and Pragmatism

in a Changing World
HIS BOOK HAS SHOWN THAT THERE IS NO SINGLE
East Asian model. Rather, die eight high-performing
Asian economies have used different and changing
sets of policies to achieve rapid growth with equiy.
In this chapter, we evaluatesix 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.

Foundatonsof RapidGrowth-
Getting
theFundamentalsRight

T HESTRESS
ONPOUCYFUNDAMENTALS
ISA HALLMARK
OFALL
eight HPAEs. These widely shared, market-friendly policies are
the foundation of East Asia'seconomic success. In chapter 2 we
identified several broad policy areas as fundamental, and in subsequent
chapters we described their relevance to the strategic finctions of accu-
mulation, allocation, and productivity change. Here we briefly review
some key aspects of these policies. Although implementing these fun-
damentals successfilly is often difficult, the benefits are substantial, as
demonstrated by their crucial role in the rapid growth of the HPAEs.
Moreover,despite changes in the global economic environment, these
policies remain viable options for developing economies today.

347
 #PACI RACLE

Ensuring
LowInflaionandCompetitive
Exchange
Rates

Macroeconomicstabilityand low inflationwere necessaryprecondi-

tionsfor rapid growth in all eight HPAEs.The keyelement was the man-
agement of the fiscal deficit. Fiscal policies havc been generally
conservativein avoiding inflationaryfinancing of budget deficits, and
some HPAEs, such as Singaporeand Taiwan, China, have run sustained
budgetsurpluses.While Indonesia,the Republicof Korea,and Malaysia
acquired considerableforeign debt, their debt to export ratios were
lowerthan other developingeconomies,so creditorsnever lost faith and
they never faceda fuil-blowndebt crisisrequiringrescheduling.More-
over,budget rules and legislationrequiringcharsovereigndebt be used
for developmentexpenditureshelped to ensure that borrowingfunded
investnent rather than consumption.The successfulrecord of low in-
flationand sustainedgrowthof the HPAEsreinforcesthe viewthat there
is no substitutefor fiscaldiscipline.
The HPAEs' successin export-drivengrowthwas predicaredon com-
petitiverealexchangerates.Prudent fiscaland monetarypoliciescoupled
with flexibleexchangerate managementwereusedto keep exchangerate
movementsalignedwithdchangingstructuresof trade protectionand in-
flationdifferentialswith tradingparmers.The HPAEsavoidedstrategiesof
macrocconomicstabilizationthat stressedthe roleof the exchange rtc in
breakinginflationaryexpectations.Rather, exchangerate adjustments
weresupportedby expenditure-reducingfiscalmeasures.While Indone-
sia,Korea,and Thailand have had periodsof realappreciation,they did
not try to maintain a fixedexchangerate regimein the faceof inflation
ratesmuch greaterthan thoseof their tradingpartners.
The HPAESrespondedto externalshocksmore quicklyand effectively
than other developingeconomies.East Asian policymakershave been
more adept at managingfiscalcontractions-cutting expendituresand
increasingrevenuesin a growth-orientedframework-to attain viable
deficittargets.In contrast,governmentselsewherecommittedto funding
largeentidementprograrins-incometransfers,untrgeted subsidies, -
cessivepublic sectoremployment,and loss-makingpublic enterprises-
often lackedthe flexibilityto deal efficientlywith budget contractions.
Rather than cutting spendingon entidements,theYhave often resorted
to cuttingpublic investment.
In eachHPAE,a technocraticelite insulatedto a degreefiom excessive
political pressure supervised macroeconomic management. The insula-

348
 POLICIES CI

tion mechanisms
rangedfromlegislation,suchasbalancedbudgetlaws
in Indonesia,Singapore,
and Thailand,to customand practicein Japan
and Korea.All protectedessentiallyconservativemacroeconomic poli-
ciesby limitingthe scopefor politiciansand interestgroupsto derail
thosepolicies.

Building
HumanCapital

Wehaveshownthatthebroadbaseofhumancapitalwascriticallyim-
portantto rapidgrowthin the HPAES. BecausetheHrAEs attaineduniver-
salprimaryeducationearly,
literacywashighandcognitive skilllevelswere
substantially
abovethosein otherdeveloping economies.
Firmstherefore
had aneasiertimeupgradingtheskillsof theirworkers
andmastering
new
technology.
In addtion, rapidhumancapitalaccunularionreducedin-
comeinequalitybyincreasingtherelativeabundanceofeducatedworker,
therebyloweringthescarcty rents associatedwith cognitiveskils. These
benefitswereparticularly
evidentin thecountryside.
Basiceducatonin theHPAESis highlyorientedtowardtheacquisition
of generalacademicskills,while post-secondaryeducation tends to be
oriented towardvocatonal skills.SomeHPAEs have alsobeen unusually
large-scaleimportersof educationalservices,particularly
in vocationally
and technologically sophisticateddiscplines.In addition,enterprise-
leveltraining,muchof whichwassubsidized,accelerated technological
acquisition.
The benefitsof dosingthegap betweengirlsand boysin primaryand
secondaryschoolenrollments havebeensubsmntial, evenin economies
in whichfemalelaborparticipationratesarelow.Bettereducatedmodt-
ersraisebetrereducatedchildren,complementing publicinvestmentsin
basiceducation.The rapid increasein the educationalattainmentof
womencontributedto Ehe earlydeclinein fertiiy, whichin turn re-
sultedin a lessrapidlygrowingschool-agepopulation.As a result, mDre
resourceswere availableper child, in homes and in the dassroom.

CreaingEffective
andSecureFinancial
Systems
Financial sector policies in tie HPAEs were designed to ficilitate two
crucialfunctions.First,theyencouragedfinancialsavings.Second,they
channeled savingsinto activitieswith high social returns. There was
great varietyin the policiesand institutions used to accomplishthese

349
 _RALLE

ends, but severalflmdamentalapproaches were important in all eight

economies. These included generallypositive real interest rates on de-
posits and the creation of secure bank-basedfinancialsystems.
Positivereal interest rates on depositswere the most important f&ctor
in encouraging financial savings.All of the HPAEShad periods, usuaily
early in their development, when rcal intcrest rates on deposits were
negative. But in general real interest rates were more stable and more
consisrently positive in East Asia than in other developing regions,
wherehigh inflation has often meant that volatileand highly negative
real inrerest rates prevailed.
HPAE governmentsalsoencouragedsavingsby fosteringscableand se-
cure banking systemsthrough a combination of protection and regula-
tion. Protection of establishedbanks induded governmentlimitson the
creation of new domesticbanks, the enuy of foreign banks, and the op-
eration of competing non-bank financial institutions. Regulation in-
cluded enforcement of prudent behavior, such as limnitingspeculative
lending. Although the HPAEs had no explicir deposit insurance, they
promoted confidence in the banking system by devising rescue pro-
grams as needed, sometimes,as in Japan and Korea, pressuringstronger
banks to absorb weaker banks. In addition, many of the HPAEsencour-
aged savingsby creatingpostalsavingsplans to lowertransactioncostsof
savingsfor small savers.
Balancingthe advantagesof protection against die need for compet-
tion has been a major challengefor HPAE bank regulators. East Asia's
banking sector structures,as well as some measures of bank efficiency
(profitabilityand borrowing-lendingmargins),suggestthat banksin the
HPAEs have been at leastas efficientas those in other developingregions.
Evenso, innovation and efficiencyundoubtedly sufferedfrom limits on
competiton. Pardy in response, HPAEs gradually have opened their
banking systems to increaseddomestic and fbreign competition. Gov-
ernments have also promoted bond and equicymarketsto increasethe
depth of mamring financialsectors.
SeveralHPAESsupplemented commercial banls-and met the need
for long-term capitl-by creating specialized dvelopment banks.
These banks fiuthered the growth of the financial sector by establishing
proceduresfor project financingand monitoring that commercialbanks
then copied. In contrast to developmenrbanks in manyother develop-
ing economies,which tend to be subject to extensivegovemment med-
dling, H1AE development banks controlled project selection and

350
 P O L I C IES AN Dn:rIcv

monitoring independently,albeit within broad governmentguidelines.

As a result,their bad debts were usuallyassociatedwith economicdown-
turns ratherthan die willful default common in economieswheregov-
ernment officialsare closelyinvolvedin day-to-day management.

Limiting
PriceDistortions

Despite the HIAEs'many market interventions, reiativeprice distor-

tions were limited and indeed smaller than in most other developing
cconomies.Flexiblelabor marketsand capitalmarkets with positivereal
interest rates meant that wagesand interest rates reflectedmore closely
true scarcitiesof labor and capital. Thus resourceswere generallydrawn
into labor-intensiveproduction in earlystagesof development, shifting
to capital- and knowledge-intensiveactivities as physical and human
capital deepened.
A combination of competitivereal exchangerates and moderate pro-
tection levelsmeant that domestic pricesof traded goods were closerto
international prices in the HPAEsthan in other developingregions.The
largeweight of exportsin total output, duty-freeimports for exporters,
and extensive domesticcompetition meant that all the HPAEshad higher
proportions of their manu&acturingsectorsoperatingat or near interna-
tionl prices than the import-substitutingeconomiesof atin America,
South Asia, or Sub-SaharanAfrica.

AbsorbingForeign
Technology

While none of the HPAEsexcept Hong Kong and Singaporewas en-

tirely open to international trade, al were open to foreign technology.
Japan, Korea,and to a lesserdegreeTaiwan, China, relied heavilyon li-
censing, imports of machinery, and reverseengineering during their
rapid growth periods. Hong Kong, Singapore,and the Southeast Asian
HPAEs,in addition, welcomeddirect foreign investment that came bun-
dled with technical, managerial, and sometimeslabor force skills. Li-
censing has become more and more problematical because owners of
technologyare reluctant to share it, even for a fee. However,the success
of the SoutheastAsian HPAEswith other forms of technologicalacquisi-
tion suggeststhis need not be a seriousbarrierto growth.
The alternativepath of self-reliancehas litde to recommend it. Some
developingeconomies,such as Argentinaand India, have restrictedcap-

35'
 0 A: -X1RACLE

ital goodsimports to promore thegrowchof domesticmachine-building

industries.In thesecases,however,domesticmachinerygenerallyfalls
short of internationalstandards.Such machinerybecomesyet another
high-costimportsubstitutethat inhibitsexports.

Limiting
the BiasagainstAgriculure
Agriculturehas playeda morecentralrolein the H1IAEsthan in many
other low- and middle-incomeeconomies.Wide adoption of Green
Revolutiontechnology, highinvestmentin ruralinfrastructure,and lim-
iteddirectand indirecttaxationof agriculturemeantthat ruralincomes
and productivityrosemorerapidlyin EastAsiathan in other regions.
As in othereconomies,agriculturalsectorsin the HPAESwerea source
of capitaland laborfor the manuficturingsector.But in EastAsiathese
resouroesweregenerallypulled into manufacturing by risingwagesand re-
turns, ratherthan squeezedout of agricultureby hightaxesand stagnant
or decliningrelativeincomes.Asa result,rural-urbanincomedifferentials
weresmallerin the HPAEsthan in most other developingeconomics.

Creating
Institutionsto PromoteGrowth

rT-%HE SUCCESSOF THE EASTASIANECONOMIESSTEMSPARTLY

Sam the policies they have adopted and partly from the insti-
.1trutionalmechanisms they created to implement thenL AUlof
the HPAEscreated secure institutional environmentsfor private invest-
ment that led to very high levels of private sector-led growth. HPAE
civil servicesrange oromthe highly meritocratic and insulated bureau-
cracies of Japan, Korea, Singapore, and Taiwan, China, to the less
effective and less insulated public administrations of Indonesia and
Thailand. Nevertheless,each of these economies has a core of techno-
cratic managers. In Indonesia and Thailand, their scope is limited to
management of the macroeconomy. In the other HPAEs,competent
civil servicesadminister a much wider range of policy instruments.
How did these economies create a reputable bureaucrac? First, pay
mattered. The salariesof bureaucrats (exceptin Singapore)usually fel
short of those for equivalent positions in the private sector, but they
were sufficientlyhigh to attracr and retain good economic maragers.

352
 POLICIES` N

Second, in the HPAEswith high-quality bureaucracies, rules and proce-

dures governing public sector employment were institutionalized and
insulated fromn political interventions. In particular, recruitment and
promotion were merit based. Third, public employment was accorded
high status. These factors improved the quality of the bureaucracy, dis-
couraged corruption, and created an esprit de corps among civil servants
that helped insulate the bureaucracy from political pressures.
Each of the HPAE governments created institutions to improve com-
munication with the private sector. Formal deliberation councils esrab-
lished in five of the economies-Hong Kong, Japan, Korea, Malaysia,
and Singapore-included government officials, journalists, labor repre-
sentatives, and academics. The economic and political benefits of these
councils, and of the more informal mechanisms in the other HPAEs,are
impossible to measure systematically. It is likely, however, that econom-
ically they improved coordination among firms and improved the flow
of information between businesses and government Politically they
helped establish a commitment to shared growth and reduced rent-
seeking. Information sharing made it more difficult for firms t curry spe-
cial favors fiom the government and for government officials to grant
special concessions. Thus the deliberation councils helped check oppor-
tunistic behavior. The deliberation councils in Japan, Korea, and Singa-
pore also performed an important monitoring function, assessing
performance at the industry or economy leveland administering contests.
Consultative mechanisms in Indonesia and Thailand were less for-
matl, and their role in facilitating business-government communication
was much more limited. Malaysia's experiment with deliberation coun-
cils is quite recent but appears to hold some promise in the area of
macroeconomic management.

Inteirvening
in Markets

M OST OF THE HPAEs, ESPECIALLYTHOSE IN NORTHEAST

Asia, intervened in markets in an effort to hasten growth.
All interventions carry with them costs, either in the form
of direct fiscal costs of subsidies or forgone revenues, or in the forrn of
implicit taxation of households and firms-for example, through the
structure of protection or interest rates controls. One of the main char-

353
 MIRACLE

acteristics of interventions in the HPAEs is that in general they have

been carried out within well-defined bounds limiting the implicit or
explicit costs. Thus, price distortions were present but not excessive,
interest rate controls generally had as benchmarks international interest
rates and were binding at positive real levels. Explicit subsidies were
also kept within bounds. Given the overriding importance that each of
the HPAEsascribed to macroeconomic stability, interventions that
threatened to undermine thar policy fimdamenmally,were modified or
abandoned-for cxuample,the heavy and chemical industries drive in
Korea or the heavy industrialization push in Malaysia. These limits to
intervention stand in sharp contrast ro many other developing econ-
omies, where interventions have not been consistent with macro-
economic discipline.
Whether these interventions contributed to the rapid growth made
possible by good findamentals or detracted from it is the most difficult
question we have tried to answer. It is much easier to show that the
HPAEs limited the costs and duration of inappropriately chosen
interventions-itself an impressive achievement-than to demonstrate
condusively that those interventions that were maintained over a long
period accelerated growth. In this book we have evaluated three sets of
policy interventions: the promotion of specific industries or industrial
subsectors, directed credit, and the export-push strategy. We conclude
that promotion of specific industries generally did not work and there-
fore holds little promise for other developing economies. Directed
credit has worked in certain situations but carries high risk. The
export-push strategy has been by far the most successful of the three
sets of policy intervention and holds the most promise for other devel-
oping economies.

PromotingSpecific Industries GenerallyDid Not Wark

Industrial policy narrowly defined-that is, attempts to achieve more

rapid productivity growth by altering industrial structure-was gener-
ally not successful. In Japan, Korea, Singapore, and Taiwan, China, pro-
motion of specific industries had little apparent impact. Industial
growdt tended to be market-confbrnning, and productivity change was
not significantdy higher in promoted sectors. Although governrents in
these four economies were undoubtedly trying to alter industrial structure
to achieve more rapid productivity growth, with the exception of Singa-

354
 POLICIES AE 1.A

poretheirindustrialstructuresevolvedlargelyin a mannerconsistentwith
marketfoirces and factor-intensiy-basedcomparativeadvantage.
It is nortaltogethersurprisingthar industrialpolicyin Japan, Korea,
andTaiwan,China,tendedto be marketconforming.Theseeconomies,
althoughselectivelypromotingcapital-and knowledge-intensive indus-
tries,still aimedto createprofitable,internationallycompetitivefirms.
Taiwan,China,placedgreatemphasison pricecomperitiveness in eval-
uatingthe performanceof its largepublic investmentsin intermediate
and capitalgoods.Moreover,in Japan especially, but alsoin Korea,in-
stitutionalmechanismsthat stressedcooperativerelationsand exchange
of information betveen firms and governmentintroduced a largc
amountof marketinformationinto the processof industrialpolicyfbr-
mulacion.And the yardstick used to evaluateindustrial policies'
success-mainlyexportperformance-provideda markettestof the suc-
cess or failure of the policy instrumentschosen. Thus the seeming
successof industrialpolicyin thesethreeeconomiesprobablyrestsnor
on pickingwinners-that is, on the adroitselectionof whichindustrial
subsectorto promote-but on the setting of export targetsfor pro-
moted industries and the use of export performanceto assesspolicies.
Our testsmayhavefailedto captIe one importantaspectof indus-
trial policyin Japanand Korea:informationexchangein a competitive
framework-in short, contests-may havelessenedinvestmentcoordi-
nationproblemsand resultedin earlierand betterinvestmentsin what
ultimatelybecamemarket-conforming sectors.But among the HPAEs,
onlyJapanand Koreahad the instiutionalcapacityto organizeand ref-
ereethe kind of contestsusedin formulatingand implementingindus-
trial policies.
Despitetheir potential benefits,collaborativearrangements,espe-
ciallywhencombinedwith accessto scarceresourcessuchas foreignex-
change and cradit, have a high risk of capture by the participants.
Capture was avoided-and resource allocation may have been
improved-in Japan and Koreaby the combinationof repeatedrela-
tionshipsbetweenbusinessand governmentand the useof exports,not
domesticGDPgrowth,as the yardstickby which the successof the in-
dustrialstrategywasmeasured.In other HPAES,such elaboratecontests
were not used-for example,in Hong Kong, Singapore,or Taiwan,
China-or wereunsuccessfil,as in the casesof the heavyindustriafiza-
tion drive in Malaysiaor recent attempts at technologicalleapfrogging
in Indonesia.

355
 .X4 N-MIRACLE

Repressing
Interest
RatesandDirecting
CreditWorked
Somefimes

All HPAEscontrolled interesr races on deposits and loans during much

oF their rapid growth. Low real interest rates on deposits and loans in-
creased corporare equity by transferring resources from savers (mostly
households) to borrowers (mosdy firms). At the same rime, by limiting
the spread between deposit rates and loan rates, govemments prevented
banks from earning excessiverents. On the downside, such controls un-
doubtedly discouraged banks fiom taking on riskier, nontraditional
dients.
Interest rate repression was key to the HPAEs'directed-credit pro-
grams. Low interest rates on loans resulted in excess demand for credit,
which permitted governments to take an active role in credit rationing.
This in turn gave governments strong leverage with which to influence
the beihavior of firms. In this section we discuss why directed credit was
relatively successfid in hie HPAES but often had large negative conse-
quences elsewhere. WVethen consider whether such polidtes could be ef-
fectively adopted in the increasingly open capital markets of most
developing economies.
Interest rate subsidies were smaller in the HPAEs than in other devel-
oping economies. In several HPAES, government-directed credit served
primarily to signal commercial banks and private investors which sectors
enjoyed govemment support Banks and private investors then selected
projects within those sectors and offered them additional funding at
commercial rates. Moreover, credit was frequently allocated according
to export performance. Our evidence leads us to condude that credit
programs directed at exports yielded high social returns and, in the cases
of Japan and Korea, other directed-credit programs also may have in-
creased investment and generated imporcant spiliovers. Even within
East Asia, however, when interest rate subsidies were large, or export
performance criteria were not imposed, directed credit often went to in-
efficient ventures and those with low social returns.
The financial sector policies used by the HPAEs in the 1960s through
the early 1980s took place in a world economy in which it was possible
to close domestic capital narkets. Since the early 1970s, developing
economies have become ever more integrated with the global capital
markeL While this has reduced government? already limited abiity to
restrict capital flight and speculative capital inflows, it has also given
them accessto less expensive capital and offered them better possibilities

356
 POLICIES AND:P

to hedge externalshocks. At the same time, increased integration has

constrained policymakers,who are finding it more and more difficultto
rnaintain interest rates at variancewith international markets.
The developingeconomies'growingintegration with global financial
markets can be seeniin the evolurion of their stock markets. Until the
mid-1980s, foreign investors'accessto domesticsecuritiesin developing
economies was often restricted. Since then, fourteen of the twenty
largest developing-economystock markets have become substantially
more open to foreign investment. The influence of international inter-
est rates on domestic rates reflecrsa similar trend towardglobal integra-
tion. Indonesia,for example,began to open its capital accounts in 1971
and freeddomestic interestrates in 1981. Sincethen, interest ratesin In-
donesia have followedinternational rates.
Although capital outflow restrictions are generally ineffective,they
do hinder the movement of some institutional capital, for example,do-
mestic pension fimds. They may alsohinder the flight of household and
individual savings in instanceswhen the transactioncosts of sending
capital abroad are greater than the incentivefor doing so. Becausethese
barriers do marer, developingeconomies continue to be less open fi-
nancially than they are conmmercially. This can be seen in the ratio of
gross capital flows to CDP compared with the ratio of exports-plus-
imports to C;DP. For the aggregate of all developing economies, the
former is only 13 percent, while the latter is 45 percent.
Nonetheless, the limited impact of capital restrictions in deterring
capital flight has persuaded a growing number of developing-economy
governments to open to global financial markets. In some cases, for ex-
ample, Chile and the Philippines, the removal of capital outflow regula-
tions has had the paradoxical effect of stimulating a net inflow by
assuring investors they can expatriate their money at will. Even where
govermnents have not liberalized restrictions, the globalization of the
economy, particularly increases in trade and advances in transport and
telecommunications, is rapidly eroding the isolation of closed
econormies.
Under these conditions, the choice for developing-economy policy-
makers is not between financial integration with the global market and
isolation. Rather, it is between two-way integration that permits out-
flows and thus facilitates inflows, on the one hand; and de facto one-way
integration that consists primarily of unauthorized capital flight, on the
other. The HPAEs have recognized this dcanging global environment; all

357
 '&1N IR A CLE

are moving-albeit at different speeds-to liberalizetheir financial sec-

tors and integrate them more closelyinto internationalfinancialmarkets.
Financialsector interventions-specifically repressionof interest rates
and contest-based direcEionof credit-may have contributed to rapid
growth in such economiesasJapan, Korea,and Taiwan.,China. Bur eadi
of these economies had competent and insulated bureaucracies and
banks to select and monitor projects, and each applied export perfor-
mance as the main yardstickfor credit allocation.In addition, thesegov-
ernmentswere able during the earlystagesof rapid growth to closetheir
financialmarkets to the outside world. Few developingeconomiestoday
have the institutional resources to consistendy impose performance-
based criteria for credit allocation. Moreover,the capital markers of all
economiesare becoming moreand more integraced& Thus, successfulre-
pressionof interest rates and contest-basedcredit allocationmay only be
possiblefor the very few developingeconomieswith strong institutions
and then only to a very limited degree.

Export
Push:A Successful
Mixof Fundameitals
andInterventions
The export-push strareg;y-the use of findamentals and interven-
tions to encouragerapid export growth-was the HPAEs' most broad
based and successfulapplication of selectiveinterventions.Furthermore,
of the many interventions tried in East Asia, those associatedwith their
export:push hold the most promise for other developing economies.
While the changing trading environmentwill limit the use of highly tar-
geted policies-similar to those of Japan and Korea during their early
export growth-the broad pro-xport stance characceristicof all eight
HPAEs remainsviable.
Trade policiesin all the HPAES (exceptHong Kong)passedthrough an
import-substitutionphase with high and variableprotection of domestic
import substitutes.In allcases,however,policiesthat stronglyfavoredthe
production of import substitutesto the detriment of exports were aban-
doned, and HPAE governmentsadopted strategicpro-export policiesthat
establisheda free trade regime for exportersand offereda range of other
incentivesfor exporrs.This export-pushapproachprovided a mechanism
by which industry moved rapidlytoward international best practice, de-
spirehighly imperfct world marketsfor technology.
Export-pushstrategieswere implenented in three verydifferentways
in the HPAES. Hong Kong and Singaporeestablishedfree trade regimes,

358
 POLICIES AN

linking their domesticprices to international prices;the export push was

an outcome of the very limited size of the domesticmarket coupled with
neutral incentivesbetween producing for the domestic or international
market. Both economicsmade export credit available,although they did
not subsidize it, and Singapore focused iEseffortson attracting foreign
investment in exporting firms.
In Japan, Korea,and Taiwan,China, incentiveswereessentiallyneutral
on azvergebetweenimport substitutesand exports.But within the traded
goods sector,export incentivescoexistedwith substantialremainingpro-
tection of dte domestic market. Export incentives,moreover,were not
neutral among industries or firms.There was an effort in Japan, Korea,
Singapore,and TaiwanChina, to promote specificexportingindustries.
Protection was combined with either compulsionor strong incentivesto
expor- In Korea,firm-specificexporttargetswereemployed;in Japan and
Taiwan, China, accessto subsidizedexport creditand undervaluarionof
the currencyacted as an offserto the protectionof the local market.
In the HPAEs that intervened selectively to promote exports, contesrs
based on performance in global markets played the allocative role that is
normnallyascribed to neutral exposure of both import-substituting and
exporting industries to international competition. But these contest-
based incentive strucnures required high governmenr institutional capa-
bility. One of the keys to success of the export push in some of the HPAES,
especiallyJapan and Korea, was the govermnent's ability to combine co-
operation with competirion. They were able to do this first, because their
civil services and public institutions were largely staffed by competent
and honest civil servants, and second, because rirms and bureaucrats
knew that there Nvasa single yardstick for performance: exports.
Export targets provided a consistent yardstick to measure the success
of market interventions. When protected sectors interfered with the
exports of other sectors, exporters could successfilly seek redress. For
example, domestic content rules imposed on foreign direct investors in
Taiwan, China were suspended when they interfered with exports. The
emphasis on exporc competitiveness gave businesses and bureaucrats a
transparent and objective system to gauge the desirability of specific ac-
tions. Interventions could not be made arbitrarily, as these could be ap-
pealed at a higher level of government if they interfered with exports.
The more recent export push-efforts of the Southeast Asian newly in-
dustrializing economies (NIEs) have relied less on highly specific incen-
tives and more on gradual reductions in import protection, coupled

359
 S'IWN'M[RACLE

with institutional support of exportersand a duty-freeregimefor inputs

into exports. Recent strategiesto attract direct foreign investmentin In-
donesia, Malaysia,and Thailand have also been explicitly export ori-
ented. Direct foreign investmentin manyother developingeconomiesis
roo often directedat servinga protected domestic market and hedged by
restrictionson exports.

lhe Export-Push
Strateg in a Changing
TradingErwiroment
The early HPAE export drives-those of Hong Kong, Japan, Korea,
Singapore, aiLdTaiwan, China-took place amid the cxpanding world
economy of the 1950s to the mid-1970s. Economic growth in the in-
dustrial economies was rapid, barriers to trade were declining, and
global effortsto spur free trade under the General Agreement on Tariffs
and Trade (GArr) were the order of the day. Times have changed. Eco-
nomic growth has slackened in the mature economies thar traditionally
provided the main markers for developing-economy exports. Regional
trade groupings are becoming more and more prominent. Industrial
economies have begun to retaliate agnst allegedly unflair export prac-
tices by imposing quotas or countervailing duties. In such a trading en-
vironment, can today's developing economies successfi&llyemulate the
export-push strategies of the early HPAEs?
In the following section we break this question into two parts. First,
will there be adequate, open markets for developing-economy manufac-
tured exports? Second, will the need for market access constrain the de-
veloping cconomies' use of export-push strategies? We conclude that
markets will be adequate and open and thar an export push, therefore,
remains possible. However, global realities will limit developing
economies'abiliry to adopt rhe more interventionist instrumcnts of ex-
port promotion. The experience of both the northern HPAEs,with their
mixed export-push regimes in which protection of the domestic marker
coexisted widt incentives or requirements to export, and the Southeast
Asian NIEs, with their gradual trade liberalizationscoupled with im-
provemencs in export insitutions and availability of export credit, sug-
gests that economies thar are in rhe process of trade liberalization would
benefit from providirg specific incentives to manufactured exports.
Modest subsidies to exports could be linked, for example, to the bias
against exports in the domestic economy and bound by strict time lim-
its linked to the pace of trade policy reform.

36o
 POLICIES ANDP

Moreover, if there are important externalitiesto exporting, a firther

tilt toward exports can be justified on welfire-economicgrounds, even
if the trade regime is neutral between import substitutes and exports.
This might argue for functional export promotion measures such as
guaranteeing access to export credit, providing tax incentives, or im-
proving market accessfor smalland medium exporters. But, these mea-
sures will necd to follow internationally acceptable rules to assure
market access in importing economies, and must be of limited dura-
tion. Fiscal discipline, moreover,wiUlrequire that the costs of promo-
tional programs be kept in check.

WhyThereWillBe Adequate Markets

The success of the southern-rier HPAEs' export drives in the late

1970s and 1980s, along with the continued rapid growth of exports
from the northern HPAEs,indicates that export drives can succeed evenr
when industrial-economy markets are not expanding rapidly. In con-
trast with the successof earlierexport drives,which depended primarily
on increased consumption in the industrial economies, the later drives
have depended on consumers i industrial economies buying goods
made in Asia instead of goods made elsewhere. If industrial-economy
markets remain open, developing-economy exports can continue to ex-
pand in this manner.
Contrary to concerns in some high-income economies about a
flood of exports from the developing world, these exports remain tiny
relative to the size of the major, industrial economies. In 1980,
developing-economy exports of manufactured goods accounted for
just 2.4 percent of total manufactured goods consumption in the Eu-
ropean Community (Ec), North America, and Japan. By 1988, despite
the success of Asian and other exporters, this had grown to just 3.1
percenL Even if exports from all developing economies had grown as
fast as those from Korea during this period, their total share would srill
have amounted to only 3.7 percent of total consumption by 1988
(World Bank 1992b).
The scope for export expansion remains substantial. Indeed, indcid-
ual developing economies, particularly smaller economies currently
contemplating an export-led expansion, could safely assume that de-
mand for their products is infinitely elastic This is even true for Hong
Kong's exports, which in 1990 accounted for a hefty 2.4 percent of

36i
 RASW&NMIRACLE

world merchandise trade (Riedel 1988). Emerging exporters are un-

likelyto approachthis level for many years to come. In 1989, for exam-
ple, Sub-SaharanAfrica'sexporcsof manufactured goods amounted to
just 0.43 percent of world exports (United Nations 1992).
Evenif industrial economy marketsgraduallybecome lessaccessiblc,
changesin the global trading environment are creating new opportuni-
ties for South-South trade. As recently as the mid-1980s, import-
substitution strategies in most developing economies forced
developing-economyexporters to rely heavily on industrial-economy
mark-ets.Since then, formerly closed economies in Asia, Sub-Saharan
Africa, the Middle East, and Latin America have responded to debt
problems, slow growth, and the need to raise fresh capital by becom-
ing more and more outward oriented-and therefore more open to
imports. They have assumed greater obligations under the GA1T and
through numerous regional trading arrangements.' Under these cir-
cumstances, the potential for Soudt-South trade has increased
considerably.
But while global markets are growing, they are also becoming more
and more competitve. Each economy seeking to emulate the East
Asian axport-push model will encounter difficulties because many
other economies are trying to do the same. For example, Collins and
Rodrik (1991) estimate that Eastern Europe and the republics of the
former SovietUnion will increase their share of world trade from 10 to
23 percent, relying initially on agricultural conmnoditiesand labor-
intensivegoods that compete direcdywith the products of other devel-
oping economies.
Despite this heightened competition, severaltrendssuggest that ex-
panding global mnarketswill remain open to developing-economyman-
ufactured exports. Onc of these is the continuing effort to open
markets under the auspicesof the GAIT (see box 7.1). The other is the
growing role of regional trading blocs,which present a major opportu-
nity for increasedglobal trade (see box 7.2). Who providesthe mnarket
if all economies push exports? The dramatic growth of trade within
Asia makes clear that export-oriented economiescreatemarkets for im-
port. To compete internationally, exporcersmust meet global stan-
dards of quality and price, for which they need accessto technology,
capital,and intermediate inputs at world prices.As a resulr, production
in Asia has become more and more international, with capital goods
coming from Japan, labor-intensiveassemblybeing perfbrmed in low-

362.
 POLICIES AN: D.

Pb~~ ~~~~- s7fFiTr er' he;|G

Box7.1 Increased
Access underthe GAiT.
.'THE GA7rTHAS BEENINSTRUMENTALIN FACITATING FREER
trade.Since the end of World War II, tariff rates around the world
havcplummeted.Among the major induistrialeconomics,weighred-
-averagetanff lcevlshave been lowered to blow 5 percent. The cur-
rent UruguayRound of trade negotiationsis addressing -
.~~~~~~~o . .. .V.i s is ad ci.r.
.

*Increasing market access,inudingmore-open trade in agricul-

ru.--
s texties, ad
na l rsourcs and e remo or reuc-
ofrmainin tariff and nonrariffbarriers
-. tion
a Extending the GATTto new areas, principillytiradein services,
standards,and the enforcementof intellectualproperty rights,
and trade-relatedinvestmentrequirementsimposedby govern-
; -rnents
.: -Strengthening
- CAT rules, especiallythose coveringantidump-
ing subsidization,
productstandards, import licensing,"safe-
guards" (temporary import protection) 3 ard dispute selemenr

In short, the Uruguay Round is covering a- broai4 ambitio

agenda.Evenat thislate dareiris impossibleto knowhow successfill
the round will be. Successcould offer inip&rmnt-ropportuiitiesfbr
:developing-economy exports,particularlyin sucb arcasas agriculture
alandtropical products.Progress in eliminating the Muiltifibre
Arrangementwould also bring major benefits;particularlyto new en--.
-trans into textiles and apparel-alth ough here some economies
:- -might lose eisting,rents. -Ar the same time, however, a successful
UruguayRound miy erodeprefrntial reaent thusEfr accorded
..som developingeconomies.Developing-economygovernmentsmay.
'be:obligedtto ughihen-their7:nforcement of intellectal propeAiy
-rights, while easing rade-related investment.reqwremens They'
:couldalso -fi;e.touglie rules of oigm :and anricircumventionmnea-
-.sures-concerningdumping alhgarions. On the whole, hwe 6the
?WUruguay>Round' holds considerable pornise forte, develon
economies.;
-
--Conversy,developing CononMies-wouldlose imporntopporni-
-nies if the round fails.Thcsc would be pardq. lary seriousminagn-
.. :culturebecause the global natuie of gn&ilrure markets,makes
_*regiosounsineffectve Devopi economis udao be-
ffecd byie increased use of unilateral and bilateralre-
- . adversely
.- stictions in te' event oh GArile.
'. U I

363
 ASAM- IRACLE

Box:7.2 Regional.TradingBlocs
' -

''ROGRESSTOWARD GLOBALWINTEGRATONCOULDcapital flows. Stronger regional integraton need not

.be reversediffdworldconory splits into regional be' assdciaredwith higher exteral barriers. Indeed,
trading blocs, each centeredon a major currencyand regionalblocs could havepositive globaleffects,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 integrnwionwithin Europe, for example,will
imposc-subtle and nor-so-subtle protectionist' mea- &cilitate trade with the rest of thc world. A common
esurs'.
Inboth teEurpean Comrnunity; (EC) and -set ofsrandards nakes it easier forallwho wish to se[l
the :iNorth Ainerican'.Free Trade. Ara (NAFMA),for in Europenor just insiders. Tough rules to prevent
ex'ample, fins' are already lobbying Forules olisnit -governments from subsidizing 'domestic firms .aid all
the marker 'access of extra-regionial competiors. 'competitors, not only those located in the European
Typically these rules require that products entering Community.
one memrber economy contain ai inimuma regional -Eierging regional arrangements may provide an
' content to quaJr for entry to the markets of oiher . incentive -for some economies to. open, more rapidly.
, members. (In--the,United Stares-Canada Free Trade than they would havcr otherwise. For example, the
,Area, a 50 percent regional value-added content re- formation of the Ecwas an imortant inipulsc for the
. quirement was required for eligibility for regional Kennedy Round (L.wrence 1993); As Mexico
access. .In the NAFTA,tihere a`re special rules. for auto- moved into the' NAFTA negotiations, it simukmne-
mobiles and tetiles.`The textile provision is particu- . ously made overture toward the Pacific and towaid
.lady protectionist-it requires triple uansformation, Cential and South America. Mexico is 'now seeking
such that'finished:foihing.products have'to becuiit :to join the OECD, is negotiating to form a free trade
..- "..'.and
sewn flom &ibric spun rom Noth American. area wthVenezuelaandCie,andhas. signed agree-
1
- fibers. This could resiilt in considerable trade diver-. :mcnts for freer trade with seven Central American
sion.) Regiona:l blocs'may also,-tighten fbrcemen economes. . -.
'.ofrules against -Ddeiiumping.
as: another means of,' oDevelopingeconuisib' differently
,-prorecion. .by th -Eco
e n of rgn b tha
,- .; con.erns
But -. about regional .protectionist -enden-
currendy enjoy prefrrentiaJacc to'the,U.S. and EC
des are easily overstaie. Unlikc-their predessors,m..arkets under schem such as thc gene ed S
en 5 wblcs go beyondliberaizationoftradeto
th: 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 -of, thee prege erod& 1f- hoever,w 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

Howthe Needfor MarketAccessWillConsbtain

intenweuions
Developingeconomicsseekingto increasetrade and attract invest-
ment will experiencegreater pressuresto open their economiesand
bring their practicesmore closelyin l ne with those in industrial
economies.Those that attempt to usemoreinterventionistversionsof
che export-pushmodel will risk retaliationfrom industrial-economy
markets or punishment under the GATT.For some policy options, such
as financial repression, increasing integration with global markets will
limit the scope for intervention.
Over the years, developing economics have daimed and received ex-
emptions from GATT rules (Whalley 1990). By invoking Article XII, to
safeguard their balance of payments, or Artide XVIII, which allowvspro-
motion of infant industries, developing economies have been able to es-
cape GAIT disciplines more or less indefinitely. Thus developing
economies have been free to implement trade policies of their choice at
bome while benefiting from the openness of industrial economies. But
as GATT shifts from a narrow focus on tariffs to a broad approach that
seeks agreements on rules, developing economies will find it more and
more difficult to hang onto exemptions. Under Uruguay Round pro-
posals, rules governing intellectual property, subsidies, countervailing
duties, and trade-related investment would eventually be the same for
industrial and developing economies. Although developing economies
would have up to eight years to bring their practices into line, the trade
obligations of developing economies clearly are increasing. Changing
rules about dumping illustrate the case. Industrial economies are in-
creasingly applying antidumping and countervailing duty measures to
developing economy exports (Lawrence 1991). These measures will
make it more difficult to follow the example of Japan and Korea, which
protected domestic industry in order to subsidize exports.
Developing economies that seek access to the emerging regional
trading blocs may find that it comes tied to increased obligations that
could inhibit adoption of the interventionist policies. For example, in-
terventionist versions of the export-push model could not be imple-
mented within the EC with its stringent rules on stare aid and its
provisions allowing for the fiee movement of capital and for the com-
munity to implement competition policies. Moreover, no economy ex-
posed to European competition could operate an effective infant
industry strategy.

365
 MASVAN
IRACLE

Does this imply the Asian export-push model is a historic relic?

Clearly not. Indeed, an essential feature of the East Asian export push
has beeatthe 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 everywherelibemlizeand the
NIEs shift to the production of more advanced products. Moreover,
Asian exportershave prosperedduring the past two decadesdespite a de-
cline in global growth and a rise in protectionist actions.
Exports can be promoted by a var.iet of means that are consistent
with developingeconomies'emerging obligationsfor market aco-rs and
limited subsidies. Improved institutions-mak-ing customs services,
duty drawbacks, and fiee trade status for exportcrs and their suppliers
work well-are all consistent with agreements on intemational tradc
and arc effectivemechanisms for export promotion. Access to export
credit, while morecontroversial,wasan ePffctiveinstrument in all of the
-PAEsand remains feasiblefor other developingeconomies.

.* C

We began this book with the objectiveof understanding East Asia's

success.We found that the diversityof experience,the varietyof institu-
tions, and the great variation in policies among the high-performing
Asian economies means that there is no East Asian model of rapid
growthwith equity.Rather, eachofthe eight economieswestudied used
variouscombinations of polides at different times to perform the func-
dons needed for rapid growth: rapid accumulation,efficient allocation,
and high rates of productivity improvement. Most of the policies that
the HPAES used reflectedsound economic fundamentals:they enhanced
the working of markets, helped prices communicate information about
relativeeconomic costs, and fosteredcompetitivediscipline.
Some economies-notably Japan, Korea,arid Taiwan, China-went
beyond fundamentalsand intervened in markes with industrial, trade,
and financialsectorpolicies.On balance,some of theseinterventionscon-
trbuted to their extraondinary
growth, but this was only possiblebecause
of highly unusual historical and institutional circumstances. The govern-
ments in theseeconomieswere more successfi than most in combining
the benefitsof economic cooperationwith the benefitsof vigorouscom-
petidion.They did this bycreating contestswith exportsas the yardsick of
success,with subsidizedcredit and other government-controlledfavorsas
prizesand with governmentofficialsas competentand usuallyhonest ref-

366
 POLICIES AND PRAGMAS

erees In other economies,induding some of the HPAES,similar interven-

tionist policieshave uiled,largelybecausegovernmentsofferedincentives
without a clear link to economic perlormance and strong institutional
support. They offeredprizesbut lackedthe rulesand referees.
What can other developingeconomieslearnfrom the EasrAsian mir-
ade? While there is no recipefor success,there are some positivelessons:
keep the macroeconomystable; focuson earlyeducation; do not neglect
agriculture;use banks to build a sound financial system;be open to for-
eign ideasand technology;and let relativeprices reflecteconomicscarci-
ties. And there are some negativeones: promoting specificindustries or
attempting to leap srages of technological development will generally
fAil;strongly negative real interest rates and large subsidies to borrowers
debilitate the financial system; and directing credit without adequate
monitoring and selection of borrowers distorts allocation. Finally, we
found that a successful export push, whether it results from an open
economy and strong econornic fundamentls, or from a combination of
strong findamentals and prudently chosen interventions, offers-high
economic gains. Of all the interventions we surveyed, those to promote
exports were the most readily compatible with a wide diversity of eco-
noamic circumstances.
East Asia'sown responses to changing domestic and international cir-
cumstances put these lessons in perspective. T7h HPAEsare themselves in-
volved in a continuing process of reform, adapting poliy inscruments
and institutions to achieve the objectives of continued growth with
equity. In many cases these reforrns involve reducing& modifring, or
abandoning policy instruments thar were judged to have succeeded in
the past. Korea's financial sector reform, Indonesia's trade reforms, Thai-
land's promotion of breign investment,and Malaysia'sprivatization pro-
grams are cases in point. The outcome of these initiatives will provide
further valuable lessons on how successful policy instruments shift over
time, as the relative roles of markets, the public administration, and the
private sector change in response to economic and social development
The experience of the HPAEs broadens our understanding of the range
of policies that contribute to rapid growth. it also teaches us that will-
*ingnessto experiment and to adapt policies to changing crcumstances
is a key elemenit in economic success. What we have not discovered fully
is why the governments of these economies have been more willing and
better able than others to experiment and adap . answers go beyond eco-
nomics to indude the study of institutions and the related fields of pol-

367
 1rtCA'*T3
M 1 R A CL E

icics,history, and culture. Taking such large realitiesinto account com-

plicates rather than simplifies the caskof development. The challenge
for policynakers becomes twofold: to uisepolicies compatible with ax-
istillg institutions and to avoid policies tha. exceed the economy7sinsti-
tutional capacity,while at the same time improvinginstitutions in order
to make a wider range of policy options available.

Note
1. In the past, regionalarmngementswere motivated
by the desireto pursueimport-subsrirutionpolicieson a
largerscale and were thereforedoomed to failure;the
morc recentinitiativesare motivatedby a desireto con-
soildate more outward-orientedstrategies.Subregional
arrangementsareflourishing.In Latn America,for exam-
ple, theseincludethe North AmericanFreeTradeArea;
Mercosur(Argentina,Brazil,Paraguay,and Uruguay),
whichaimsat a common marketforgoodsand services by
1994; the Andean Pact (Bolivia, Colombia, and
Venezuela,with Ecuadorand Peru slated to join); the
CentralAmericanCommonMarker(CostaRica,El Sal-
vador,Guatemala,Honduras,Nicaragua,and Panama);
Caricom;anda FreetradeareabetweenChileandMexico
that aims at reducingtariffs to zero by 1998. While
progresscowardformalarrngements witrin Asia have
beenlessapparent(althoughthefreetradeareawithinthe
Associationof SoutheastAsianNationsis an exceprion),
incra-Asian
tradehas increasedrapidly.

368
 S

Bibliographic Note

THISREPORT HAS DIRAWN ON A WIDE RANGE OF MA-

terials,includingspeificallycommissionedbackgrund
studies, outputs of related World Bank-sponsored re-
search, economic and sectorwork carried out by Bank
operationalstaff;and the errensiveacademicliterature
_L on the EastAsian economies.Citations in the text and
footnotesrefer to specificsourcesused. The principalsourcesand specific
contributions to individualchapters arc discussedbelow.Lists of back-
ground and other commissionedpapersand refrences for workscited are
at the end of this note.
Many members of the World Bank Staffassistedin the design of the
researchand made helpfidcomments on one or more drafts of the report.
They indude: YoshiakeAbe, Ramgopal Agarwala,Harold Alderman,
Paul Armington, Ahmad Ahsan, Mark Baird, Gerard Caprio, Ajay
Chhihber, Yoon-JeCho, Jaime de Melo, Shanta Devarajan, Ron Dun-
can, Wllliam Easterly,Gunnar Eskeland, Delfin Go, Jeffery Hammer,
JamesHansen, GregoryIngram, Paul Isenman,EstelleJames,Emmanuel
Jimenez,Christne Jones, Koji Kashiwaya,Homi Kharas,Miguel Kiguel,
Elizabeth King, Michael Kline, Brian Levy, Samuel Lieberman, Will
Martin, Deepak Mazumdar, Saha Meyanathan, Vikram Nehru, Arvind
Panagariya,Martin RIavallion,D. C. Rao, Anandarup Ray,SetsuyaSato,
Andrew Sheng, Lyn Squire, Vinaya Svaroop, Shekhar Shah, Jee-Peng
Tan, Jacques Van der Gaag, Dominique van de Walle, Dimirri Vrtms,
Andrew Warner,David Wheeler, Shahid Yusuf,and Heng-Fu Zou.
Those outside the World Bank who contributedwith comments and
materialindude: AliceAmsden,New Schoolfor SocialResearch;Masahiko
Aoki, Sranfrd University;Jere Behrman, University of Pennsylvania;
Robert assen,Oxford University;Gay Fields,CoomellUniversity,Richard

369
 ,JG
A1':ipKNOTE

Freeman,Harvad University;RichardHanushek,Universityof Rochcster;

Shig e.lruikawa, AyoamaGakuin Universitr,HirohisaKoharna,Univer-
siy of Shizuoka;KazuoKoike,HoseiUniversity,LawrenceJ. Lau,Stanford
University;DavidLindauer,VassarCollege;KazushiOhkawa,International
DevelopmentCenterofJapan; KenichiOhno, UniversityofTsukuba;Moi-
sesNaim, CarnegieEndowmentfbr IntemationalPeace;PeterPetri, Bran-
deisUniversity;GustavRanis,YaleUniversity;ChangyongRhee,University
of Rochester,Helmut Reisen,OECD; Hilton Root, SranfordUniversity;11
Sakong,Institutefor InternationalEconomics,Seoul;Hermanvander Tak;
Shujiro Uraca,WasedaUniversity;Rovert Wade, Institute for Advanced
Srudy,Princeton;and Toni Yanagihara,HoseiUniversity

Chapter1
The data for this chapter are taken fom. the data base compiled for a
researchproject, How Do National PoliciesAffectLong-TermGrowth?
directed byWiliamr Easterlyand Ross Levine.It is an expandedversion
of the data contained in Summers and Heston (1991). A summary of
the resultsof the resch projcct are contained in a symposiumissue of
the Journal of Monetary Economics (forthcoming). Additional data are
drawn from the World Bank Economicand SocialData Base(BESD)and
from the WorldDevelopment Reportdata base. The TF estimateswere
made possiblebv accessto preliminary estimatesof constant price capi-
tal stock series developedby the Banks International Economics De-
partment, under the direction of Vikram Nehru (Nehru 1993). Stanley
Fischer generously provided his country-specific estimates of TFP
growth, which underlay the resultsreported in Fischer(1993).
Maurice Schiff (1993) provided calculations of agricultural incen-
tives in East and South Asia, based on the methods developedin Sciff
and Valdes(1992). The discussionof export performanceis drawn pri-
marily from Bhattacharya and Page (1992). Analysis of the demo-
graphic transition is drawn primarily from Birdsalland Sabot (1993).
Ed Campos (1993) provided most of the backgroundwork on income
distribution. The discussion of private investment draws heavily on
Madarassayand Pfeffermann (1992). Box 1.1 draws on Birdsall and
Sabot (1993). Box 1.2 draws on Pack (1993) and Packand Page (1993).
Substantial research on the economic tmnsition in China has been
conducted by the Macroeconomicsand TransitionalEconomies Divi-
sion of the World Bank'sPolicyResearchDepartment. This researchef-
fort which is led by 1.J. Sirwh is summarized in Chen, Jeffrrson, and

370
 BIBLiO'

Singh (1992). The Asian Miracle ResearchProject has also funded re-
search by professor Shigeru Ishikawa and his associateson the rapid
growth of China'scoastalprovinces.Preliminaryoutputs of that research
will be availablein the fall of 1993.

Chapter2
This chapter draws primarilyon Pack (1993), Pageand Petri (1993),
and Stiglitz (1993). The description of contests also draws on Leipziger
and Kim (1993), Uy (1993), Okuno-Fujiwara (1993), and Ito (1993).
WorldDevelopmentReport1991 provides the basis for the discussionof
the market-friendly approach. Box 2.1 draws on WorldDevelopmentRe-
port 1991. Box 2.2 draws on a paper prepared by Nancy Birdsalland
John Page for the Colloquium on Lessonsof East Asia, February 1993.
Box 2.3 draws on material provided by Jaime de Meld. Helpful discus-
sions with VimodThomas, Robert Wade, and partcipants in the Devel-
opment Economics Study Group, Tokyo, and the Trade and
Development Seminar at StanfordUniversitycontributed to the chapter.

Chapter
3
This chapter draws on Bhartarcharya and Pangestu (1993); Chris-
Dar. Sa±nvwdla,andlu
A.. 'vicnyanona i993; Corcien (I993J;
Dahlman and Sananikone (1993); Leipziger and Kim (1993); and
Salleh, Yeah, and Meyanadhan (1993). Many of the data on macro-
economic perfomance are drawn from a researchproject, How Do Na-
tional PoliciesAffect Long-Term Growth? directed by WilliamEasterly
and Ross Levine.The concept of "export push" is due to Coln Bradford
(1990). Villiam Easterly,Lant Pritchett, and Michael Walton provided
major inputs. Box 3.1 draws on Corden (1993). William Easterly pre-
pared box 3.2. Box3.3 is based on Ira C. Magaziner and Mark Patinkin,
The Silent War (New Yorlc Random House, 1988), by permission.Box
3.4 was prepared by Varuni Dayaratna.

Chapter4
Chiah Siow Yue, Oh-Hyun Chang, Chaianan Samudavanija,Hadi
Soesastro,and Bruce Tolentino provided information and analysisfor
their countries. Hilton Root and MharyShirley made imporrant contri-
butions. Discussion of the role uv small and medium-size enterprises
hac drawn heavily on the resultsof a researchproject, Support Systems
for Small and Medium Enterprises in East Asia, directed by Brian Levy.

37'
 R&fl(c OTE

Chapter 5
Material on human capital accumulation is drawn primarily from
Birdsall and Sabot (1993). Robert Gassen provided the material on
vocational training. Data and analysis on savings and investment are
drawn primarily fiom Uy (1993), Stiglitz (1993a, 1993b), and Sriglitz
and Murdock (1993). YoonJe Cho, Dimirri Vitras, and Andrew Sheng
made important contributions to the analysis of financial repression,
based on the results of a research project, Effectiveness of Credit Poli-
cies in East Asia. The section on tax policy draws on Shah (forchcom-
ing) and on preliminary results ofa research projecr, Tax Policy and Tax
Administration in East Asia. Discussion of the relative price of invest-
ment goods is based on data drawn from Pfeffermann and Madarassay
(1992). Lant Pritchett made important contributions to the chapter.
Box 5.1 was writen by Myriarn Quispe. Box 5.2 was prepared by
Leora Friedberg. Andrew Sheng provided the material on which box 5.3
is based. Box 5A was writren by Jennifer Keller.

Chazer6
The discussion of labor market performance is based primarily on
BirdsalUand 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, 1993b), and on con-
tributions by Andrew Sheng, Yoon-Je Cho, and Dimitri Vittas. The
sections on industrial policy and exports arc based on Pack (1993) and
Pack and Page (1993). Data on axport performance are drawn from the
BESDand World Development Report 1991 data bases. Discussion at the
1993 Canegie-Rochester Conference on Public Policy were helpful in
improving the sections on export extemralities and growth.
Box 6.1 was written by Brian Levy. Box 6.2 was prepared by Leora
Friedbeg. Box 6.3 draws on a recently cornpleted internal World Bank
report on Indonesias industrial seaor. Box 6.4 was prepared by Leora
Friedberg. Box 6.5 was prepared by Danny Leipziger.

Chapter 7
Discussions of the international trading environment and boxes 7.1
and 7.2 are drawn from Lawrence (1993). Discussion of the interna-
tional capital market is largely based on Global Economic Prospectsand
the Developing Countries (World Bank 1993a). Vinod Thomas made a
number of helpful comments.

372.
 BIBLI9CAi

Background
Papers Samuudavanija, Chaianan."HighSpeedGrowthand High
Pertfrmancein a TechnocraticPolity:The Thai Case."
Behrman,JereR., and RyanSchneider."AnInternational Shah, Shekhar,and SubodhC. Marhur. 1992."Sources
Perspectiveon SchoolingInvesrmentin the LastQuar- of LaborMarkecr Flexibilityin EastAsia:A Reviewof
ter Century in Some Fast-Growing Eastern and SelectedIssues."
SoutheasternCountries." Stiglitz,JosephE. "SorneLessonisfromtheAsianMiracle."
Birdsall,Nancy,and RichardH. Sabot."VirtuousCirdes: Uy, Marilou."The Roleof CapitalMarketsand the Role
Human CapitalGrowthand Equityin EastAsia." of Government"
Campos,Ed. "The InstimtionalFoundationsof High-
SpeedGrowchin the High-Performing AsianEcon- AsianEconomies
Studiesofthe High-Peforming
omies: Part I, Insulation Mechanismsand Public
Sectr-PrivateSectorRelations." Thesepapers willbepublished shortlya WartldBankCoazn-
. "The Institutional Foundations of High-Speed seriesdteThe Lessonsof EasrAsia.
tryStudfiesu rsh;lje
Growthin dteHigh-PerforningAsian EconomiesPart Bhatracharya,Amar,andMan Pangesrn.'Indonesia:De-
II, The Bureauacy asan Instruent of Development." velopmentTransformationsince1965and theRoleof
Chia,SiowYuC.'The Roleof Institutionsin Singaporc's PublicPolicy."
EconomicSuccess." Chau, Leung Chuen. "Hong KongjA Unique Case of
Corden, W. Max. "SevenAsian Miracle Economics: Development."
Overviewof Macroeconomic Policies.- Christensen,Scott,DavidDollar,AmmarSiamwvlla, and
Dekle, Robert."RaisingSavingRates:Lessonsfrom the PakornVichyanond."Thailand.The Insirtuionaland
JapaneseExperiencc." PoliticalUnderpinningsof Growth."
Fields,GaryS. "ChangingLaborMarketConditionsand Dahlmnan, Carl,and OusaSananik-one. 'Taiwan, China:
EconomicDevelopmenrin Hong Kong,Korea,Singa- Policiesand InstitutionsforRapidGrowth."
poreandTaiwan.' Kim, Kiwan,and Danny Leipziger."Korea:A Case of
Freeman,Richard. "Does Suppressionof Labor Con- Government-Led Development."
tribute to EconomicSuccess?Labor Relationsand Leipziger,Danny,and VinodThomas."CountryLessons
Markersin EastAsia.' from EastAsia."
Kim, Chang-Shik "Government-Business Relationship Petri,Peter."GovernmentPolicyandEastAsia'sSuccess."
in Korea:SelectiveCaseStudies." Salleh, Ismnael,Kim Leng Yeah, and Saha Dhevan
"Trade Policyand Industrial Transformationin Meyanathan."Malaysia: Growth,Equity,and Struc-
Korea." rural Transformation."
Lawrence,RobertZ. "The GlobalEnvironmentfor the Soon,Teck-Won,and C. SuanTan. "Singapore:Public
EastAsianModel." Policyand EconomicDevelopment?
Murdock,K., and JosephE. Sdiglim"The Effectof Fi- Thomas,Vinod,and YanWang. GovernmentPoLicies
nancialRepressionin an EconomywithPosiriveReal and ProductivityGrowth:Is EastAsiaan Exception?"
Rates."
Pack, Howard. "Induscrialand Trade Policiesin the Studiesof PublicPolicyinJapan
High-Performing AsianEconomies."
Pack,Howard,andJohn M. Page,Jr. "Accumulaton,Er- Thesepaperswerepresented
at a SeminaronPublicPoliy
ports and Growth in the High-PerformingAsian duringJapanf?RapidGrowth. Tolyo,Jzanry 1,993.
Economies.' Horiuchi,Akiyoshi."GovernmentControluJ Financial
Change
Page,JohnM.,Jr.,andPeterA Pctd. Productivity Mechanismin PostwarJapan:A TentativeReport."
andStrategicGrowthPolicyin theAsianMiracle." Itoh, Motoshige. "Industrial Policy and Corporate
Ranis,Gustav."LaborMarkets,Human CapitalandDe- Growthin the AutomobileIndusaryr Japan'sPostwar
velopmentPerformancein EastAsia-" Experience."

373
 IflCNTE
Koike, Kazuo. "Human ResourceDevelopment Systems
in Japan.'
Kuroda, Masahiro,Takanobu Nakajima,and Kanji Yosh-
ioka. "Industrial Policy and Economic Growth in
Japan 1960-1985: PreliminaryReport."
Kuroda, Masahiro, and Shoko Negishi. "Comparison of
the SrructunmlChangesin AsianCountries Outline."
Maki, Atsushi. "The RelationbenveenHousehold Saving
Rate and Economic Growth: Consumer Stimulus to
PurchaseDurable Goods."
Ohno, Kenichi. 'Dynamism ofJapanese Manufacturing.
Evidencefrom the PostwarPeriod."
Okuno-Fujiwara,Masahiro.'Government BusinessRela-
uionship in Japan:A ComparativeInstitutionalAnalysis.'
Shimomura, Yasutami. "Industrial Policy in the Rapid
Growth PeriodcEvidencefrom Early Writings."
Teranishi, Jurn. "Emergenceand Establishmentof the Fi-
nancial Svsremin Postwarjapan: Government Inter-
ven;ion, Indirect Financing and the Corporate
Monitorinp System."
Ueda, Kazuo. "The Industrial Bank of Japan and Public
FinancialInstirutions."
Urata, Shujiro. "Technological Progress by Small and
Medium Enterprisesin Japan.'
Yoshikawa,Hiroshi. 'Demand-Led Growth and Income
Distribution: The PosnvariapaneseExperience."
Yoshino, Naoyuki. 'The Low Interest Policy and Eco-
nomic Growth ofjapan.
References
DcherOJud. 1985. dssa .
Aheuwal,
New Delhi: Oxford UniversityPress.
Aigner, D., and P. Schmidr,eds. 1980. "Specificationand
Estimation of Frontier Production, Profir, and Cost
Functions," a specialissue of Journal of Eoometrics
13:1-138.
Amsden, Alice H. 1989. Asia'sNext Giant- South Korea
and Late tdusriaLzaeon. New Yorlc Oxford Univer-
sity Press.
Ando, Albert, and Franco Modigliani. 1963. 'The Life
Cycle Hypothesis of Savings:AggregateImplications
andTesrts."American Economic Review53(1).
Arrow, KennethJ. 1969."ClassificatoryNotes on the Pro-
duction and Transmission of TechnologicalKnowl-
edge."AmericanEconomicRetiew 59(2):29-35.
374
Asian WallStreetjolrnaL 1993. 'MatsushimaBuilds Ex-
port Basein Malaysia."May 18.
Aw, Bee-Yan,and Amy R Hwang. 1993. "Productiviry
and the ExportMarker:A Firm LevelAnalysis."Penn-
sylvaniaState University,StareCollege,Pa.
Aw, Bec-Yan,and Hong W. Tan. 1993."Training, Tech-
nological Capaciry, and Firm Level Productivity."
WorkingPaper. World Bank, PrivateSecrorDevelop-
ment Department, Washington, D.C.
Balassa, Bela 1982. Development Strteies in Semi-
Industrializing Economies. BalLiliore, Md.: Johns
Hopkins UniversityPress.
- 1991. EconomicPolciesin thePacifxAreaDeveloping
Countrie.tNewYork. New York UniversityPress.
Balass, Bela,and Associates.1971. The Structureof Pro-
tectionin DevelopingCountries.Baluimore,Md.: Johns
HopkinsUniversityPress.
Balassa,Bela,and Marcus Noland. 1988.Japan and the
WorldFconomy. Washington, D.C.: Institute for In-
ternationalEconomics.
TheBanker.1990 Uuly):155-85.
Barro, Robert J. 1989. 'A Cross-Country Study of
Growth, Savings,and Government? NBERWorking
Paper 285 5, Cambridge, Mass.
_. 1990."GovenmmentSpending in a SimpleModel of
Endogenous Growth." Journaloflolitical Economy98
(5, pt2): SI103-25.
. 1991. 'Economic Growth in a Cross Section of
Countries." QuarterlyJournal of Economics105(2):
407-43.
Barro,Robercj., and Jong-Wha Leer1993. 'Incecmaonal
Comparisons of EducationalArtainment." Paper pre-
sented at a conference on How Do National Policies
Affec Long-Run Growth?World Bank, Washingron,
D.C. February.
Behrman,Jere P. 1987. 'Scholing in DevelopingCoun-
rries: Which Countries Are the Under- and Over-
Achievers and What Is the Schooling Impact?"
Economia ofEducaSn Review6(2): 111-28.
Behrman,Jere RP,and Ryan Schneider. 1991. 'How Do
PakistaniSchoolingInvesunentsComparewirh Thosc
of Other Developing Countries?" Williams College,
Williamstown,Mass.
. I992. "An International Perspectiveon Schooling
Investment in the Last Quarter Century in Some
Fast-Growing Eastem and Southeastern Countries."
Backgroundpaper for TheELatAsianMiracle.World

Bank, Policy Research Department, Washington,
D.C.
Bell, Clive, Peter Hazell, and Roger Slade. 1982. Projecr
Evaluation in RegionalPerspective.Baltimore, Md.:
Johns Hopkins UniversityPress.
Berry, R. Albert, and William Cline. 1979. Agrarian
StntutreandProdiuctiviy in DevelopingCounzriesBal-
timore, Md.: Johns Hopkins UniversityPress.
Berry,R Albert,and BrianLevy.1993."Indonesia'sSmall
and Mvledium Industrial Exportersand Their Support
Systems."Report preparedfor the World Bank, Policy
ResearchDepartmcnr,Washington,D.C.
Berry, R. Albert, and Richard H. Sabor. 1981. "Lubor
Market Perfbrmancein DevelopingCountries:A Siur-
vey." In Paul Streetenand RJolly, eds., RecentIsuesin
WorldDevelopmentOxford: PergamonPress.
Bhagwari.JagdishN. 1978.Anatomyand Consequencas of
Exchange Control Regimes. Cambri"-e, Mass.:
BallingerPress, for the National Bureauof Economic
Rescarch.
Bhagwati,JagdishN., andT. N. Srinivasan.1979.'Trade
Policyand Development.' In RudigerDombusch and
Jacob A. Frankel, eds., InternationalEconomicPoliey.
Baltimore,Md.: Johns Hopkins UniversiryPress.
Bblacachatya,Arnar, and John lvi. Page,jr. 1992. "Ad-
justnnent,investment; and Growth in High-Perform-
ingAsian Economies."Paperpresentedaraconference
on the Nev DevelopmentEconomicsin Global Con-
text, American Universityin Cairo, May 28-30.
Bharacharya,Amar, and Mari Pangestu.1993. Indonesist
DevelopmentTransfirmationsince1965 and theRoaeof
Public Policy.World Bank Country Study, Washing-
ton, D.C.
Biggs,Tyler, and Brian Levy. 1988. "StrategicInrerven-
don and the PoliticalEconomyof Industrial Policyin
Developing Countries." In Dwight Perkins aind
Michael Romer, eds., RefirmingEconomicSystemsin
DevelopingCountries.Cambridge,Mass.: Harvard Jn-
sticute for International Development.
Birdsall, Nancy, Dav.d Ross, and Richard H. Saborz
1993. "Underinvestment in Education: How Much
Growth Has Pakistan Forgone?"Paper presented at
the Ninth Annual General Meetng of the Pakistan
C..ocietyof Development Economist, Islamnabad, Pak-
ismt, January.
Birdsall,Nancy, and Richard H. Sabor, eds. 1993a. Op-
portunity Forgvnc Education, Growth,and lnequaiq
in Brazil World Bank,Washington, D.C
BIBLIO-GR
. 1993b."Virtuous Circles: Human Capital Growth
and Equity in East Asia." Backgroundpaper for The
EastAsianMiracle.World Bank, Policy ResearchDc-
partment, Washington, D.C.
Black, Stephen L., and Lance Taylor. 1974. "Practical
General Equilibrium Estimation of Resource Pulls
undcr Trade Liberalization."joUrnalof International
Economics4(1): 37-58.
Blomstrom, Magnus. 1989. Foreigninvestments and
SpilloversLondon: Roudedge
Blumenthal,1. 1970. "Savingsin PostwarJapan." East
Asian Research Center, Harvard University, Cam-
bridge,Mass.
Boadwvay, Robin, Dale Chua, and Frank Flatters.Forth-
coming. "Investment Incentives and the Corporate
Tax Svstcm in Malaysia."In Anwar Shah, ed., Fiscal
Incentivesfrrinvestmentin DevelopingCountries.New
York: OxfordUniversityPress.
Boneili,R. 1992. "Growth and Productivirvin Brazilian
Industries:Impacts of Trade Orintation." Journal of
Development&Economics39(1): 85-111.
Borrus, Michael,Laura D'Andrea Tyson, and John Zvs-
man. 1986. "Cr- ir Policyand International Comnpe-
tition." In PawII rugman, ed., Srategic TradePolicy
and the New International Economics.Cambridge,
Mass.:MITPress.
Boskin,MichaelJ., and LaurenceJ. Lau. 1992. "Inrerna-
tional and Intertemporal Comparison of Productive
Efficiency:An Application of the Meta-Producton
Function Approach to the Group of Five Countries."
Economi StudiesQucarterly 43(4): 298-312.
Bradford,Colin L, Jr. 1986. "East Asian Models:Myths
and Lessons."In John P. Lewisand ValerianaKallab,
eds., DevelopmentStrates ReconsideredU.S. Third
World Policy Perspectives 5. Washington, D.C.:
OverseasDevelopmentCouncil.
. 990. Policy Interventionsand Markets: Develop-
mentStrategyTypologiesand PolicyOptions."In Gary
Gereffi and Donald L. Wyman, eds., Manufacturing
Miracles:Pathsofindurialization in Latin Americaand
EastAsia.Princeton,N.J.: Princeton UniversityPress.
Brimble,PeterJohn. 1993. "IndustrialDevelopmentand
ProductivityChange in Thailand." Ph.D. diss.,Johns
Hopkins University,Baltimore,Md.
Bruton, Henry J., G. Abeysekera,N. Sanderatne,and Z
A. Yusof. 1993. ThePoliticalEconomyof PoverJ Eqa-
uiy, and Growth:Sri Lanka and MalaysiaNew York
Oxford UniversityPress.
375
 OG4R~KflflC-NOTE
Calomiris, Charles W., and Charles P. Himmelberg.
1993. "Directed Credit Programs for Agricultureand
Industry: Arguments fiom Theory and Fact." Papar
presented at the Annual Bank Conference on Devel-
opment Economics,World Bank, Washington,D.C.,
May3-4. May3-4.
Canning, David, and. Marianne
= .
Fay. 1993. 'The
. _ Effect
,, _ of
Infratucture Networks on Economic Growth." Co-
lumbia University,New York,January.
Caprio, Gerard, Jr., IzakAnyas, and James Hanson, s
Forthcoming. Financial Reform. Theoryand F.erl-.
Fnot.Camrnidge:Canbridge URrivrrsityPress.
Cic, Caridge:, Camidg e, Uni Press. SInterpreton
Carroll,
Crspe David and LawrenceH. Sp.ca
-imers.193. "savings and Growth: Arcgicerpreca-
dion." Paper presented a1tthe Carnegie-Rochester
Public Policy Conference, Bradley Poliqy Research
Cen_ r April
Center, Api 23--24.
23-2.
tChan,Steve,and Cal Clark. 1992. FA&xiii Fior
(Than,
Stuev an CaizCark 1992. Fesgt
andFotun,i
Thwa pment Navigatingheo
tweenScyllaand Chanybdis.London: Roudledge.
Chant, John, and Man Pangestu. Forthcoming. An As-
sessmentof FinancialReform in IndonesLiz
s t- 1983-90."
' T
In Ge_rd Caprio,Jr., IzakAti, andjames Hanson,
eds., Financia Rqornm: Theoryand Eaperince. Cam-
britG: Canibrc;ge 1-jniversity l ress.
bridg Cdamrd U erst Pres.
Che,
K. Ewad
979 Hpergrow:h in Asian Ec_n,
omier: A Comparative Sturvey of hong k'on~g
JpaP,
Korea,Singaporeand Taiwan.London: Macmillan.
Chen, Kang,Gary H.JefFerson,and InderjirSingh. 1992.
"Lessonfrom China's Economic Reformn."Joumna of
ComparativeEconomica 16:201-25.
Chenery, Hollis B. 1959. The Interdependence of In-
vestment Decisions."In Moses Abramovit ecL, The
AllocationofEconomicResource.Sianford,Calif.:Sran-
ford UniversityPress.
. 1960. "Patterns of Industrial Growth." American
EconomicReview50(4): 62454.
-. 1986. "Growth and TransirM2aton.' In HoloisB.
Chenery, Sherman Robinson, and Moshe Syrquin,
eds., Industrialzation and Growth: A Comparative
Study.New York Oxford UniversityPress.
Cheng, Hang-Sheng,ed. 1986. Finncial Policyand Re-
fonn in PacificBasin Countries.Lexington,Mass.:Lex-
ington Books.
Chia, Ngee Choon, and John Whalley. Forthcoming.
"Paterns in InvesrtmentTax Incentivesamong Devel-
oping Countries.' InAmwarShah, ed., FiscalIncntive
376
frInvestmentin DevelopingCgAnntries. NewYork Ox-
ford UnivcrsityPress.
Chia, SiowYue. 1992. 'The Roleof Institutions in Sing-
apore's EconomicSuccess." Backgroundpaper for The
EastAsian Miracle.World Bank, PolicyResearchDe-
~~~~~~~~~parmacnt,
Washington, I).C.
PautnCnr,
~~~~~Chiu, Paul C.WH. 1992.
i onenD.C.
"Money and FimnancalMarkets:
The Domestic Perspective."In TaiwanfromDevelop-
ing to Mature Economy.Colorado:Westview.
Cho, Yoon Je, and Thomas Hellmann. 1993. 'Govcrn-
menr Intervention in Credit Market: An Alternative
of Japancse and Korean Experiences
from the Ncw Institutional Economics Perspective."
Paper prepared for a World Bank researchproject on
Th Efetvns*fCei-Plce nEs sa
World Bank, FinancialSrecitPoriDevelmEt Dcpa.r-
Bankl, FinancialSectorDvcvdopmenc
Worlci Washington,
mentc, Depart-
D.C.
m-siWahigon DC
Cho, Yoon Je, and Deena Khatkare. 1939. LessonsofFi-
nancaalLiberalizationin Asia: A ComparativeStudy.
World Bank DiscussionPaper 50. WashingEon,D.C.
Cho, YoonJe,andJoon-Kyang Im. 1993.'Cedit Poli-
T I
- YnJ
A v^ ttFa~~cesand IndJoonKyg
Industralization ofKim.1
Korea."CWedt Poli-
World Bank,
Finanal SectorDevelopmentDepartnent, Washing-
t, Fcbra.
Chrisrensen,Scott R., David Dollar, Ammar Siamwala,
Q
and Pakor Vichyanond.1993.sInstitutional andPolit-
ical Bases of Grwth-hiindztng Policies in Thailand
Lessonsof East Asia World Bank Counnty Study.
Washington, D.C.
Carke, Gerge. 1992. "More Evidenceon Income Dis-
rribution and Growth." Policy Research Working
Paper 1064. World Bank, Policy Research Depart-
ment, Washington. D.C.
Cohen, S. I. 1985. 'A Cost-BenefitAnalysisof Industrial
Training." Economics of Edication Review 4(4):
327-39.
Collins, Susan M., and Won-Am Park. 1939. "External
Debt and Macroeconomic Performance in South
Kora" In Jeffey D. Sachsand SusanM. Collins,eds.,
DevelopingCountryDebt and EconomicPeifrmance,
vol. 3. Chicago: Universityof ChicagoPress
Collins, Susan M., and Dani Rodrik. 1991. Easter Eu-
ropeand theSoviet Unionin the WorldEconomy.Wash-
ington, D.C.: Insitute for International Economics.
Corbo, Vittorio, and Klaus Schmidt-Hebbel. 1991.
"Public Policyand Savingsin DevelopingCountries."
Jownl ofDevelopmenitEconomics36(1): 89-115.

Corden, W. Max. 1971. The Themryof Protection.Lon-
don: Oxford UniversityPress.
-. 1974. TradePoicy and EconomicWelure. London:
Oxford UniversityPress.
-. 1993. "SevenAsian Miracle Economies: Overview
of Macroeconomic Policies." Background paper for
The Est Asian Miracle.World Bank, PolicyResearch
Department, Washington,D.C.
Dahlman, Carl J. 1990. "Technology Strategy in the
Economyof Taiwan:ExploitingForeign Linkagesand
Investing in Local Capability." World Bank, Private
SectorDevelopmentDepartment, Washington, D.C.
Dahlman, CarlJ., and Peter Brimble. 1990. 'Technology
Strategy and Policyfor Industrial Competitiveness:A
Case Study in Thailand.' Industry Series Paper 24.
World Bank, Industry and Energy Department,
Washington, D.C
Dahlman, Cad J., and Valdarcs Fonsea. 1987. "From
TechnologicalDependence ro Technological Devel-
opmenr: The Case of the Usiminas Steel Plant in
BraziL" In Jorge Kamz, and others, eds., Technoloy
Generationin Latzn American ManuracinUg frig1nd.
tries.London: Macmillan.
Dahuman,CarlJ., and Ousa Sananikonc. 1993. E:conomic
PoalcesandInstiutions in theRapidGrowthof Taiwan,
China. Leson of Eas Asia. World Bank Country
Study. Washington, D.C
Dasgupta, Partha, Amarty Sen, and Stephen Marglin.
1972. Guidelinesfir PrjecetEvaiuajon. New Yorlc
United Nations IndustrialDevelopmentOrgization.
De Long,J. Bradford,and LawrenceH. Summers. 1991.
"Equipment Investment and Economic GrowtdL
Quareer4journalofEconomics10C6(2): 445-502.
1993. 'How Robust Is the Growth-Machinery
Nexus?' Paper presentedat a conference on How Do
Natonal PoliciesAffect Long-Run Growh? World
Bank, Washington, D.C., February.
de Melo,Jaime,Carl Hamilton,and L AlanWinters. 1990.
YVolunwiyExportRestminm :ACaseSrudyFocusingon
EflecrsinExporting Countries."Paper 464, Institute for
InternationalEconomicStudies,Stockholm.
de Melo, Jaime, and ShermanRobinson. 1992. "Produc-
tiviry and Externalities: Models of Export-Led
Growth."Jounl ofJnternwtionalTradeand Economic
Development1:4169.
Dearon, Angus, and Christina Paxson. 1992 'Savings,
Growth, and Aging in Taiwan?' Paper presentedat an
BIBLAjt
NBERConference on Aging, Cancel Bay, U.S. Virgin
Islands,May 7-10.
Dekle, Robert. 1988. "Do the JapaneseElderly Reduce
Their Total Wealth?" ReischauerInstitute of Japanese
Studies, Harvard University,Cambridge,Mass.
__. 1993. "RaisingSaving Races: Lessons from the
JapaneseExperience."Backgroundpaper for TheEast
Asian Miracle.World Bank, PolicyResearch Depart-
ment, Washington, D.C
Dollar,David. 1990. "Ounvard Orientationand Growth:
An Empirical Study Using a Price-BasedMeasure of
Openness."World Bank, EastAsiaand PacificRegion,
Country DepartmncetI, Washington,D.C.
_. 1992. "Outward-Oriented Developing Economies
Really Do Grow More Rapidly: Evidence from 95
LDCs, 1976-1985.' EconomicDevelopmentand Cul-
turd Change 40(3): 523-44.
Dollar, David, and KennethSokoloff.1990. "Patterns of
Producivity Groth in Soudl Koreai MaencEcturnng
Instnes, 39-7.
cr33(2): 309-27.
Dombusch, Rudiger,and StanleyFischer.1993."Moder-
ate Inflation." World Bank EconomicReview 7(1):
1-44.
Easterly, WVlliam.1993. "Explaining Miracles: Growth
RegressionsMeettheGangofFour." Paperpresented
ar the NBER'sFourth Annual East Asian Seminar on
Economics,San Francisco,June 17-19.
Easterly,William, Michael Krerner, Lant Pritcherr, and
LawrenceH. Summers. 1993. "Good Policyor Good
Luck?Country Growth Performanceand Temporary
Shocks." Paper presented at a conferenceon How Do
National Policies Aficcr Long-Run Growth? World
Bank, Washington, D.C., February.
Easterly, fiiam, and SergioRebelo.1993. "FiscalPolicy
and Economic Growth: An Empirical Investigation."
Paper presented at a conferenceon How Do National
Policies Affect Long-Run Growth? World Bank,
Washington, D.C, February.
Easterly, William, and Klaus Schmidt-HebbeL Forth-
coming. "The Macroeconomics of Public Sector
Deficits A Synthesis."In Wlliam Easterly,CarlosRo-
drigucz,and KlausSchmidt-Hebbel,eds., PublicSector
Defiit and MacroeconomicPerjirmance.New York.
Oxford UniversityPress.
Edwards, Chris. 1990. "Prorection and Policy in the
MalaysianManufacturingSector." PolicyAssessmnent of
theMaysian Indaried PolicyStmu (Mw) and the In-
377
 LOGR ATPMIC--N O T E
dustrialMasterPlan (tw1), Voieme3. Vienna: United
Nations IndustrialDevelopmentOrganization.
Edwards,Sebastian.1988. ExchangeRateMisalignmentin
DevelopingCountries.Baltimore,Md.: Johns Hopkins
Universiy Press.
-. 1992. "Trade Oricntation, Distortions and Growth
in Developing Countries." Journal of Development
Economics39(1): 31-57.
Elias, VictorJ. 1990. "The RoleoFTotal Factor Produc-
tivity on Economic Growth.' Background paper for
the Worlk DevelopmentReport 1991. World Bank,
Office of the Vice President, Development Econom-
ics, Washington, D.C.
Erzan, Refik, KiroakiKuwahara,SerafinoMarchese,and
Rene Vossenar. 1989. "The Profile of Protection in
DevelopingCountries." UNCZADReview1:29-73.
Fanelli, Jose Maria, Roberto Frenkel,and Lance Taylor.
1992. "The World Development Report 1991: A
Critical Assessment."Preparedas part of an UNCTAD-
executed project to assist tizc Intergovernmental
Group of Twnet-Four.
Far EasternEconomicRevitew.Various issues.
Farrell,M. J. 1957. "The Measurementof ProductiveEf-
ficicncy."Journal of tht RoyalStadsticalSociety,Series
A 120 (part 3): 253-8 1.
Fecher, M., and S. Perelman. 1992. "Industrial Effi-
ciency in OECD Countries." In R. Caves, ed., Indus-
tialEEfficiernyinSilNaionss Cambridge, Mass.:Mrr
Press.
Federazion of Bankers Association of Japan. 1991.
'Japanese Banks."
John Pci, ShirleyW. Y. Kuo, and Gustav Ranis. 1981.
The Taiwan SuccessStory: Rapid Growth with Im-
provedDiributon in theRepblic ofChina,1952-79.
Boulder, Colo.: Westview.
Fields, Gary S. 1992. "Changing Labor Market Condi-
dions and Economic Development in Hong Kong,
Korea, Singapore,and Taiwan, China" Background
paper for TheEastAsianMiracle.World Bank, Policy
ResearchDepartment, Washington, D.C.
Fischer, Stanley. 1993. "Macroeconomic Factors in
Growth." Paper presented at a conference on How
Do National Policies Affect Long-Run Growth?
World Bank, Washington, D.C., February.
Foundation for Advanced Informarion Research. 1991.
"The Development of Financialand Capiti Markets
in the Asia-PacificRegion.' Tokyo, Japan. September.
378
Freeman, Richard. 1992. "Does Suppression of Labor
Contribure to Economic Success? Labor Relations
and Markets in East Asia.' Backgroundpaper for The
EastAsianMiracle.World Bank, PolicyResearchDc-
partmenE,Washington, D.C.
Freeman, Richard, and Martin L Weitzman. 1987.
'Bonuses and Employment in Japan." Journal of
Japaneseandf!nternanoonalEeonomies1(2): 168-94.
Frischtak, Claudio R., Bita Hadjimichael, and Ulrich
Zachau. 1989. CompetitionPolcesfor Ina¼strializing
Countries.World Bank Policy and Research Seriec
Paper7. Washington, D.C.
Fry, MaxwellJ. 1988. Money. Interest,and Banking in
EconomicDeveiopment.Baltimore, Md.: Johns Hop-
kins UniversityPress.
Fukasaku, Kiichiro. 1992. Economic Regionalization
and Inta-Industry Trade: Pacific-AsianPerspecaives."
TechnicalPaper 53, OECD DevelopmentCentre, Padis.
Gelb, AManH. 1989. "FinancialPolicies,Growth and Ef-
ficiency." PRE Working Paper 202. World Bank,
Washington, ''.C.
Gelb, Alan H.,Ju..n Knight, and RichardH. Sabo: 1991.
'Public Sector Emp!oyment, Rent Seekingand Eco-
nomic Growth." Economic Journal 101(408):1186-99.
Gerschenkron,Alaxander. 1962. EconomicBackwardness
in HistoricalPrspecive. Cambridge, Mass.: Harvard
UniversityPress.
Ghose. T. K 1987. "The Banking System of Hong
Kong." Burterworth (Asia)Ltd.
Giovannini,Alberto. 1983a. "Essayson ExchangeRates
and AssetMarkets." Ph.D. diss., MassachusettsInsti-
ruteof Technology(Mrr),Cambridge, Mass.
-. 1983b. "The Interest Elasticityof Savingsin Devel-
oping Countries: The Existing Evidence." WorldDe-
velopment11(7): 601-07.
Gordon, David L. 1983. DevelopmentFinancCompanies,
Stateand PrivatelyOwned-A Review.World Bark Staff
Working Papers578. "X'orldBank, Washington,D.C.
Goto, Akir, and Ryuhei Wakasugi. 1988. "Technology
Policy." In Ryutaro Komiya, Masahiro Okuno, Ko-
taro Suzumura, eds., IndustrPolicy oflapan. New
York-AcademicPress, Inc.
Granger, Clive W. J. 1969. 'Investigating Causal Rela-
dions by Econometric Models and Cross-Spectral
Methods."Economerica37: 428-38.
Greenwald,B., and Joseph E Stiglirz. 1986. "Externali-
ties in Economieswith Imperfect Information and In-

complete Markers." Quarter4'journal of Econonic
101(2): 229-64.
_. 1938. 'Pareto Inefficiencyof Market Economies:
Scarch and EfficiencyWage Models." AmericanEco-
nomicReview78(2): 351-55.
Gregory, Peter. and Deepak LaL.1977. "Demand, Sup-
ply, and Structure:A Cross-SectionalAnalysisof Man-
uficuring Wages.' World Bank, Employment and
Rural Dcvelopment DivisionWorking Paper, Wash-
ington, D.C.
Grossman, Gcne, and Elhanan Helpman. 1991. Innova-
onvand Growth in the Globe:!Fconomy.Cambridge,
Mass.:mT Press.
Gupta,KanhayaL. 1984 a. FinancedamlEconomicGrowth
in DevelopingCounties.London: Croom Hclm.
-. 1984b. "Financial Liberalizarion and Economic
Growth: Some Simulation Results." Jouirnalof Eco-
nomicDevelvPment9(2):25-34.
H-adad, Mona, and Ann E. Harrison. 1991. "Are There
Dynamic Externalities fiom Direct Foreign Invest-
ment? Evidence From Morocco." Indus-ry Series
Paper No. 48. World Bank, Industry and EnergyDe-
parument,Washington, D.C.
Forthcoming. "Are There PositiveSpilloversfrom
Direct ForeignInvestment?Evidencefrom Panel Dara
for Morocco."JournalofDerelopment Economics.
Haggard, Stephan. 1990. Pathwaysfrom the Perzphery.
Ithaca, N.Y.:Cornell UniversityPress.
14alvorsen,Robert Forthcoming. 'Fiscal Incentives for
Investment in Thailand." In Anwar Shah, ed., Fiscal
IncendvesJfrInvestmentin DevelopingCountries.New
Yortc Oxford UniversityPress.
Hamilton, Carl B. 1989. "The Political Economy of
Transient 'New' Protecrionism." Weirschafi'cs
Archive,Reviewof WorldEconomics 125: 522-46.
ed. 1990. Textles Tradeand the DevelopingCoun-
ins: Eliminating the Maid-Fiber Arrangementin the
1990s Washington, D.C World Bank
Hanson,james, and Roberto Rocha. 1.986.'High Inter-
est Rates, Spreads, and the Costs of Intermediation:
Two Studies." Induscryand Finance Series8. World
Bank, Washington, D.C.
Hanushek, Eric A..,and Richard H. Sabo. 1991. "Notes
on Changes in Educational Performances."A paper
preparedfor PEERMeetngs. Boston,Mass.
Harris, John, Fabio Schiantarelli,and Miranda G. Sire-
gar. 1992. 'Thc Effort of FinancialLiberalizationon
BI BLOO,G.X
Firms' Capital Structure and Investment Decisions:
Evidencefrom a Panel of Indoncsia Establishments,
1981-88." Paper presentedat a conferenceon the Im-
pacaof FinancialReform, World Bank, Washington,
D.C., April 2-3.
Harrison, Ann E. 1991. "Openness and Growth." PRE
Working Paper809. World Bank, Washington, D.C.
Hendrie, Anne. 1986. 'Banking in the Far East: Struc-
turesand So:,rcesof Finance." FinancialTimes Busi-
nessInformation, London.
Hill, Anne M., and ElizabethKing,cds. 1991. "Womcn's
Education in Developing Countries: Barriers. Bene-
fits, and Policy." PHREE Background Paper Series.
World Bank,Washington [D.C.
Hon, Wong Seng. 1992. "Exploiting Information Tech-
noAogy.ACase Study of Singapore." World Develop-
ment20(12): 1817-28.
Hong, Won-Tal. 1989. "Korean Economy at Cross-
roads." SasangQuarterly1 (Summer).
Horioka, CharlesYuji. 1990. 'Why Is Japan's Household
SavingsRate So High?A LitcratureSurvey."Jounalof
JapaneseandInternaional Economics4:49-92.
. 1991. The Determinants of Japatn'sSavingsRate:
The Impact of the Age Srructure of the Population
and Other Factors." The EronomicStrdies Quarterly
2(3): 237 53.
Horiuchi, Aiivoshi, and Qing-Yuan Sui. 1992. 'The In-
fluence of the Japan Development Bank Loans on
Coporare Investment Behavior."
Hou, Chi-Ming, and Gee San. 1990. "National Systems
Supporting TechnicalAdvancein Indusry. The Case
of Taiwan."Taipei,Taiwan, China.
Hughes, Helen. 1991 'East Asian Export Success" Re-
search School of Pacific Studies, AustralianNational
University.
Hulten, CharlesJr. 1989. "The Embodiment Hypothesis
Revisited."Universityof Marvland,CollegePark, Md.
Huseinga, Mark, and ChrisEosKostopolous.1992. "Op-
tions for Reform of the BusinessRegulation System:
Nepal." Country Report 8, Center for Instirutional
Reformand the Informal Sector,CollegePark, Md.
Hwang,Y. Dolly. 1991. TheRiseofa New WorldEconomic
Power.PostwarThiwan.New Yorlc Greenwood.
Imai, Ken'ichi. 1988. "Industial Policyand Technologi-
cal Innovation." In R. Komiya, M. Okuno, and K.
Suzumura.1988. IndustriidPolicyoffapam New Yorkl
AcademicPress.
379
India. Various issues.ReserveBank ofindia Bdlletin.
Indones;ia.Various issues.IndonesianFinancia Statistics.
Jakarta: Bank of Indonesia.
Ingramn,James C 1971. EconomicC&angein Thailna
1850-1970. Scmnford, Calif.:StanfordUniversiryPress.
Inoki, Takenori. 1993. "Retrcment BehaviorofJapancsc
Bureaucraciesand rhe Role of Public Corporations:
An Economic Interpretation of Human Resources'
Mobility from Central Government .o PublicCorpo-
ration." World Bank, Economic DevelopmentInsci-
tuteWorking Paper,Washington, D.C.
IMF (Inernational Monctary Fund). 1993. World Eco-
nomic Oulok. Washi.gton, D.C, May.
International Road Federation.Variousyears. WorldRoad
StatisticsWashingronjD.C.
rsu(Indonesian Economists'Association).1990. The Un.
demandng ofEconomicDemocracy. a.
Ishikawa,T. 1987. 'Savings: Structural Determninantsof
Household Saving auid the Financial and Tax Sys-
temns? In Koichi Hamada, Malcato Kuroda, and
AkiyoshiHoriuchi,ed., Macroeconomic Analysisof.the
Japanesetmx.Amtrdam: North Holland.
Jto,
Takatoshi1992Thejapanese
RncAnoerdamNorthrHolland.Katseli,
Io Takatoshi. 1992. TheJapaneseEconomy.Cambridge,
_. 1993. "Industial Policyfor Development Japanese
Experience and Replicabiliry." John F. Kennedy
School of Government, Harvard University, Cam-
bridge,Mass.
Itoh, Motoshige, and KazuharuKiyono. 1988. "Foreign
Trade and Direct Investment.' In Ryuraro Komiya,
Masahiro Okuno, Kotaro Suzumura, eds., Inditfil
Policyoffapan New York:AcademicPress.
Itoh, Moroshige, Kazuharu Kiyono, Masahiro Okuno-
Fugiwar, and KataroSuzumura 1988.'Industrial Pol-
icy as a Corrective to Marker Failures." In Ryutaro
Komiya,Mashiro Okuno, and KoraroSuzumura,eds.,
IndustiaPoficy offapan.New York:AcademicPress.
1991. EconomicAnalysisof ndustrialPoiy. New
York:AcademicPress.
Itoh, Motoshige, and S. Urata. 1993. "Small and
Medirun-Size Enterprise Support Policy of Tapan.
World Bank Policy ResearchDepartment, Washing-
ton, D.C.
Jaff, Ad-amB. 1986. 'Technological Opportunity and
Spilloversof R&D: Evidence from Firms' Patens,
Profits and Market Value."AmericanEconomicReview
76(5): 984-1001.
380
JDB/JERI (Japan Development Bank and Japan Economic
Researrh Institute). 1993."Policy-BasedFinance:The
Expericnceof Post.ar Japan." Tokyo.
Japan. Various years. &otnomicStaisticsAnnual Tokyo:
Bank ofjapan.
-- Neitio
val IncomeAccount ofjapan. Tolgro:Japanese
Economic PlanningAgency.
Jitsuchon, Somchai. 1991. "Rctrospecusand Prospectsof
Thailand's EconomicDevelopment." Working Paper
2, JapaneseEconomicPlanningAgency,Tokyo.
Johnson, Chalmers. 1982. MITI and theJapanrseMiracle.
Stanford, Calif.: Sranford UniversityPress.
Jones; Leroy, and Fadil Azim Abbas. 1992. "Malaysia:
Background, Malaysian Airlines, Kelang Container
Terminal, Sports Toro." Paper prcsented at a World
Bank Conference on The Welfare Consequencesof
SelhingPublic Enterprises, Washington, D.C., June
11-12.
Jorgenson, Dale W., Masahiro Kuroda, and Mieko
Nishimizu. 1987. 'Japan-U.S. Industry-Level Pro-
ductivityCorrparisans 1960-70."joumalofJap
aIo .
adlteniodEooisI()
LoukaT. 1992."Foreign Direct Investment and
Trade Interlinkages in the 1990s: Experience and
Prospects of Developing Countries." Background
paper for United Nations Symposium on Globaliza-
tion and Developing Countries, The Hague.
(Reprinted as Ccnter for Economic Policy Research
DiscussionPaper 687, London, July 1992.)
Katz,Jorge,ed. 1987. Techno Generation in LauinAnmr-
icanManufacturngInduarie. London:Macrillan.
Kaufman, George G. 1992. Banking Structuresin Major
Count7is. Boston: KluwerAcademicPublishers.
Kawaura,Alihiko. 1991. "Identification of the Sizeand
Direction of the Post-War Credic Allocation in
Japan. World Bank, Financial Sector Development
Department, Washington, D.C.
Kim, C. 1990. 'Labor Market Development." In Jene K
Kwon, cd., KoreanEconomicDevelopmentNew York
Greenwood.
Kim, Chang-Shilc 1993a. "Government-BusinessRela-
tionship in Korea: Selectve Case Studies.' Back-
ground paper for TheEasrAsianMiracLcWorld Bank.
PolicyResearchDepartment, Washington, D.C.
. 1993b. "Trde Policy and Industrial Tranfrma-
dion in Korea" Backgroundpaper for The EastAsian
liiracle.World Bank, Policy Research Department,
Washington, D.C.
Kim, Ji Hong. 1990. "Korean Industrial Policy in the
1970s: The Heavy and Chemical Industry Drive."
Working Paper 9015. Korean DevelopmentInstitute,
Seoul, Korea.
Kim,Jong-l1,and LaurenceJ. Lwu.1993. "The Sourcesof
EconomicGrowth of the East AsianNewly Industrial-
ized Countries." Paper presented at a Conferenceon
the EconomicDevelopmentof the Republicof China
and the Pacific Rim in 1990 and Beyond, Stanford
University,Stanford, CaliE,April.
Kirn, K, and J. K Park 1985. Sources of Econonic
Growth in Korea. Seoul:Korea Development Institute.
Kim, Kihwan,and DannyLeipziger. 1993. Korea:A CGst
of Government Led Development Lessonsof East Asia.
World Bank Country Study, Washington, D.C.
Kim, L., and J. Nugent. 1993. 'Korean SMEs and Their
Support Mechanissr: An Empirical Analysisof the
Roleof Govermrent and Ore:erNon-Profir Organiza-
dions.? Worid Bank, Policy Research Departrment,
'Washington,D.C.
Kim, S. 1987. In Service Trainingas an Insmnrentfbr the
DevelopmentofHwman Resourcesin Korea.Paris: OECD
DevelopmentCenter.
1988. Japan Civil Service Sys Structre,
sonne andPolitict New York: Greenwood.
ICm, Y. C., and J. K Kwon. 1977. "The Utilization of
Capital and the Growth of Output in a Developing
Economy: The Case of South Korean Manufactur-
ing.?Journd ofDevelopment Economics 9: 265-78.
King, Elizabeth, Katherine Anderson, and Yan Wang.
1993. "Feedback Effects on the Labor Market on
SchoolingChoice in Malaysia? Paperpresented at the
April 1992 Meetingsof the Population Associationof Kunio, Y. 1988. The Riseof Ersaz Capitalism in South
America,Cincinnati, Ohio.
King, Robert,and RossLevine.1993. "Financz,Entrpre-
neurship and GrowdLhPaper presentedat a conference
on How Do National Policies Affect Long-Run
Growh? World Bank, Washington,D.C., February.
Knight, John, and Richard H. Sabor. 1991. Education;,
Produciviy, and Inequality. New York.cOxford Uni-
versityPress.
Komiya, Ryutaro. 1990. The Japanese Econoyn: Trade,
Induatry, and Government. Tokyo: University of
Tokyo Press
Komiya, Ryuraro. Masahiro Okuno, Kotaro Suzumura,
eds. 1988. IndusialPolicy ofjapan. New York Acad-
emic Press.
Konishi,Y. 1989. A Quantitative Anadis of Eduational
Policies in Postarjapan ResearchInsticute for Eco-
nomics and BusinessAdministration, Kobe Univer-
sicy,Kobe,Japan.
Korca, Rcpublic of. Various years. Economic Statistics
Yearbook.Seoul: Bank of Korea.
_. Various issues.The FinancialSystem in Korea. Seoul:
Bank of Korea.
-. 1978. Major Statistics ofthe KoreanEconomy. Seoul:
EconomicPlanning Board.
Krueger. Anne 0. 1978. "AlternativeTrade Strategies
and Employment in LDcs."American SconomirRe-
view. 68(2): 270-74.
-. 1984. "ComparativeAdvantage and Development
PolicyTwenty YcarsLater." In MosheSyrquin, Lance
Taylor, and Larry E. Westphal, eds., Economic Struc-
tureandPerformance.New York:AcademicPress.
_. 1993. "East Asia: Lessons for Growth Theory.'
Paper prcsented at the Fourth Annual East Asian
Seminar on Economics, National Bureau of Eco-
nomic Research,San Francisco,Calif.,June 17-19.
Krueger, Anne O., and Jagdish Bhagwati. 1973. 'Ex-
change Control, Liberalization Attempts, and Eco-
nomic Development." American Economic Review
63(2): 419-27.
Krgrnan, Paul R 1986. Introduction: New Thinking
about Trade Policy." In Paul R. Krugman, ed., Strate-
gc Trad Polcy and the New Intenational Economics.
Cambridge, Mass.: MIT Press.
Kumar, Nagesh 1985. Cost of Technology Imports:
The Indian Experience.? Economic and Political
Weekly 20: Ml 03-114.
EsAsia. New York.Oxford UniversityPress.
Kuo, S. 1983. The Taiwan Economy in Transition. Boul-
der, Colo.: Westvzew.
Kuo, Wan-Yong. 1976. 'Income Distribution by Size in
Taiwan Area. Changes and Causes." Indmut of Free
Cbina 45.
Kuznets, Simon. 1955. "Economic Growth and Income
Inequality. American EconomicReview 45(1): 1-28.
_. 1959 Six Lectures on Economic Growth. Glencoe,
Ill.: Free Press.
. 1966. Modern Economic Growth. New Haven,
Conn.: YaleUniversityPress.
Lal, Sanjaya,and G. Wignaraja. 1992. 'Foreign Involve-
ment by European Firms and Garment Exports by
381
 >NTE
Developing Countries." DevelopmentStudies Work-
ing Paper 54, Queen Elizabeth House, OxfordlCen-
tro Studi Luca d'Agliano,Torino, Imlia.
Lanyi, Anthony, and Rusdu Saracoglu. 1983. "Interest
Rate Policies in Developing Countries." Occasional
Paper 22. International Monetary Fund, Washing-
ton, D.C.
Lawrence, Robert Z 1991. "Scenarios for the World
Trading System and Thcir Implications for Develop-
ing Countries." Tcchnical Paper 47, OECD Develop-
meentCenter, Paris. (Paper prepared for the Rcsearch
Program on Globalizationand Regionalization.)
1993. "The Global Environmenr for the EastAsian
Model.' John F. Kennedy School of Government,
Harvard Univcrsiy, Cambridgc, Mass.
Lee, Chinaboon. 1985. "FinancingTechnical Education
in LDCs Economic Implicatons from a Survey of
Training Modes in the Republicof Kora? Education
and TrainiingSeries 6. World Bank, Education and
Training Deparmnent, Washington, D.C.
Lees,FrancisA., James Botts, and Rubens Penha Cvsne.
1990. Banking, and Financial Deepen in HraZ6;lw Little,Ian M. D., Richard N. Cooper, W. Max Corden,
New YorakiSt. Martin'sPras Depening in
NewYo& Press.Adyjustmen
Sr. Martin's
Leipzrger, Danny.
Danny. 19
Lzpzlgcr, 1987. Korca Mnnugmg the
87. Ko-reicManaging tJycIndustrial
SnivZ
Transition.Washington, D.C.: World Bank.
Leipziger,Danny, David Dollar, Anthony F. Shorrocks,
and Su-YongSong. 1992. The DistributionofIncome
-and WeaLrhin Korea-Washington, D.C.: World
Bandh
Levy,Brian. 1993. "Obstaclesto DevelopingIndigenous
Small and Medium-Size Enterprises: An Empirical
Assessment.? World Bank Economic Review 7(1):
6585.
Lewis,W. Arthur. 1954. 'Economic Developmentwith
Unlimited Supplies of Labor." ManchesterSchool of
Economicsand SocialStudies22(2): 139-91.
Li, KI T. 1989. The Evolutionof PolicyBehind Taiwun's
DevelopmentSuccessNew Haven, Conn.: Yale Uni-
versiryPress.
Liddle, William. 1991. "The RelativeAutonomy of the
Third World Politician: Soeharto and Indonesian
EconomnicDevelopmenr in Comparative Perspec-
ive." Interational StudiesQuarterly35: 403-27.
Lim, Chee Peng 1992. "Managing Structural Change
and Indusrrialization-Heayy Industrialization: The
Malaysian Experience" Paper presented at the isis-
382
HIID Conference on the Malaysian Economy, Kuala
Lumpur. Institute of Strategic and International
Studies.June 1-3.
Lim, Linda Y. C., and Pang Eng Fong. 1991. Foregn Di-
ret:InvestmentandIndutrializtioninMa(aria.Sing-
apor, 7Taiwnand ThaikladnOECD, Paris.
rim,Youngil. 1991. 'Upgrading industrial Base and
Trade: The Role of Government Policy in Korea."
United Nations Industrial Development Organiza-
tion, New York.
Lindaucr. David L., and Barbara Nunbcrg, eds. 1993.
"Rehabilitating Govcrnment: Pay and Employment
Refrmn in Sub-Saharan Africa."World Bank, Policy
Research Department, Washington, D.C.
LiEte, Ian M. D. 1979. UTheExperienceand Causesof
Rapid Labor-IntensiveDevelopment in Korea, Tai-
wan,Hong Kong, and Singapore,and the Possibilities
of Emulation." Working Papers, Asian Employment
Programme,ILO-ARTEP, Bangkolk
_ 1982. Economic Development New York: Basic
Books.
and Sarath Rajapathirana.Forthcoming. Boom,Cri4s,
BbMatroeconondc
Th EEpeneneofDevelop-
AduenThMarcomiEprece/Dvl-
mg Counties. New York.Oxford UniversityPress.
Lltle, Ian M. D., Tibor Scitovsky,and Maurice FG.
Scott. 1970. Induniy and Trade in Some Developing
Countnies.London: Oxford UniversityPress.
T
Liu, Alan P. L. 1987. Phoenixand the LamneLion Mod-
Bakentioznin Taiwanand iMainlandChina, 1950-80.
Stanford, Calif: Hoover Institution Press.
Liu, Ligi. 1991. "Entry-Eir: Learning and Productivity
Change-Evidence from Chile" Policy Research
Working Paper 769. World Bank, Country Econom-
ics Department, Washingron,D.C.
Lockheed, Marlaine, and Adriaan Verspoor. 1991. Im-
proving Primary Education in DevelopingCountries.
NewYork: Oxford UniversityPress.
Lockwvood,William W. The EconomicDevelopmentof
japaw Growth and Structural Change, 18681938.
Princeton: Princeton UniversityPress.
Long, Millard. and Dimirri Vita 1991. "Changing the
Rules of the Game." Policy ResearchWorking Paper
803. World Bank, Policy Research Department,
Washington, D.C.
Maisom,Abdullah. 1992. 'Pattems of Total Factor Pro-
ducrMty Growth in MalaysianManufaucring Indus-
 tries, 1973-89." Paperpresentedat ISLS-HIIDWorkshop
on theMalaysianEconomy,KualaLumpur,September.
Maki, Atsushi. 1993. "The Relationbetween Houschold
Saving Rare and Economic Growth: Consumers'
Stimulusro PurchaseDurable Goods."Paper prepared
for a World Bank Asian Miracle Project Workshop,
Tokyo, Feb. 1-2.
Malaysia.1990. QuarterlyBuledn. KualaLumpur: Bank
Negara.March-June.
Mankiw,N. Grego.y,DavidRomer,and DavidWed. 1992.
'A Contributionto the Empiricsof EconomicGrowth."
QuaterlyjJornalofEconomis107(2):407-38.
Mayer, C 1987. 'The Assessment:FinancialSystemsand
Corporate Investment. Oxfrd Review of Economic
Policy3(4): i-xvi.
Mazumdar, Deepak. Forthcoming. "Labor Markets and
Adjustment in Open Asian Economies The Republic
ofKoreaandMalaysia." WorldBankEconomicReziew.
McKinnon, Ronald I. 1973. Money and Capitalin Bc-
nomic Development Washington, D.C.: Brookings
Insttution.
Mellor,John, ed. 1993. "Agriculntreon the Road to In-
dustrialization." International Food Policy Research
Institute, Washington, D.C.
MiddkUton,J., A Ziderman, and K van Adams. Forth-
coming. Skillsfor ProdictivicV.Baltimore,Md.: Johns
Hopkins UniversityPress.
Modigliani, Franco. 1970. "The Life CycleHypothesisof
Savingsand Intercountry Difirences in the Savings
Ratio." In W. A. Elis, Maurice FG. ScoEt,and J. N.
Wolfe, eds., fnduccion Growth and Trait Essaysin
HonorofSir RoyHarrod Oxford. Clarendon Press.
Mody, Ashoka,R. Suri, and J. Sanders. 1992. "Keeping
Pace with Change: Organizationaland Technological
Inperatives." WorldDeveopment2o(l2): 1797-816.
Monetry Authority of Singapore.1991. "Savings-Invesc-
ment Balances in Singapore: Determinanrs and
Medium-Term Outlook" Singapore: Economics
Department.
Morris, Felipe, Mark Dorfmnan,Jose Pedro Ortiz, and
Maria Claudia Franco. 1990. laan America5 Banking
Systems in the 1980s: A Cross-CountryComparort
World Bank DiscussionPaper 81. Washington,D.C.
Mukai,Yurio.1963."DevelopmentofPOStalSavingsand Its
Factora In Japanesr.IGnywI(eiza no. 83. December.
Munasinghe, Mohan. 1987. Rural Electnficationfor De-
velopment PolicyandAnaij sisandAppication. Boul-
der, Colo.: Westview.
Murdock, K., andJoseph E. Stiglirz. 1993. "The Effecrof
Financial Repression in an Economy with Positive
Real Races." Background paper for The EmFtAsian
Mirac& World Bank, Policy ResearchDepartment,
Washington,D.C.
Murphy, Kevin, Andre Shleifer, and Robert Vishny.
1989. "Industrializationand the Big Push."Jotrnal of
PoliticalEconomy97(5): 1003-26.
Mutoh, Hiromichi. 1988. "The Automotve Industry."
In Ryuraro Komiya, Masahiro Okuno, and Koraro
SuzLunura,eds., IndustrialPolicyofJapan. New Yorlc
AcademicPress.
Myrelka,Lynn. 1978. "Licensingand TechnologyDevel-
opment in the Andean Group." World Development
6(4): 447-60.
Nagaoka,Sadao. 1989. 'Overview ofJapanese Industrial
Technology Developmenrt" Industry Series Paper 6.
World Bank, Industry and Energy Department,
Washington, D.C.
Nalebuff,B. 1983. "Information,Competition and Mar-
kers."AmericanEconomicReview72(2): 278-84.
Nalebuff,B., and JosephE StigLirz.1983. "Prizesand In-
centives Towards a GeneralTheory of Compensation
and Competition." Bel Journal of Economics14(1):
21-43.
Nam, Sang-Woo. Undated. "Korea's Financial Reform
since the Early 1980s." Korea DevelopmentInstitute,
Seoul.
Neal, Craig R 1990. "MacrofinancialIndicators for 117
Developing and InldustrialCountries." PR£Working
Paper 58. World Bank, Office of the Vice President,
DevelopmentEconomics,Washington, D.C.
Nehru, Vikram, and Ashok Dhareshwar. 1993. 'A New
Databaseon PhysicalCapital Stock. Sources,Method-
ology, and Results."World Bank, Internatonal Eco-
nomics Department, Washingtn, D.C
Nelson, Richard R.. 1973. "RecencExercisesin Growth
Accounting: New Undesmtnding or Dead End?"
AmericanEconomicReview63(3): 462-8.
Nishimizu, Mieko,and John M. Page, Jr. 1982. "Total
Factor Productivity Growth, TechnologicalGrowth,
and TechnicalEfficiencyChange: Dimensionsof Pro-
ducriviy Change in Yugoslavia,1965-78.' Economic
Journal92(368): 920-36.
383
 p_E" .
. 1987. "Economic Policies and Productvity Change
in IndusErr.An International Comparison." Paper fior
a World Bank research project an Productivity
Change in Infant Industry. World Bank, Policy Re-
searchDepartment, Washington, D.C-.
-. 1991. "Trade Policy,Market Orientation, and Pro-
ductivityChange in Industry.' In Jaime de Melo and
AndresSapir, eds., TradeTheoryand EconomicRefonn:
North, SouthandEast-Essays in Honor ofBe/ Eatsa.
Oxford: Blackwell.
Noguchi, Y. 1985. "Nenkin, zeisci to kokumin no
chochiku kodo [Pensions,the tax system,and the sav-
kg behaviorof the people]."In KenkyuHakoko Gaiyo
[-Sumwy of ResearchFindings].Kokuci Nin'i Seimei
Hoken no Shorai Tenbo ni k..isuru Chosa Kenkyu-
kai [Reserch project on the Future Prospectsof Gov-
einment-ManagedVoluntaryLifeInsurancel,Tokyo.
North, Doulas C., and R Thomas. 1973. TheRiseofthe
WesternWorn!c A Nev EconomicHistory.Cambridge,
UX.K:Cambridge UniversityPress.
Ogura, Sciritsu,and Naoyuki Yoshino. 1988. 'The Tax
Systemand FiscalInvestment and Loan Program." In
Ryutaro Komiya,Masahiro Okuno, and Kotaro Suzu-
mum, edmi
P ndustria PolicyofJapan.New York.Aca-
dlemncPress.
Okimoto, Daniel. 1989. BetweenmiT! and the Market
JapaneseIndstrial Policyfir High Tecbnolg. Stan-
ford, Calif.: StanfordUniversityPress.
Ono. 1992. Jissenrteki SangyoSeiskurn. In Japanese.
Tsusho SangyoChosakai.
Pack, Howard. 1971. Structurl Change and Economic
Policy in isreeLNew Haven, Conn.: YaleUniversity
Press.
-. 1988. "Industrializationand Trade." In Hollis B.
Chenery and T. N. Srinivasan, eds., Handbook of
DevelopmentEconomics,vol. 1. Amsterdam: North
Holland.
19Ha2. n a P C AmericanDevelopmentBank,Washington, D.C.
_. 1992. 'Leamningand Producdvily Change in D:evel-
Oping Countries." In Gerald K Helliner, ed., Trade
Policy Industializaton and Development. Oxford:
Clarendon Press.
_. 1993a. "Exports and Externalities:The Source of
Taiwan's Growth in Taiwan." Universityof Pennsyl-
vania, Philadelphia,Pa.
_. 1993b. Indtstrial and Trade Polcies in the High-
PerformingAsian Economies."Backgroundpaper for
The ast Asia Miracl. World Bank, PolicyResearch
Department, Washington,D.C
384
. 1993c. "Technology Gaps between Industrial and
DevelopingCountries:Are There Dividends for Late-
comers?" In Proceedingsof the World BatnkAnnual
Bank Conferenceon DevelopinentEconomics,1992.
Washington, D.C.: World Bank.
Pack, Howard, and John M. Page,Jr. 1993. 'Accumula-
tion, Exports, and Growrh in the High-Perfirming
Asian Economies." Background paper for The East
Asian Mirade. World Bank, PolicyResearchDepart-
ment, Washington, D.C.
Pack, Howard, and Lary E. Wcstphal. 1986."Industrial
Stratcgyand TechnologicalChange: Theory vs. Real-
ity."fournalofDevelopmentEconomics22(l):87-128.
Packer, Frank. 1992. 'The Role of Long-Tarm Credit
Banks within the Japanese Main Bank System." Co-
lumbia University,New York.
Page, John M., Jr. 1990. "The Pursuit of Industrial
Growth: Policy Initiatives and Economic Conse-
quences."In MauriceFG. Scott and Deepak Lal,edts,
Public Poticy and EconomicDevelopment Essaysin
Honoroflan Litte Oxford:Clarendon Pres.
Page, John M., Jr., and PeterA. Peti. 1993. "Productiv-
ity Change and StrategicGrowth Poliqc in the Asian
Miracle" Backgroundpaper for The EastAsianMina-
c& World Bank, PolicyResearchDepartment, Wash-
ington, D.C.
Panagariya,Arvind. 1993. Unravelinj the Mysteries of
China's Foreign Trade Regime." WorldEconomy
16(1): 51-.
Pempel, T. J. 1984. "The JapaneseBureaucracyand Eco-
nomic Development Can the Model bc Copied?'
Economic Development Institute, World Bank,
Wsigo,DC
Washington, D.C
Perry, Guillemo, and Ana Maria Herrera. 1993. "Finan-
zas Publicas,Esmabilizaciony Reforma Estructural en
America Latina." Working Paper Series 165. Inter-
Perm,PeterA 1988. "Korea'sExport Niche: Origins and
Prospects." WorldDevelopment16(1):47-64.
Pfeffermann, Guy P., and Andrea Madarassy. 1992.
Trendsin Pivate Investmentin DevelopingCountries.
1992 Edition. IFC DiscussionPaper 14. Washington,
D.C.: World Bank and IFC.
Psacharopoulos,Georgc 1993. "Returns to Invesunent
in Education: A Global Update." Policy Research
Working Paper 1067. World Bank, LarinAmericaand
the CaribbeanRegion,Washington, D.C

Psacharopoulos, George. Jee-Peng Tan, and Emmanuel
Jimenez. 1986. Financing Educaton in Developing
CoanmierAn Exploration ofPoliqyOption. Washing-
ton, D.C.: World Bank.
Ranis, Gusrav. 1973. "Industrial Sector Labor Absorp-
uon." EconomicDevelopment and Cultutra Chlange
21(3): 387-408.
. 979. 'Industrial Development." In W. Galcnson,
cd., EconomicGrowti and Sructural Changein Ta)-
wan. Ithaca, N.Y.:Cornell UniversityPress.
, ed. 1992. TaiwdanFromDevelopingto Mature Econ-
omy. Boulder,Cola.:WstviCw.
_. 1993. "labor Markets, Human Capital, and Dcvd-
opmenr Performancein EastAsia."Backgroundpaper
for The East Asian Miracle.World Bank, Policy Re-
search Department,Washington, D.C.
Ranis, Gustav, and John Fei. 1961. "A Theory of Eco-
nomic Development." American Economic Reviewt
51(4): 533-65.
Ranis, Gustav,and Chi Schive.1985. "Direct Foreign In-
vestrment in Taiwan's Development." In Walter
Galenson, ed., ForeignTrad and Investmen Economic
Developmentin &ieNewly Industriizng Asian Couen-
tiim Madison, Wis.:Universityof-WisconsinPress.
Ranis, Gustav, and Frances Stewart. 1987. "Rural Link-
ages in the Philippinesand Taiwan.' In Frances Stew-
am, ed., Macro-policsfor AppropriateTechnolog in
DevelopingCountries.Boulder,Colo.: Westview.
Rebelo, Sergio. 1991. "Long Run Policy Analysis and
Long Run Growth!. journal of Poitdcal Ecotnomy
9(3) ~50>21~ ~ ~ ~ ~ .
99(3): 500-21.
Reid, Gary. 1992 "Civil ServiceReform in Iarin Amer-
ica: Lessons of Experience." Occasional Paper 6.
World Bank, Latin Ameica Technical Department,
Washington, D . and High PerD.Ciance in a Technoc.ric Poliy: The
Remolona, E., and M. Lamberre. 1986. 'Financial Re-
f[ormsaand the Balanceof PaymentsCrisis:The Case of
the Philippines, 1980-83." Working Paper 86-02.
Philippine Institute of DevelopmentStudies, Manila.
Revel, Jack. 1987. "Mergers and the Role of Iarge
Banks? Institute of European Finance, University
Colege of Norch Wales, Bangor,Wales.
Rhee,YtmgWhee. 1989a."ManagingEntry into Intcrna-
tional Markets: Lessonsfrom the East Asian Experi-
ence? Industry SeriesWorking Paper 11. World Bank,
Industry and EnergyDepartmne Washington, D.C.
1989b. "Trade Finance in DevelopingEconomies."
PolicyResearchSeries5. WorldBank,Washington,D.C.
BIBLO'4
Rhee,YungWhee, and Lary F. Westphal. 1977. "A Mi-
croeconomicInvestigation of Choice of Technique."
JowofDevelpmenrtEonomia4: 205-38.
Riedel,Jamnes.1988. "Demand for LDc Exporcsof Manu-
fictures: Estimatesfrom Hong Kong.' Economiclour-
nal98(389): 138-48.
Rob, Rafael.1990. "The Evolutionof Markcts niderDe-
mand Uncertainty." University of Pennsylvania,
Philadelphia,Pa.
Romer,Paul M. 1993. "Two Srategies for EconomicDe-
vclopmenr Using Ideas and ProducingIdeas." In Pro-
ceedingsof the Word BankAnnual Bank Conferenceon
DevehopmmtEnomŽs 1992 Washington D.C.
Rosenberg, Nathan. 1976. Perspectiveson ecohnoloV
Cambridge, U.K.: CambridgeUniversityPress.
Sabor, Richard H. 1979. Eonomic Development and
UrbanMigratin. London: Oxford UniversityPress.
Sakakibra, E, R. Feldman, and Y. Harada. 1982. The
JapaneseFinancialSyttem in ComparatvePerspectiva
Washington,D.C.: Joinr EconomicCommittee, Con-
gressof the Unitcd States.
Sakong,1. 1993. Koreain de WorldEonomy. Washing-
ton, D.C.: Insritute for International Economics.
Salleh,IsmaeL1991. The Privatisaion of PublicEnter-
prises:A CaseStudy ofMalaysia."In Geer Gouri, ed,
Pi vatisa n and PublicEnterprise.New DclhLOxford
Universiy Press and IBH Publishing Company,
Sallc, Ismacl, Kim Leng Yeah, and Saha Dhevan
Manatan 1993. Growth, and Srcr rat
Meyanathan. 1993. Growab, Eqity, andf SDxrd
Trsmj$rmazionin Malaysi The Roleof the Pubic cc-
tor. Lessonsof East Asi World Bank County Study,
Washington, D.C.
Samudavania Chaian. 1992. "High Speed Growth
ai .n
P acegin a TocreastAic o irT
Thai Cae." BackgroundpaperoforTec stAsarnMir-
acs. World Bank, Policy Research Department,
Washington, D.C
San, G- 1990. "Enrerprise Training in Taiwan: Results
fromdieVocationalTrainingNeedsSurvey? Econam-
ic ofEducationReview9(4): 411-8.
Saxonhouse,Gary. 1983. "What Is Al This about Indus-
trial Targeting in Japan?" World Economvy6(3):
253-73.
Scdiff, Maurice. 1993. "Agriculture Pricing Policy and
Growth in Asia.' World Bank, Policy ResearchDe-
partment, Washington, D.C
385
Schiff, Maurice, and Alberto Vald&. 1992. Tle Political Singapore-Variousyears. YearbookofStatsticsandanEco-
Eonomy of Agriulural PricingPolicy Volume 4: A nomic Surveyof Singapore.Singpore Department of
Synthesisof the Eamomicsin DevelopingCountries.Bal- Statistics.
rimore, Md: Johns Hopkins UniversityPr. . 1983. Economicand Social Statistics 1960-1982.
Schultz, T. Paul. 1988. "Education Invesrmentand Re- Singapore:Department of Staiics.
turns." In HoUlisB. Chenesy and T. N. Srinivasan, Siqgh.Ait. 1992. "'CIose' vs. 'SErategic'Integrationwith
eds., HandhookofDevelopmentEconomis voL 1. Arn- the World Economy and the 'Marker-FriendlyAp-
sterdam: North Holland. proach to Developtrnent'vs. an 'Industrial Policy':A
Scirovsky, Tibor. 1954. "Two Conceprs of Extemal Critique of the WorldDevelopmentReport1991and an
Economics." Journal of Politcal Economy 62(2): AltemativePolicyPerspective."Facultyof Economics,
143-51. Universityof CanibridgecCambridge,England.
Scott, Maurice FG. 1979. "Trade." In W. Galenson, ed., Singh, Ajit, and Javed Hamid. 1992. 'Corporate Finan-
EconomicGrowth and Stral Canga in Taiwan cial Structures in Developing Countries." Technical
Ithaca, N.Y.: Cornell UniversityPress. Paper l InternationalFinance Corporation,Washing-
Scully,MichaelT., and GeorgeJ. Viksninsa1987. Finazc- ton, D.C.
ing EastAsias Succesr ComparatveFinancal Develop Solow, Robe M 1956. "A Conterbution to he Theory
menctinEghtAsian Counris. London: Maanilan. ofEconomic6Growth." QuareeryJornalofconomics
Shah, Anwar, ed. Forthcoming. FiscalIncentivesfor In- 701)65-9.T
vetcmenttin Devloig Coun Ne York Oxford -- 1957. "Tehnca Chang and die Agreae Pro-
Uesnivrersity
PressUniverilty Prem ducton Function.' Reizew of Economicsand Sttstics
~~~~~~~~39(3):
3 12-20.
Shah, Shekhar, and Subodh C. Madiur. 1992. Source Soon, Tek-Wongand Suan Tan. 1992. Public Policy
of Labor bMarket Flexdbilityin East Asia:A Reviewof andtEcnomic Development in Singapore.' Manu-
SelectedIssues."Backgroundpaper for The EastAsin script prepared for a Wodd Bank Conference on the
Mirrd, World Bank, Policy Research Department, Role of Goverment and East Asian Success,Wash-
Washington,D.C. ingon, a.C.
Shapiro, Helen, and LanceTaylor. 1990. "The State and Spence, M. 1981. 'Te Learning Curve and Competi-
Industrial Strategy." World Development 18(6): on BeffJourd ofEconomics 12(l): 49-70.
861-78.
Shea, J.D.Fothoin.DeeopenadstuSquire, Lvn. 1981. Employment Policy in Developing
Shea, J. D. Forrhcoming. 'Developmcnt and Strruaural CCounS: A Surveyof lssuesand Evidence.New York-
Changeof the FinancialSystem."In Patrick Hugh and OxfordUniversityPres
Yung Chul Park, cds., Financia DevelopmentJapan Srinivasan, T. N. 1992. "Datbas for Development
Korea, and Taiwan. New York: Oxford University Analysis:An Overview.' Paper presentedat a Confer-
Press. ence on Data Base of Dcvelopment Analysis,Yale
Sheng, Andrew. 1989. "Bank Restructuringin Malaysia, University,New Haven, Conn., May.
1985-88." PRnE Worlkng Paper 54. World Bank, Fi- Stem, Joseph,Ji-Hong Kim, Dwighr Perkins, and Jung-
nancial Sector Development Department, Washing- Ho Yoo. 1992. "Industrialization and dhe Stare The
ton, D.C. Korean Heavy and Chemical Industry Drive."
_. 1990. "The Art of Bank Resuucturing2Isues and Research reporr Harvard Institute for International
Techniques."World Bank, FinancialSector Develop- Developmenr,Cambridge,Mass.,and KoreaDevelop-
ment Department, Washington, D.C. mentInsitute, Seoul.
Shouda,Yasutoyo.1982. 'EffectiveRatesofProteciion in Stevenson H., and J. Stigler. 1992. The Learning Gap.
Japan." Nikd7onieza Kmkyu 11 (March). New York.Summir.
Shoven,John. 1988. GovernmentPoicy TowardsIndustr Stewart, Francis, and Ejaz Ghani 1991- 'How Signifi-
m the United Stares and Japan. Cambridge, U.K: cant AreExternaities for Development? WorldDevel-
Cambridge UniversityPress. opmentl9(6): 569-94.

386
 Stigicz Joseph E. 1993a. "The Role of the State in Fi-
nancialMarkets."Paper preparedfor theAnnual Bank
Conference on Development Economics, Worldc
Bank, Washington, D.C.
. 1993b. "Some Lessons from the Asian Mirade."
Backgroundpaper fbr The EastAsian Miracle.World
Bank, PolicyRescarchDepartment,Washington,D.C.
Summers,LawrenceH. 1985. "Issuesin NationalSavings
Policy." NBERWorking Paper 1710, Cambridge,
Mass.
Summers, Lawrence H., and V. P. Summers. 1989.
"When Financial Markets Work Too Well: A Cau-
riousCase for a^
uious Cse for a Secuwriics
SecuritiesTmnsaceian
T ransactionTax."
Ta." In
In Regv-
Ret-
&no'yReforrnof Stockand FuturesMarkeet,a special
issue of the Joumal of Financial ServicesResearch.
Boston:KiuwerAcademicPublishers.
Sumnmers,Robert,andAan Heston. 1988. "ANew Set of
International Comparisons of Real Product and
Prices:Estimatesfor 130 Countries." Reviewoflncome
and Wealth34(1): 1-25.
. 1991. 'The Penn World Tables (Mark 5): An Ex-
panded SeEof Inenational Comparisons,1950-88.'
Quane4yJournalofEconomics105(2).
Sun, C., and M. Liang. 1982. "Savings in Taiwan,
1953-1980." In Kwoh-Ting Li and Tzong-ShianYu,
eds., Experencesand Lessonsof EconomicDevelopment
in Taiwan.Taipei: Institute of Economics,Academia
Sinica.
Sundavarej,Tipsuda,and Tniiranvorakul,Prasarn.'Experi-
encesof FmancialDistrss in Thailand." PRE Working
Paper283. World Bank,Officeofthe VicePresidentfor
DevelopmentEconomics,Washington,D.C
Sussangkarn, Chalongphob. 1988. "Production Struc-
tures, LaborMarketsand Human CapitalInvestments:
Issues of Balancefor Thailand." Thailand Develop-
incur ResearchInstitute,Ofijprinr#7, Bangkok
Suzuki,Yoshio.1987. ThejapaneseFinancialSystem.Ox-
ford: Clarendon Press.
Syrquin,Moshe,and Hollis B. Chenery. 1989.Paterns of
Developmenr,1,950to 1983. World Bank Discussion
Paper 41. Washington, D.C.
Taiwan, China. Various issues. Financial Statistics
Monrh4. Taipei: Cenial Bank of China.
Variousyears. Statistcal Yearbookof the Republicof
=Chinz
Taipei: Directorate General of Budger, Ac-
counting, and Statistics.
1.988. Satistical Yearbookof the Republicof Chinz
Taipei.
1992a.Ststical Yearbookof the Republicof Cbina.
Taipei.
1992b. StaSist of Sll and Medium Brainesin
T1iwab,R&asissic of mina. Taipei: Minisumyof Eco-
nomic Afneirs.
Tan, JeePeng. and Akin Mingar. 1992. Edcation in
Asia: A ComparativeStudy of Cost and Financing.
Washington, D.C.: World Bank.
Tankingoki. 198. "Alu m Rfni s
Tanaka,Naok. 1988. AlumnumReFiningIndustry In
Ryuraro
mur, ed.Komiya, Masahiro
Indrutnl PvoliyOkuno, and
of/Japan.NewKoraro
YorkcSuzu-
Aca-
dmim pes.
mc rss.
Tanzi, Vito, and ParthsarathiShome. 1992. "The Roleof
Taxation in the Development of East Asian Econ-
omies." In TakaroshiIto and Anne 0. Krueger,eds.,
hser Economyof Tax RPfrm Chicago: Uni-
Poliicagl
versiry of Chicgo Press.
Teranishi, Juro. 1986. "Economic Growth and Regula-
tion of Financial Markcts:Japanese Experiencedur-
ing the Postwar High Growth Period." Hitotsubashi
Journal ofEcnomics27: 149-65.
Thailand. Various issues. QuarterlyBulletin. Bangkok:
Bank of TliailancL
Thomas, Vinod, and Yan Wang. 1993. "Government
Policiesand ProductivityGrowti: Is EastAsia an Ex-
ception?" Backgroundpaper for TheEastAsian Mira-
cle. World Bank, Policy Research Department,
Washington, D.C.
Thurow, Lester. 1992. Head to Heatd The ComingEco-
nomic BattleAmongjapan, Europe, an/Amria.New
York:Morrow.
Trela, Irene, and John Whale,y. Forthcoming. "Taxes,
Outward Orientation and Growth Performance in
the Republicof Korea." In AnNvarShah, ed., FiscalIn-
centivesfor Investment in DevelopingCountries.New
York: Oxford UniversiryPress.
Trezise, P. H. 1983. "IndustrialPolicy Is Not the Major
Reason for Japan's Success."BrookingsReview 1(3):
13-8.
Tullock, Gordon. 1988a. "Why Did the Industrial Rev-
olurion Occur in England?"In C. Rowley, R. Tolli-
son, and G. Tullock, eds., The PoliticalEconont of
Rent-Seeking.Boston: KluwerAcademicPublishers.
387
: .1988b. "Rents and Rent-Seeking."In C. Rowley,R.
_ The WallStreetJournaL1990. "Taiwan,Long Notd for
Tollison, and Gordon Tullock, eds., The Political
Economyof Rent-Seeking.Boston: Kiuwer Academic
Publishers.
Turnham, David. 1993.EmploymentandDevelopmenetin
New ReviewofEvidence.Paris: Development Center,
OECD.
Ueda, Kazuo. 1992. Insitutiona and Regrulatory Frame- Warr, Peter G., and Bhanupong Nadhiprabha. 1993.
worksfortheMain Bank System.Tokyo: Universityof
Tokyo.
UNESCO. Variousyears. StatisticalYearbook.Paris.
United Nations. Variousyears,a. DemographicYearbook.
New York.
_. Various years, b. Eneg Statistics Yearbook.New
York.
_. 1976. WorldEnergySupplies195o-74. New York.
_- S1992.WorldEconomicSutnvy: Current Trendsand
Policiesin the WorldEconomy.New Yorkl
uNop (United Nations Development Programme).1991.
Human DevelopmentReporr.New York: Oxford Uni-
versityPress.
Van Adams, A. 1989. "A Study of the Economic Envi-
ronnent's Influenceon Human Capital Formation in
Korea and Egypt." George Washington University,
Washingon, D.C
Viner, Aron. 1988. InsideJapaneseFinaial Me.
Homewood, Ilk: DowaJoneserFin.
Homewood,
Dow Jons--Invin.CArr
IL:
Vitts, Dimitri. 1991. 'Measuring CommercialBank Ef-
ficicncy:Use and Misuse of Bank Operating Ratios."
Policy ResearchWorking Paper 806. World Bank,
Country EconomicsDepartment, Washington, D.C.
-_ .1992. 'Poliy Issues in Financial Regulation."
World Bank, Country Economics Department,
Washington, D.C. April.
Vogel,Ezra F., and DavidLindauer. 1991. "Towarda So-
cial Compact for KoreanLabor." In David Lindauer,
Jong-Gie Kim, Joung-Woo Lee, Hy-Sop Lim, Jac-
YoungSon, and Ezra F. Vogel,eds., Korec The Strains
of EconomicGrowt. Cambridge, Mass.: Harvard In-
stiture for InternationalDevelopment.
Wade, Robert. 1989. 'What Can EconomiesLearn from
EastAsianSuccess?"Annab oft heAmericanAcademyof
PoliticalScience505: 68-79.
. 1990. Goening the Market EconomicTho?y and
the Rolevfthe Governmentin EastAsian Industriai-
tiomzPrinceton, NJ.: PrincetonUniversityPress.
388
Cheap Imitations, Becomesan Innovator."June 1.
Wang, Fang-Yi. 1990. "Reconsidering the 'East Asian
Model of Development': The Link between Exports
and Productivity Enhancement, Evidencefrom Tai-
wan, 1950-87." Ph.D diss.Universityof Californiaat
Los Angeles.
"MacfocconomicPolicies,Crisis,and Growth in die
Long Run: Thailand." Backgroundpaperprepared for
Litde, Ian M. D., Richard N. Coopcr, W. Max Cor-
den, and Sarath Rajapathirana,cds., Boom,Crisis,Ad-
justnent- The Macroeconomic Expience of Developing
Countries.New York:Oxford UniversityPress.
Wei, Shang-Jin. 1993. 'Open Door Policyand China's
Rapid Growth: Evidencefrom City-LevelData." Pre-
pared for the Fourth Annual East Asian Seminar on
Economics, National Bureau of EconomicResearch,
Cambridge,Mass.
Westphal,LarryE. 1978. "The Republicof Korea'sExpe-
rience with Exporr Led Development." WorldDevel-
opmnent6: 347-82.
Westphal, LarryE., YungWhee Rhee,and Garry Pursell.
1988. KoreanIndustrial Competenre:Wher It Came
FmronWorld Bank Staff Working Paper 469. Wash-
ington, D.C.
Whalley,John. 1990. "Non-DiscriminatoryDiscrimina-
don: Special and Diffierential
for Developing Treatrnent
Countries." under
Economic the
Journal
100(403):1318-28.
WiboonchutikulaPaitoon.1987."otal FactorProduc-
tivity Growth of Manufacmring Industries in Thai-
land." In Productivity Changes and International
Competitiveness of ThaiIndustries.Bangkok Thailand
DevelopmentResearchInstitute.
Winkler, Donald R. 1990. Higher Edruation in Latin
Amerca: Issuesof Efficiencyand Equity. World Bank
DiscussionPaper 77. Washington, D.C.
Wolf, Charles. 1988. Markets or Governments:Choosing
betweenImpect Altenzetives.Cambridge, Mass:Mrr
Press.
World Bank. 1983. "SpecialEconomic Reporc Policies
fior the Financial Secaor." Europe, Middle East, and
North AfricaRegion,Washingron,D.C
. 1987a. Korec Managingthe Imndrial Transition2
volt Washington, D.C.
_. 1987b. WorldDevelopmentReport1987 New York.:
Oxford UniversityPress.
_. 1989a. 'Indonesia: Strategyfor Growth and Struc- Trade Policyand 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:MatchingRisk and Rewardsina Calif.: Stanford UniversityPress.
MixedEconomy.Washington, D.C. Yamawaki,Hideki. 1988. 'The Steel Industry." In Ryu-
_. 1989c. WorldDevelopmentReport1989. New York: taro Komiya, Masahiro Okuno, and KomaroSuzu-
Oxford UniversityPress. mura, eds., Indusrial Poliy of Japan. New York:
l990a. Indonesic SterategyJfra SustainedReduction
1. AcademicPress.
in Povety. Washington, D.C. Yarnazawa,Ippie. 1988. "The Textile Industry."In Ryu-
. 1990b. WorldDevelopment Report1990. NewvYork: taro Komiya, Masahiro Okuno, and Kotaro SuZU-
Oxford Universiy Press. murm, eds., Industrial Polic of Japan. New York:
1.l99a. Colombia:IndustrialCompettionand PJefor- AcademicPress.
-mance. Washington, D.C. Yan, Lin See. 1991. 'Savings Investment Gap, Financing
-. 199lb. WorldDevelopmentReportThe Chalengeof Needs and Capital Market DevelopmenL"Bank Ne-
Development. New York:Oxford UniversityPress. gara,Malaysia.September.
-. 1992a. "China: Reform and the Role of the Plan in Yaron,Jacob. 1992. SuccessfidRural Finaee Insitutions.
the 1990s.' East Asiaand the PacificRegionalOffice, World Bank DiscussionPaper 150. Washington,D.C.
Washington, D.C Yonezawa,Yoshie. 1988. YThe Shipbuilding Industry."
-. 1992b. GlobalEconomicProspects and theDeveloping In Ryuraro Komiya, Masahiro Okuno, and Kotaro
Counties. Washington, D.C. Suzumur eds., InduserialPoliy ofJapa;LNew York:
. 1992c. World BankStpportfor Industridizaton in AcademicPress.
Korea,Indi, and Indonesia.Washington, D.C. Yoshino,Naoyuki. 1993. "The Low Interest Rate Policy
-. 1992d. WorldDevelopmentReport1992. New York: and EconomicGrowth ofJapan." Department ofEco-
Oxford UniversityPress. nomics,Kelo University,Tokyo, Japan.
1993a. GlobalEconomicProspectsandtheDeveloping Young,Alwyn. 1992. "A Tale of Two Cities: Factor Ac-
Counties.Washington, D.C cumulationand Technical Change in Hong Kong and
. 1993b. WorldDevelopmnt Report1993:Investing Singapore."Paper presentedat a World Bankseminar.
pmenthRrport1993:
PFrbebruary.
rdzwfnginii
Health.Ncw YorkmOxfordUniversityPress. Yusof,Z A., A. A. Hussin, I. Alowi, S. Singh, and L C.
Wporld
EconomicForum. 199eneWorldvCompetitiveness Sing. Forthcoming&The Impact of FinancialReform
Geneva.
Repor- in Malaysia" In Gerard Caprio Jr., Izak Ariyas,and
Yaamamura, Kozo. 1986. "CaveatEnmptorThe Industrial James Hlanson,eds., FinancialRefrnrn TheoryandEx-
PoliCy of Japan." In Paul Krugman, ed., Strategic perience.Cambridge:Cambridge UniversityPress.

The completebacklistof publicationsfrom the World Bank is shown in the

annual hIder which containsan alphabcical tide list and
ofPublications,
indexesof subjects,authors, and countries and regions.The latest edition is
availablefiee of chargefrom the Distribution Unit, Officeof 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
 < ~~~~ ~ ~~~ C,,, ~ ~~~~~C

J~~~~~

I] $
r ~~ ~ ~ ~~~~~~~~~~ClHINA6
r N - 1)

C't ,.. 6r- - j,

>~~~~~~~~~~~~~~~L , _-. - r

%L:0 befI/ t

j2kmmx~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

SfDN
I

Ais,~~ ~ ~~~~~~- 1
. - - -;, ;; D D- .. ~~~~- *.