Public Disclosure Authorized

Public Disclosure Authorized

Public Disclosure Authorized
Public Disclosure Authorized

44

L,' 1.

*tT*I
O'i;

EAi

.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4!~11

#Mk A C l.,
("'t

,,~~~

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

it,;t4

4
~~~~~~~~~~~~~~~~~~~~~~~~~

NE

~~~~~~~~~~~~~~~~~~~~~~~~~~~1

~ ~~~~~~~~~.''

( rOS

Cll

f)

s~~~~~~~~~~~~F

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

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

Lul=lpph.

Jakarta

;
*

_' d

;.

.r.

*NWLI
X\

I)

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
or

Drwmm&mc4
in any farm

rewonmg&
ore.z

se,iw

or by any meas,

merhank4

etvepdorpermrnuonfxf0uirdUriviy

Mamftiazdn
she Unisd
s
Fon paning&Spemner 1993
Ciogun
ph

l .woredina rrtrudqsfem

deamn,

/Ameria

s-At the wp. a ricefud inJava, Indoniaig;c

r beraomI, the por ofPuan,

pumwpy4

Prm

Rephfir

ofKar=cowty

resy
ofMatricrAiser.
At

ofJindre

&pjDA

The map
tnappars o n the izsiderova,
bok wsprqmred soyfortke
onvenie

wbib shws t6c eight eom wigs diiwed in Mis

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
23. Asi Sawi&s-Ecenamicpole
L. WorldBank f. Se,ie
HC4625.A275 1.993

338.95-c20

tE a-Econidc cendo-,
4. Ask

CitiomS

9S33046&

ismv1G20-0851

Torprepaper
P1pfr1Jiwd

anom
MreLw
ialr

theAmeria Natnax SxdrdfrPawzw*w

Z39.48-1984

wf

Foreword
EBATESON THE APPROPRIATE ROLE OF PUBLIC POUCY IN

economic development have occupied policymakersand

bcgan in
economnies
scholarssince the study of developing
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 developmentpolicy.ForthesercasonsI announcedat the timeof the 1991AnnualMeetings of the Board of Govemors of the World Bank in Bangkok,
Thailand,that our DevelopmentEconomicsVicePresidencywouldundertakea comparativestudy of economicgrowthand public policyin
East Asia.

ch. appears
This volumeis the summary
arc
as the firstin a seriesof PolicyRcsearchReports,which intendedto
on developmentpolicy
bringto a broadaudiencethe resultsof research
issuescarriedoutbystaffof theWorldBank.Asreportson policyissues,
we intend that theyshouldhelp us to take-stockof what we know and
cIealy identifirwhat wedo not know;theyshouldcontribureto the debat in both the academicand poli-y communitieson appropriatepublic policy objectivesand instrumentsfor developingeconomies;and
Becausetheysunmarize rethey shouldbe accessibletX nonspecialiscs.
search,wealsoanticipatethat PolicyResearchReportswill provokefarther debate,bothwithinthe Bankand outside,concemingthemethods
usedand the conclusionsdrawn.
The reWhat doesthisreport tell us about the EastAsianmiracle?
seawhshowsthat most of EastAsiasextraordinarygrowthisdue cosuperior 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
functions ofgrowthbetter than mosrother economies.
The eight economicsstudied used very differentcombinationsof
policies,firomhands-ofTtohighlyinterventionist.Thus, themisno single 'East Asianmoder of development.This divcrsityof experiencereinforcesthe view that economicpolicies and policy advicemust be
country-specific,
ifthey areto be effectivec
Buttherearcalsosomecommon 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 resources.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 abundandlydemonstrated.
The reporr alsobreaks some new ground. It condudes that in some

economics,nainly thosein NortheastAsia,someselectiveinterventions

contibuted o grwth. and it advancesour undersmadingof theconditions required for interventionsto succeed. The authors argue thar
whereselectiveintervcntionssucccededthey did so becauseof th eso
sential prcrequisites.First, they addressed problems in the functioning
of markets.Second,they took placewithin the contex of good,findamentalpolicies.Third, theirsuccessdependedon the abilityof governmerts to establishand monicor appropriateeconomic-performance
criteriarelated to the interYcntions-in the authors' terms, to create economic contests. These prerequisitessuggesr that the institional context withLinwhich policies are implemented is as important to their

successor failureas thepoliciesthemselvcs,and the reportdcvotessubstantial artention to the institutional basesfor East Asi-a rapid growth.

While thesefactorshelp to explainwhy apparentlysinilar policies

did not succeedin mranyothereconomies,the report alsoleavesunansweredmany 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 interactions between policy choices and inscitutional capability and between 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
Policiesfor Rapid Growth in a ChangingWorld Economy
Note
26

1 Growth, Eqrity, and Economic Change

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 SavingsRares40
Creatng Human Capital
43
Rapid Productivity Growth
46
Appendix 1.1:Accounting for Growth
60
Appendix 1.2: What Do Tess of Cogniive SkillsShow?Notes
76

Public Policy and Growth

PolicyExplanations
81
The Functional Growth Framework
Notes
103

8
23

27

70

79
87

Macroeconomic Stabiity and Export Growth

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

105

ix

4 An Insitutional Basisfor SharedGrowth

LegitimacythroughSharcdGrowth
Achicving
Technocracy 167
theEconomic
Insulating
WooingBigBusiness 181
Notes
189

157
158

5 Strategiesfor RapidAccumulation
HighHumanCapitalForrnation
EastAsia's
Explaining

191
192

203
ExplainingEast Asia'sHigh SavingsRates
221
ExplainingEast Asia'sHigh Investment Rates

Appendix5.1:GrangerCausaliyTestsforSavings
RatesandGrowthRates
242
beween
Noteon theRelationship
Appendix5.2:Technical
245
InterestRatesandGrowth
Notes
255
6 UsingResourcesEfficienty:Relyingon
259
Madretsand Exports
259
Use
EfficientResource
EastAsiais
Explaining
261
UsingtheMarkenLaborMarketsin EastAsia
theMarkctFinancialMarketsandAllocation 273
Assisting
292
Policy
andIndustrial
MarketTrmde
UsingtheInternarional
Productiviy 316
ExportsIncreased
HowManufactured
Appendix6.1:TestingdheImpactofIndusaial
326
Change
Policyon Productivicy
berween
TFP
Appendix6.2:TestsoftheRelationship
337
ChangeandTradePolicies
342
Notes
7 Polices and Pragmatismin a ChangingWorld
ofRapidGrowdh-Gettingdhe
Foundations
347
Right
Fundamentals
352
toPromoteGrowdi
CreatingInstitutions
353
Intervening
in Markets
Note
368
Note
Bibligraphic

369

347

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 majorcontributions.LawrenceMacDonaldwasthe principaleditor.Casestudiesof
the seven devclopinghigh-performingAsian economieswerc undertakenunder the directionof DannyLeipziger.
The tcamwasassistedby
MariaLtiisaCicogniani,VaruniDayaratna,LeoraFriedberg,Jay Gonzalez,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 effectivenessof crcditpoliciesin EastAsia.BrianLevydirectedthe projecton
supportsystemsforsmalland medium-sizeenterprisesin Asia,andJohn
Pagecoordinatedprojectson lapaneseperspectives
on publicpolicyduring the rapid growthperiod, and tax policyand tax administrationin
EastAsia.

Manyindividualsinsideand outsidethe WorldBank providedvaluable contributionsand comments.(Specificacknowledgmcntof their

efifrts is madein the BibliographicNote at dceend of the book.) Particularthanksare 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.CommentsbyCad Dahlmaan,
Kemnal
Dervis,MarkGersovitz,RalphW Harbison, 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 IndonesianEconomistsAssociation,the EconomicSocietyof Singapore,
the Monetary Authority of Singapore,StanfordUniversity,and the
WorldInstitutefor DevelopmentEconomicReseareh.The financialassistanceof 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-income, middle-incomc (subdividedinto lower-middleand upper-middle), 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-incow 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 experiencingsimilar development or that other economies have reached a preferred or
finalstage of development. Classificationby income does not necessarilyreflectdevelopment status.

'V

Groups
-Anabtical

OR ANALYIICAL
PJRPOSES,
THISSTUDY
GROUPS
ECONOMIES
into severalregions,defined as fIllows

*High-perfirmingAsian economies(HPAEs),led byjapan, arc identified by severalcommon characteristics, such as very rapid export growth. The HPAESare subclassified roughly according to
the duration of their succcssfiulrecord of economic growth:
igersusuallyidentifiedas Hong Kong, thc Republic
TheF
:
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.
.
vur

* 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-income Europan economies of Bulgaria, the former Czechoslovakia, 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 analytimlgroups.

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 Development (Australia,Austria, Belgium, Canada, Denmark, Finland, 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

of East Asiagrew fasterthan all
twenty-threeeconomliies
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 Southeast Asia, Indonesia, Malaysia, and Thailand. These eight highperformingAsian 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 academic and development policy communities to the relationship between 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 different 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 Deparment ofthe World Bank (see Bibliographic Note).

-- :

~~~~~~~~

1jRRL

CLE

Figure1 AverageGrowthof CNPper Capita,1965-90

East Asia
|

_____

HPAEs
EastAsiawithoutHPAES
South Asbi
Midae EastandMeditermean..
Sub-SaharanAfrica
OECD ec__

5
2
3
4
GNPpercapit growthrate (percent)

andCarlbbean
LatinAmedcm
0

Skwre World Bank(1992d).

Singapore, and Taiwan, China-and rapid growth. And in part it refleas 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 domestic product (GPD)per capita between 1965 and 1990 and changes in
the Gini coefficient, a statistical measure ofthe inequality ofincome distribution. 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
Gabon
Greece
Brazil
Syria
Portga

:wA

Malaysia
Yugoislavia
China
Thailand

__
__

Norway
Itay

__

~-.-.-.---

Spain

Israel
Jordan

___

Germnany..

Mexico
Pakistan
Colombia
UnitedStatbes

20:

.Middle

United Kingdom
-----

Venezuiela
Keniya

________-

-----

Philippines

,.-----.-----------..

.-.

-.

Peau
Argentina
Uganda___
India
Benin__

__

__

_
_

_
_

_
_

Senegal
Haiti

Uberla
Sudan
Somalia

I___

Zaire_

__

NIgerfia
Afghardstan

~-*

'~~~~~~~~Bottom20

Midi

CentralAfricanRepublic
Ghana
Guyana

..
-

___

Madagascar

.-

Chad__

Zambia
Angola_
Mozamibique
Kuwalt

H A s--

----

-------

__

-2.

__

Soarae:.
Summers and Heston (1988).

3
2
1.
PercentagechangeIn GDP

45

-RACLLE

F.gum3 Change in Inequityand the GDPperCapita GrowthRate

Change In avenage Glni coeffcient (1980s minus ISGOs)

----

0.2--------

RankingbyGIni coefficient1980

* Most equal third ofsample

0Q15

0.05

--

U Middle thibd
A Least equal thlrd

------------

A Chilet

--

Brazil A

0-

.Rep-

of Korea

Argentina

40.05

______________

!Mexico

______A_Colom

Colombla
bia

Thailand

* Venela
.
* Philippines

__

.Taiwan.

China
_ _
_

Malaysia

Srigapore

Hong
Indonesa
Kong

-J1-

APeru

-0.20

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 dramatically. 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 percent in India and from 50 to 21 percent in Brazilfronm1960 to 1990. A
host of other socialand economic indicators, from education to appliance 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 relativeimportance, experienced rapid growth and productivity improvement. Population 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 effecrive 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 nontraditional savers, raised the levels of financial savings. Education policies 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 intervenedsystematically and through multiple channels-to foster development,
and in some cases the development of specific industries. Policy interventions 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, establishbig and financially supportng government banks, making public
investrmentsin applied research, establishing firm- and industry-specific

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 methodologicalproblem. 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 interventions 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 virtuallyimpossibleto measurethe relativeimpact of findarnentals and interventions 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 counterfactualscenano. 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, governments 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.Indeed, when selectiveinterventionshavc threatenedmacroeconomicstabilty, HPAE governmentshave consistentlycome down on the side of
prudent macroeconomicmanagement.Pricedistortionsarisingfrom selective interventionswere also less extreme than in many developing
economies.
In the newly industrializingeconomies of Southeast Asia, govemment interventionsplayed a much less prominent and frequently less
constructiverole in economic success,whileadherenceto policy findamentals remained importamt These economies'capacity to administer
and implementspecificinterventionsmay have been lessthan in NortheastAsia.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 relevanceoutside the region than thar of Northeast Asia.
The book isorganizedas follows:chapter 1 describesthe distinguishing 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 explore 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 manufactured exports. Chapter 4 disLcusses
the role of institutions. Chaprer 5
looks at the role of public poliqc in the HPAEs' unusually rapid accurnulationof 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

-' .--

IICACLE

overview parallels the organization of the book, highlighting the central

arguments and conclusions.

TheEssence
oftheMiracle:
Rapid
Growth
withEquity

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 unusual 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 income distributions. They also all have been characterized by rapid demographic transitions, strong and dynamic agricultural sectors, and
unusually rapid export growth (see chapter 1).
The HPAEsalso differ fiom other developing economies in three factors that economists have -traditionally associated with economic
growth. High rates of investment, exceeding20 percent of GDP on average 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 twothirds of the growth in dte HPAEs.The remainder is attributable to improved 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 allocating capimt to high-yielding investments and at catching up technologically 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 produc8

tivity growth?There are severalexplanationsfor EastAsiassuccess.Geography 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 neoclassical
viewstressthe HPAEs'successin gettingthe
basicsright. They argue that the successfulAsian economieshave been
better than othersat providinga stablemacroeconomicenvironmentand
a reliablelegalframeworkto promote domesticand internationalcompetition. They also stress that the orientation of the HPAEs toward international 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 neoclassicalframework,and its adherentsstressrhe importanceof human capital
in the HPAEs'success.

Adherents of the renisionistview have successfillyshown that Easr

Asiadoes not whollyconform.to the neodassicalmodel. Industrialpolicy 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 economieswere diverseand flexible.Revisionistsargue that East Asiangovernments "led the m.narke"e
in critical ways. In contrast to the
neoclassicalviev, whichacknowledgesrelativelyfew casesof market&ilure, revisionistscontend that markersconsistendy fail to guide investment 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 convincingly the scope of governmentactions to promote industrial development in Japan, Korea,Singapore,and Taiwan,China. But, in generalits
proponentshave not claimedto establishthat interventionsper se accelcratedgrowth. 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-intensive industries by boosting wages,policieshave dearly backfired.Thus
neither view fiullyaccounts for East Asia'sphenomenal growth.
The Market-Frendly
View. In describing the policies associated with

rapid growth, WorldDevelopment

Report1991 (World Bank 1991b) expands 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 enterprise, keep the economy open to intemational trade, and maintain a stable 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 developing 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 outcome of many different policy instnuments. And the instruments chosen, particularly in the nordteastern HPAES,Japan, Korea, and Taiwan,
China, sometimes induded extensive government intervention in markets to guide private-sector resource allocation. The success of these
northeastern economies, moreover, stands up well to the less interventionist paths taken by Hong Kong, Malaysia, and more recendy Indonesia 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 selective interventions. Among the most important fimdamental policies are
those that encourage macroeconomic stability, high investments in
I0

human capital,stableand securefinancialsystems,limitedpricedisrordons, and opennessto foreigntechnologySelectiveinterventionsindude 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 higherprnducrivity growth.
We maintainas a guidingprinciplecharfor interventionsthat attemptto 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,especiallyin earlystagesof development.Wethen interpretsomeof the interventionistpoliciesin East Asiaas responsesto these coordination
problems-responsesthat emphasizedcooperative
behavioramongpnvatefirmsand clearperformance-based
standardsof success.
Competitivediscipline is crucial to efficientinvestment.Most
economiesemployonlymarket-based
competition.Wearguethat some
HPABs havegone a step furtherby creatingconteststhat combinecompetitionwiththe benefitsof cooperationamongfirmsandberweengovernmentand the privatesector.Suchcontestsrangefromvery simple
nonmarketallocationrules,such as accessto rationed credit for exporters, 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 insulatedfrompoliticalinterferenceis 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 functionsof growth.Rather,theyusedmultiple,shiftingpolicyinstruments
II

RAC

MIRACLE
LAN

in pursuit of more straiglhtforwardeconomic objectivessuch as macroeconomic 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 singlepolicy 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 generally 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 private investment.
Many of the policiesthat fosteredmacroeconomicstabilityalso contributed to rapid export growth. Fiscaldisciplineand high public savings 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 encourage manufacturedexports.In Japan,Korea,and Taiwan,China, govern12

ments establisheda pro-exportincentivestructure that coexistedwith

moderatebut highlyvariableprotectionof the domesticmarker.A wide
varietyof instrumentswas used, includingexportcredit, duty-freeimports 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 protection 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 "developmental 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 econornicdevelopment wasimpossiblewithoutcooperation(seechapter4).
ThePrinciple
ofShardGrowth.
To establishtheir legitimacyand win the
supportof the sociey at large,East Asian leadersestablished
the principle of shared growth, promising that as the economy expanded all
groupswould benefit.But sharinggrowth raisedcomplexcoordination
problems.First, leadershad to convinceeconomicelitesto supportprogrowth policies.Then they had to persuadethe elitesto sharethe benefits 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 ferfilizerpricepoliciesto 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,governments assistedworkers' cooperativesand establishedprograms to encouragesmall and medium-sizeenterprises.Whatever the form, these
programs demonstrated that the goverment intended for all to sharc
the benefitsof growth.
Cualmg
a Business-Fiendly
EnvironmentTo cacidecoordination problems, leadersneeded institutions and mechanismsto reassurecompeting groups that each would benefit from growrth.The first step was to
recruit a competent and relativclyhonest technocratic cadre and insulate it from day-ro-daypoliticalinterference.The power of thesetechnocracieshas 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 Singapore 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 anodter, the deliberation councils are intended to make allocation rules
dear to all participants.
UsingDeliberation
Council.In Japan and Koreatechnocratsused deliberation councilsto establishcontestsamong firms. Becausethe prvate
sector participated in drafting the rules, and because the processwas
transparentto all paricipants, privatesector groups becamemore willing participantsin the leadership?s
developmentefforts.One by-product
of these contestswas a tendencyto reducethe privateresourcesdevoted
to wastefulrent-seekingactivities,thus making more availablefor productiveendeavors.Deliberationcouncilsalso&cilitatedinformationexchangesbetwren the private sector and govecnment,among firms, and
between management and labor. The councils thus supplemented the
market'sinforrnation transmissionfunction, enabling the IPAEs to respond more quickly than other economiesto changing markets.
'14

communicationhavenot bem staInstitutionsofbusiness-government

tic in theHPAJ.The roleof the deliberationouncilis changingin Japan

role,alongfincand Koreato a moreindicatveand consensus-building

thecouncilsappear
lines.InMalaysia
tionalasopposedto industry-speific
in importanceandscope.In Thailandthe formalmechato beincreasing
becnusedto presentbusinesses'
havegenerally
nismsof communication
positionstogovermentand to reducetheprivatesector'ssuspicionofgovdevelopment,
asineconomicpoliymakingEast
ermnent.In institutional
havechangedwithchangingcircumstances.
Asiangovernments
AccumulaingHumanand PhysicalCapital

Drawingon the strengthof theirinstitutions,EastAsianeconomies

used a combinationof fundamentaland interventionistpoliciesto
achieverapidaccumulationof human and physicalresources.Fundamentalsincludedsuchtraditionalgovernmentobligationsas providing
adequateinfrastructure,
education, and securefinancialinstitutions.Interventionsinduded mild repressionof interest rates, state capitalism,

ofrisk(seechapter5).
and socialization
mandatorysavingsmechanisms,
Bilding Human
Capiti. The East Asian economieshad a head start in

termsofhumancapitalandhavesincewidenedtheirleadoverotherdevelopingeconomies.In the 1960s,levelsofhumancapitalwerealready

economies.
higherin the HPAEsthan in otherlow-and middle-income
Governmentsbuilton this baseby focusingeducationspendingon the
lowergrades;firstbyprovidinguniversalprimaryeducation,
laterbyincreasingthe availabilityof secondaryeducationsRapid demographic
these effortsby slowingthe growth in the number
transitionsfacilirated
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 postsecondaryeducation focused on technical skills, and some HPAEs
importededucationalserviceson a largescale,particularlyin vocationally and technologically
sophisticateddiscplines.The result of these
inclinedhumancapitalbvewell1policieshasbeena broad,technically
-stited to rapideconomicdevelopment.
.IHPAE
educationpoliciesalsocontributedto 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

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.
The HPAEsincreased savings and inbcreasingSavniugandImnestneuL
vestment with a combination of findamental and interventionist policies. 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

Figu 4 SavingsRatesof HPAEs

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 reforms In Japan, Korea, Malaysia, Singapore, and Taiwan, China, postal
savings systems lowered transaction costs and increased the safey of saving while making substantial resources available to government. These
initiatves promoted rapid growth of deposits in financial instiutions

(seefigure4)

4)

andaSelected Eiconomies,19704

(se fiur

HPAEs

nisms to increase savings. Singapore and Taiwan, China, maintained unusually high public savings rates. Malaysia and Singapore compeled high
private savings rates dtrough mandatory provident fund contributions.

Some governments also used a variety of more interventionist mecha-

- -

Othor _
0

2.0
20
30
'Percenta,ge ar GDP

40

Nenar.Finlind,uds AustriaBthee
um.

Dcrnmark,
Finland,Fmncc,theE:ederol
Republicof Gcmanybebore
reunificaion,
Greece, Iceland,Iredand.Imly, LuxemNorny, Portugal.
bour, dte Netherlands,
Spain,Sweden,Switxrland,
and the
UnitedKingdom.'Orie indudesthese
&-veoping
cconomies:
ArgenonaBrazil,T
developin eColombia.
Carenina
SEl,
Chile,Cdlombia,COredivoire, Egyp4,

Ghana,India,Mexico,Morocco,
NiScria,
Pakistan,Pem,SriLnka, Turkey,
Urugua.
the
upEra
Urgan
y Ycnzuels
Venezuela.
the fone YugosIiaI

and Zaire

SawreuSummer
andHeson (199l).

I6

'Japan, Korea and Taiwan, China, all imposed stringent contrls and high
interest rates on loans for consumer items, and levied stiff taxes on socalled 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

ivoring investmentand of policiesthat kept the reative pricesof capimalgoodslow,largelyby avoidinghigh tariffson importedcapitalgoods.
These fiundamentalpolicieshad an important impact on pnvate investment. Third and more controversial,most HPAEgovernrents held deposit 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 avoidingfluctuations (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 borrowers 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 governmentitselfor by privatesector banksworkingwith goverrunentguidance.This
heightensthe risk that capitalwillbe misallocated.Thus there is a tradeoff betweenthe possibleincreasein investmentand the risk that the increasedcapitalwill be badly invested.There is some evidencethat in
Japan, Korea,and Taiwan,China, govemmentsallocatedcredit to activities 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 negativerelationshipbetweeninterest 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.
in the northeastemAsian tier,
Finally,some governments,especially
risksto the
haveencouragedinvestmentby spreadingprivateinvestmcent
'7

A?
m
vLRC
y
~~~~~~~~V. ZiN [.IA
RC
LE

s;~

-o

public In some economiesthe govern-mentownedor controlledthe intitutions providinginvestmentfunds,in others it offeredexplict credit
guarantees,and in sdll others it implictly guaranteedthe financal viability 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 Taiwan, China, signalcddirectionsof governmentpolicyand providedimnplicit insuranceto private banks.
EffcientAllocationand Pmductity Change
Somepoliciesthar favoredaccumulationin the HPA,s,induding financial repression and the socializationand bounding of risk, could
haVeadversely
afFectedthe allocationof resources.Similarly,industrial
targetingcould have resulted in extensiverent-seekingand great ineffidency.Apparentlyrheydid not The allocadionalrulesfolowed by H4PAE
govemments-particularly the devicesusedto shift marekeincentives-are thereforeamong the mostcontroversialaspectsof the Eat Asiansuccessstory (seechaprer 6).
Like policies related ro accumulation, policies affecting allocation
and productiviy change fall into findamental and interventionistcategories.Labor marker policies tended to rely on fundamentals, using
the market and reinforcing irs fiexibility.In capital markets, governmentsintervened systematically,both to control interest rates and to
direct credit, but acted within a frameworkof carefulmonitoring and
generallylow subsidiesto borrowers.Trade policieshave indcudedsubstandalprotection of local manufacurs, but less than in most other
developing
countries; in addition, HPAEgovernmentsoffiet some disadvantagesof protection by activelysupporting exports. Finally,while
interventionsto support specificindustrieshave generallynot been successfil, the export-push straregyE-themix of fundamental and interventionist 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 successul Asian economies contrast sharply with the situation in most
I8

other developingeconomies.HPAEgovernmentshavegenerallybeen less

governvulnerableand lessresponsivethan other developing-economy
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,adjustmenr to macrocconomicshockshas generallybeen quicker and less
painfiul in East Asia than in other developing regions.Rapid adjustments 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 incomestwice 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.
CaprtalMarkets
andfllocatio. Most HPAEs influencedcredit allocation
in threeways:(i) by enforcingregulations
to improveprivatebanks'project selection;(ii) by crearingfinancialinstitutions,especiallylong-term
credit (development)banks, and (Li) by directingcredit to specificsectors 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 ensure their solvency.In Indonesia, Malaysia,Singapore,and-Thailand,
banks areprvately ownedand exerciseindependentauthorityoverlendbig. 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 priorityactivities. All East Asian economies except Hong Kong give automatic access 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 shipbuilding, chemical,and automobileinduscries.
was
The implicit subsidy of direcred-creditprograms in the HPAEs
generally small, especiallyin comparison to other developingeconsupport to ifomies, but accessto credit and the signalof goverunment
vored sectorsor enterpriseswereimportanL In Korea,the subsidyfrom
preferentialcredit waslargeduring the 1970s,resultingin a largegap between bank and curb market interest rates. This gap has dedined
sharplyin recentyea, as Koreahas shifted awayfiom heavycreditsubsidiesto selectedsectors.In Japan imnplicitsubsidiesweresmall, and the
direction of credit may have been more important as a signalingand insurancemechanism than asan incentve.
AlthoughEastAsiasdirected-creditprogramsweredigned to achieve
policy objectives,they neverthelessinduded strict performancecriteria.
In Japan, public bank managerschoseprojectson basiceconomiccriteria, 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.Recenrassessments of the directed-creditprograms in Japan and Korea provide
microeconomicevidencethat directed-creditprogramsin these economies 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,perfbrnance-basedalloation and monitoringand thereforehavebeen largely

unsuccessful.Evenin the norther-tier economies,the increasinglevelof
financialsectordevelopmentand theirincreasingopennessto international
capimlflowshavemeant that directed-creditpograms havedeclinedin irmhaveliberalizedtheirfinancialsectors.
portance,as the economnies
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 speededtechnologyacquisition in Hong Kong, Malaysia,Singapore,and, more recently,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 acquisitionofforeign knowledge.Often they have been preoccupiedwith supposedlyexcessivepricesfor licenss They have refised to provideforeign exchange
for trips to acquire knowledge,been restictive of DFI, and have attempted prematurelyto build up their machine-producingsectors,thus
forgoingthe advancedtechnologyembodiedin imported equipment.
PrmofingSpecificlidusies. Most East Asian governmentshave pursued 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 imported 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 haveaffected 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 knowledgeintensiveindustries,they also took stepsto ensurethat they were fostering profitable, intermationally competitive firms. Moreover, their
21

rRAA
CVLIACLE

industrial policies incorporateda large amount of marker information

and used performance,usually export performanc, as a yardstick. Efforts elsewhereto promote specific industries without better informadion 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 alsodisappointing effortsto build an aerospaceindustry in Indonesiaand to promote heavyinduswtries
in Malaysia.
Export
Push:AWinning
MixofFundamentals
andInterentionsOne combination of fundamental and interventionist policies practiced in the
HPAEs has been a significantsource of rapid productiviy growth: the active 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 liberalized, and currencies frequently devalued, to support export growth.
Overall, these policies exposed much of the industrial sector to international 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, Chinahalted 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 subsiclies 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 cxport would come.
22

East Asia'ssectoral policies were usually geared toward export performance, in contrast to the inward-oriented policies of less successful developing economies. Japan, Korea, Singapore, and Taiwan, China, all
relied on economic performrancecriteria, usuallyexports, to judge success.
For example, in Taiwan, China, the government suspended domesticcontent requirements that interfered with the exports of foreign investors. 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. Because 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 informationreated 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 significance of exports to rapid productivity growth.
These experiencessuggest that economies that are making the transition from highly protectionist import-substitution regimes to more balanced incentives would benefit from combining import liberalization
with a strong commitment to exports and active export promotion, especially 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 macroeconomic management, and limited price distortions, there would have
been no basis for growth and no means by which the gains of rapid productvty change could have been realized. Policies to assist the financial
2-3

-XIAIEMI

RAC LE

sector capture nonfinancial savings and to increase household and corporate savings were cental. Acquisition of technology through openness
to direct fbreign investment and licensing were crucial to rapid productivity 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 secondary 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 interest rate controls. Unlike many other governments that attempted such
interventions, HPAE governments generally held costs within welldefined 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 importance ascribed to macroeconomic stability, interventions that becamnetoo cosdy or otherwise threatened stability were quickly modified
or abandoned.
Whether theseinterventions contributed to the rapid growth made possible by good fundamnentalsor detracted from it is the mosc difficult question we have attempted to answer.It is much easier to show that tie HPAEs
limited the costs and durtion of inappropriately chosen interventionsitself 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 combined 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 northeastern 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 mar24

Icetsto the outside world. Furthermore,strong bureaucraciesand a genemlclimate of government-privatesectorcooperation meant that their
restrictionson capitaloutflowswere more effectivethan similarrestrictions in many other economies.In today'sincreasinglyglobal economic
environment,fewgovernmentshavethe abilityor desireto closetheir financial markets. Indeed, many East Asian governments are in the
-proes of liberalizing restrictions on capital flows. In such circumstances,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 partners. Furthermore, like financialrepression,thesehighly directed interventions 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 environment. Creatinga free trade environmentfor exporters,providingfinance
and support servicesfor small and medium-sizeexporters, improving
trade-relatedaspects of the civil service,aggressivelycourting exportoriented 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,allocation, and productivity ,rowrh. The sheer divenrsiyof these policies
precludesdrawingany simplelcssonsor makingany simplerecommendations, exccptperhapsthat pragmaticadherenceto the fundamentalsis
central to success.These market-orientedaspects of East Asia'sexperience can be recommendedw;ithfew reservations.More institutionally
demandingstrategieshave often filed in other settingsand they clearly
are not compatiblewith economic environmentswlherethe fuindamentals are not securelyin place.The use of contestsin Japan and Korearequired competenrand insularedcivil servants.In parts of Sub-Saharan
Africa and Latin .Amnerica,
and elsewherein Asia where such institutional 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 circumstancesisa key element in economicsuccess.In the followingchapters we explore more fully the contribution of fundamental and
interventionistpoliciesto East Asiis remark-able
growth, 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-

trialeconomics
arguablyforallofthiscentury,mayat first
seemto be an inappropriatesubjectfor study.However
manyof the policyinstrumentsusedby theJapanesegovemmnentduring the period of rapid growth, 1950-73,

26

have been used subsequentlyby devdoping economies.

Thus, notwithsranting
Japan'slongerhistoryof modem
economicgrowrh,it may providesome usefulinsights
into the relationshipbetweenpublicpolicyandgrowth.

Growth, Equity, and

Economic Change
ml .HE EIGHT ECONOMIES OF OUR STUDY ARE HIGHLY

diversein natural resources,culmre, and political institutions. Japan, unlikethe others,wasalreadya relatively 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 distinguishesthem 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 economic 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

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

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 catching up with the advanced economies; more than 70 percenr of the de-

velopingeconomicsgrew more slowlythan the high-income-economy

2 More disturbingly, in thirteen developing economies, per
average.
capita income actuallyfell. Growth among the eight HPAEsis quite different. Their growthrates are significantlyabove the high-income-economy 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,particularlybefbre the 1980s, but few others have sustainedhigh growth rates
for three decades.3 Figure 1.2 shows the growth rates in per capita income 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

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

Taiwan,China i

~*Rep.f

*Rep.-f

'aHongKong
Kor

Ko

--

ma~~~~~aa
lballand 111i

hOOh8SIB*
i
2-----O----

-*--_

0 **.

*S

-2

......

-----

-PEs

*A

--

. ...

..-

-.

U HIgh4I.come
econondes
-* HPAEs
S Other developingecononies

*~~~~

.
0

0.2

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

1.0

1.2

- 0.061RGDP602. N = 119: R7 = 0.036.

Neo-This figuroplotsthis egession equation:GDPG = 0.013 + 0.OG2RGDP60
(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

6%-

6$|-

_.!._

._ d_>_______

~~~Singapore**

of Korea

*Mai*a.siaRep.

Taiwan, China

*Thland

Japa

4% -w*,*f

0%~~~~~

*.
4%-

-6%--

.4%

U High-Income economies
*HPAEs
a Otherdeveloping economies

!
-2%

0%

2%

4%

GDP growth rae (percent),

6%

8%

TEM

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 highgrowth, 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 declining 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 middleincome 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

EQUITY.

GROWTH.

AND ECONOX-rO

Figure1.3 Income
inequalitandGrowth
of GCP,1965!89
GOP

growtb per capita (percent)

* Botswana

Rep.of Korea
7

Taiwan.China

Singapore*. Hoagon
' Gabon

----------

apan----------

.M

e Indonesie

:- Malaysia

i Th.lland

_ _ _ _

--

_*I

;___ Brazil
.

IItaly
* Austria

.Spain
o France
* Belgium
AuS ilia
United Kingdom
e Sri Lanka
* iine
Switzerland'
Ia Pakistan

Mec,

'0Kenya

.A.t--

*NeChIIe
*PBolivia
*

Bangladesh

Colombia

i*

* Venezuela

. India
lMalaw_

1-._

;.

Argenttna'j
*eruge!tn

- Coted'lvoErr
* Maurita
nia
*. Ghana

I
*

a Sudan

-1

--

. __--...

- .-

Za~~mbia

$_

10

15

20

25

30

I
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 increasedvariation in incomes. Second, reductionsin inequalityin Thailand havebeen relativelyminor comparedwith chosein the other HPAEs,
although Thailand's performance is still better than that of most developing economies.
Givenrapid growthand dedining inequality,theseeconomieshave of
coursebeen unusuallysuccessfulin reducingpovertyTable 1.1 compares
the declinesin poverry,definedas the inabilityto attain a minimal scandard 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

iYPICALLY,

AS AN ECONOMY DEVELOPS,AGRICULTURE'SSHARE

tof
he economy dedines. The six HPAEswith substantialagricultural 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 relativeimis not becauseagriculturehas lacked
portance of agricUlture in the HPAEs
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 Revolutiontechnologies. Their substantial investments in irrigation and other rural infiastruct hastened adoption of high-yieldingvarieties,new crops, and
the use of manufacturedinputs, such as fertilizerand equipment, to cultivate dthne In Taiwan, China, during the 1950s, 45 percent of rhe
32

GROWTH,

EQUITY.-

AND ECCOiL

Table1.1 Changesin SelectedIndicators

ofPoverty

A&vnenq

PTerntage ofpoprdauion
below thepoverty line
F.rst
Last
year
yar
aAnge

Year

N mber ofpoor (millions)

Fis Last
Percent
'ar
year
chang

HPAEs"
Indonesia
Malaysia
Singapore-Thailand: `

197242
1973487
1972-82
1962-86

58
37
31
59

17
14
10
26

-41
--23
-21
-30

67.9
4.1
0.7
16.7

50
41
45
30
54

21
25
24
31
43

-29
-16
-19
1
-9

34

-9

23
27

-31
-10

:36.1
8.9
0.8 .
-3.1
311.4
.
6.6
26.5
3.9

30.0
22
0.
13.6

-56
-46
-71
-18

Ot,ers
:Bra4YAb
:Wolombia
:. CostaRica
C6reds'Iire
I-ndia

0 1960-80
-1971-88
197146.
1985-86-1
1972-83

Morocco-R;
Pikismn0;
.-. >

1970-84.
.

Sri-Laik

1962-84 .
1963-82

-43
;

54
37

25A
7.5
0.6
3.3
315.0
7.4
21.3.
.

4.1

29.6.
-15.7
-25
.6.4
1

:12..

-19
5

:NoeThiis ableuseseconomy-specificpovertylines
Officialorcommonlyusedpovertyrlineshavebeenusedwhenav2iblec In other
hasbeensetas 30 percentofean incomeor expeaditrc.Theenge of povertylincs,epressed in tems ofewendi. perhouseholdmemberandin tersofpursingpowrpari
(P) dollars,isapproximnatdy$3004S700aycarin1985ececptforCosaRic (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).
C3SC5die poveyie

growth of agriculure was due to rising productivity, much of which resulted 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 comparisons 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, transportation, electicity, water, and sanitation. Table 1.4 shows that there
has been a more even balance between rural and urban public investment 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 effectively provided electricity to rural areas. Since the early 1980s, elec-

33

I :1R.ACLE

Table1.2 LifeExpectancy
at Birth,1960and1990
Lif expeeane,
:-&onoalrfreon
-;HPAEs-

1.990

:,--

HongKong
Indonesia
-K(orea,
Rep.of.
Malaysia
Singapor .
Nailand: - -:
Th:

birthnh
6a)

-1960

;-.

OtherAsW:
-Chin.

-: -

64

78

46
53
58
65

59
772
71
74

- 52

.68

.0

.62
-69

43

In&dia

.47

-Sub-SaharnAlea
.
Lad Ame7ia and Cribbean
:Low-Inome econom

58.
52

43
54

.s

-Middle-lacnc
ieconomies
-Industnaleamenimie

36
. 49

70.

6266

70

N- TAherCgionalpavsarweghreda.;
-tExdudes Chinaandndi:L
SountWorld Bankdra.t
--

77:.

tricity has been universally available in the rural areas of Korea and Taiwan, 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 disparity in econorries with approximately the same per capita income (Bolivia and Liberia) or the same population (Brazil [see table 1.4]).
Table 13 GrowthRatesofAgculturalIncome,LaborForce,andSharein Output,1965488
(percent)

:,,-;,:
t

:'

-irc,--:

SouhAsa
asrArn

:~xree..LLL2jt....

198

41

22

Decline,Average
A

2.

~~c:~;'nA2,
34,.,

34

.;-:-;:e
196
,3

1965-88l.

:10"

21~~3

~..

anased'growth rv&eof
Atcrds
A'd'-:.-.:'..'
kiborfirce
'

:-t

34
S

t,i d

37

LO)

17

"
<2., 6 1.l9U.::

rru8,,,A

2.2

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 mdustrial growth, have funneled surpluses from agriculture to industry
through taxes,food price controls,and pro-industryallocationsof public 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

1-

101
102
113

S PA Es
Korea,Rep. of
;54
-26
Thailand
Indoncsia -168
OtcrAia
Latin America
SubS
- bnibwAfrica

:-

64.6
58
43

38.6
4665
30

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

co~namy
' PAEr
Indonesia,-1984
:M.a:ilays.;1;983 Thailand,1984

Ur

: -:0: g 5-

Rura

39
85

10

55

78

:40

-OrhlerAsik:

-Bangradash, 1981 :20

:India, 1981
-S -Lakna,j982
LeninAme-ic

<25
35

-..

15
8

*~ArcndinaJ
1982Bolivia, 1981:
.. Bra;Ail,
1981

->95
72

Senegk1982

Soercertoppanel:
UNDP

^
*

~-;

5
9

>95

..........
:0 0- - :

:-.i
N

;C6tc -dvoir, 1981

i iberia,1982
-

-~

19.

93.
86
83

20

:
;
.

('r
fariousyis);Bottompan. Muasii

~0~

.12

(1987)

.; -

..

3S

_RACLE

manufacured imports reduces thc domestic currency proceeds of agricultual exports. Industrial protection acts as a hidden tax on agrculture,
raising the price of agricultural inputs to subsidize industry. Figure 1.4 illustrates the lower taxation ofagriculture by contrasting three HPAEswith
large rural economies with three South Asian economies. Direct interventions include export taxes and price controls, while indirect interventions also take account of industrial protection policies and real exchange
rate overvaluation. Both Korea and Malaysia have substantially lover taxation 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'
__.

K.

PSilippines

_,

.........

_. _---_

-.

_ _ +_
_ _ _-1_
_j___________________________
Sri lanka '::H-~

-60

-50

---

-'i-~~R

19709~~
1
j

. .

-40

-30

*;

__I___._________

30

40

50

-20 -10
0
10
20
Levelof piotection (percent)

:~~~I

60

Rep.
of Korea
alys

...

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

Thaland .

--

PsidalsiU

=96
___

____

i
___

~~~~~N1970s
3---197ks
I
k
I

PhEippines

Sri lanka IE~

-2

Sour= Schiff(1993).

36

.~

~ ~~ ~ ~~
~~~
l_

~~~~~~~~~~
I

2
4
6
Agrutural growthmte (percent)

__

10

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 resources from agriculture to industry. Rarher, in contrast to ocher developing 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

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 (Indonesia, Malaysia, and Thailand), which had been historically dependent 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 evidence at the outset that East Asian economies would achieve such spectacular 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 increase in the openness of thcse economies, the share of exports plus imports 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

i965-90
SelectedEastAsianEcononies,
Table1.5 ExportPenetration,
Share in wuorldeports
Eeonq
-

:1965

1980

S-are iL developing-economy xporm

1990

1965

-1.980

1990

9.0
6.7
2.4

-7
6.0
6.2

13.3

33.9

7.8

12A

18.2
198
100.0

12.2
100.0
NA

21.1
100.0
NA

56.3
100.0
NA

13.2
1.1

44.9
3.8

61.5
12.0
73.5

-Totalmg,parn

Japian
Fourflgcr-

5.0

-7.

NEs

SoutheastAsian
HPAE subtoalAll developing economies
World
:

Japan> -.
--

Fourr

--i

-Nor

2.2
-13.1
28.7
100.0

7.9
.242
100.0

=.- ;-

7.8

rsa

AIIdevc1opingeconomies
*Wodd

1.5

a-u.-:;

SoutieasAsian NIEsbHPArEsubtotal
*

3.8

1.5

1.5
0.1-

9.4A
11.1

100.0

11.
5.3
O
17.3

7.9
1.5
21.3

11.8

12.9

11.8 -

1;00.0

-100.0

14.-

48.6

100.0

100.0

100.0

NA

NA

NA

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

.1.

and dearh rameSbegan in Europe and North America with the

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

HPA&
'Hong Kong.

Indonesia- .
-lorea,Rep. of:
Japan
Malysia
Singapore

-Taiwan,China

1980

1.50
0.25
- 0.32
0.19
0.89

'0.25
.1.00

2.12

3.70

-.0.53

0.95

1.52
0.46 :
0.63

1985-

1.988

1.78
2.82
-0.38
-0.42
:0.66
0.66
0.23
0.11
0.85 -1.09
-2.77

3.47

...0.82

0.90

049

0.44.

0.35

SubR
-SSI,arnAfr-k
- 0.24
: 0.30
5out6Asia
-;
; :; 0.17
-: .-;
A-men-Amc'sa
ad Caribbema 0.20
025

0.27
- -0.11
0.16
0.22

0.45

Th1ailand

0.28

0.19
.0.23

N..-eTotlrade .-valueofxportsandvalueofiimpomr
dividedbygrassdomesdc
-

-.- produce :-.

- -t'
-.SowrwWodd'Bandata. Taiwan,China,vaniousissues.Nadionial
Accounts
SMtaddcs:AnalysisofMainAggegates, 1988-89 (Unietd
Nations).
-

During the period 1965-80, all the developing regions of the world
experienced a marked decline in crude death rates (see table 1.7). The decline 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 explosivegrowth in population from the rapid mortality decreases. In SubSaharan 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 sufficient to reduce the rate of population growth in Bangladesh but insufficient 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
cr-de

-ange in
crude

birth

eadl

rante,

Eonomy
-

Aventgeannul
grovtt of
popadation

rate,

19650

1965-80

160-70

1980-90

--52
-40
-54.
-25
45

-54

-46

-58
-16
-30

2.5
2.1
2.6
2.8
2.3
3.1

1.4
1.8
1.1
2.6
2.2
1-.8

-31
-40
-33
-31

-36
-55
-50
-38

-2.8:3.3
2.9
3.8

2.2
2.0
2.3
2.7-

EaswAsia

* HongKong
Indoncsia
. Korea,Rep.of
Malaysia
Singapore
Thailand

-55
-

LatinAmen'ca

Bazil Mexico
Peru

Venezueda
Sub-Saaran
Afri-.
Ghana
Kenya
SierraLeone

-6

Tanzarta
SouthAj~a
Banglad
India
Nepal
.-Pakisran.
I

:-2

-3A-

-50
--29-

-3.2
-1.7

3.8
2.4

-2

-22

2.7

5.1

--33

2.5

2.3

-45

2.3 1
2.8

2.1
2.6
31

-27
-33

-13

-:

-13

-Nor avaibbk.So5Wrrw
WorldBa dara

23

-13:
-2

42

-43
--

:-

HighInvestment
andSavings
Rates

HYSICALINVESTMENTINCLUDESALLOF THE ECONOMY'SOUT-

Jput that is not eitherdirecdyconsumedor usedup in the producuionof 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,savings is the only sourceof invesrmenr,and the two, by definition, must
be equaL But in an open economy, investment can be financed by bor40

GROWTH,

EQUITY.

AND ECOQi

rowing from abroad as well, that is, with foreigners' savings. Even so, inadequate domestic savings will eventually pull down investment rates,
either direcdy or through constraints on the continued build-up of foreign liabilities, which must eventually be repaid firomdomestic savings.
Between 1960 and 1990, both savings and investment increased
markedlyin the HPAEs, outstripping -',e performanceof other developing 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 levels 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 SubSaharan 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

Figure5 Savings
andInvestnent
as a Percentage
of GDP
Gross
domesticinvestrnt
Developing
HPAEs
LatinAmericaandCauibbecn

SouthAsla
Sub-Sahaan
Africa
Gross
domesticsawings

Developing
HPAEs_-_

_|

i .1-...-&----I --_____________________________

LatinAmericandCaribbean
SouthAsia
Africa
Sub-Sahoran

: - .

g.

e10

20

-30

40

N-eo:The regionalavages arcunweiglhted.

SepurcWorld Bankdarn.

4'

RAC L E

Level,
1960
andGDPperCapita
ofGDP,1960-1985,
InvestmentRateasa Pementage
Fgut 1.6 Average
AverageInvestmentrate (percentageof GOP,190-8S)

40

30-----

bnga*Japan

0 *Malaysia

Rep. of Korea

Talwan.Chilna
207

-6-- -

'.-

'

SXhalland

111gb-Income
economnies
* HPAEs
* Otherdevelopingeconomies
U

1'

0--

0.2

0.8
0.4
0.6
Relativeper capitaGDP(percentageof U.S. GDPpercapita, 1960)

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

.0

2.2

119: R2 = 0.295.

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

the top quartileof all low- and middle-incomeeconomies,but sincein5 What is

vestmentgenerallyriseswith incomes,this is not remarkable.
remarkableis their high share of private invesrment.Figure 1.7 compares averageprivate and public investmentshares in GDP for 1970-89
in the developingHPAEsand other middle-incomeeconomies.Private
than in
investment is about 7 percentage points higher in the HPAES
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 recoveredby 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 developingeconomies diverged.In other economies, the fiscal contraction of macroeconomicadjustment 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,public investmentin 1980-37 in these East Asian economieswas countercyclicalto 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 education 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 highincome economies(seeappendix 1.2).
Figures 1.8 and 1.9 present a stylizedsummaxyof the resultsof regressingprimary 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 secondaryenrollnmentraceof 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

1984

1979

1989

Public Investmfent/GDP
(percent)
20

------

5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

10

-U111FPAEs(excluding Japan)
0

1969

II

' }

1974

~~~~i
;

1979

-4i

OtherLMIEs
I

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 incomepredicted 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

120Brazil(108)

.. ......... .

ioo

.n lindonesia (1118V
[37.0] HongKong(106)

--

Singapore(1053

Rep.of Korea-(101)u
[29.2,1
. Malaysia(90)
Thailand(78) [17-9_

Brazil(103)

~~~~~~~~~~~~~[-7.5]

[21.9] *

HongKong(103)
[15.61

_-

Tbai95)
[7.01

i7aysb(102) Rep.of Korea(101)

[-7.8
[4.3]

8 [10.8]
Indopes15

72)1

-;

I60 ~
pesI5ji2y
d
-- _ _ _

*Bangladeh(4.9)
4-Palkistan (40) 6$2...........
1-24.5]

_Banigtadesh
'[8

,Pakistan

1965

40-its

(59)
(52)
[O79]1
-.

Lw

20--

250

500

750

. ------- 197---

.
1,750

2,000 2,250

500

Per capita Income (198

U.S.dollar)

1,000

1*500

3,500

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 increasing 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 expenditures 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 between 1970 and 1985 (Birdsall and Sabot 1993b). These dramatic differences 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, standardized across economies. In the relatively few international comparisons available from such tests, East Asian children tend to perform
better than children from other developing regions-and even, recendy,
be'tter than children from high-income economies.' 0

45

& 4TIRACLE

Fimre
1.9

Cross-Economy
Regresin for SecondaryEnrollment
Rates,1965and1987

Secondaryenrollment rates

100

-.

,U

.
so

60

,|

',

-;

Rep.of Korea(88)

_-

Indonesia1

Bangladesh(13)

0- 2 1

akstan

250

500

*!k [19.1]
*Malaysla (28)

77--{
~ir3jidnf

Malaysia(59)

.119.5]

40
40~~~~~~~~

-..

Singapore(45)

_-

--

--

,~-

~slhailand

.-

9)

750

BraA

9*-[18.

ril3s

(28)
I

.BalI (16)f
(12)

>I

Rp're\35 --lt-

Bangadesh
(18)Ft8

119
1,750 2,000 2,'
Percpl

500

1000

1500

3,500

Income(In 198 US. dollrs)

NomtFigures
in parentheses
arcenrollmentraes;bracketed
numbesshowv
residuals.
SeurceBhrman andSchneider(1992).

How much a childleams isalsoinfluencedby the nature of the leaming 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

HE FINTAL
OUTSITANDING
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

rT-77

MT VT!"12

Ze.

w,Was-,FAA

,i Mz

11AAHI'..MRICALLY-k6---,----!:
eci c.- ene tv o cu!,adng:V
.9d.
es Lrsuto - oyt i'an
biin subiiiiia 7
4.
lop.1
o y a=,;-.,us ygra uaUj;'..'ib;gdsj gender'
J--Th
d
il
iffii:ntii
f6cyiii-,.,
aryan sc6ndzrj'k:&
-rnenriv
n";'evc
1.o-ca-d6i.
in
-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[
rnaryp_
c
ccorioriuq,-,gnder
V V
WM Swl"cr-,,E=ticxpected.,,-Bye'--HPAB_
e.--homm`e:gcn-':-'.,-',;;,'
'j!aI[70Wid-,i" The,,;,.t-',"Jt)r,outst
e,--.tlii
'1987'thi,&ur
jan!eqonoa,Lies-f6i.W.W&6ca.
pmg
ot er eV 0
aM
r; ;cconomjqs,' splteZwr'u''iat,norms-,,ttl2t.
put, greater--, achieveduniversaip6mary
cation
C, :r
Wqq01Lucation-ofs6diaahi,
tisci;
-Ov-iiiiirii'n'aungthe''gendir p-at'th'a-'t-1e'%'rib7SLLbi6iL
C
dau
-d6cationl.,
ii
ai
dvt:
I
gender-gaps,M'Mm um mtescontinik46"
C CXCCPtI0r4.-in-liow-tncom"C'
M
e; ra er.
SUCCCSS
0r-t1MVCrsJid[7-,
Is
PUS
7-,C ...
P-M
'. -,; Z:. - , 'yucauon
Sphe'
thi! m c3q%=ono e,,
qarrip
oni
ive4rgu -64c4,,th
f G61
:F6v
=&T.1fij,'-1991
S31S
7
lecon6nues, GNPpq.qLpibi
'$500'-%
-in
aTsmoolv -o
0'ff
m Miienr, etwccwiboys
as,'no-r4n
fs-,--=,a,
ges-2 6fita- i pio,in,
.AD"Mak-Yea 6f'Edudfialfl
IF.AJRL''

,i'.

10

-'.

E.

-;-

"

-, -

-.

- , --

-. 4.-, -"s.

I.

..........

Sl

t Ca-1r0m-1

64i

'I.

In!;1965,

-.44
nAmi

C5

70,

1-ir '(6c,cq 6 Thi

gcn..d&,gaps-,-avtqe-k'di
Iiidcc'd'-"
-16iill:
aryschool-pp
increastrig.second
.2-C
as;mu
e;M
'an
Ij
-,q80s!2,S1t-.1S
rhl --v-166'.1-970icl
ey opmg.-wo
&
SUCCCSS Ilh laL%ngj
-Ccon6mli
.98
cntsto
gn !cr:ga,p
V4.1
-wJ-`6ii6,
ii
I
Ja
er
1)
capim-41
-6
A-siadeconornies.5nour,stu
PV

4-. -jk

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

betrer organi(TFP) growth." Among its sourcesare bectertechnology,

zation,gains fiom specialization,and innovatons on the shop floor.
As we demonstrate below,about two-thirds of East Asia'sextraordinary 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 actributable to increasedefficiencyor TrP.This is largerelativeto other economics, both absolutclyand as a share of output growth, and therefore
partly explainswhy theseeconomieshave been catchingup wriththe industrialeconomies,whilemost other developingeconomieshave not. In
box 1.2 we reviewtwo broad interpretationsof the relationbetween accumulation and output growth and the link between productvity

changeand technologicalcatchingup.
Regressions
Evidence
fromCross-Economy
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 schoolingcontributesignificantlyto economicgrowth.'3 An increaseof 10 percentagepoints of
GDPin the rate of investment-about the differencein privateinvestment
rates between HPAEsand other low- and middle-income economieswould 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__

)NOMISTS,

ROAD,

1-4

mmmenr

onj'am Murpur
-w-V

VC.'.

h
e --o

Flj
V4,1.

3K

umanan

er:

accumuw-

.,Jf.

ton;t,&I

3Vlcwlls"_b"l

_T'.
yIC

um2n ppi

ajt

rexten
-

ait

r=en

ini

p
nonluesiysi
&pfil&72ACt

irk*,

-o'wth' an
1.06

-,,-,,.amo4ccon6n3jcs_cr
ccononiic,_-or

".w

rz-k iE66-Er;

esame,
-4
assem-

MR

6i&bffinva=
nonuc-.,gr

P tLon

_-llFAE:fa
"d

gan,=qon;,-jnnovau
o ogyarc
o arm y.14666i

-0

'lik"

zin

iAiii '446
ccum non-,
ii'ibilbimer
vict-, ruo

ll,z,

Om

,p

gm.

2ssum

rec

(SOIW49-5.7)

LW2'CX

0 f -T,
P,P'.
v

coiiiii.econoun

r
Oj_

,-,'2ni

----

i=cs 0
-nu e-mcoifi

.. ........
economc

oil ficior-

Z.-

have

ccum

con-._t-,,, -,-tc
Y

"W o
lporpcmicnVor

Om

_9M

31i

cKami)

CoDrifth

M,
07Ww"Ah

cromki&m

Cj;V;.

aupajmong

J:i

onspi
de

-J

crcmeat&jj06,
er,
6ibasji-,tVoutpu poon
f
ItaL

econo

;_2

gt

_7

c
19P

'gro

0
.4;_

V4

OD

'p

t
Sd'

49

i-MIRACLE
L

':

--.

':~~~~~~r~

~i.

l'--'rr'

r:*rt

',,

,..

:.;,..:,,

?%e 7 aWe to acqwre technologyhniore~ =ie4iy; daiinical efnaen4 chan ge- Bo6htho y -andempi ic
prgres&more rapidlj than induiial-- economics. evidence:leadusto, ncludechat tol&cr prucK>Thisiis
thdeindusirialecmomiioicilrer
.
mov dc
radbe- up..
i"
a vy growthl
;
<
U-pnid4crrybasedcatcbiiig:
j--

r'BuLiiiwrnadbsn4a1 be

targetTow
i---

fpraiceis uiseif amoving,

fiictor.producvity cbangeii; el.igh

t. technologicacg

-nonal
t
e practic4Whileinflow-and middle-icome

.-di

OM '-Co

>~~~~~~~~~

"ltec-noloejcl
r
pracice,
whitWChe
temi
bcwce
slp
ni#eicrsKn
~do
O&

banmcemanon
iia bhs'd

4'

9; >.~movemenrs)owardbestphiice,whicWweolraech'-

-sc toieto, oc

Z =

g5wrkXChenry 1986;B.n an
192
m
-liI;pm~~~~~~n
a &uncV~A9)--4
i,
pendix
in resents acformiantan4is"oEfi rea~onow-bs
;

the-. m

iic

j(Pacad
l .nocr;

mov4iidj
-moIn

-. must

11993XForpductrv-

ccInomits, cam-b
occr,
Ass*

.:

on, 6te
gaicing
'lo-Cand

zzG*

.- Ceposk,iv---:

Economies that were relatively poor in 1960 grew significandy fister

than relatively rich ones, controlling for levels of education and investment 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 technological 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 important 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 producrivity 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
* Variable

O&Osedns

Inte'rcept
RelativeGDi' to U.S.,1960
Priuary enrollment,1960
Secondaryenrolmenr, 1960'

4).0070
(0.0079)
-0.04O
(0.0118)

-0.0034.

0.0264M
(0.0065)

0.0233
(0.0062)

(0.0075)
-0.0293~*-0.0320
(.15)(0.0110)
~

0.0262
* (0.0139)

*Growth of population, 1960-85

Averageinvestmendr/CD.
1960-85

113

obsenmign

0.1015,
(0.2255)
0.0578
(0.0224)

0.0160

0.0069
(0.0131).

~~~(0.2095)

0.0998
(.03

0.450.0285
(0.0211)

(0.02.07)

0.0201

-(0.0056)
*LatiAmrca

0.0272
(006)

(0.0132)-

0.0230

I{HPAFS'.

~~~~~~~113
obsen'aions
0.0042
.(0.0081)

0.0171
(0.0056)
t

*-0.0131

(0.0039)

Sub-aharan
Africae
Adusted
R.

0.0o9.oo
0.48211

0300424

~Stadrisda1y.signfl=wacrtiC
0.0f lVeL.
-'6.

nnsdca1y.sniflc~nt2tt die0.051V&
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
It alsoshowsthe share of acg'rowrh,and relativeincome for eac-hH-PAE.
tald growthpredictedby thesevariablestExcept for Hong Kong (44 percernt),60 percent or mare of the actual growth rate in the HPAESis
predicted by the accumul-ationof physicaland human capital,initial income levels,and population growth, ranging as high as 87 perce-ntfor

FML~"xN
I R-A-CL E

Table1.9 Coutnbution
of Accumulation
to theGmwthof theHPAEs,
196045
,~~~~~~~~~~mm
11os

-aaf hiv

Pear
*VWzh&

mlal

~40676
.

Ian~~~x - car
US.2560
-f-mapa,Imau
15%6
S
5
md
-cmXular
. 19S0i
- -:
Gcpq&u.n2104
GM*
AwpiPdUwunt
160.85

:~

AaW Ewd.A
- U6
P- M&
gn
F-

.X7e

4.030
O-.Jf
e.o2 0.1025

~~~~~J
-L0N12

Mg.

4270
.e70
-ISe
20o
0.3
0X3
lex57

JaUsihIa-

Iha

aeihsh
6

MlUe

uik(si
(.26)
{.Nz
(7)
(a5)
(231
s1) -.
(4)
-1q

--

-..

&vh
5W.

(rf-)

476
(.1)
402
wu78
-141
4 32
(.242
1.77
-9)
0.16
(7)
2
(16) 2.10
It1._D.108 {1918
5.7
223

w .

ral
prial
gush

69
2A

pdiicS

* .vAwM

.6

(vi

-- fa-d
PedSI

4710
.1.1
-2.16
.72
1.4
0.20

If-d
pen&tu

fl,.~A,
P0f +:'

rmd
ugaqh

pia

guhw.
4.151
(1
(sw
1421
(2l
(.1)la
M

=.6P

r_

5
4.69

s
473
-W48

(-191
17-191
(10

+X4

243
e, a19)
0.79
.4

1.40

Lin

--

30

(44)(
cmibdn

yr.

wpz.n

-:dbc

vacsylwdwwAaeesI

.iwnyim
XI

.pi

nbdal

enIa dwbEnul(nr

dc bienps.

flqnJul

rend Grcceyqlsdwwm

.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), followed 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 secondary enrolhmentrates, Indonesia,Malaysia,and Thailand, have the
smallestproportion of their predictedgrowth attributable to secondary
enrollments(lessrhin 15 percent).Investment,conversely,is most important in Hong Kong,Indonesia,Malaysia,and-Singapore,whereit accountsfor 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. Controlling 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),controllingfor the same variables.Thus, the expectedgrowrh rate differential between the high-performing East Asian economies and
Sub-SaharanAfrica or Latin America,even if they had the same accumulation and initial income, is nearly3 percent.
To assessthe contribution of the HPAEs'superior accumulation relative to other groups, we can predict growth rates in four groups of
52

EQUITY.

GROWTH.

... P
-"I'

-. *

a.77

-X

:.f
..

QOb

'-;)-

0,0284:

- 0 ,- . 1141

25 .S
W;t2
0624
uJ

lwt

1473

'.;4

-. 57

13z

1;

Ax-

-; S

3.70
'-R'

.-

C.gi7

.(6)

(69'

(tv
-

fi)ls.

(36)
6.03
3310

638c

'(3m
3.45;SC e

j19J
i)

.0.7

(-4q
-2-

-. 70.

''

.fw

,^c>*d*.s_

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

; ,-

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

4.~~~'~

7
.L4
231;
04
Om

-~i

.Jimut.-

G-

p-

a('
die
442.

(.)3

ANDND

*rr
4NB

t 21)
Om72

.%t

oil

(231

MIS2s{

IA9

(431

au

151
IG

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 groupdifferencesin accumulationor initial income)and to an unexplained
residual(seetable 1.10).Accumulationexplainsonly someof the differencein 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 SubSaharanAfrica.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 difand Sub-SaharanAfrica
ferencein predictedgrowthrates btween HPABs
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 Differences
inGrowthDueto Differences
in Accumultion:
TheHPAEs,
Latn America,Africa,andthe OECDEconomies
:

. . .:;

-:

ff

..Pa r -e .
twst

Ab~~~-

G4CPIZdISIW
I CtLS..9l60
IrInay.enmIm-.-. 1960
- r,tuy a9uIbnnv.1910
-

-- AI
;UAaSt

K-

--

.-

.lar

-013-

-.
Afie.41.

.
-0-

O.

..

Q25
-

iHr:E.
am. I.3

0*3 2
.0*015
M.U095
*73
.

-(.

2427
4003
.I3:

;~~~~

:
-

X4774
774
66
l

.-

:4o

.
*fa

-DiPkmr ad posh

-:w

et

40

Irk

dLmImtMIbimriu
tjwdkm3cW.rhto
hkm6dundanJSwhA-.
c
: ;--T M,mtilhedSehthcrCn.lnfoawsehnwairlEa

:.

Irk

& m.uqe wdwin

nrlm

EbnkrX

,:
.wsqewmbldlucnlueln

DNbuuhudaSdnb.wbmebn&

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

Asian high performers.
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 betweenSub-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 expiained).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
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
iTP

54

O
OEC :AAim-0D

0a7m7
y

4tE574

-:D612

Aq.whFdpla.
156045
*-- PAEn
Dl.
39.--AS
- b- mtME

: -.

GROWTH,

iup,2

*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.ad

UFAA.

wu,f

t-

li2

..

pwe pfnwi

Rel-"

,s''0'

CM!

(11i2

fm.

3,0'..

130
0.2
.-

L9

9C1H
.

- -"' '';'

~~~~A4i.ORD
(;0)

AND ECONONIIMC

.,.

A.Mm

0.17
136)

EQUITY.

H -1,
.<

~~~~4.IM

rs
,,;:.-.
()
(4).

;-

(
4b

., ,h

M-. P11

-013

15i200
;

(I 4

2.32

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

techniques are discussedin appendix Li..) Figure 1.10 plots the resultingrTr growth rstimates for 196049 against relativeincome levelsin
1960. There is greatva-riancein the rates of n.' growth among low-income economies.A number of develo'pingeconomicsshow higher rates
of TEP growth tha industrialeconomies,consistentwith the possibilityof largecatching-upgains.Many others,however,have lowor even negativi 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

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

Taiwan.China

~~*Hong
Kong

*Rep. of Korea

*Japan

Thai!rd
3

2 -U

Indonsia*Malaysia

*5

-2~~~

-2
0

0.2

0.4

0.6
Ratioof GDPto U.S.GDP,1960

0.8

1.0

1.2

Source.WorldBankdama.

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

we.recatching up
offer some speculations concerning which of the H-PAEs

56

GROWTH.

EQUITY.

AND

ECONOMICC

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 allocativelyefficienteconomieshave 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 essentially just keeping pace with technological progress in the highincome 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 (Singapore) a year (see table A1.3 in appendix 1.1).
When the HPAEs are contasted with other developing regions, however, 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 benchmarks, 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, between 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-HPAElowand middle-income economies have contributions of TFP growth exceeding 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 economiesIndonesia, Malaysia, and Singapore-conform to the developing-econ57

&EAAS4
ksA N M I RAC L E
Dueto
Growthand Part of Growth

Figure1.11 TotalFactorProducti
Gmwthof FactorInputs,196049
TFP Growth1960-89 percentpar yer)

,\

4.

4.
4

4.

'. 2966

.4
2

64

.4

.4

Hong Ki

.4

.4

.4.

Highincomeecon

-2

.4

N .4

om.

''

1E

4.__

*.

'

TOtalbector hipLrt growth (percent

Note:Dashcd
line:s tcprcscnc:

*Ta"n.Ch1na

4.an

4.

3[

4.npan

'4

HPAEs
Othiordevelopingecornomles

.*

'2%'

4.
':

0%

D4

-4~~~~~~4
r _
0
4

4.

".

.4
4.

10%

s~~~~~~

' 8%

4
per year)

toul 2vcngF GDP growch raus, 19650 89.

Soli rays reprewnsi

EhecontributionEotoclJgrowthby TFP growth.

S#tirees.Nchruand Dh2rcshw:ar11993):World B3ankd2m.

omy pattemnwithalowTFPcontriburion. Conversely,theproductvitydriven 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 EIPAEs
are unusualarrong developingcconlomies
becauseof the

relativelyimportantrolc of TFP growth.Evenso, we find, consistent

=:

~~~~~~~~~with
our previous findings with cross-economyregressions,that be-

tween60 and 90 percentof rtheiroutput growthctcrivres

from accumulation of physicaland humanacapital. Producdvirychange has been
higher than in other clevelopingeconomicsand is importat to the East

Asiansuccessstor. But it
''- s8

is nor

the dom-inantfactor.

GROWTH,

EQUITY,

AND

,~~~~~~7
-:

-..- -1
;:.tc~~~~~CtW

,111a'S;!

Ati:wud-btcmle
dout.'i cosdeaontf,:
-, - - o- a~v>s i--

I;anoa45!"26

r-. i-.o;.

...- 7

'

,jp,

Y:

=;^

B-soxn TnheChnese M ac

EN.FECNMI(.;OWFI.'EAST..o;
lnim.est:-an-astoundiiien3-...
1 men

;--tic t
'@-'

;!.

smc
-

G..P..

17.ase

ii.

rkia

S
.
C6ns1ump'non.ldf-I

em

da

improvement:

e.-

-.

in

---on..of.m
4.

:-

op.

NO--.
ASSESSMENT
tECns-Ofi
OFapn=enGNOMiC
OdmsIlas
T
nvest aatu
n 30erearonspnit
su''
i y
-?-ea;.e;
.. -aea.
y
.n-ca...r o"'
.& ow.. ub
-b&&mpledy,tcu.wir-thourconsideadolsm.he
71- 1
4-49
f4..r.;
.
metZ' sinledo
17hasoa
see a 1978:t.v
lnlfunaln
akdimrvmnt
1, Asia.ewofd:
:. v_of e19i8_8 oft
t-,
- ,/~t.o.
g~~~~~~~~~~~~~~~1gtCnM-;..
_..;ie-.t
ze. ^ ,5m ;9 ,P;A,E_
A

in

!-

mwlgg
-wcm<tj~
ra
.tame~~nt
pronne:
fyn
,
=-r {to 't;--'v
'-..
- a ->- gpal
x- i ..
>;
T
WrtL

.9.s

-~~~~~
Ghina's

-,: ;

-j

fsh

_.

-te.-

-; ,:gr.d.m7
-rpDi

gntao
wlde .a>mos',-s -o tuc,:2E,
- ;...s
_ JWiflg'staiildsgo
f
r. thia.za,utOr19
comaierdmds.

e.orublihgnedpopuation oFalleigh-sbonomf
mC
- l sour;& W:nimal2'peror inp
the
>
CWfiyP
w;
.i-heiAEwdbledYfroi
198
.
!,
,
:^W3- . ,-s

%AAuciMAdannbaawaL.I
4
.4<. . .;;>t;!-<,.+<-=.<-2.
.&zt
,'4-

be-at
termeo d an'

Jdp

'"-f

;!n

eo

government-fbszeehxci&

evemen
;'ba
and
..recorde.av-nna
gewdoflocDeta;ose
. Bt'.*.urdiW4oremteonents
d. GNPud

iesebqdir alsOna

't?

4'

>-.-

;i'9

slXRoiinsgwing
'g

gnll;6roucy

ar>a.a-.
lled,c-;l

non ti
tliey

ton'
'health,
and ie Cinacrnoeuhhiiiicsr.'bili'ty,
.,China --- ~~~gecrber
wdt
2 jrcenr
ihcreincrTpjancs
w.:erenr'n~lso0,o
:onowicjmeifChina
nuiFxceibk4f%lorX
televi ia ksit
?r 9.leibusexiosnS.-'
iScr&Did1t0
-r.carmEa-relm.t
.;:
eoetsn:
1d16&arnvtlper.
alddyinnually
ill
aulEzdiD
S
t- fro<i;vasmessof,CltinE.lIiaks
B-,
edswmgatn,oterenmdyrai&y.avo&hlltaed
ftrnEm
l ta baarmcsinvsted
T in.ms.
Uk8 Jx:o5
in 1990.-uds,-,
inXrw
nsip-nd
1u.4;>pcrct}en.~R-<t,;saviing
1979
Asih'soCre
H'Fi
growdi-n;
-t zessment
ass, tonenili7
ph-:h^S
'&rf or.ii iah&isi
Af&0MF1982wjt
198~Fo
bier-retreii.chmin
aeti~se.Czi
il
-~

'

-'n

8iehi
tharS.comc.cauamlgyihsEhEr
Droblcmanc.
rcM'4&44eii&
p

mg

3)1125

onpmentn,,T-i.rW
s

4-breald&&ckfTt2r|n

rapid

anL
ndshi
h~iidM suFpn4dealbi.tiowih
c

v-~~~~~~~~~~~~~~~frerLNqs
gnrou,
, ,-wzA;
cottE
tro
{tsIea'te
Ex
St*n
ovr the Ian.1,!
,e5Lrv
Ttpo*k-^nm~uvcs
as
-- ixbiJqjig

n iY-.U:-.-e.,;nomy;Tas

-~ X-.-&WZ-

-Ci
-si
our-p

eee

e&oiiozfies
r
s, A

dof
high
terms
reductionsind ininequali
un rates
captala
a.nd~ W
...

DeoDkallewedEom
Dvmosmce-EF

ofsms
Ch-inniaksj
4.-

.:y~~~~
~~

&n2

flSnr

'.

tt

-ikA'r.-.--oer
oftheecopnomies in-inour

stud_

t_.h.._

i aeynwichiuecnvinwg
cmpa- od wheno %ro r
4-

ip
noive

wererrm
ar-:S,

~~~~~~~~~~~~~~p~~~~~~6~~~~--Ie
,sfxpd

p"-'u

-:

jDiledyLbr.i of

t,s.

-C$Qeiiid

oAtes
r..

nu"C-S
--. .

";- iebhn
comesstIi

tbknaii&
chea

ha
ghptr,wth

--

'

mb
et44 efiuiiveN~rpsgy grwhanx-du&v

aIst..aa'~~~~~~~~~~~~~~~~~~~~~5
r in achieving
Asianeconomxies'
success
We havedescribedtheEast
hlighrates of growth and redluctions in inequality- These were accompamied by-.agricultural
transbrmnaions,
rapid
fertility
dedinci, andmau
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-.

.w

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 described 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 Summers 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 income 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 enhanced neodassical growth model with human capital.
Table Al. below reports our basic results on the relationship between accumulation and growth. The estmated equations compare fiavorably with other studies using similar specifications. The overall fir of
the regressions is good, and the coefficients of the variables are ofthe expected 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
Nwnberofobsennrtianr

-965
1413

Intercept-,

GDPrelativeto U.S

1960

.19160-70 1.970-85
.11.3
11.3

1960-5
61

-1960-85

-0.0070.

0.0064.

-.

-.

04

004

0.0243*

*(0.0079):

(0.0092)

:(0.0109),

(0.0113)

0.8):

(0.0094)

--0.0430"

:-0.0444*

-0.0422*

0.0408**

(0.018)
* Drw.xynrolircnt, 1960

0(0.0103146

0.0264*"
0.0169'
(0.0065) (0.076)

~Secondary
enrollment,1960

0.062
(0.0139)

16

0.0192
(0.01G2)

0.034"
0090

.:

. - -

.00578*
,

0.1153"'

~~~~(0.0224)
.(0.0260)

0.25
016)

0.09
.07
(0.0.192) -. (0.0180)

0.1015, 013
0.2738.
K ~~~~~(0.223)
(0.2592) 7(0.3083)

AverageInvestmendGD,.1960-85'

.196-55
5

.0.0247"*
0.0082)

0.03
0000)

of population,1960-85
-

-0.0292'
(0.0133)

.Educational
attimemnt,
1960
Grot

"98

-0.0450
0281

-0-3135'
(.251
0.086"

0.0201
(0.0309)

-0.0002

0.0013)
.-0.4290

(O.64
1'.-

(0.0245)-

Avrge'qupmentinvestment
* GD?,.1960-85

0.3100"

-0.3050"

(0.0721)
2
Adjusted
R

~~~0.3480 0.3424

0.1921

0.36

~
0,1893

(0.0743)

0.2614

Safl1ilM
sgtfintt the0.01II~
Staisdeallyiignicattth
0.5 lve
* Ne
Coffiien istopnumbermn
eachceIL.Smiinkdanirrris
in parahee
SOUIrcZWorld
Bankstaffestiniares.

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 investment, 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 reduction iniresidual variance is not sufficient to justifyra change in our basic
specification'. Because school enrollment may not be a good indicator of
Gi

,*

C LE

1IRA

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 physicalor 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 backwardness. Per capita income is an increasing function of the share of investment. Observations are plotted relative to tle estimated regression
line, and the HPAEs are identified The estimated productivity performance 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*:
*
0.02 -

I*

O002-

-0.04

indon

..

~~

ia
e ,

*Japan

ngpo

__

s*!

.halwidi

--*
- - --

''*'

H*
HPAEs
* OUuerd.velopIugeconomies

.0.06.
-15

-10

-5

0
5
Orthaganalcomponentof I/GDP,1960-85

i0

i5

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

20

GROWTH,

EQUITY,

AND

EGo

estimates. All seven of the economiesare positiveoutliers.A dummy

variabletaking on the valueof I for HPAEs is positiveand significant.
TFP

HowWeEstimate
TFPChange
Suppose that every economy has access to an international crosseconomyproduction 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 dlasticities of output with respect to physical and human capital.The shamweightedgrowth 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 workerafterdeducting 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, crosseconomyproduction 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 cconomy63

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 estimated TFPgrowth rates. One striking finding of our analysis is rhe low
elasticity of output with respect to capital from the wholc crosseconomy 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 capital 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 results 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

to CapitalISKI,Labor(SL),
TableA1.2 Elasticity
of OutputwithRespect
Economies
andHumanCapital(SH):FullSampleandHigh-Income

Fullsample

2,093

High-incomeeconomics
Ruling

460:

rtaO

SL

(t-)m

(f-stat

0.178

10.895

0.669

6.411

0.154

1.49

0.399

10.237

0.332

1.679

0269

1.476

&ximafi r t HPAEs
toasaifacuerprodsadvi4ygtou*h

; - - - -Eonomy/region.

TFPrwsA
; ; -0t;t(filsa5Jmpk,
.
parpameer

Japan

3.6470
1.2543
3.4776

Kore,.Rep.of

3.1021

M2;ysia
sing2pore1
Taiwa,China -

1.0755
1.1911
3.7604

-ThaiLnd

2.4960-

Hong Kong
Indoncsia

~+ - ~0
LatinAmerica
~
--SSub-Sahv fric'

-0.9978

TFPwtb1
(ldg-im
v:j
ates)

.te

2.4113
-0.7953
1.4274
0.2355

-1.3369
-3.0112
1.2829
0.5466

0 0.1274;

a. Alsoincludes
TunisiaandSouthAfNic:Sor- World Binkda:L---

64

. SH

SK

Obsations

-0.9819
--3.0140

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 presented 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 constant price capital stock, Elias (1990) calculates economywide TFP
growth rates for a subser of 73 economies using standard growth accounting 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;

*Singapore

*S.ong Kong
Thailand

i @

*Japan

*MalaWsia-----'

, --------

Indonesia

is
MS
'-__

0-

2 26--~

- 4i t' ;

9---

--

Se S
Fl
0D
sa.k

Capita-growth
-pL_

et

----

.06@(2*

510

rwh

_
- _ _
N4igh-Income
economies

*NPA~~~~~~~~~~~~~
tes eelpn economies

1520
capita growth(percent)

--

High-Income
economiesGDPgrOwth= 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.

0*

~~~~~~~~~~~~~~Singapore

2-

Ohina*

-Taiwan.

0--

a Hwighicomneeconomies.
A
* Otherdeveloping
econiomies

-2 -----

-3

-2

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 ranking 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 crosseconomy, one-sector setidng.One way of separating the concepts of tech-

66

GROWT1,

EQUITY.

AND

ECCN

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

----

3 -- __

-3-*
-2 ---O
Ch_

,:in
I~~~~~~~~~~~~~~~~~~~~~~~~~~~awn

O-

~~ ~

~~

econoiesn
~~~~~~~~~~~
Hlg-,h4ncldoesau
of Korea

t.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Rp

Whetdevlfopingeconomies

/~~~~~~~~~~~~~~~~~~~~~

-3

-2

-1.

1
0
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 bestprIacticetechnology;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-

F(.) satisfies
puts in natural units at time t. We assume that the frmncrionx
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 techniques 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 inputs 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 element 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
in output nor explained by input changes. Thus for any observation, i:

Figure
A1.5 TheGrowth
of InpuisandOutputs

(1.8)

Output-labor

A-

:: fflz
qo
f- -0

68

_
S ;-

TFPJ(t)=

F+ u -(t).

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

change in efficiency observed at the individual level.
These concepts are represented in figure Al.' 22 We assume confh
stant returns to scale in capital and labor. The international bestpractice production fitnction, fJ,relates output per worker to capital
(induding human capital) input per worker. Economies that are tech- k
capitalout-u nically inefficient operate along functional relationships such as f in

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.Industrial economies,which employ international best practice, are limited 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 provide 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 alternative is to measure TFP change direcdy by growth accounting. Since
measured TFPchange consists of both technological progress and technical efficiency change, we impose the strong assumption that technoTableA1.3 TechnicalEfficienq
logical change, the movement of best praaice, is constant and does not
ChangeEstimatesforthe HPAEs
2
5
I enbikid
vary across economies. Under this assumption all of the variance in
effdciency
rates of TFPchange derives from variance in the rate of technical effiw
-60-89
ciency change.
'
lkEcoJomylrqov/g t-.1960L9
HongKong
:
9714
We have argued, however, that allocative mistakes were partly reIndoneia
-1.2352'
sponsible for the low output elasticities in the cross-economy producJapan -.
9876.
ton function estimates. Hence the high estimated TrrPgrowth rates of
-0.2044
Ko Rep. of
those economieswith rnid factor accumulation may represent allocaCM
a~Mlaysia
1L7767-3;4510
Eive efficiencvmore than either technical progress and technical effi0.8431K Tanvan, China
ciency change. To address this we use the parameter estimates of the
high-income economy production function on the grounds that these
Thaiaind
0-OI67-i-t
LatinAmeca- -- i
:17
economies are the most allocatively efficient. We then sub tract from the
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

1.2:'WhatDoTests
Appendix
SkillsShow?
of Cognitive
Tests provide a convenient quantitative measure of diflFerencesin performnance.By providing a consistent instrument, it is possible to compare the performance of students in different learning environments.
The key is objective versus subjective assessmenr. It is important, however, to recognize the weaknesses of standardized tests. First, rests evaluate 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 cognitive 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 variety 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

ea;-.RpNewf wkC

N-w Bf`Ilik.

|^s m&mac!;,t,

,,,,m/nvnc

-Biish
,Irelirdant,

100--.
100 .5-

lh

;100

(Frenc)

J Onrado

.99

tquebe(Fwnch)

69
55
-

6-

8t
. 95;
95
92

two-step'
ab/s

:-78

Ar
- .-nria'

Udrtns
-o.c-ptP

p0kn

24-

- :-

78 -

e14
.

58

5
1

:4

12
.16-

3222

2
-.

18

-65

55

:::-87

70.

~100.

;
!

,'problems

100 ;a 98-.

JE~igIish~

.tg

simple

::and

-.
1

-*U

GROWTH.

EQUITY,

AND'E-

mathematicsadministeredby the EducationalTestingServicein a number 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 Thailand 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

____

.-_______

Sweden

______

-.

Thailand

--

NowZealand
England and Wales

Canada(Ontado)

_;

UnitedStates
Scotland
Hong

Kong

Finland

Israel

_.

Canada(Br Columbia)

Belgium(French)

Hungary

____

Netherlands

_ __

____

Belghirn

_*- _

__

__

France
Japan

20

40

60
Mom score

80

100

SouweHauvsheh and Sabor (1991).

7'

- M 1RACLE

These results reinforce the conclusion that in some East Asian

economies cognitiveachievementlevelsexceedthose in other developing 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 children also lagwith respectto levelsof cognitiveachievement.

FigureAL7 #-Lii
Coefficient
andCW perCapitaGrowth
Rate,1965-70
G1nlcoefficint
0.7

;-

-16--- .

'

~~~~

i Peru

* Gabon

L-eio

'Brazil

*Colonmbia
j

0.5

Venezuei au
,-

Chile

Malaysia

Sudanm

Sudanu

*lnldia

0.4A--4---

Singapore

U~~~~~~~
-~~~s~Phiil;Dlr
~
U Philoianes

-~-

EZamba

E Argentina

*SrlLankau

EHong Kang

N~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
ThailandI
*

Olcoe-

I * Pakistan '

TaIan, CaJapanU
Chinau
*Rek'rwa

X3---X--i

0.2-1

* O .

-0.01

SaurceWorld

72

-----

----

Bank darm.

0.01

'-----

0.02

0.03

0.04
0.05
GDP per capita growth rate

'

'

0.06

0.07

0.08

:~~~~~~~~~*

0.09

0.1

GROWTH,

EQUITY.

_|O

AND

Fire Al.8 GiniCoefficient

andGOPperCapitaGmwthRalte,
197180
rlnl coefficlent
0.6
----- -

-enya

EZambla

NCOI

IPeru

iBrazil

Nepal CWIie

MeVdcof

enezuela
*Malaysia

* Singapore
E~~~~~~~~~~~Horg
Kong

EPhillippines
I

it-OA
-

0.4j

, *ArgenU;tinda
,_,,_hdiaE

.:

Indong
Idnea
* Rep.of Korea
Taiwan,China.

------

1111halland

* Banfjadesh
*~ L.~
02

--1 --

--

|1apa--

0.2-h

.
0

-- ,
.0.04

-0.03

-0.02

-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

EChile

XMexiao

Venezuela

!
0.4

Argentina*

. -

l-~----------EPeru-O

alyi

Peru-

- ;>

* Philippines

_-ngapore

Ehilane
EThailand

M Hong Kong

Rep.of KoreaU
0.3

--In-,donesiaE
r - _ --**Talwan.-China
-

I~~

.I

~ ~~~~~~

~~~

.-

'-

~~~~~~~~~~
.

U.1~~~~~~~

-0.02

-. 01

Soun:Wo rid Bank data

74

0.01

0.02
0.03
0.04
CDP per capitagrowth rate

0.05

0.06

0.07

0.08

AND

EQUITY.

GROWTH.

EC

for GenderGapin PrknaryEnrollmentRates,1965and 1987

Regression
Fgure Al.1G Cross-Eonomy
* Pakistan
120.61

1987

I
I

987

1965

di

E Indonesia
[-5,

~~~~~~~~~~-~~~~~~I

>-

*Pahistan

~~~~~[19.21

.-

maliand 1nd
%ralaysna

[-10.t6]

N
[5.1]
jRlof Korea
t as21 Singaporeu;

'A-ndnei
[8

\
I

^ Hong kng * \

-~~~~

,I

0
o
500
2,0070 2,750
P^ercapita Incomer(1958 U5S doflars)

2,50

14-31:3

[-2.2]

1,000

1,S0

3,500

Note Figue in bracketsshow resiEdua

Sowre: Behrnmanand Schneder (1992).

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

.
6,

a
Figre 0

di

,,

1 ~~~~~~

[7-7]-

* -Rep.of KoreaK,

~-

~~~~1965
I,

.a,

*__Indonesia
E37o.'j].
At1

250

Soo

750

CosEnoIy

.L[.1, -

*-61

RersinfD

edrGpin 2,M5
:1,750 2,000

noletRts
SOD

eodr

1,000

1,500

30500

95ad

J-

1987

*P'akistani ,,

5ware.
S
Bdhmnanand Schncidkr(1992).
IndonesaKo3g.
913

~~ ~

~ ~ 29 P.r cap6a

residual
Not- F4urcs in brucker show
E
~~~~'Thailand

1h

laysa)
Mcame
US. do
PoakKontan
[-.2

18.2]I

-J~~~~~~~~~~~~~

.~~~~~~~~~~~~~~~~~~~16

.98

* 7077
E

Sn

apore__

N
Av~~~~~~~~~~~~26
1,750 2,000
0 250
750 500

Maayi

2,250

500~

~~~~~~~
1,003,0
~~~~~~
1,00

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

1[2.91

~ ~ ~~3.4
rigN~~~~~~~~7

SI

A~~~~~XRC L E

Notes
1. As we pointed out in the Ovcrview,a strong argument couldbc made for indudingChina amongthc "mirade 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 makesartempts to understandthe links bctwen public policyand
growchon a cross-economybasiseven moreproblematic.
We lookatsome aspectsof China's rapidgrowthin the rmair.der of the book but have not tEratcdit as systenatally 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, Luxembourg, 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 industrial economies, is defined as 'unconditional convergence' in the new literature on growth dteory (see box
1.2) and would be expectedto occur only under very restnricveassumprions,for example,equalsavingsratesand
technologicallevels. Below we exarminethe concept of
"conditional convergence,"catch-up controlling for differencesin variablescontributingto growth.

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 official 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.Appendix1.2 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
3 This issue,termed"persistence,"
is exploredby Easterly to address different questions-equipment invesrment,
and others(1993),who find evenmorestiking evidenceof
De Long and Summers(1991); trade orientation,Dollar
the uniquenessofpersistentgrowdtamongthe HPAEs.
(1992); endogenousinvestrment,Barro (1990). We have
- not includedthesevariablesbecausetheyare not centralto
4. Hong Kong P-:d Singapore,which almost entirely our theme.This approach,whilegenerallyassociatedwith
restsof endogenousgrowth theory is alsoconsistentwith
lackagriculturalseaors, are exdudedfrom this discussion.
neoclassicalassumptions (Mankiw, Romer, and Weil
5. Staristicalanalysisindicatesno significant (in the
1992).Arecenrconference,HowDo National PoliciesAfstatisticalsense)differenceof theseeight economiesfirom fectLong-RunGrowdt, summarizesthe state of the art in
other cconomiesin their share of total investment,conthis field(Eastely and Schmidc-Hebbel,forthcoming).
troling for relativeincome levels.

76

GROWTH,

13. We have also estimated the basicrelationshipfor

two subperiods, 1960-70 and 1970-85. The fit of the regrsion 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
amasreo educationstock,
. b a ron
san Lee.19)
and Lee 1993), mcasurcof educationStOCK,based on
population censuses,failsto perforn better than the pmaty enrollmentrare.

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)wsoefthfittonetedvergence 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-

16. FiguresA1.8 and AI.9 (in the appendixesto this

oping countries.
of
chapcr)r
<:simatcs
chaptercompare
compareour
oursrmaLes
e
of TFr
r growh vvidir
ItwihTO. 22. Mankiw, Romer, and Weil (1992) also employ a
Solo
other independentlyderivedestimats for a largesample
wtype neoclassicalmode. Proponents of endoge
of countries (Elias 1990; Fischer 1993). The pattem Of
pr.oductivitygrowth rates in figur 1.10 is remarkablyronous growth theorywoull not acceptthe depictionof the
,
.
~~~~~~~production
functionvAr. Aminishingreturnsto cpita.
- i ,^,
^
busr to the specificationof the growth accountingequa23. A comprehensivereviewof the largeliteratureon
tion and to the capital stock series used. IMF (1993),
contrasting Asia with other devcloping economies,
te measurementof best-practiceor frontier producton
reachessimilar condusions, both with respect to the esci- fiunctionsand theirrelationshipto traditioralestimatesof
the production funcdion is contained in Aigner and
mated magnitudesof TFPchange and the relativecontriSchmidt (1980).
bution of TFP change to output growth. Thomas and
Wang (1993) and S. Edwards (1992) also reach broadly
24. This is the approach adopted by Nishinmizuand
similar results concerning the pattern of productivity
Page (1982) and recently applied to high-income
changein the HPAEscomparedwith other economnies.
economiesby Fecherand Perelman (1992).
17. Young (1992) finds similarlydisappointingresults
25. This assumption,whilestrong,may not be fir fiom
with respectto Singapore's Tr perfiormancc.Some caution is neededin interpretingtheseresults,however.Much
the truth. Industrialsectorestimatesof TFP change in inof Singpore's investmrentbetween 1960and 1990 was in
dustrialeconomiicsgenerallyyieldacompactdistributionof
housingand socialinfrastructure,outputs of whichare noras with a meanvaluenear 1.5percenra year,both within
ntoiously
difficultto meas;:e. It is possiblethat we have: and acrosseconomnies.
This may thereforebe a good first
thereforeundervalued the rate of growth of output and
approximationofthe rateof technologicalchange.
hence the rate of rvp change. Similar derailedcriticsms
couldbe made for other economies,both HPAEand non26. The rests
wereadministeredto a random sampleof
HPAE,aLwour TFP resultsare rherefore
best regarded(asin2,000 students in each county. The test questionswere
deedare the :ross-economyregressionresults)asindicatve
drawn from the National Assessment of Educational
of broad internatonal trends. (On problens of estimates Progress,a test designedfor U.S. students. See Hanushek
of cross-economyregressions,see Sriniasan 1992).
and Sabot (1991).
.

77

R.TE

Public Policy and Growth

HAT ACCOUNTS

FOR THE REMARKABLE

record of rapid income growth and declining 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 potentially 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 chapter 1-accumulation of resources, efficient allocation of those resources, and productivity growth, that is, increasingly grearer output
for given resources.
How much of East Asia's success is due to geography, common cultural characteristics, and historical accidenit Certainly some-but definitely 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 benefited from the kind of informal economic linkages geographic proximity encourages, induding trade and investment flows. For example,
79

S_

MIRWACLE

throughout Southeast Asia, ethnic Chinese drawing on a common cultural 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, Myanmar, 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 traditional trade networks, reinforced in the nineteenth and twentieth centuries by surges of emigration, have fostered elements of a common
trading culturc, including two lingua francas, Malay and Hokein Chinese, thar remain important in the region today.
In our own cenrury, key Asian ports were integrated into the emerging worid economic system as the result of European military and trade
expansion. Cheap ocean transport and shared historical experiencesfirther knit together a far-flung, culturally disparate region. U.S. assistance
in rebuilding Japan after World War 1i, followed by massive U.S. economic 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 involved in the combat to focus on economic development Perhaps as important, the U.S. military's exensive purchases in Asia provided a ready
market for emerging export industries. Japanese industry receiveda substantial 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 trading companies and directing the structure of industry; Malaysia focused 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, particularly in the past decade, as Northeast Asian manufacturers of laborintensive 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 Japanese yen following the Plaza Accord)-flowing first from Japan and later
from Hong Kong, Korea, Singapore, and Taiwan, China-has contributed significantly to the dynamism of Indonesia, Malaysia, and
Thailand. From the policy viewpoint, these linkages have been encoutaged 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 imposed 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 adequate 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 Philippines 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

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

-8i

SvA

MIRACLE
LSYXN

TheNeodassical
View
In the neoclassicalinterprectaionof EastAsia'ssuccess,the markettakes
center stagein economic lifeand governmentsplay a minor role. For eaample,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 imperfections and shortcomings-to exercisea decisiverole in determining resource 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 provided 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, governments in all of the HPABs provided a relatively scablicmacrveconornic environrment, 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 importsubstituting 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 foreign trade regime and domestic factor markets. In the neodassical view,
East Asia's export success is due to similar rewards for selling in the domestic and foreign markets. iBecause variation in incentivesacross sectors, 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 aeporters have iced an n.ternational price regime in making their deci-

-St

PUBLtC

P OTC

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

real rates of interesthave prevailed,and thcre has been an absenceof duality 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 intervened 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 industries in the first fifteen yearsafter World War II, offeringprotectivemrifis 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 intervennion 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 governments can foster growth by "goveriing miarker and "getting prices

83

; *

-1-

- RA:

-advantage,

~~~~~L
E

wrong" and by systematically distorting incentives in order to accelerate

catching up-that is, to fadlitate the establishment and growth of industrial sectors chatwould not have thrived under the workings of comparative advantage. Amsden(1989, p. 14), for ex-ample,asserts that all
"economic expansion depends on state intervention to create pricc distortions that direct economic activity toward greater investment State
intervention is necessay even in the most plausible cases of comparative
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 comprehensive 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 deitarticasMarrkea-tendly"suategy
limited government aciivism. In the
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 internationally uncompettive sector, and resulted in widespread rentseeking. 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;
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, opedu-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
rriLokvole. Injapan an.dKorea,foruinstace,gov~Y
-;appro&;
0'T eReporrcnotedtrhtareeconomicufronim.
ement :agen and-indut asaton
co' bot investng mi peopplcan be high,bur manygovern-. rate to gahier and. disseminate i:normationon:.-J,.
'no..ton]
ar
rd-'anddeop quaity control'Ibrp& t.. '
-ntarget thcir
snbapkpmropacrod
ecoanro,-in':',-Ale'rocr
micc foundtionoi. on ofthe
5 did'
Brazil2nd Pakistan,'f
alone
m,sost i
'
oosihagormenrs
OOpublic
ca
- derto; .unprovt.
hie sot
.rovide,- as th neotpule.ot
o. Wt%an
~fl indicators;i Ce.and
jamnaica,howevr
to c
CarS inprov eVe. i
mnt ocr
th
o
times,ofKislogw.wth..Governmentsmust ais
overspens,
Mr
terdui
senoonowinaj
st
'monetayryl
n fnniscordfiute,'
opnin
.-prov- th 4ini
fhes
To,
hled
by inflation,choni
i
'''o% pvisios,
o
for -recurrent: pen
are
' inadeof
y alssof
,thel
o 'p
? i,
quam
te rEsuldjiawasrefilu undemrutli2ation>
Thdeconomies
with a:bs-orof macn&o'c
u"r>
Motsessfuuonlei
nudhJai;Si-"
iaiitprruesleionthaiidd
ArgeCntn.'olivia
<
=ic salihdter~o
o'
m on
s~~~~~~~~~~~~~~~n
~~~iposr-,adoc~hv
-sf-tabedili
.y>ir iel
d
hnOh&nu~rj
IN-THE;AST;-TWEN

YYEASA

CONSENSUS.HA -;trenologi

'

9'

tt

'w

ItcP%%Y :avarae.a through rhen riors,- of~'

ke,'..' doeia Koe,Maas;
nTalidhiemn
KL.markeslcqi.k%versely; retuctionson competition'
gecLtod
"keep.thefir rmacroononic:. polid&son.'I
,
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.,
(fo e5pe;Indsiadcensn
o nappopriate.
course adteir-.bioaUednoic
jerfbr~ceha
ihes

?P

.biddk'.b
>
'betarizlyo

opeoi)

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
Asia's

...

,,.

.-.

aspects

of Eastcr

succe:m.These economies are macroeconomnicallystable,

havevEry

Asia-ssuc=ss These
-,

captures important

X
' - ' ' 0'

.oniies are macroe-onomiay stabl, he

:'~
~
~
~ ~ ~~~8

v85,

*,~~~~~~~~~~~vr

high leves of human capital, are thorughly integrated into the world
tmc
ocure
economy, and have very high levels
of
competition among firms. Moreoe,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 expectations. Evenso, other interventionscombined with export targetsapparently 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 northeastem economies compares favorablywith the successesof Hong Kong,
Malaysia, and more recently Indonesia and Thailand, where policy
choiceshave been less interventionist.

-NMultip
Pathsto Gmwth
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 central 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
with which govenmentstried -policyinstruments
pragmatic-flexibility
in pursuit of economic objectives.Instruments that worked were retained. Instruments that failed or impeded other policy objectiveswere
abandoned. Thus, the Korean and Malaysiangovernmentsreduced intensive 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 privatizel 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, finallyconcluding by coordinatingthe increasein wageswith rising productivity.
And Malaysiashiftedfrom policiesthat resultcdin business-government
conflict to highly structured mechanisms to encourage businessgovernment cooperation. Chapter 3 discussesthe pragmatic flexibility
of HPAEgovernments in pursuing two goals, macroeconomicstabiliry
and growthof manufacturedexports.
Obviously, our three functions of growth-orientedeconomic managemnentwere not objectives in themselvesfor policymakers in the
rapidly growingEastAsian economies.Instead,they are simplya framework 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 import 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. Indeed, as we illustratein the next chapter, the flexibilityand pragmatism
ofpolicyrnakingin most of the HPAEs
reflectedgovernments'willingness
to try new approachesto achieveparticularpurposes,discardunsuccessfuilones, and keep those perceivedas successful.

Th Functional
Growth
Frameworlk

_ IGURE 2.1 PORTRAYS THE FUNCTIONAL APPROACH TO UNDERstanding 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
Figure21 A Functional

:owh

Comnpetitive
discipline

Policychoices

Hihsa.g

macroeconomy

* Highhumancapital
MdHgll
andsecure
: Effective
Metetransti
financialsystems
Export
isin
Um.t-igpricedistortions
capital competiin
* Hlghhuman
Openness to foreign
It

SOpennessto foreign
*technolrtY

InsUtUdons
W
. _

* H gh-ualfty

laborrnart
_

>

Monitoring

on

Ilmestmrent

<

>Equal

Ineeme

Contest4edtu

clstributbon

*Export credit

. Reduced poverty
.

* Lfwestmnent

civil service

of

use
capital in.
~~~~~~human

credit-Reduceyfective

coordination
* Information
exchange

Productty change
* PmduftJvpbased
catchingup
R

Insulation
o Technocratic
0-

industalizadon

ron

* High reurts

Export push
vz Financial repression
* Seiectivem prormotion

rI\

* Directed cedit
I

agricultural
RFoapid

Alboatin

I
I

growthof
Rapid

ho
ai
Rad
transformation

I
~~~~~~~~~~~~~capital

SelectiveIntprventionsI

Rapiddemographic

: Directedcredit
:

capital

sustained growth

human

*Hg,aigsexports

Contit

~~~pollicis

Increasing

Domesc

AFicultuial depelopment

Rapidand

Accnnmulatn

Fundanmentabe
. Stable

Outcomes

Growthfunctions

-~~~~~~~~~~~~~~~~~~~

+_______________________.______________

tec_

Improving sorlal

inditrs

&
____

managemenrt and tie outcomes of growth and equity discussedin chapter 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 contributed 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 rechnology was a major vehide for productivity-basedcatching up, and agricultural 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 distributions 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.

parSimilarly,they helped in the creationand protecton of instimtuions,

ticularlythe dcvi service,by reducingincentivesfor corruption.
The sixpolicieslisted as fumdamentalsare so defined in the sensethat
they affectthe attainment of growti functionsprimarilythrough marketbased 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 inrervenrions 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 intermediare 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 public 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 mediumsize 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 agriculT
ture, small and medinm-size enterprises, and, in Japan and Korea, largescale imvestments.

By intervening
-

in credit makets,

- - -

governments

- -

- .

~~~~~~~~~~
-

selectively promoted industries, ownership groaps, and, in some instances, 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 invested directly in technologically sophisticated industries. Singapore and
Thaiand have attempted to attract foreign investment in capita- and
knowledge-intensive sectors. Malaysia initially undiertook public investmenus 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 undertaken 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 intermediate goods producers what outputs are required by final goods producer.
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$DtN`

':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-rativerelations are ensured because of benefits from maintaining a good reputation. 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 exchange in developing economics are weak-er than in industrial econominis,yet the needs for these forms of coordination are undoubtedly
gre'ater. When economic change is.siow, making predictions about future 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 markers can provide; hence there may be a greater role for govermecnt to
create institutions and facilitate coordiination.
MissingInfonnatio and CreditMarkets. Among the markets most affectedi by information problemnsare capita markets. Even in industrial
economies, equity markers generally do not finance much new investment, largely because of information asymmetries (Stiglitz 1993a). In
economies where'equity markers are weak- or absent, credit mechanisms 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 characterizes 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 confidence. 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 directedcredit programs,.and through explicit and implicit guarantees or other
forms of interventions, reduced the risks borne by investors.
A'

ScaleEconmnies,
Externalities,
andCiooperation.
Coordination may yield
substantial benefits when large indivisibiliries in investment lead to
economniesof scale. One such case arises when investments are interdependenc. 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 happens. Market failures due to incomplete markets, such as the absence
of capital and risk markets, exacerbate thi~ssituation. With such largescale 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 mnechanisms by which these risks can be divested.
Economists have traditionaly found exteralities a persuative argunent 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 situations 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 hghtechnology servicesand nontraded intermediare goods.
In the earlier developent liteature, the coordination problem arising from interdcpendenc investments, economies Ofscalesand exteralities 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 investments were in monopolistic public sector companies, severe problenmsof economIiicperformance arose related to- corporate govemance
and market powec:
92.

PUBLI

Some

HPAEs

L14

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 relationships that facilitatedthis process.In doing so, however,these gorernments did not attempt to supplant cxisting information networks.
Rather, thcy built on the superior information that firms had as dccisionmakers Their approach to these problems, emphasizinginformation and cooperation through formal and informal interactions
1 is
discussedin chapter 4.5
Creating Contests

0;

Cooperation raises severalproblems, however.First, cooperativebehavior may become collusion,if firms act together to raise prices. Sccornd,cooperation may inhibit competition,leading to managerialslack
or a more generallossof efficiency.Third, business-governmentcooperation 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 interdependent.investments-with competition by firms to meet well-defined
-conomicperformancecriteriaThey developedinstitutionalstructures
in which firms competed for valued economic prizes, such as accessto
- credit, in some dimensions while actively cooperating in others; in
short, they created contests (see box 2.2). Market-basedcompeiition
*and contest-basedcompetition both included prohibition of monopo
lies, although the number of competing firms was sometmes small.'
EventhoughJapanand Koreahavetended to have high levelsof consector (seetable 2.1), domestic comf centration in their manufacturing
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
with many smaller rivals,a principle that is well-recognizedin athletic
- -competitions

Table2.1 Concentration
RtoinManufacturing
Ratiosm

in-

7%em0cewhflia-.

i8

K&R9:oi962A

56Z'.

4;

nI9

KBraz41980

AtiZN

'--

s4
.

(NalebuffadSii193.v

Contest-basedcompeution included dear well-enforced rules, and

prizesforwinners.Table2.2 brieflydescribessome of the contestsorganized by HPmE
governments.6 The simplestand most widey used were

&

~~~~~~

93

C
-RCLE
R RA

j:~~~~~~~~~~~~~~~~~~~~~~~~~~~'t.'

..

'X47

,.'.

'-

Y'Oi'3m2Running
a Contest,
Rewais;,Ruies,n
. JT-H-lE,ECONOTC
CONASTS.IT

GOVERNMENTS.

:i.japan,.Korea, and Taiwan,.Chi'Ta,-serup-amoi g

b&eherts
co*flins
wri;orfcohpetidod andian
J diemontheo'combime
sQi ,
::Cj_ -!j;--

-.>..;--:

'''.;

.Y

....

?Rdees

Contestsin NortheastAsian.economicshave succeeded pardy becauserthe prerquisits were right.

1.:Preferentialaccess~to -d
.credit .and-Ibrdgn excchange:
.
*

.. !.;,.
..............
e5..,|
1_...-

7
'caconomicpertormance,primarilawull arrl&hi'
base,however,these :-centere
coh:
montns
ies-. land somchmes
rcontestis,anddiir-role-mthe
economy resmble chl7-"
understoodd:mpcrauve
to ,expor; Referees,he gov-ren,sprc gaes
officials-wh&hbave
design'ed~andsprie
ti.- .O:ganiihgconrc
s ists a more pccaned way
to .
,ihetonresrs
t5m
-have:been
gFnerally
comperciuandfir.,
run!;a e:,oomv Esez-fi*i;
than,el:'i.i
r
r:i W.
, -jt a
.'
."'
. been
- economy
.L
rC nng.on
ssz-talre, justas
nA
same time,;parcipation has nortbeen ,;
-b'f , p
s
r
r:,Complcate,
anatoq.
c at prce cd
rrtosit
out'a, contest
g.
!
Thul
IC
ffi.l1iezi
dobeeni
phdinr
igo.; In some inpunished
nave not
etunga-hcxiuo
pAovA
KaA~ieana-p ayingfid
tnAy-fied,
L
both'vk. . -.-srancessuchas
Hondasshifr
toHau-;
l easc., i-fa
ie
-fi;
s;-l--vm1
*i*o
-*e-toau
ut7laeyp"'
and-u&ec"oplay
game,:e-. 6ioniobile'
maniffctrang
.at a time whenzfJapines
.p
ugb
pgarta
a well-mrri
:^Bu--a-wdll-run.contesr,
.: .
-

.''ernmenc

,,

s'

luce

; Xi can:erite6er nibec&r ul&by prpVidhigi

i~

.-..offiaiaIs
consideredthe
eautnomotiv

fieldsatruraced th i

ffhtolerance
haspald offandsomely
ofz5nconfdrmss
P.;or iLn
0
- l,
m
.
.Cnet.
won't w
cvcryw
er,.4
, c:
-as..
mnzatr
onttr.
steprpoi
m o:dvec6mga.
lioihjmstances
6I
therearc.threcprerequisitesfor.. derainceot:uuois
mutts`tic
be reoo
econiorreine'tat:
tck.a.e
la'
r,,,veJ-.;l-st .ep.of~uak-clear orov
ec
0~~~~S
. success:~
; rewards,'rules,p
f -g _6 andr&rees.Rewairds
-r:- o-u-h/t
i-.; .eliit .broa'
.partiipationan
'r.ues.and,retre m
. M
e in
instirutions,anwith-strong
government
ebe
dearcut so -; economies
:Rulesr
.oCnbeutidon
eA&rt&c
k .iW
,vih:fehvllb
e
, >;6 otestu
'e-ing contst -.
becd.es hardea, fs growand ther,:
X

aidnducingpartipantsto
forcwii.c
peduiioh~nL
; cooper-!..
qX~~cop
ta es
.pnt
--icac:llE
CYb

'

-J

-wardedL-ahcL.
whih
puiserh'FnulyL_ a
+;:z
iy&.wardyAandL.who
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

.-

a morefinuonal

'conrescard

-: ;pevs
rel- ghavdicoverd.runmng
niz thed_`hi
:..cd
&h :pa1ntnzr
Uiildren-Si -9 ,lgdi.byrfessil'-- -peten'to!-W/nLZhaeFidZ...-.&voin.dow d,ild.
;. -w

pakygaineTor".
,C'r

rexport contests fior access to credit and sometimes foreign exchange.

scr

rum-t
-:
. :

_.

ow,

aWhile
eaprt credt schemesvariedwidely,all madecreditavailableto
~~~~~irms
with confirmed export orders. For many small and medium-scale

p-

ts

entrprises, export creditswere their only accesstO the fbral financialS

:sector
- .- (Levy1993).Monitoringperfomiance
wassimple;firmswerere-i;quired to show evidence of export orders to receive credi

In Koreaand Taiwan,China, exportcontestsweremore elaborate.

char ex-s
Nor onlydid exportprojects er easieraccessto fids, but firms
had an easiertimegettingfiure loans.Becausesuch
portedsuccessfilly
creditwasfungibleand creditgenerallywastight, firmssometimesdil
nonexportinvestments. its
vertedexportfinancingto higher-yielding
'.4-

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~P
U B Lil-C.
'2

. - - -;

exports became a performancecriterion for projects that were not directly 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 licenses.Table 2.2 shows two of these contests,based on Japaneseexperience. 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

Noon ep..r1
7.1..,;:-:-.
-Ctnin.

t- niaja lmeroroulbe.C).
1. Nomadidnbvn
lndrnnpons Cnrsqjdknm spnenndmqr

UCaMmnd
MaIqt3 bonge po.lpmeoe u
an.tita,-vo'.4

~~~,;1*
io

an.

zb

.ed-

* 6 F-v, ;
ee.r

,e

,-ef ,

,inInn

pre
edm

,n

,pntoe

-;

nt,htMepmpm-

. *,

Oneith

'~~~~~~~

,s'.

obetves.deFl

t.,. ,- ..

.
.

'J

pei, l

I, ,

cmin

iR

...::............ :

',

gonuesu

eem ankmpnda-I

'.

7'-7''
"'.'

DOz

.
'4

...
'

-rr

'i

~~~~

.....

:C

'.-'''

-.. -. : ... ' ... .~ ,

4Z~-

CL{~,

:,7:

'

.j--g.

esnndafdSepne&u
.

AYf

'

.7

.g

..

'

'M

yoteJvbgrnnueppn.

,e-,:
. -.

'

UPbnie
n

'JI,

dkmi
.;-t

~~~~~~~~~~TegIdaPund.,,'-:'..'
-.

ea ditM in.

a(Ja

onSk-

Tuyo;e.ayMemrnao

-0pcmr'ti156
.''

.c
.. .r

. .

*tep

.w

dexpnplocmfpeypm

-&hudue
An

hymn

Muhinewrder Pamnodom
kfim
tow (194141

. '

a,nno

...
..-

: cadSwt-e
pnepnder-idsbanwud.......................r-hsnnnccenahe
is etttitre'-'-\
'>biw
.....

danh"okma

:--- - : - .

. o

'mut

hav
;seadeomnu

.~~~~~~~~~~~~~i~f

: C,

-:

Cana bob
ofkrlnnL,

7'_,>:
.

i1

es
*a-;k.
u

n-iai

.-

b.;-.. :

kl

~~~~~(siwtubmncnwCa)ocscaonI
t

''

.C

.,boglCaeg

<

n,

to.m
n

_.

at

a0

eX.einIses

lftvo.epena&f
<' iI1 mOtmuanmatipn- deump
- r.h
*11.CUU~~Lordto

RM11112NEW411W

unto
LC. &smk
C. -

to;Fctdmi

... .

,$

7hr.r

. :- -

I'

.9

-"

,.-.-7.

'.:

Ct~.

, :,:

-;5

RA C LE

'Table2.2 (contfinued)

IN-

JtrP o u a u h d S i d h m :~
lnabSInInxhlnhm.um
udncmhhAnadit

~~~~~~~~~~~~

44-.

Tiw hnk
k ufj p uimppnwkteu
m hatd mh uu m u
-,two?"

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

Am* rudpmiadl
*

h~iwIUcDlvsnn
andvbmmrad. Miseh.a-.
lie. *.

~kiW*rnnd
~~~~Th
~~~~~~~~~~
I

A.)..

-...

. ..

(dr
I~

zdlIc.M

-~~~~~~~~~~~~~~ko
rd

~~

.1

Iwam

-'hZ-

(1d

~ ~ ~ ~ ~~~~d
iime

initu
cm mm, dwhmmsd&dtuW
Damrmguhcflsp-dykdLn
p"c
apwtnruaiiitli
uilidwaI
*lugemlmerpand
ammafc SlpI.nsio.la
dto mul.

~~bmm.Kq~
.1
;'L,

bmulmyrnL

mas, huh alps ais

rJrmmck
yKrjymn&

wte

d,nknm

o"

rW61

uhidgl*p*lluuca
ddrrtmu
aaa

aS .

ade
Nlazm afirms SsAflCs.

.,.

lca

PCf'Wi Wdurn
.. tmunymhiepiRaIMacMftWOI.
1
tapumumnnsuuchatmaecdlmsrniiza~~~rnaum..

.'Z

uilobmalbr
*1~~~~~~~~~~i

(C-

it.

-',

mapnhdma.*

'VJ

Mm*mm:armshV,.a
lafaryolavcn
ItS nheppamatn4talmUd.
tedon Amaik

Nandi ladmim'strativ.

'.

'"Mr

guidoMancemf

Cumitla
cm dehpmimmmihqClr3lesc-usuanmkn.y

hev-n

thcauirau
S

ulending

hniclidstis

banksotbar

failneydIm oco

90

pl

19

-s

admsinistrpatsivegudacpithei
of
ledig:bakthatwas
feqired
to complye
wirsthaguidaceiabou the spreatsdubetweedepositiand lni-rngrtes werenrepermitteorexpandt
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 protected
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 contest, one that combined licensing of capacity, access to credit, and protection against competing imports with both physical and economic
targets fbr performance. Its economic success is still not clearly established, 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 structure 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 acdusive 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 government support during business cycle downturns.
These incentives have been the source of disastrous economicperformance in other economuies.The check on poor perfornance in Korea
wasthe result of rwo performance-based rules. Fi-st, government estabIished a timetable for the attainment of international competitiveness in
eachindustrial sector Firms that failed to maintain the pace of cost recductionsin 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 beneficial as their rules. If perfbrnance criteria channel effort into unproductive 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 yardstick ainst which the success of other allocation decisions-for example, credit allocation, domestic content requirements, and industrial
licensinwere
judged. Accordingly, there was a high degree of competiion among finns, in spite of the &fctthat in certain domains and
'97

LVRcE

rttC

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 interventions 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 economic 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 compethiivethan domestic markets. Even if the firm's success is based on finding a niche in a foreign market, its contribution to the domestic
economy is still as large as if it had succeeded by developing new production processes or othenvise boosting efficiency9

'~

''-~~~~~~~*

W;'X

'~t

~~~~~M

5
~R

l'i

zmo.g

irk'~~~~~~~~~~~~k

C}:*ahezspmrheJsccPIAaEasJencdtixar&icellzngSowoFuha*cm
YE;uats~amcRA
'n.amamiarllot
~"
~
~
p_no

u -~~~~~~~~~~~~s

: ~~~~~~
R

FrsRER>^*z.zbV.F

tioo

__tij.;!str*

f
~~~~~~~Isoii

^gg;2ii5X;t>

a~~~goyrmmen<

Eiwa

, -. -

ortherslP.asyV

S p8

or

amodo

Zsa&UO%%OOi5UDnaV3flCPeSw
9
,*>w~~~~M.14o<tjf, -- rSi,'22...{
V

reihogn.

.r emuncess,U

e Sz >o c uc nt areusofdlereamorun
.'-~~~~~~~

.'~4r~

,g

:
..

gs,s

-0

BB,..N.X>.^*,>t

lR

a*t;

*.co

S~~~

3 ncollnpSaEpovtcotowDhsnRneurApmanceMDo6luonsqo
t>;I:6Sy

p8~~~~~~~~~~~~~~~~~~~~~~~M
9s-

oXBg,.sp

ttXs

+;X:j

a~~~~~~~~~~~~

PUB LIC

PO LI C.YAbrD4rO

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 spillovers related both to mnarketingknow-how and to producrion knowhow,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) typicallypreferreddomesticlending to foreignlending and for a simplereason (Stiglitz 1993b). Banls are likely to be less informed concerning
externalmarketsthan they are concerninginternalmarkets,and from the
banks perspective,there isgreaterrisk associatedwith lendingfor exportorientedprojectsthanwith lending for productsfor the domesticmarket.
HPAEgovernmentswere also lessheavy-handedthan some others that
haveattempted selectiveinterventions.Though they made mistakesof
judgment, they generallydid not forcedecisionson otherswho werewilling 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.
Running
the 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
and theeconomicadministrationmonitoredperformance.
leadership
The.Liits and CoustaintsofCootests
usedto promore
Whatdetermineswhethercontestscanbe effiecively
betterallocationand fitstergrowth?Thereappearto be two important
elementsthe relativebenefitsof coordinatedbehaviorand the institutional costs of implementingnonmarket,contest-baredcompeiitive
discipline.
Wherethe benefitsof coordinatedbehaviorarenot great-for example,in small,highlyopen economicswith goodentrepreneurialskills
and smallnontradedgoodssectors-the benefitsof coordinationof investment decisionsor sharingof informationare likelyto be small.
Internationalprices conveysufficientinfiormationfor marketsto be
efficient,and socialand privatereturnswillcoincide-This mayexplain
suchasHongKongand Singaporehavenot foundit worthwhyHPAEs
while to run contestsand establishthe types of highly structured
councils)foundin Japan
mechanisms(deliberative
information-sharing
and Korea,despite the high quality of their bureaucracies.

Evenin caseswherecoordinadionbenefitsare potentiallylarge-in

nontraded goods, externalities,or interdependent investmentscoordination
maynot be ableto enforceperformance-based
govermnents
rules.Taiwan,China,is an example.Despitebenefitsfiom coordinated
behaviorthat werepresumablypotentiallyaslarge asin Korea,and the
service,contestsofthe types employedin
presenceofa high-quality-civil
Japanand Koreawere not used.This may have been due to Taiwan,
China's,differentindustrialstructure.With a largenumberofsmalland
firms,in contrastto Japansand Korea'srelativelysmall
medium-scale
numberofmajorindustrialgroups,the institutionalcostsofattempting
to coordinateall but the simplestactivitieswereprohibitive.Instead,
Taiwan,China,addressedthe coordinatonproblemthroughpublicinvestmentin large,indivisibleindustries.
The absenceof effectiveconrests-exceptfor the exportcontest-r
the SoutheastAsianeconomies,Indonesia,Malaysia,and Thailand,can
alsobe interpretedasa caseof highinstitutionalcosts.Civilservicesin
theseeconomiesare norwellinsulatedfiompoliticalor economicinterferenceIt isnot surprising,then,that effirtsto coordinareeconomicact-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 docurnents 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.-

etdf6i-r.iu

4waiJ.wISYt~OFuOsDEt

;-

luc>S

;v,==auSmat
.

as<ol^

nuL.

e eonos(ni
or-moseb
thnoi tsucr-asi--.:o;

-- et"N'Vith.
^c.

"mtsincl-u)

~~~~~~~~~~~~~~~~~~~~~&~iggmii

'

w3

m usgaousmfluestrXn^SDriomeg-w

98,'

yjmr&|w[1Sim;-W*t;S>-rt-4r4---R,-F,3

m&m
R ;uncustue,Xandh8D~silldadrmcdl
t n>-

~~~~

'

;;~-

'e

C q

8,'.

4,^_s 4~

H:
I-

g'4

g:(~~~~~~~~~~~~~~~~~~~~~~~~~

t- sovv

inreqmnemure,nnn;otdaterneationalreemisvaiSSnieEt

i;>snonaerDqns
tua~an-dc

FSl.gr.-otXZ
^se-i>t>tzr~<<;Qv

-%w,7
-t4i

litel\97osaaI-S-lsineznotXiiaN
iM
w-t
g -;

zC

- )sD~~~~~~~

t->.<,<-3
; V ; <gX*
jb,

-vS
-iissS

--

%~~~~~~~~~~~~~~~~~~~-Wr

gi

nnNaarnlnen,-\nttniioegn
cmPulnscV
a
ura.N-Udj#et.uiibyi
71 P1
2-.
l$stp;.,

,v-.v;

3 <m .

-s

lRitFl-r|J-t

LtR

9stX-lr

g2J;i-*;

-9Oro!E,

5fS=tX

fZpno;MTp
ZEnSLEool>qcao^adeeomnhpnhat
.>+; e 5s- ,>:~"rn|y;E>.fi#.ana
iC
g>z.
sF^
,>:*
3;is iR 9+j=,-t%rinz+mmal DoMan;.was~~~~~
sQaxoamolaI4- ~~~~~~~
mos<Xsor;cell!loLeaElDr-u=Es{
8 -u^'.^&a>
t J-8 *iX45rW .+5=i
9s.+N
$t05Ar
~
=>. >
'>4XAS-rW3^DRbRX
2*^t
ttti
>,==
84tSitt
Drexnod7sw;:nEnUte,uedM
ms;n.
iotlkhEdasaloyEt
~ ~~~~~~~~~~~~Pq
.

-g>l~-leimR^ow
f~~~~~~~~~~~1~l

i1L

:-~

'gS

95,>uW,-^Q*

Z,oAaodnra:$rDvdcpcyTa-ytt
5W.^grLtg'@eS=< @X<'

: i

Q.
a =

...............................

ut9.
|R

,e

'

%M

Br
Ostbrecnsructo,lSycmDeesacm
n ea

Rr".wrl,>,,oE.isi,4e,a.2*,sw044>
i'-z-'tt.Z'^:
[^sl^ttt>'}Y'-'^F4Ea
ngn
'-ru
le.d>ndu l
>i
=.E$*-.w'xjl2>(g,.<t,'.X>an

blcne

ncg4cot.2-Xti.>C{vlsgt/4't.

W
t

~~~~~~~~~~~~~~~~~~~~~~io

l*jE;^;s

>

coinasoofiie
-

->4n4:,.,-

Yu*

or

d-

ots

MI~RACLE

.;e xe_^,r,>,U.le.x-U..
-t
i,_

riod had diminished as-the economy grew and became more complex,
and that the insttutional costs of administering contes had increased
as power shifted from government admininstmraors
to enterprises. Simnilar
concerns appear to be driving recent changes in Korean industrial polices (Km and Leipziger 1993).

The searc hfrpolicyexplanationsforEast AsiA succss hasnotbeen

completey successfulLEach of the broad views of the relationship between 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 hree
centml Sanctions:acinulation, allocation, and produc growth-neocl.
Markcr-orienitedpoicies, 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,insulation, and relativelackof corruptibilityof the public administrationsin
Japan and Korea.The historicaloriginsof theseeconomies'high-quality
economicadministrations and the mcchanisms that they used to maincain

quality and limit political influence and corruption are the subject

of chapter 4. The second is the pragmatism and flexibilityof governments 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

PUBLIC. PO'LICyA

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 estimatesof nominallevelsof protectionderivedfrom comparisonsofdomestc 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 period afer 1969, while the other two presentevidenceon
ERPS
forjapan and Korea Conversely,Wadedoesoffercarefrfly documentedqualittivc evidenceofthe extentofselective intervention.Despite the frequent use of Taiwan,
China,as a modd of limitedintervention,thercareno estimatesof whichwe are awareof effectrive
ratesof 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 aluminum refining.
4. A formal expositionof this concept is containedin
R%gand Peuri(1993).

5.Popular discussionsof the successofJapan and severl other economiesof East Asiahavestressedthe cooperanve 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 popular imge caggeractes.Like all governments, governments in EastAsialack the control t cnforcea common
set of goals.
6. The list is by no meansexhaustive,nIordoes it indude 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 nature 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 commodity 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 accumulation, ficent allocation, and strong p roductivity growth thiatformed the basisfor EastAsia's
suiccess. Governments achieved macroeconomic stabilityby adheringto orthodox policyprescriptions-in particular, by holding budget deficits to levels that could -be prudently
fin2nced.They achievedrapid export growth through an export pusha 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 economic circumstances.Policiesto achievemacroeconomicstability and
rapid export growth differedover time and acrosseconomlies,partly because governmentswere responding to a changingeconomic environment 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 macroeconomuc 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 develOpig coonomes,particularlyin Sub-Saharan Africaand Latin America.
1105

Export push isa more complex phenomenon than macoeconomic stbility 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 contests, 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 diversity of approach, as a group the HPAES
are unique among developing countries 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 inflation 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 eaternal shocks, govermments quiddy implernented orthodox solutions,
reducing the fiscal deficit and, when necessary,devaluaring the currencyIn 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 confionted 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 imbalances and generally lower variance in key indicators, including teal exchange 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 consistendy 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.
have perforned better than many
In both these dimensions, the HPAEs
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 restore 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 relainvelystable, and we contrast their performance with the macroeconomic instabiliy that plagues many other developing economies.
Wethen rum brieflyto four HPAEcase smudiesthat illustte thevariey
of rapid and effectiveresponsesto macroeconomicshocks.

Adhering
to Macroonomic
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 private 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 findarnentals budger deficits. inflation, external debt, and exchange rates.
The HPAES' budget deficitsare not
Budget
Deficits
Manageable.
Keepi.g
dramatically better as a group than other developing economies'. But
there are two distinctive things about the HPAEs. First, dtey almost always 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 consequences of pubLicscaor deficits depend on how they are financed. 2 Excessive 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
kept each type of financing within bounds, avoiding
aises The -PAFEs
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 Indonesias balanced budget law and Thailand's exchange rate management 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 budget 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. Thaieconomies in
lands budget deficits were about average for devdopi
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 successfully financed their deficits.This was possible for the foilowing reasons:
io08

MACROE.ON-O,

Table3.1 Consolidated
Public
SectorDeficits,
Selected
EastAsian
andOtherEconomies
pliRaikmo

-Averag

*Eir .,~ ~

1^89

Korea,Rep.of
;Mn.Iaysia
-nd e

;u0

10980'
5,8023

.*

.$Thai

ng

4O

def20tcii
pcet
0>;0vj6wP
dkevelpngncotrnries
'age:
-.
:
-gi
^.
ofDP~
D,a9r
'980-88; .(
(I
ehi
4stefx)
~~~~O .,
-.
.

'

034-

6.39,

Auemge~~~
-viag. OECDeronomnies- :

2.82
:.

-- Oderieconomi&

~~ Az~~cndna

~~9.62-

.BIUI

:4.02Yhilippines

Sewtcasticr1y

Rdgc

~~~~~z.4.30
-and S

Hcbbdl
-

(lirrhcomi-g).

:* Frst, dierewasfeedbackfromhigbgrowh. Sincegrowthwashigher

in Malaysia and Thailand than in less successfil economies, a
higher budget deficit could be financed. Becausehigh growth increases 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 savingswent into the domesticfinancialsystem(asopposed to realassets 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 resortng to inflationary financing. In Malaysia, the government's

Provident Fund mobilizeddomesuc savingsfor the goverrment's

use in noninflationary financing of the deficit
109

~~~~~~~LE
C L E-

_~~~~~rIR

- -

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

levetof externaldebt to GDPwas very low, which meant that external financingwas availablewhen needed.

Table32 InflationRates
-

Aveiag;'
*Econonre

tt96iAOi

75

,.- HPAt.
Hong

1'

.ong

tjQAKorei~iRe~of

&z^4J,l

.;

8.8
1^2.2 Athe

-I i ,

$3

nism. Other self-imposed constraints on fiscal policies and borrowing

-;

--..
6 > :;.
-

M-%

lie-.............
:.
-

.A

~
2005l|wzX,
^ .
* i s .>^7^
-1.1

- - -:
MiK&
i

__*>'1120Z:

no.

have avoided the inflation-inducingburscs

Becauseof this, the HPAEs
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 financing dried up. The impact on inflation has been equallydramatic.
Moderateto LowInflation.Unlike many Latin American
maintaining
economies,the HPAEskept inflation from spinning out of control. Table
3.2 shows the low inflation in Malaysia,Singapore,Thailand, and Taiwan, 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
and Fischer 1993). Low inflamacroeconomicinstability (Domrnbusch
tuon is a corollaryof fiscalprudence: East Asian governmentsnever had
to rely heavily on the inflationt becauseteir deficitswerewithin financeablelimits.
Asian governmentshave made fornal commitments to
.
.Several
i. . r East
low inflation that constrained their options fr acivism. Historically,
most important of these was the commitment to a fixed exdhange
rate; however,aU but Hong Kong have since abandoned this mech-

ld^y^ l

0.$g11i;ES

induding balancedbudget lawsand variousinstitutional checks,remain

in place and have generaly contributed to fiscaldisciplineand low infladion.(We discussrthesein chapter 4.) In general, HPAE govenments
been
>^;have
strong enough to alter public spending and foreign borrow~ .;
ing asneeded,althoughin Thailandthishasbeena continuous struggle
(Warr and Nadhiprabha 1993).
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

grewout-of traumatic inflationaryspiralstiat accompaniedeconomic

and politcal crises.In Malaysiaand Singapore,the successof colonial-era
fiscalconservatsm seemsto have helped shape postcolonialpolicies.In

MACRQ ECON~

;:.~~~~~~~~~~~~~I
I: asa Percentage
Figure
3.1 Revenues
fromMoney
Creation
ofGDP.Examples
fromEastAsiaandOther
Selected
Economies
(Percent)
4
4

12 -

:: ..10

iMalayhia

IIa-

-a

~~~~~Thafland

Rep.of Komea

1970

1972

8
2
190
12-

22

NReene
ie
-

1974

1976

1978

1976

194

iw moe c:ato

ZaireM
Arge.tnna

10 -

198

1378f-

Ji a pecng

2982

988

1986

M2fy1984

-sde

afined
asroo

oaIl

-2_

1986

198.2-

1980
of G'

1984

--

GD--

6-uce World^
Ban data.

--

-I

j,

__

frode

ansfrgincrase goenmn

1970

1972

1974

1976

spending.#

1978

1980

1982

--

1986

is"4

Nofte.Rev:nucs from moncy crmcaonas a pucrcntagep

of GDP is ds fi ned as ra2do
of nominal

nge in hirgu-powerdmony monominalGDP.

O

iwcff Weridl
S
Bankc
damc

: Thailand,the traclitionof responsible

fiscalpoliciesdatesfrm t:heniine: teenthcentury,whiena strosngcurrencyhelpedthffe
kingdomretainiftsin-

v dependence.In HongKong,

coloniial

rWfie

has

insffJated
the

government

; from demandsfor-increased
govemmrent
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
is rcmarktble.In the East
Mexico.Aswith moneycreation,thecontrast
Asiancases,lowinflationand flexiblefinancialpolicieskeptrealinterest
rateswithina narrowrange.For the comparators,the combinationof
nominalinterestratecontrolswithhighand unstableinflationwasdeadly.

wildgyrationsin realinterst ratescreatedsevereuncertaintyfor investorsm

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

20~ -

JI

-70
SO-80._R.~~~

1970

1972

1974

20-

Rep.ofKorea
Malandu:
Malaysia. ,~~~1~

j
I

1978

197E

1980

296

2988

1984

40

o4-A-.
-30

-70
"-80

-- . '

-.

-1

!-

!_
i7[7

400

400

19-70

'

+_

'-

1972

1974

1976

197

2S80

Ghana

Argentina

_
1i982

_]

1984

_=

1986

Nte. Realintercstratesaredefinedas diedepositrate deflatedby the consumerpuce index.

WorldBankdata.
&-wuree:
.

11-2.,

--.

MACROECQ

Keeptmg
Extenal
Debtunder
ControL
Of the sevendevelopingHPAES,

-M

only

Indonesia,Korea,Malaysia,and Thailand havepublic or publiclyguaranteedforeigndebt. The governmentsof the others-Hong Kong,Singapore,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 adjustment. 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 finance 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
;t

4g-."

lndebtedness

-~S '

Pea pt

WStFrnom'rgo

199 d

HPE
x0

--

hdonesia

690

qv478r

,:

- *,

'A

;;f-

Dc';-.

r4+44;-;-t'
7:2
0hte;nme
*
~~-.-is~,t-

A..4.

-,'."-'

47"-

U' -:9

POW

-455
2

47

-`3408'

Pk

s66A

~~MaI~jysxa
~865

>Akioftvnal
deb
to

,xen
.- Ra,. ofor

1384

390T98joCr'.;

54

177

''
'''0

4:

r.

-'!r-sA

38

..

-.

026B0

tx

-1762II

"3

AaNt 1VRAC

burgeoning international reservesgenerated by exports to make payments aheadof schedule;by 1990the debrtCNPratiowas down-to14
percent (In contrast, when Mexicofaced scvereproblemswith its creditorsin 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 inadequare rare of nominal depreciation)in the faceof continuing inflation
but depreciated when necessary,someumes quite sharply. Fiscalprudence 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

.-

. f.

.Ecoonomy
..
: H

Indx
;t
:
f(highier
vau means. .
mareapp recr

Kong

*Indonesia.
Koreo,Rep. of

,- ' "' ' '''

~~

0 ' ''

--

.8 D-

..-

.r.s

25
41
1

87;

I-

e-co
;;Slidiend
"nomiei-s

114' ,

'-

,,

~~~~~~0;

...

4a

ap.-p. at0a&O

~~~88

Singapore-

Bolivia

.e

98
1

Malaysia

Oce - .

Pe;rcent -.- aamgerankr>:

( O: si.an10
most.-

,-

90

.2- J--";

MACRORC

:
.* -:

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

Thailand had the longestfixedrate regimc the baht was fixedto the dollar during 1954-84, with a singlesmall devaluationin 1981. In 1984
the baht wasdevaluedand floatedas part of an adjustment program. Indonesiafixedthe rupiahto the U.S. dollar from 1971 to 1978. Because
inflationwas higher in Indonesia than the United Staes, the rupiah appredated vis-a-visthe dollar,necessitatingmnajordevaluationsin 1978
(51 percent), 1983 (38 percent),and 1986 (45 percent),afterwhich Indonesiashiftedto 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-butadjustablerate 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 managed 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 inflation (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

Rep oKorea-

25D 0200

SD- -

__

0
1970

I__

I_

I'

1972

mo

__I.,

__

1976

1974

1978

I980

--

1988

-- Arg ntarM
e-i.o

Z.

100 -

~~~

10

1972

!1 i i

-5-

1970

1986

Tr.

MO-

1984

1982

300--

-------

300-

I___

!I

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 describestwo more typicaldevelopingeconomies,Cc-e d'Ivoireand MexWia

ROECNO:
MACRAC

-I

gik"I!xL~R

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

&q i"-,

J5tczur

g'%k

dafiii "

c7

'-#-

yx;rvn

*1 ~~~~~~~~
:tt

K>e,;efiile(nu
*-;--

tgl

g$>

>0

*t>

10

"'1"t.>-iJ

-F'

cnsxilngwod4sing

mn

is a

n boxa3&

ath-:
anca frgr ttSF

. variety of
-i to.ea
-,=- j*fd
l- E' Nh, ; + <succdi,
uya
athe HPAESinrsonde
in whidi*&e.mtvt'w>.ttiS.................-i>
rheismdy4bun&m ihakd&Rod&M->
-13eeoiOmlecs
,^fto
action in rs,'l;lponsewr; uhshcsanc-v
fMaulettaocltaket'imEely
that
<6_. f $X
,,3ii'-g\;-,..)-sVl,<w;-;='i--*,te>;t,r<,.,
=tr
2,;
j,
.;

t'F"'vhe
awer"'ou

sjeoomamr

term.asOft2dcbrnzoi.EhcEastArsiaccn
voabe
.,as

cfa-vscin

ohocsf

dvlo96u-t.aggl

'.-s

sods:Indonesia,
maCrOe:oneomic
hva rge
SomxenIobsrvr

-Bld-ow

*_

-..

-wlWtora--

wstas
mi*>>icshavebeoinaica

ico,din flledto tak toiml ac5tidoninepos to

jt, wini-omm CZ>-

.m4e for caes

-lci

'a,d

t-;gTl,

or

r*T.

Uroole'Eithti

io*eeldvomgm

b&AkfrAhe

thinaUe.Weinri

hcdo

aia

bbiisl'btt`~i~

o-n'
uq`c1ie.aRAsianiconj
x1B

rm'n

m;

dXc~~~~~6aecord4icsi.&acrh.

.;

?YvsK:r~~

'

we

i-tand
s
uc shockl

unfaexa--

*su

Thand..
Siengapoenloand
rather
Asia has bee +lucky7
Ea'noooS>ast
ta ore,h,

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

$
to aaieyo
n ou ae in whc th HPE respode sucesll
Kra,Singi;Dapore,andThailand.
Indonesiai,snc
macroeconomicshomcks:
0

%-"7J
2.eex14f-;RS,.-~*S.-4,~-tk;k.te-fSKe-t.w,WNit..hej

''-'~~~~~~~~-.'i'-:a~~~~~~n~;'

..

~_t-j
~~

~ ~~~~ ~

:1

~.

,,,',PO

_~

:fifl

~0'

*.

_V

..

O-MalwAfl&q8w

..........................................

d4x,O

--

no

fd

ipA

Aursou1s4

ntoJran
; -o errcostlsXnomy<naewdroDalln
':*mrC&jp.

XanonianadiLiebanksav-w1rkt1iaru&oPmnsm;amapcemad
trto

;DeB

.4l~~a

^n~~~~~t

5tf

~~~~~

XR2SOZSFU.uas=g

1986 Inonsi *fuced

From 1982ctoitls'l
OilPries hi.sIndonesias.
: > w>W~~~~~~~~~ein;iug
decli%ni;ngi
traecuedrmriyb
ters
oqf
rarpily
worsenuing
- tMcreaeq-puD^ltMndgwith avremarkabl
r esponde.di
jgovernment,
thej!
Strtn in_ 1983,4>
t t~~~~~~~~rcs
DinUestlrsrlvs~~omplrehelnsieanF-dvsucesuladjustmenstpro2gDram.Idealuedthe
..
.
minl

an cur exedtus
1986>=
in 198 awntd
.~~~~~~~~~~il
.~~~~~~~~~Ddnn

inilntacil90

adarLsIndonesa. OF
aalcs

oDmff
teachieveen
. remmentSemMewolskomeffmeasurem
>erStalfizaonx DastalpriXeoa.Srng

*-. . -~~~~~~~n
"8~~~~~~~~~E

~~~~~~~~~iprso goos adnnactoservices

.~ ~

~~

ca'lto

ollow

bthe
i
fosqrom
98tillPustated

n18,tegvranenTaa-nit

robinssnaett)eltSmitocapiml-intcnsivepoDjeC

--

faed

to 196,Iadronosi

rom.ut1982

lreOspondkled.u with a.eaml

othes orthaocomng). The exces of

fo
;Littlean
calculoytaEnViamSoaetion

SS

~~~~~1

by recedln

(fo

Li'

frormtab
overannon-oltexots f*i:3ls

OIhened
t eue

anc

otes

93

hegvrmn

account deficit

foromn)

192no
scdvices
F ove
si
~~~~~~~~~importsof
lslpiaiyb
td
rapidlyw~~~~~~~9orsligtrso
rosSarigi

urn

xeso

A96
exportsi facedfi

epoddw8

elnn

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 interest 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 government pushed ahead with the adjustmentprocess,which actuallyacceleratedin 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
Crisis
inKorea.In 1979, Koreaencountered
a varietyofproblemsthat threatened to undercut the 1970s'impressive
gro
wthu Risingoil pricesbattered Korea'stermnsof trade, the world recessiondampenedexporcdemand, and high interestrates boosted debt
servicecosts. Koreawas not unique in these troubles,of course; these
outsideof
werethe samewoesthat led to debt cises in many economnies
East Asia.

Koreahad plenty of specificproblemsbesides.Realappreciationduring the 1974-79 fixedexchangerate regine had made exportslesscompetitive,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 largeandgrowing budget deficits.And while privatesavingshad dropped due to decdining 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 perCentr,inHfationsoaredto more than 25 percent, and the currentaccount
deficitapproached9 percentof GDP.The strong medicinewaspardy responsible for the economys worsning symptoms: the devaluation
spurred inflation, while tighter aggregatedemand polcies excerbated
:i9g

~~~~~the
drop in output. Even so, and despite politicaloutraiesoverKords
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
Th
mcmeh
secondoilshockand the risein
worldinterestrates.By 1980-81,thetcornlidatedpublicsectordeficit
was 7 percent of GDP,nearlyhalf of which-wasthe deficitof nonfinan-

dal publicenterprises.
Thecurrentaccountdeficitwasalsoabout7 perctentBecauseforeignbo rro inghad been moderate-the debt-GDP

radtowasonly35 percentin 1982-Thailand wasnot &cinga dbt crnsis and continuedto borrow.Evenso, the new governmentthatrook
wover
in 1980 perceivedthat macroeconomic
adjustmentwas needed.
Monetry policyoptionswere
lifmitedbythe fixed xchangerateandthe
relativelyopencapitalmarke. The governmenttherfciretookthealternativepath,fiscl contraction,movig gradualybut consistentydudrg
the next severalyearsto cut expendituresand boost revenues
Policymakerssteeplycur deficits of the nonfinancialpublic enter-

prises,then gradualyreducedthe centralgovernmentdefici. As a rtsuit, the consolidatedgovemrnentdeficitdecined from 8 percentof

GDPin 198142 to 1.6 percentin 1986-87,whenadjustmentwas
es-

;- . : -

120.

--

sentiallycomplete.Meanwhile,steepertaxrates and toughercolection

efTorts
boostedcepntalgavernent 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 weresustained into the late 1980samid an exporctand foreigninvestmentboom.
Sin 19878Tailand has becenaccumulatingfoeign excamngerewerosu
fan
hesanqconev
~~~~ecnomy
serves,
and the governm eficlpiiswreosset
has
ent rcgularlyrecordeda fiscalsurplus.
ThaCocing
PolicyFailwesh Snpor. Singaporeescapedthe 1980-81
world recessionwathsecacelyadip in its robust growth.But in 1985the
-economyencountereda sudden and severerecessiongrowthfel hom

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 laborand the
intensiveproduction but backfiredby erodingcompetitiveness,
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 emnployers
compulsorycontribution 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 recoMy-c
In 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 Inth CenrMProvi
th
contibutio
emploers' tOcmpulsoy
.
alsocut
donesiaand Koreahave also been long-termlow-inflationeconomies,
while Indoncsia and Koreafal into the moderatelylow-inflationategoesvLowor modate inflationfor lbongperiods providesa fivorable
growth.
envirorinmentfor
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 conaolled,
and hence real interest rates become negativeand volatle; depreciatons of the exchangerate lag behind inflation, so that rel
appreciatonsand exche
rate variability
rsult.
As wc
satw,
the
grelinGompeiitiveneSS domestic
andsumulated
demand.
The
governmenr

7---7Pen

MAL

terestrateand the real exchangeratewereunusuallystablein most

HPAEs,compared
withotherdeveloping
economies,
whichmeantthat
at guidingresourceallothesecrucialrelativepricesweremoree&fctive
cation(a topic to whichwe shallreturn in chapter6).
Second,realtax collectionslaginflation,becausecollectionsarebased
on nominalincomesof an earlieryear(the Tanzicffect),andpublic utility 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 investmentand 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,scarcemanagerialresourcesin 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 example,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 modcrateinflationmay 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 macrosuccess, these
economicshockswere not the whole story of the HPAES'
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. Macroeconomic stability helped exports by easing the liberaization of restraints on tade and by facilitating realistic and, in some cases,
undervaluedexchangerates.We discussthese issuesbelow.Few generalizations 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 summarizedin 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 construeguidcddevclpmcnc,
do'a,1950-O0
194846'

Mket-led developmet .150-70

Land reformnanid
reconstruction,

~~~~~~~

Natural-resouirceae prs

~~~~~~1949-.521900

Impori-substinicing
industrialirzatiLon-,
1953-57

AF:K~~~~~~~~~~~Eport
promotion

1958-72

ZTilting.twr
~ incur
73 1967
tiC .;;J*-Export

promotion.
tSelectiv6inter:vendion>-2andirnPortssubst-~
Omlandcommoitydioi4hii:heavy2nd
-boomIn,974-81
cciakuwe
.

Fvoring import sb
Indi'srial. con-solida- sio
197180:
no
stontut7-o5
growth,1973-080

~~Functional

*
AdjstmntEto aex-

abdlbcraizaon,
1809
shodci`
~ ~ ~ ~85~ ~

..

.-

.':

~~~~~~~~~~~~icntvs

Hihtecnlgn
Adjw~~~mciz-andh- moeniztion,
19iiod-ors
cAIimno,

Derc~~~u1adon:and
oriU.nrarion,T
* outward
d
A_

~~~~~~~Refborm
ad epr
-1980-

present

,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 account balancesneve-rfaced serious long-term deficits. ndonesa and
Korea. with more troubled current:accounts, might have imposed restrictions 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 accommiodating 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 account deficit. Often trade liberalizationhas been part of a policypackage 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 liberalizationin 1958 with a policy packagechatincluded a 25 percent devaluation,a unifiedexchangerate, exportincentives,and the widespread
remnoval
of quantitative import restrictions.In Korea,devaluationsand
tradeliberalizationhavegone hand in hand, beginningwith a dtrastkc
devaluationin 1961 (Collinsand Park 1989). Indonesiabegantrade liberalization 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 policies to off-setthe adverseimpact of trade liberalizationson producersof
import substitutes.A few went beyondthis objctive, however,and used
* deliberatelyundervalued exchangerates to assistexporters.In theseinstinces, 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 Indonesiaalsodelibecrately
undrvalued 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 percent in 1986), resultedfrom governmenteffortsto managethe exchange 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 potential 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 anticipatoryexchangerate protection.No immediatebalanceof payments 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 export 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 export 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
From 1986 to
and Taiwan, China, for developingexport indlustries.
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 volIume
of 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
*
X

Korea'sdevelopmenthas passedthroughthreesmgesand is currently

in the midst of a fourth Unlikeseveralof the largereconomiesin East
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 partitioning 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,thesouthwasalmostentirelydependenton 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
Takeoff
1961-73).
UnderPresidentPark,aggressive
promotion
of exports was combined with-classicimport protection at home. Koreanpolicyakers nmainLiained
dose control over trade, exchange, and
- 1financialpolicy,as wel as aspeccsof industrial decisionmaking.
In contrastto othercontrolled
economies,theyusedtheseinstrumentsto pur127,

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 discretionary.Exporters were supported with multiple exchangerates, direct 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,
exchange rates determined by
but it also gave access to he firavomble
scarcity rents in heavily protected domestic markets. Even as discreincentives were gradually replaced by more automatic instruments, exporters received significant exemptions from import controls.
tionary

for Imports
andExports,
Table3.6 Effective
Exchange
Rates
Republic
of Korea
'

'

'

''

' ' S' '''' ^'#ren'ccs due to:

'"~

'

-~~' '

Ecnge-

; 4 f-,--'-,"-,
Yen

:.ax ax a n.'
':D'treet
ji.
materb-- - '.' u a
equwa..
inSurceMt)
(percent)

p'remiums phs:

~d6r tcashX

of2&i<E&YZY.5.
suGbsud

. wE

-1958

1.79

1959
.19 0
-1961
1.962

97

10.7173

`19638

19

...196.,
,1970

:.~~~~~~~~~16

1964 114

'~~

.4

-225

118

.44

44
.-. :

113

-69
;

r'c onnrts;E

xz8

~~29: 21

33

'0

23
29
.5K

148

3:

,-

Efct. 'i=bunos

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

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 protection 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 explicitly earmarked.by the government for particular activities or industries, lent passively by banks at preferential interest rates. Following
explicit government directives, banks increasingly used export performance 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 incentives of the takeoffperiod to a commitment by govremment to use all
its levers to steer resources into specific sectors to rapidly alter the industrial structure. Special legislation singled out six strategic industriessteel, petrochemicals, nonferrous metals, shipbuilding, electronics, and
machinery-to receivesupport, induding tax incentives, detailed engineering, subsidized public services, and preferential financing. Thcegovernment 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-industry sector, Korea was at least partly successful. One reason is that the
government made dear fiom the outset that these industries were expected 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 devalued, and credit allocation policies switched, with a termination of
large-scale preferences to the HcI sector. The five-year economic plan
129

-*

:.

~ -%.

-Cj
_

,-cmpnc

B EW4

P1...L

L E' .

fr:tU.C

rE

i4B
ySps
tioizhzr-nontss

,,mdopmentBaEcL:

)ctytpscueebeRayEng,:mlcK0wav,es0vens,3nfNthe:Earler

l ;!l:
ha&obtide3".*>mt'i
mt.'w"gWW $8!stft47
.PM.wodsbMwoucra'ovcnsaSo
t

* .5
i.-1OI6W4lKLQJon

loan-

2 5

,,-,-

033

apancsg-'co
ramadcaelz

ays-ng.cheap

s-Cot

2t m.~

oi

l.F.Sdcparunrnjstwe4*
&lSCca4O :ioo
cn

__~~~~~~-w

S-It?'

-EsSw-S^h
jwtiglrttm&ask*Glitbaej&fl

Moi~~~~~~~~~_S
an

___

a~-

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

f echolgyand
scupport for h develokmopmcent

don
competitiont.ti
of in it apprs:oa
poSosoulcylis
ME-Z><dustM-iDsSoxrxSStesrtrialed

o fnaLncasetor

the promo-

rneform.The overn-si

oh bndcartiuepn
glarlyaact
eateinbaleofo
susete mand wlesacdcu in
dustries. ith remnantsoffnancsr liaeanr port-oitepizaie Ko grais
paymfedhaien
nt1has
been
thersdeat
of theonesomy
ws and
and
of the ex
th 1970m
was
incgnwized
th reideopei theompexcesse
mdn*e
197 opngithd
industries, sidpdetfnna h devsector.en of t-chnology and the promoparclarlyn andv in brailorsrole sunrt and oveaseveragedlly
-1309.y, ssovemmenn1.officms Wwortionhas bendciv
ofcompctiton.
progreamis.p
remants ofpsfinancies are ipart-libtepraiceio
; . ~~~~~~~~~~utrieslathedita
*Infoprointervention
iue tof
oest
ablishgandkme
edbishresse
olcies foruedo
ad forthe
ther failucure
for pror sicp9
ntrvention apolicies79and
-- ~~~ ~ ~~~~~m
adtepoo
fiupportfoste dvlpetofrc.lg
~~~~~~industrident
: -

MACRO: EcoNP.

of Export
PushinTaiwan,China
TheEvolition
Developmentpolicyin Taiwan,China,has consistedof fivestagesin
which the governmenthas implementedcomprehensivebut changing
poliqcpackages.Throughout,low inflationand macroeconomic
stabilityhavebeen a foundationfor growth-enhancingpolicyinitiatives,and
sincethe late 1950s-exportgrowthhasalso beena fundamentalgoal.
LandReformandReconstruction
(1949-52). When the Taiwanese authorities 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 imports of machinery,equipment, and industrial raw marerials.These in
rapid development
turn helpedto make possiblesubsequent cxport-led
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, expanding transportation and power necworksbuilt by the Japanese;U.S.
aiidwas an important sourceoffiriance,funding 49 percent of public investment 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 stimulating the import of capital and intermediate goods while penalizing

exports, imporc substitution contributed to a growing trade deficit, financedlargely by U.S. aid. As the small domestic marketbecamesaturated, 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 idendifjypromisingindustries for export promotion and development. On the basis of Taiwan, Chinas, comparative advantage in low-cost labor and existing
technical capabilities, the institure chose plastics, synthetic fibers, and
electronic components. Other indusrries subsequently promoted induded 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. Although 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,economicdev-elopment.It shifted the economy from a relativelydosed to an open
economic system and exposedit to the forcesof international competition 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 lepercussionsfor 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-relianrdevelopment 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 capabilities,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 production of raw materialsand intermediatesfor the use of export industries.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 internacional 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 revealedstructural weaknesses,particularly the financial system'sinability to match
the increasingdemands of induscialization and external trade. Externally, 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 government decided to focus on high-technology industries: information,
biotechnology,electro-optics, machineryand precisioninstruents, and
environmentaltechnolo&rindustries.Theshift to ahigh-technologyeconomy has necessiated the close coordinaton of industrial, financial, science 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 diversifyr
and improve production 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 infrastructure projects, including expansionof the energy,telecommnunications, and transportation networksand developmentof water resources
and national parks. The government'sdevelopmentplan for 1991-96
callsfor 330 billion dollarsin public sectorprojects.To support theseeffbrcs,the governmenthas adopted an overallstrategy of economic liberalization 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 produce incrcasinglysophisticatedand higher-value-addedproducts. For
most firns, however,the transition to hi-tech industrieshas been difficult. 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 skillstrainingneeded to shift toward high technology.As a result, most Taiwan, 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
aldtough it intervened extensivelyto promote mral develgovemrnment,
opment and providesocialand physicalinfrastruacure.While the government protected import-competng industies, protection wasgenerally
less strong than in other developingeconomies.Government was re-

strainedfiom heavily biasing incentives against agriculture by the economic 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 imporc substitution was the same as that in odter developingeconomies:
reducingimportsof consumer goods and increasingprocessingof natural resourcesto create industial employmentopportunities. The government 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 Malaysia'sgrowth of manuFacuringoutput (Salich,Yeah,and Meyanathan
1993).
Combining
Export
Promotion
andImport
Substituion
(197145).Malaysia's
1969 ethnic conflictstriggereda reexaminationof developmencpolicy
The New Economic Policy(NEP) launched in 1971 had many dimensionsintended to promote growth with equity. In trade policy,the govemmnentbegan more acive promotion of natural resource exports,
particularlyrubber, timber, palm oil, and petroleium,and light manufactured 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 export 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 manufacturing warehouses that permitted duty-free import of materials to be
assembledor processedfor export were crucial to the successfillcombination of import substitution and export promotion. Foreign investment, 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 dedined to about 40 percent in 1989.
(1986).A combination of terms of trade
Adjustment
andLiberalzation
shocks and fiscal imbalances (describedabove) prompted the govcrnment to move in 1986 away from state-ledindustrialization.Promotion
ofprivateminvestment
acrossa broad rangeof sectorswas combinedwith
macroeconomicadjustment and continued effortsto increasemanufactured 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 reductions and changesin export incentivesincreasedthe variation in cffectiveprotecCton
acrosssectors.
[135

-ALCL
I

Indonesia
TriesIt BothWaysandOptsforExport
Push
Indonesian trade policies have swung from protectionism to openness.The current open phase has gone fiurtherthan any earlier efforts
and obtained the most impressive results.
Nationalism
audGuidedDevelopment
(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 Indonesianimporters 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 exchange 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 previousowners,and a newForeignInvesamentLawthat provideda thirty-yearguaranree of nonnationalizationwas enacted.The exchangerate was adjusted
throughlargedevaluations.In 1970,the governent establishedthe mclianismfor a market-drivenunifiedexchangerate and abolishedcontrolson
capitalmovements.Sweepingchangeswereintroducedin the tradeand incentiveregimes,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 annual 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 resources 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 social 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 investment- 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 Pertamma 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 capitaland resource-intensive sectrs as oil refining, liquid natural gas, chemicaL, 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 pressure from domestc interests caused the trade regime to become more
protectionist and variable.
AdMjusuenttExtenalShocks(1982-85. The high-cost nature of the industrial scaor and die end of the oil boom sparked a debate and reassessmentof industrial strategy in the early 1980s. The govermnment
initiated a broadly based adjustment program designed to maintain balance of payments and fiscal stability while reducing the economy's dependence 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 financial 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 dedined, 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, reduced 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 resources 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.
In 1986, the economy sufOrietaton(1986P--.
DerigflationandOutward
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 macmeconomic stability, the goverrnent launched a program of broad trade
and regulatory reforms that opened the way for a decisive shlfif o export
promotion. Exporc-oriented deregulation has gathered momentum

Figum3.4 EffectiveRatOs

since 1986-8 7, culmninatingin a series of major reforms in trde policy,

licensing and transport regulatons.
The first step toward trade reforrn bad actualy been taken in 1985
with steep tariff curs. In 1986, pushed forward by extrnal shocks, the

...
Import-:X
Impoflb 2sminvestment
Tramnston

_ l
:Exportpush-dI

1--+,i

o-- 20 40
- 6o so 100
Effatve rtedf potectJon

(percent)

31Indonesia Thabnd
and Mcyanadmn
-cSalieb,
Xcah.

* Mlaysia

-mcr
507zSo&k,

138

ment announced a package of exporc incentives, a major devaluation,

and the first of several programs to simplify imprt and export procedures. Major exporters were given unresticted, duty-free accessto im-

porcs (see figure-3.4). Later the government deregulareddomestic and

foreign private investment, and for the first time permirted the private

Yeah and Mcpnathan

(1993); BbarracharyaandPangx
Brinmbc (1993).

Iprocess gathered momentum. Between May and Ocaober, tde govern-

(1992w);

sector to invest in power, telecommunications, ports, and roads. It also

firher deregulatedthe financialsector

'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.Foreigninvestmentprojectsapprovedby the govcrnmentincreasedtenfold between 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 export growthdue to increasedaggregatedemand promptedby a prior
easingof monetarypolicies.This led to a wideningcurrent account
deficit and a surge in foreign borrowing.As with previousmacroeconomiccdistortions,
the governmentrespondedby tighteningmonetarypolicy and curbing borrowing. Meanwhile, the obvious successof
export-pushpolicies can be expectedto generatemomentum for firTher

export-oriented
reflrm. If this is the case,future swingstowardprotectionism,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
rent associated
tradepoliciesemphasizedtaxationof the resource-based
with theproductionof theseexportablecommodities.Tradewasheavily
controlled.Rice exports, Thailand'smajor commodity export, werecon-

trolledby a statemarketingmonopoly,exporttaxeson othercommoditieswereheavy,and a multipleexchangerateregimediscouragedexport

producton (Jitsuchon1991).
Exports
11955-70).In 1955 the exchangerate was
Natural-Reso.rce-Based

unified,and the state marketingmonopolyon rice exportswas abolished.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 developingeconomy standards. This was especiallynotable in consumer goods
manufacruring,where nominal tariffs were in the range of 25-30 percent ad valorem.Tariffson machinery and intermediateinputs were in
the range of 15-20 percent.
following
(1971401.
In the 1970sThailamd,
FavoiingImportSubsfition
the import-substitution strategies favored by many other developing
economics, raised tarifli on consumer goods to a rangc of 30-55 percent. Capital and intermediate goods continued to enter at low duty
rates, leading to an increasein the effectiveprotection to valueadded in
protection
import-substitutingindustriesand to dedines in the eff-ective
to agricultureand traditional exports (figure3.4). The biasagainstagricultural and export production increased.
Textiles,pharmaceuticals,and automobile assemblywere particularly
favored.At times, vehide importswere banned, which resulredin limitlessprotection 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 effect 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 direction 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 protection 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 surchargeto raise revenue for macroeconomicstabization objectves. Effective 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 cx140

ACROECO?rw1

. .~~~~~~~~~~~~~~~~~~~~~~~~.

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

the applicationof investmentincentives,had long been a tool for implementing the import-substimtionstrategy.One of its keyroles was to
distribute (and often increase)the incentivesfor firms operating in protected domestic market niches through waiversof tariflis,import surcharges,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 recommended radicalreformsin the wayin which the BOIdid business.It advocated using moreautomatic,sectorailybased incentives(in contrast to
- the firm-specificdiscretionalincentivesof the past two decades) and
shiftingthe targetof incentivesto labor-intensive,export-oriented,geographicallydispersedactivities.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 investment. Other incentivesfor exportersinduded tax exemptionsand
rebates,reductions in electricitytariffs,automarticaccessto credit, marketing assistance,and promotion of tracing companies.The government 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 extended 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 offset remainingdistortionsfrom Thailand's import-substitutionera. Tax
rebates, duty-freeimports, and export-processingzones correced partiallyfor existingdistortions.In this sense,Thailand'sexperienceis similar 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

-_

ii~~~~LE'

~~ ~
.g ~.
~ .,L~ . ~ ~~ie
.

~
-~

E-or

-. ~~~~~~~~~~~~~f

wS_

.'',M

''.~~~~~~~~~~~~~~~~~

~~~i

'

ar

-"

.n

'.ape'roa9AesS>'~~~~~~~~~~~~~~~~~
'
-

=CU

roa

es

:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-:
f

NT
:~

:~~~~~~~~~

~~~~~~~~~~nstet

sm W-96

ne

:--:: ::~~~~~~~~~~~~~~~~~~~~~~~~~i3

!:g

:-

;~~~~

- l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~a

- u - ~~~~~
- ~~~
- The-outcome
of thesepol~~~~~~~~~~~~~~iy
alaacB
96*

- - : -

manufacoJres
reprcsented
30.6pelt

pora;.Ladingsecorsinclu ed
d

: - -.-

elrr, ad integrarcd is

.~~~~~~~~~P

.- 0 .

X
:xS.;

50

;
-;

.:

..

Woin,
~IO

foei
Oirect
ilVtmetpl
n

*
0--

ex-

ojla

ofa gowingvolumeofThat
ao

MACR0O

of Man.factnriug,
Thailand
: 0 -0Table'3.7
v3: Distribution
;n.uMtA

C R1O36

J~~

(Percentageof grossdomesticproduct)
Senor'
,97.
'1986

Hev inutis3.9

* . Total.mTab
3 e Banko
ofTi2n

3.
.13.3
.23.'3

42.6:

High-skilllabor-intensive
idusre
Traditionallight inutis19.2

9.7

11.1
201

100

100

100

aflTiland

TheElementsof a Successful ExportPush

These summariesof the evolution of export push revealthe plethora
of policiesconsistentwith exportpromotion. (China'srecentpoliciesfa.to axvoring manufactured exportsare outlined in box 3.5.) AssistaLnce
ports varied over rime and across economiesand induded p..ferential
.financing,promotion stubsidies,tax incentives,subsidized inf1auctLre, and foreign investment incentives.Moreover,at any given time
and in any particular economy,the magnitude of the incentivesvaried
acrossactivities.Some governmentsfavored nontraditional exports,
exportshEven so, there
some favoredparticular types of manufacmured
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
npntrin
bonmd
aed
in
aitn
psorts-tewrld prinacer
who do not face similarly expensive imports. To combine import protection 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
exemptions from tiffs. However, the
warehouses, luty drawbacks, or
img
epotesapssst
foud yia wystogan
higher thelevel of irmport-substitution protecton, the greatcerthe level

0 ovrnen:shn

wof
adminisrative

cormiptenc

for exporters. Perha

needed to cseate countrvailing imcentves

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

_5
-tCL~CE

...
..

."

~ ~~ ~ ~
. . ~
~~ ~ ~ ~ ~ ~
M

.~~~~~~~~~~~~~~~~pt,

- :

, ar, L

.*
G

.~~~~~~~~~~~~~~~~~~~~~~~~~~~

g.

..

<._

...

MC

pogamtoenux

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

ofcreit
j
ntnvru
the degre
oSsubsidizaon9
(
!;aMantS; subsi
as
a,

we oberve i

- shorr-tan),
f

dizedraresX,
theselecizviyallQqporT
ve9u

T~~~~~~~~~~~~~~~~~~~ ~
.~~~~~~~~~~~~~~~~~~~~~A-

..

~~~~~~~~~~~~~
.V.
. .-:.: . ~, .:: -~ ;:
~

ipesvartyofhye

.r,

-a

Lgo-.

T_

.5

ex
5age
, ortatirres

~~I

MACRO.ECRO:E

all governmentsrecognizedthe difficulty

Mabrket
Penebaiom. NearLy
expoers Fce in crackingfbreignmarkersbut againchosevariousmeans
to encourageexportersto overcomethese hurdles. Somedirectlysubsidized 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:Policyflexibilityprovedimportant 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 necessarily bad, as long as the mistakesare reversed.The dassicexampleof policy 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 regard. 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 HPAEsdramaticallyincreasedthe share of exportsin machineryand equipment at
the expenseof such traditional manufiacmresas food products and textiles. In 1991 the share of rnachineryexports in total manufacturd exports 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 shiftingcomparativeadvantage.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 macroeconomicstabilityand much more rapid exportgrowththan most other
developingeconomies.They achievedmacroeconomicstability by adhering to macroeconomicfundamentals-particularly by keeping fiscal
deficitswithin the linmitsof prudent financing-and by rapidlyand effectively correcting major macroeconomic imbalancesthat emerged.
They achieved rapid export growth by fostering a favorableenvironment through macroeconomicstabilityand by applyingon a trial-anderror 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 assessed 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 altering policiesas necessaryto attain thesegoals?We mrnto thesequestions
of politicaleconomyin the next chapter.

3-47

R'A CLE

Appendix
3.1 Economic
andPloflitcal
rimermes
Indonesia

M.~

OpemdanatsubhJJisa':-

~~nnhcfup
ho

mita
b m n m

&

mawiled
~~~
Ci

Wed
dkea,mtecuainl
atpinduainnmmd
rank gugjsajfraden

ccanoiny.. a.
Rdninrpaaaim
Th~Nw
ndp
Wailvs!m

~ ~ ~ ~ ~ ~ ~~Ol

McPtsUniicatone.

dm

-fr'nlto.'--totim

in.

-f
.

cak-

r.hghmafi

firms ewzuedwt
Sq 01

HjyCIfDAhtuIKI

ins-LM

c=7-

7.

148

Gicderiy

1New
bcip

ntu

aio

asI&uiamoat

-minm

iila

Su&,wont
z.kdqaSns,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
"~"_...

MACRO

i-

Rqmn.

.h.c 1

J;.Dfdb

it
4

'S~~~~~~~~~~~~~~F
,;-b

Fa t h-

appmv.4mpa;

'

'.-

;elto

~~~~~~n ,
;
,,.E,--.-.ica
.,p9&
;9?,
:cauua
t~~~~~
~~~~~~~~~~~~~-__
* ~~~~
aduap ~~~~~~~
Duicamny&mns~

m-pr
,had;,

. .rtat&

4(**

*.

.--.

__

__

EWuaJn

.:PN

.
Smwod

SB

lBS.

1982

'

ai

185

298

cpu!

rus,* ;t

Sore

Applbw
mun-of

kauga

.DcpIaa
-Oc

sar

bb-l

-iiil0

6hd

0-

'149

::::~~~~~~'W

--

Malaysia

E' XFin;Ma/ape '

RmMaj'
t,.

-; ,,hiu..mREr;Fads''\
'

.,

-t* --.
.~~~~

-!
: .,,ar

-,'1

F95

-'

1;:'

'., .as

- 'eT
:. V
Wea

'--

''H

.>::-*

a'

TaiFhvki
...;.i:AeMof
T

~ ElbIrn f

.;Me:.i

.-

CoaiI

-f-l _ak '.

1.:;

2.95

286

26.

;36

.56

150

-;
1

_o

-d -d

aidnfrminMlyCnrn

-.

'

-.

:Invutcu

-'''t Inra

C.nnhzmoc
nduth
Ai.
.. s--

P-o
-d
Dcv
F uIn

.-.edpam

Atikd

2.;6

.~~~~

:~ ~ ~ ~

'a 'za
. .-

2.6

;, -

n:

--

tr*m
=t.

PmThdc
'Z,Aa

_39

27

Flu'

-|-.:

Guarimeu
-.

_.

puawersii.linccihispendcin

P-a
;'

lnancc

-it.hl,duc.rf

-,

155

'

-_

8h

-:

F. . .UIfllf
--

19:7'2
.uc-y:-- ,"
*e

;--T

MACROE!cONOI'd

f'aasuanu
Jlae
flundd

rA

Caaveuaan

7mteaa&

fmAapaL

-.--

w?Lntae
d

w?apAL

ik.;

-~tw

Ccep0
mdon.has

flatten
hakswete,
CoalIuonAa

mm

~~~~~~~~~~~~Hwedpma
A
H~

DFMdapt

6dm;

ate

Pummlga9onc

.9

-95

ZntaAa~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

SzittqfSis~~arhz4wteutitym

alirsics~~
Fmna
,
Dndh
Avl&mmmac=:-cdtdldnrc
kbca~hsrt'

Ddpsasll
imInanics .lard

~ ~

mn

toa rsc
indusif

dvilnnm

- lctor
Praadnfpn)-

aun
RelasionsAct

apusrake
Vzmz

Epl'e
A

asnr

ad~~~~~~~~~~~~~~
tarn

rwmpato

mqtpmoD

emda.A
indusrs

a cnpi

in

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

cir

q
Cud

byp(uu
adlcldorrr

'I

p
.1

ijr Ii

I:
S.

4
5.

N
U
E.

..

.2
U

Fgfl

Liii

iI

jS

H*

*j.

I:

* lB
*

* up
U
N
i

i.irr

I i-Ii
JJ'L[
**

a
.

I
I"

'

I,

ill
-

Lr
'(U
Id

CLE

Tainan,China

~~~~~~~~Tbhir
liur-Yac;Ema,nonc
i

141gbmaaflntmt
aauvlapdcpnrIndprn

D
Pi'lt

*
LiRnClp

LaNd
to
Ddnr
1845
1550
-

'54

-.

.93'164

1952

1953

law-

~~~~~~isarml
Smax far
nd
Tnug.IJiar

~~~~~~~~~~cadlazao
lan
lWborm

'InniCal

1965

290

1571568

1359

[WI

SWiPCle

RateL

86

Stwn fa

-cnt

of~~~~~~~~~~~~~~~

'CLWO'

*'

mv"nt
Kmr

oa

panacfmcfmeat
herpm

.7.
54

Nto

.-

185-19

16

C.

MACROE

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

-.

, -RL.----

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

- -

f fd;
Urnlimdon

Dewlcpmmr a

FAmU6k,cn&

-ofIndustr[al

0
:-,f - S T.-;
Estahlubman

'

-.d zial

Ins
-

,.,,'N

T.

,'

010

. aiwan.

-.

.,Ne

e
-'','.or

l.:-:,
Wdncbu

,, -

D -l
.'

Foreign

SAkbmcnt
O~Iiutr
=,
- &ah

.d;1972:
197n

,:
t; - ' -;*< u
i-,0 ' &uU'nihmc,
-d' c

"

:infmu,-c

'J;D';;

laboriaws
.

'x

'5en;-

Initkials&.a.

_______

WoddEatak
.

pu4n

*-.'

.d

d''',''

.;

oppm-IIn

Liwnadotpl

..
.. I

m7i~~~~~~~~~~~~~~~~~~~~~~5
0
f ....

,,,me%'to

Nat
''

is, i1ce
-.

m--a-r".

..

Notes
1. Bradfordis among the first to categorizea combination of functionaland sclectiveinterventionsthat resultin
the effectiveexchangerare for exports exceedingthat for
imports as an export push. By this definition the SoutheastAsian NiEswould not qtalifl (Bradford1986, 1990).
We we dse term somewhatmore broadlyro indude sustained movementtoward parity of incentivesbetweenexport and import substitutes,combinedwith institutional
supporr for exporters.
2. The analysisin this section draws on Easterly and
Vchmidt-Hebbel (Ibrthcoming. The data on consolidated 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 becausetheyindude 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 provided incentivesfor investmentsin agriculture,industry.

and tourism.
5. The BOi's policies of selectiveinterventionto promore firmsand specificsectorsare discussedmore fullyin
chapter 6.
6. Onc impormtn exampleof these effortswas the establishmentof export-processingzones (Ens) and bonded
warehouses.Drawingon Korea'sand Taiwan,China's, ecxperience,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 theseincentivesalso favoredlargeand often foreign-ownedfirms.
7. A firmwishingto obtain accessto the schemeissued
a promissorynote, which was discoLntedby its commercial 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 answer the basicchallengeof legitimnacy,
they included apindpoIe ofsbazred
growth.The leaders hoped chat rapid, widely shared improvements in
economicwelfarewould bring legtimacyOf 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~~nomiuc

growth
'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
throughSharedGrowth
Achieving

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 equitaMlywithout stifling growh The coalition'ssolution was the New
EconomicPolicy(NEP), an attempt to improve the lot of the Bumiputera, the largestbut poorest ehinic group. The coalition'sown early descriprion 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 experiencesany 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 system of EconomiicDemocracyreceivesas m.uchpublic exposure as possible 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 redress 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 leadership 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 destruction 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 effectiveconsultativearrangementsbetween 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 significantly 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 included 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 regarding labor by limiting the power of unions and intervening to check
labor radicalism, while at the same time encouraging a cooperative dimate in which labor was rewarded for increases in producriviy.
These wealth-sharing measures have differed from the typical redistributdve approach of most developing economies. Instead of granting
direct income transfes or subsidizing specfic commodities (for exampIe, food or fuel), HPAE leaders have favored mechanisms that increase
opportunities
for upward mobiity. The frequent result is that indivliduals 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, opportunity-creating mechanisms. The provision ofuniversal primary education 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 expansion 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 ownership 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 communist 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, compensating 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 markets. 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 interest in radical activities.Former landlords, as new shareholdersin state
enterprises,had a vestedinterest in the successof the Taiwaneseauthoriies 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 enterprises (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 attracting 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

I IACCL E

inTaiwan,China
Businesses
andMedium-Size
Table4.1 SmaUl
--.
- ,.-; ;e

-.

-;

..--

,,

.d

;-u.,.nt

-i

Smell

-s

-;f

---

-- - 0E u~0--:--::-Tou
talnipX resX

Cue.ory

^Agri
culur, fishing,livstock,forestry
..Mning and:quarlying
.
MItanu-;ting
gm a
r. . .: 1t..
;Ee c,gis,nd
,
-.
Constructon
Commerces:
Tmansporr,-srorngc,and
ommunimcations
services
Funancc,.isurancerel estatebusiness
social,andpersonalservic .69,550
--.:.Comxruunit$,
Total
-';

-1,390-157,965,

8628,419
489,864
36,897
30,534

a -ni

vi - rts

3,171

97.39

.1,360
155,263

97.84;
98.29:.
90.32

168.

818,061=- :':

StmrnnTai

Percentageof

Number0

3,256

-medim--e

-.

26,456
475,106.96.99
35,818 28263
69,229..

93.09.

-794,834
0 . . --

,97.16
-- : - ,: :

97.08-;
92.56: ;
99.54

China(1992b).

(see tabIe 4.2). Other HPAEshave also encouraged small and mediumsize industries. Japan has direaed enormous financial resources toward
dcveloping small and medium-size enterprises. Public financial instirutions have allocated an average of 10 percent of lending toward
SMES.During the rapid growth period of the 1950s, as much as 30 percent of their total lending for fixed investments went to SMEs.The various 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

Businesses
in Taiwaan,
China
Table4.2 TheExportValueof SmaliandMedium-Size

(millons of U.S. dogrs)

.,.i

.,.

-V

.::

-e~tkr.f
g .

'M2d

Auqon.

Po

.436.89

64

;,23525

-2.'...L.',

674'-'4

,:

OSrTIIJ

hn

3---

42

-'

i6z~~~~~~~~~~~~~~~~~~~~~~~~~~6

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 conglomerates,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 establishedan extensivesupport system for SMEs.And, as in Japan, financial 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 housimg markers to win the support and cooperation of non-elites. By proriding low-cost housing for the majority of residents, both programs
have helped to decrease inequality and minimize social unrest, thus providing<the long-term stability attractive to investors. Moreover, the massive 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 laborintensive manufacturing.
In Hong Kong, which has generally followed a laissez-faire approach
to the economy, mass housing programs were a iesponse to a massive influx 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 social disturbances, which culminated in riots in 1967. To difflasethe tension 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 govermuent 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 territoriesand populations, the experienceof Indonesia suggeststhat public housing carefully targeted coward low-income groups may have
benefitsin larger economies as wel (seebox 4.1).

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

MM. J"F

.3>titY$~

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

-.

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

i.>rohow

.,,peratlvece;

N,:i_OPULt-;

JU,SF

unablcosafford,...

-,

conca4 ctconianrsbdI'nolidal

nri
schaSuetck&^le
r

lu;

uron,

isarbcle

d.:-oLo6.

NI-4 &tr;$
'i=rt%v(i
0E e shCazi
cisr;+f_,';i
s,the lndoneslan
,dii,i potntil
.$QyiTOi'eSpottiid2*it&
-

14PT
'jA'A

ThdlW1I&wnroSfi
mmp~~~~g
V,
.
0~~~~~~~~~~~~~'

praflvhezoeinxa e4!zng.aa-WpdcaaunOp;pflennca

.2

O' q

.rp4rEonrir.
-

..

.rcen

2 ~,
atM

4zh4overnnenthshepd,h
b__ rl

01 1onapmcwohWidin-ahd.s
l

1c64

ooen3 4'

ao0O&n1nttl&n4

tt

own4paymentt,

5L~;

a., .~~i

oaraW-X;
4 ::

mmnnns

ropenu~1auz
,- ;ORS..

AN IN4S-TQ -*

tivityin an effirt to ensurepoliticalstability.Governmentsabolishedtradebasedlaborunionsand pushedthe creationofcompany-or enterprse-based

unions. Managementand company union representativeswere then requiredto 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 influence 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 productivity" (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 investors industrial peace. To meet workers' rising expectations, the governmnent 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 economy. 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

C LE

, 1, ,t77;tS>ri

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).

ATVESORAGANIZD*.BYSINGAPORE'S
K

i-ONQO5tTHEE1R.
FIRCOOPE

LW. ,>NtwOACoA't
bf.lrodany,
a

byOf&r(

ari

.....

eforp1crncuorVrne
-r
aarryrJo

f,ruung_.

fridhbYii

80.ebrrelxelw:pa

te .Tdfoms;. iAti4niz

6N-ii!'Ode4if&1I

oflisr int

m,drNeriiBccause
n q,qfgoods41n1.p
there west limirbon
ZihcnumberErg4$-;$
;$S-E wre&nn
owners .were-ab1le to
O
&sute
htAusvresorred-ro
dnyangiotr
*
tt-.

lcenseds

hA

Ar'hAgsLen.fl

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

rt

%L&AY.=u'.&O

'j

,s

sout__ro

z6b-d_

admzn,

*~~~~
2 .,i.. Mos

tamnve5

3'r-7.aWef
D;9iS!si

,5i2&~o

,.Kt,> Zrix

~~~1miUtw Xonda;toso&.faealtrncensei
unDedarUG.
id.@ 2
r,=LTz.eiC

;~~~~~~~~~~~~~g

-~

30 X123

.:.

~~~~~~~~~~~nf.wn
-~ ~ ~ ~~
~ ~ ~ ~ ~: ~ -:"~ ~

,T.4

,5r S

nd.sotonuceSE
oarEs;;2ancasc.

,C

25R"-iTr7:aCPVSi:*'

rtt''''';>r*CA -RX 'SgBgWtt

f-

I;"

.~

..

Ei

.ii

_'

..-

:s~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Other economieshave forestalleddevelopmentof largeand politically

powerfulunions with less formal arrangements.In Taiwan, China, for
example,closemanagemcnt-laborcontact in small and medium-sizeenterprisesreducedworkers'interest in union representation(seetable 4.3).
Moreover, the emphasison labor-intensivemanufacturing generated
rapid increasesin the demand for labor. Consequendy,as figure4.1 illustrates, from 1952 to 1979, inflation-adjustedbasic wages increased
more than sixfold in manufacturing and more than fourfold in agriculture. 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 achieving both thesegoals.Foremostamong them is technocraticinsulationthe ability of economic technocratsto formulate and implementpolicies
formulatednationalgoalswith a minimum of
in kceepingwith politicaUly
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).
Activit byScaleof Businessin Taiwan,China
Table4.3 LaborUnion
(pnenetz

,p-

No .

*-3,--es

4:i30-49

.-.

.85.
29S.

.;

-2

.t

19.45

So28

,a
.

.51

:X05
6.26i_'

r'-SourerTaiwan,
clnnChial
Bure of rheTii _Proviial
tSt

&'980-84

Norrenponre

06.M

.44

51:05
,

9:
-f67-

-i-:;.
-ojo.itX6

008
015
0
37T'..

Government,
Annual
:

167

4.1 IndexofRealWags 1952-79

Figure
I

ii.

o;#{~~~~~~~

Percent(IndEx:1952;100)
:.:
l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

.::

0~~~~~~~~~~

'

~~~~~~~~~~Murnsfacturln

300
of

Fue41Inv

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 Democratic 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

~~~~~~~~~~~~~~A
N- -INS

~ ~~~~~~

.u'

-:

t.di

.____.~~~~~~~~~~~~~~~~I

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

EP,Pr
L

,psuN'_O

utyt~~~~;g

,i

dgm

:~~~~~~~~~~~~~~~~~~~~O
Ce
Ico,
u~~~~~~~~~t)r,,

V ormmg

-wsC

0:00000~~~~~~~~~ICR
a

-~~~~~~~~~~~~'

an

:P3p

3986

rcotan aswep

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

'

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

}:g_D
,-.

. f ,

, X; 0~~~~~~~~~~~~~~~~~~~~S

.
.t:

:= --

f -

f . D6d .tD r r

;0~~~c-un0000.;:0-0--0I6.:

.
:

D~~

'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 bureaucray'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 independent policymaking. This was boosted in 1961, when army General
Park Chung Hee seized power and reorganized the govemment to promote 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 political pressure. Indonesia and Thailand, where bureaucratic traditions
more dosely resemble thse in other developing economies, are instructive on this score.

Inflation
Control
and DebtManagement
in Thailand
andlIndonesia
For historical reasons, Thailand and Indonesia have been proroundly 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 monarchy's concern with avoiding indebtedness to foreigners. And a third,
-170-

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

and perhaps the most important, was the establishment of a bureaucracy 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 administration and policymaking.But it was-not until 1957, when Gcncral Sarit
took over the reins of government, chat bureaucratic reforms were introduced to address macroeconomic imbalances.Since then, keeping
inflation low and efficientdymanaging debt have been among the government'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 deteriorating infratructure. Suharto'sNew Order govemment ascendedto
power with an explicit conmitment to fight inflation and develop infratructure, particularlyin the rural areas.This led to passageof a balanced 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 Ministryabout expectedrevenues.It then deternines together with the Centtral Bank how much deficit financingthe economy can toleratewithout
increasing inflation. Having detcrniined aggregateallowable expenditure, 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'sproposals usually survivewith few alterations.After approval by the commnittee,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). Parliamentary rules offer a fiirther unusual guarantee against runaway deficit
spending nonrninisrerial membe of parliament on the budget commuitte may propose-only minimial changes and in any event may nor revise 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 adwinced 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 6inancial policy: In 1967, the legislature approve-da law limiting expendiares to domestic. revenues plus foreign assistnce. Since then, he
finance ministry has institutionalized a review that requires each ministry to justify prposed expenditures on a line-by-line bAsis. In addispend ~ing
nminisr
of priaet
onIh
ugtc

Iemer

Figure4.2 ThailandConsumerPrice Index

(percentage
changefrom previous year)
Percentagchange
30 --

20--

k.ii----

10 -i'

4j
If

1955 1958 1961 1964 1867 1970

&unr Innualtional MoneratyFund da.el

r72

1973 1976

1979 1952 1955585

S~~~~~~~~~~~~~~~~~~~~~~

ST

dion, parliamentary rules restrict the legislature'sdiscussionof thebudget 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 remained low and stable (seefigure4.3).

Indirect
lusulation
Mechanisns
SomeHPAEshaveenhancedbureaucraticinsulationindiiecdyand generally inadvertently.For example,wealth-sharingmechanismsreducethe
potential gain to interestgroups of intervening in policymakingand implementation,sincesomeshareof the firuitsof growthhas alreadybeen assuled. In addition, when wealth-sharingmechanisms reducethe size of
oroductionunits, aswithland reformand support for smalland mediumsize 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

*.

Figure4.3 Indoneia Consumer

PnceIndex
(perentage changefim previous year)
Percentagechange
-

rv--

--

30

-'-N-

20 -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

0-

; I

0~~~~~~~
~

-~01970
-5400

~~~~~~~~~~~I
1
j

~~~~~~
f

1972
l- :n eItional

197, 197
0
;our.neranaa

1974

1976

197t1

h
Monc4y Fund daa

"SO

1982

lS84

1986
.

2.97 197
29:
=8
Mocx Fun darn00. ,

196
;8

1.98S
_

819

1990
_'.

.98

-17_

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 stOppages and other disruptions and fiom lobbying the government for mandated 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 bureaucracyan 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 regulate. These ministries enjoy less power and prestige than those that resist
politicization. 5 Even well-intentioned bureaucrats may become counterproductive or lose sight of the goals they are supposed to pursue if bureaucratic 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 sufficient-in the long term. To sustain growth, a bureaucracy must have the
competence to formulatEeffectivepolicies and the integrity coiimplement them fairly.The more policymakers attempt to fine tune the economy, 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 bureaucraaes. 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 industrialized economies.
Despite the stress on culture in much popular writng about the
Asian miracle, the most successflil bureaucracies have not relied on culture alone. HPAEbureaucracies have employed numero us mechanisms to
increase the appeal of a public service career, thereby heightening competition 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-Based
Recnuibnent
andPmmotion.
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 averagerate of successwas 2.7 percent (Kim 1988). Promotion within the
bureaucracyis basedon a combination of seniorityand a host of performnanceindicators that dif&r across the ministries. Becausethe number
of personnel in the bureaucracyis fixed by law, compention for promotion 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 followed SUiL Most institute researchershave a PhD., many from a respectedforeignuniversity(im and Leipziger1993). LikeKorea,Taiwan,
primarilyto offsetweaknessesin
China, recruits heavilyfrom academnia,
the civilserviceexamsystem.Academicsare recruitedfrom major universities,usuallyfor fixedperiods rangingfrom three to sixyears.
acion
In Malaysia,civilserviceexam resultsaresubjectto affirnmative
guidelinesmeant to increasethe number of Malaysin gaovement. Becausethis reducesthe pool of eligibleapplicants,it probably hinders development 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 significandy berter in the Four Tigers than in other economies. Relative
pay in Mvialaysia
and 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

L~~~

r'A)l4A'

trta

~d.

Wotsurp'sne4ingapre,uowhidii'widely'
JapanF.t anfiS Korca
I

alsomke^6

Japn,oteNationaul

a, dcieatem;
Pcrsnne

Koreabascsldsaelwrthan
W-pine

aflownrs

hlp

effor*;t
>Jto<>

icAuthmForityHo

thepulic sectorwforsentry-level,m_id-l

bu

he

;1 tNrnsnorecdRBErmwt
oinon
1 an
ontlybaejmlayofacputHministercvdiningapordranws d:o

bCS
'fie

sector

peroivd' tohv>

match

the

annuay

posi_-io

private
sur

dif-

salnre.otarein
lare pivaslrbed
copnisto 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$1,390)
s
whl .. a n
Japan
ndKreaalo
ak
dliert
effort tomathte
privaste
s
seconn Jeapan th@eNationalt Per,.sonnel
kAu*thoit
annuallyS*,-,{F.4
surves
thu
dso
o
mpanis toderin :job
pay rates$(Kimo1988 p.) In
Koea
bs
p;saries
arcloer
than
saares
in largepmays
pr,;iFitvate
cmpaies.x
gin's mtost compentsan uaprihtabreacracyb
itasn
iburacr
ofisalar plusthly onuss
te riae sector amnd
alrypls a llowncsin
b7estT
basensalayoaflmister
iSngorerang

-~~ ~ ~
176-go'

~~~~~ne

in totasrnly compnsatione appear to biel smlerie

otcmern

pgtbracaqspy

tOhane
t

in most

ueurt

1NSTwt:ANWpNS

Table4.4 ICPEstimates
of PerCapitaGDPandRaftioof Publicto PrivateSectorSalaries,1992
(penent)

';f - i ; *;
apo*r;

n4:b

:Korea,Re.. of
;

-.

Ta aM China

98.

40
47.G.6-;
0 f;
.

Thailand.
- .Philipie

60.3
33

*.;

0:--2i7.50
.:

i277

GDPpacai

'Mddee

B~A

114.

,':o-

;65.2

0LMa1-aysia

Seir leve

HPA&; &ononq/regran
,.f.4i.
-a),

L2r;4:;.3

_B-A

;Ennyluas

1150
-

107-'.
7.
.

8 .3
63.5343
-;33.6
. -

.-43.0

25.0

325

822

83.6
60.0

65.8

920

-710
7,954

-5001
5oo

4418.0

62.

232

A-

;Tind-anToa,
.

63.5b -

,3

SnbSsjz70Afia
~~2 tmgu;
;~aLua

2416
;.

lAor

76'9
-

24
35

2.

;wCili

63.5
-

Somi- a

-LEsciae
a.

_.

779

3 1P
25.0-

11.2=, -'

2
-

'

average compdranontn

y, and other 2civa

--

122r-

'.

i~

iconomige
u

te

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
least sell
regarded among HPAES.
In economies where public sector wages are at least decent, prestige,
job:seciny,
c cll -=AE-s.
and other
advantages
of public employment will persuade
cas
reare
amn
H
se

6-g, :---06
60.0

007
3 7.1

toth UN4.
Ineraoa ComparZionPr e. ,. reth to i- ernational
indude alloaneos and boiisetDtarnibrpriarteseor comnpmsdoui ae ased

smalrilab,
on

8,51 ;0

oDfpu -

plymn

wil: peruad

;.

in
asurveyof

impranis with500or more;-

(o

-::

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 preremainsa powoccupationwith 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. Meritbased screening with economic incentives to attract a large pool of
applicantscan produce high status for civilservants in any society.Provided that the exams are sufficientlycompetitive,individualswho pass
them will be relat-ivelyrare. Thus passin the examnor winning a competitive divi servicepromotion becomessimilar to being inducted into
is special. 'While
a hail of fame; it signalsthe public that an indLividual
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 forward to a retirementpension, a benefit not normallyavailablein the private sector exceptin large corporations.
bureaucracy
AClearPathtothe Top.A final requirementfor a successfuil
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
fiom the pubfic sector to the privatesector and back
in other economides
again, descent from Heaven is a one-way tp. Moreover, rethringbureaucrats in Japan do not choose their sinecuresbut are assignedthem
178

0~

~ ~~

~A

~~~~

IN.ST-W

A NijN

,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

~~~~~

B The Iu1e.f

NXA
&A,

nL~

iR:ACkAUnENrTO` A BUW-

: IIIGHLYCOMPErITJV& MESBASED
-auaa
,

3'-

fieai

E6~~~at

ftheel

-cr

Dbutmntiad-pr
MouShaMW no

faos

witha'L,
|

cotxmpedarive

l mana

is n
thcaw
re

levimnmcnt
e iteia e

Whil there is no deyn

-.

to

:2O

idii vatue

beisreweeninstrvies,i

firns

welcome

tha firs

elom

denying
whohatv

Be--'-4

6n,isui,

i`

Whl

noiv;Enr

~
1,4u2
~

faos

-~~~~~~~~~~~ruCG-2~eP

iWZIi&
rbiidMdZl2&

int.scl
except

the

,1
unnv,zes;2ix cra~tcsanzespnt de4Ff*

ra;cmtomsfo:grslcne

roc=on

r-

Ilentrusns%lenlowingimem-w.?-

creaeswuciors

vy

etredseniceor

bu-

rtiedseiob

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

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

hoet
bya

havlue
c wull4ould
diquaif nyiapplicant,. in.rsocietiEes

acommite
wthatincthei

cminitrycThese faturs,

publicr

combined wihoanesy

179

i;RA ACLE

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

members have an incentiveto discouragecorrupt activities.This mechanism appears to be especiallycffectivein Korea, where tradition has
placed a high premium on goveniment service.
Strictly enforced dismissalpolicies also help reduce graft. In Singapore, 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 circumstances, 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 economies have an advantagein building honest and competent bureaucrades 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 applicants. 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 corruption and bribery do not occur in ithose HPAEswith reputable bureaucracies. 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 lowand middle-income economies. Indeed, most of dteHPAEs
scored significantly 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 providing 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.
Mmlh
Legaland ReguLaztoy
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
andMedinm-Size
Enterpriss Smalland medium-sizeen-

texprises
generally
havedifficultygeridngcreditto financecapitalimprove-ments, obtain better technology,and break into markets, particularly
export markets.Hence, a nontrivial SMEsector cannot be established without external assistance. Studies indicate that the northern-tier HPAEsJapan, Korea, and Taiwan, China-lhcv been able to develop relaively
effecdve financial supporc programs (Itoh and Urata 1993; Kim and Nugmt 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

HE HIGH-PERFORMING ASIAN ECONOMIES TEND TO HAVE

fbrmal instiutions that

communication and cooperadon between die private and public sectors, whereby rentsharing 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 cooperation of business elites.
fciitate

EastAsia'sDeliberationCouncils
Japans efforts to establish such institutions are the most widely recognized 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 represtentatives from the private sector-business, labor, consumers,academia,
and the press-to discusspolicy and marker trends and gencrallyto acchange 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 acplain the objectives of the new policy. Final approval by the diet is usually routinc. Without consensus from the relevant deliberative council,
however,a MITmpolicy stands little chance of success.9

Figre 4A Indusial WsionFonnuationProcess

Conductsurver

compile
data

r|

MMjuniorstaff

Hearing

studygroup

- Learned
individuals
- Interested parties
- Overseas erployees
- Local rpresentadves

Prepare draft

>

{eedbacM

- oUers

Research group
(Subcommittee)

(Outside lcturers)
Otdelcue;

(Briefings.subcomnittees' reports)

.~~~~~~~~~~~~~~

-|Deliberativecouncil
Publec r

(Report) ----

-*

Public Aions

Explanatorymngs

Lectures

-Others
S&urn-.Ono (1992).
:

'

-~~~~~

Ationr

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
andictop
businessleaders.Businessmenexpressed
theirviewsabout marets,regulations,and potentialplans.The president:
theninstructedspecific 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

and
cxperts topresenttheirviews.The focuswason informationgath- 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 infrastructueandgenerallyLeftthe 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 aslenterpriseshave 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

S~~~~~~~~~~~cni

consultative mechanisms, a determinat-ion reflected in PrimecMinister

taiv Bin
stmcFu
I
Mahathir
Mohamad's 'Look East Policy." Formal consultations
began in 199 1, when the government invited capt-ainsof industry to discuss the national budget with finance officials in a much publicized
budget dialogue. In thessame year Mahathir aeaaed the pinnace

conulttivstuctre,

of the

the Malaysian Business Council, which'i chaired

bycth pruie minister and includes more than sixty members from industry, 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. Othmermembers are top government officials, particularly from the economic agenaies,
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 members 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-resentatives fiom govenmen; business, and labor Becauseof this tripartite structumr, 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 citizenassociatiorLsto provide channels of communicanon with the government
Formal mechanisms for business-government interface are almost entirely 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 cenrualplanning plays a key role. These public, quasi-public, and partyowned firmsgivethe governmenta high degreeof control overlarge-scale
sensitiveinvestmentsand a means to influencethe generallysmallerfirms
that oomprisethe privatesector.Over the years,the largerole of the public 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 infbrmal 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 susceptible 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
suggeststhat 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's
Councils: 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

= - .

QSI.YMiI
R-AC L

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

reaucracy.Ironically,this not only strengthenedthe bureaucracyparticularly 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 foreignexchangeand overimportationthroughthe ForeignExchangeand Trade

Control Law.But intermationalpressure,combinedwith the need to
maintin an export-basedeconomy,led the governmentto graduallyliberalizethe rade and investmentregime.The processbeganin the mid1960sand continuedtill 1979,whenthe ForeignExchangeand Control
Lawwasdismandedand the ForeignCapitalIav liberalized.Of course
this meantthatMmTwaslosingkeycomponentsof administraiveguidance. By the late 1970s,guidancehad taken a distinctiely different
form-mergers of industies spicedwith financialsupport from the
government-and coerciondeclined.
and Korea.Malaysia'sexperiencewith
Ealaysia'sCouncls: mnulalingJapan
public-private cooperation through ddiberation councils has paricular relevancefor developing econqomies.
UnlikeJapan or Korea, Malaysia is a multiechnicsociety,and unlike Singaporeit has a eladvely largepopulation and
land area.Thus Malaysiais in many waysmore tpical ofa devloping economy 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 relarionship 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 ministries 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 increase the bureaucracy'sefficiency,eliminate corruption, and educate
civilservants on the value of cooperatingwith the privatesector.

*.

WhyDeberaton
Councils
Assist
Growth
From an economicstandpoint, deliberatve councilsfacilitate information transmission-They enable the bureaucracy to gadter infbrrnation 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. Selfinterested 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.Finally,becausethe rules that governan industry are effectively
esablished
- rwithin
the council, everymember is assuredthat the rules canolt be altered arbitrarily.With dear rules established,memberscan concentare
on market competiton and not worry about others tring to curry special &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.
:

. -

-. ,

Studies of the successfulAsian economies have fiequently emphasizedthe authoritarian nature of their leadership.Leadersof these soCd-

are
etiesdescribed
as benevolent
dict;ators
who,incontrast
0

0 X H -:-:

- - :.

to

mostother-

1-

I~~~~~~~~

- - -

~~~~CL
E

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

pursuit of development.This chapter ofreaucratsinto a single-mLinded

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
have to
growth. Whether by designor accident, the leadersof the HPAEs
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 fiuntre of their wealth.It introduces changesthat are likelyto open avenues
for fellow elitesand new competitors to contest control of their enomic 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 unres; unless its fruits are shard by the non-elites.

The taskof initiating and susmininggrowth thus presents a very difficult 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 operarionalizedte princple of shared growth. The NE in Malaysia,deliberation 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 bureaucracy,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 Singapore has helped coordinatewageriseswith increasingproductivity. In Japan, the establishment of the Japan
Produtctivity Center helpedi institutionalize labormanagement consutrations.
2. In fact, the smE.sectordeclinedonly when the HCI
push wasinitiated in the mid-1970s.
3. In those cases where the prime minister appoints
both vice ministers, he is constrained by custom to appoint a career bureaucratfrom the ministryto the post of
admninistrative
vic minister.
4. The debt-equityratio has beeninordinatelyhigh in
Korea. Official statisticsindicate that this is in the range
310-80, although the 'true' figuresare said to be lower,
at about 160-80 (Wade 1990). Nevertheless,this is still
high comparedwith the ratio in the United Statesand the
United Kcingdom,where the rangeis50-90.
5. Okimoto (1989) ranks the power and prestigeof
ministries as the inverse.of their degree of policicization.
* The heavilypoliticizedconstructionministryrankslowest
and the relativelynonpolicicizedMmu
the highest.
6. Senioritybearsheavilyon promotion decisions.It is

hurdledthe diEfassumedthat, onceone hassuccessfulfly

cult entrance exams,she or he is capableof doing (effiwhat is askediof her or him. This is
* ciendlyand eff-ectively)
not -to say that performance does not count. The pyramidlikestructure of the bureaucr-acy
guaranteessome c
ok oehrt
* greeofcmeiin
A neigds
ensure that its membersprosper.. Who becomesa bureau chief,a director-general,or ultimatelythe (administrative) vice minister is a source of intense competition
amongthe classesin a ministry"(Johnson 1982, p. 62).
..

8. The criteriaselectedare (a) the transpare-ncyof the

government,(b) the effectivenessof antitrust laws,(c) the
securityof propertyandtperson, (d) the faimirssin the administration of justice, (e) the extent of improper practices in the public sector, and (1) the extent to whiich
lobbying influences government decisionnmaking.The
rankinigsfor eachcriterion arebased on responsesof business executivesstationed in each of the economiesin the
survey to questionspertaining to the criteri-on.A Borda
scorewas computed for each country by summingup its
rankingsfor eachof the criteria.The overallrank score for
a group of economiesis the avenageof their Borda scores.
9. Industrialassociationshavebeen important participanesin the consultativeprocess.In many cases,theseassociations are empowered by law to represent their
sectors.T'hus,the businesscommunity has strong representationin a council.

10. Though dhe channels of communicationare still

operational,the democratizationof the polity has led to
somechillingof relations.The country is currentlyundergoingrapidpoliticalchange that has affectedthe businessgovemment:interface. There is a recognition in many
quarters, including business,fir the need to reestablish
good relations but under a new form of collaboration.
That form is likelyto reflecta more evenbalancebetween
the tWOSCectr.
I11.It is alsowidelyknown that the relativelyfeiwlarge
private businesseshavedloselinks to the party.
12. In fuct,inte'rviewsrevealedthat even.somemidsizc
firms haveestablishedrelationsof this type.

13.Johnson (1982) givesan interestingexample:"The

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
land on which to build elemnentary
schools.Whena. conthe Bank of Thailand and the BudgetBureau, havetheir
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 conthan in the rest of the public sector.
crete. The contractortook die cityto court, but the court
upheld the city" (p. 266).

189

-w~R

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 highperforming Asian economies (HPAEs) accumulated
both physical and human capital much more rapidly
and consistently than other economies and that accumulation 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 accumulation, financial savings, and investnent. For each area, explananions fill into three broad categories:
- Conditions that were not the direct result-of accumulation policy
: Policy fundamentals: areas in which nearly all governmenrs intervene, 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 resources 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 governgover9
ry
econon
fundam.nml
:~~~~~~~~~~~~~~~~~~~~~~~~~~g

ment spends on education, but the

spent their money more

wisely, emphasizing universal primnaryand, later, secondary' education.
Similarly, every governmen tcries to control inflation and secure property 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 HPAFsdeveloped, their government:s strengthened the legal and regulatory firameworks fir bond andt security markets to facilitace investment.
Besides fostering Favor0 ble general condidions and getting policy fundamenrals right, some HPABshave deliberately intervened in marker to
solve specific coordlination problems related to accumulation. These interventions aimecdat boosting savings and investment rates are the most
controversial of the HPAEs accumulation policies. Generally, interventions 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 industriaiigeconomieishave 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 neuh
tral or perhaps negative but not sufficiently negative to hinder accumulation 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.
HPAEs

Explaining
EastAsia'sHi'gh
Human
CapitalFonnation
XCEPTrFOR IN THAE AND, THE QUANTITYOF BASICEDUCATION

provided to boys and girls of school aehas been consistently

the HPAEs
tha in economiieswith similar levels of income. After having achieved universal primary schooling, thereby eliminating the gap between boys and girls at the primary level a decade or
.mren searle than most, the HPAEs
expanded secondary educaer
rapidly

iEtiigierin

192.-

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 transidions, and more equal income distributions. Each of these greatly increasedthe 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 allocation 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 education 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 percent, in Sub-SaharanAfrica. In Indonesiaand Malaysiathe annual earnings 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 demographicuansition,and the educationsectorwasa majorbeneficiary.
One important outcomeofa decelerationof the rate of populationgrowth

is a deline in the rateofgrowthof theschool-age

population.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 thepopulationup 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:
-- f-.-

-rcaRPf -

-i'o n

143r4o-

'4

'

p"t.of."

d'

1 Banglaesh.X'
...
1~~~~~~~~~~~~~~~~~~~6

:,..:a'
'

'

' ' ..
'C:

T-

1:

402

ong-Ko

iio:' 26--r:'0.7.

- _- 37

'194.

Brazl ; t,

-0,.,3'

0.2

'

'.4-4

'-.'43

'.47

''.,

0
2 9'

47s51_3
igrn.~~~~~~
46 ~~.
~~~~~~

35

. ...-...

9~11

* I94

120
-

4
r! 7 t

' "1-

STRATEGIES

FOR. R-A:P-D`

Accordingly,declining,stagnant, or slowlygrowing school-agepopulations have allowedsubstantial increasesin per pupil expenditure in all
of the HPAEs.
Table 5.2 shows the magnitudc of the savingson expenditure 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 Percentage
of CNPAllocated
to Education
Saved
Dueto Lower
Fertility
Rates
_f

;\;

X~-

Expendtre

dne torowthm;s#e
rhuol-aepopnltdon
tower an.

'.on

basic eduation: --ere

a prcenage
-a
= :of GNP:
.
fi!ny'

:Economyp- :.-;

Ho,gKon'
1975
- 7 t180-1

Mcio

#t

X.n

;1975 .
1. 98849
.
.*-

12

1.0

1.7

1.5

4.

38
8

3
3.9

~~~
04

~~.
04-

4.2
2.8

- --

w;~~~~~~~~
-..

2.0

:4.8
w

l9-;

1975

13
28

1988-89:

3.
16

* ''J1975

; .21
''i

-9 8

4t,'-,'^ .iNnc'

Fr.

, .

ofcul

:ijS"

<

t'' ''05i'!

1.

..

0~~
6
.

1980-81
~~~4:4
11
;1988-8 4.40

"_J

1.0

20 %8
:14
0 4'0
04

08

20

'.'

','

268

lan o. [Cur

on basic ed

0.3

'

an an asapccara2e

08
aof )

'

ErAsiznan4orhcrccononucs
inscihoolWsg
popadi)'n2u',_ -+,L{';
zpercencddference
s S * q ;~~r between
;2,-,;j:u;f
t
*, , 1;-th.iii
(cxpcn
mme
on
on .a;p ied ;oFGN)an cCai
=a
f
addal
itht
would havte
been rqe
h
p
gw
e
m
rp
s. A 0C-d
.
.
- t - r.
LTvuuut1accoarcstuiooup.
4 %Th%isanndImrrofd1Saviui
Thris
ra ndxue
n
asic dlihearo,usso(iouy",brn"~

':95

.~~~~~~~~C

of GNP on education instead of the 2.8 percent actuallyspent to achiieve

rates.Thissuggeststhatlowerfcrtilityratessaved
Korea'shighenrollment
school-agepopulahad Pakistan's
Korea2.8 percentof GNP.Conversely,
tiongrownat the relativelylowKoreanrate, thegovcrnmentof Pakistan
couldhaveincreasedenrollmentratesby as muchas 50 percent.

Equality
inDistributon
ofIncome
The moreequalan economy'sdistributionof income,thehigherpriof more
maryand secondaryenrollments tend to be. In a crosa-section
significantnegthereis a strongand statistically
than eightyeconomies,
ativecorrelationbetweenbasiceducationenrollmentratesand the level
of incomeinequality,asmeasuredby the Ginicoefficient(Clarke1992).
As we discussbelow,educationalexpansioncan havean equalizingimpact, so the causality in this relationship could run from enrollment

ratcsto the distributionof income.However,thereare a varietyof reasons 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 educanon), verypoor householdsare unableto investin theirchildreneven
whenthe returnsarehigh:the pressingneed to subsistcrowdsout highThis is lesslikelyto occurin EastAsiabecausethe
return investments.
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 basiceducationenrollmentsof 0.40, if the distributionof incomewasasequalin
Brazilasin Malaysia,enrollmentsamongpoorBrazilianchildrenwould
be roughly33 percent higher.Moreover,the incomeelasticityof demand for basicschoolingamongpoor familiesis likelyto be substantiallyhigherthan averageand mayexceed1.0.In this case,enrollments'
amongpoorBrazilianchildrenwouldbe 80 percenthigher. . . .
-96.- -

STRATEGIES

FORR-R

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 investments in education contribute significandtyto economicgrowth. But
neither of thesepoints necessarilyjustifiespublic provicionof education.
Indeed,
if the privatereturns art as high zs socialreturns on theseinvestments, individualsand householdsare likelyto make adequate (from an
economywidepoint of view) investmentsin human capitalon their own.
However,two kinds of problems generatea gap betweenprivate and sowhich-if not countered by governmentaction-will result
cialretumrns,
in

fnilies investing less in their children's education than is in society's

interesLFirst, theremay be filures in the capitalmarket and in information, 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 embodied human capital (as they can for other types of collateral).Evenpoor
&milieswho might be willingto borrow,becauseschoolinghas high priratcreturns, 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 immediate implicationsfor what type and levelof schooling the government
should subsidizeand for whit groups.There are severalsourcesof externalities.An educated person may increasenot only her household'sincome 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,coordinanon probleiims
ast, and governments acting on society's behalf have reason to choose a higher level of education than families acting alone.
I97

. -- MIRAC

LE

onInveshnent
Figure5.1 Returns
in Education
byLevel(Latest
Year)

Policy responses generally consist of making schooling availableand

freeand in some economies
compulsory.The differencebetweensocial
and private returns from educautonis probably higher at tle primary
and secondarylevels than at the university level (see figu 5.1). Many
positive spilloverscome from literacyacquired at lowerlevelsof schoolfhing,
while the retu:s from training at the universitylevel are almost

Ratreofrtuwm(percent)
Pri_

:Primary

c- ndia1

---

--

I-i

Kr

SondarJ

-Hi4her

05 ID2may L
0~~~'
5 10 15- 20 25 30 3!i
SocIalreturn U Pilvato
Social retrns
N Nvaf mt

*;

fuly! captured by te higher

i
income of university graduates. Vocational
training
also havc high social payoffs, if it improves worker productivity not only for the trainee but also for her co-workers. More importantly, evidence suggests that vocational training is most

cost-effectiveif trainees have a solid baseof primary and secondaryedu-

Soww Pswchaopoulos(1993).

cation. All of this argues for universalprimary arid broadly based secondary education as a means to improve economic efficiency and
income distribution.
Table5.3 PublicExpend4tnre
on
Education
as a Percentage
of GNP
;E con>msrnyI +.
if

5 .-

-1960 196.i9

oegg
:~5

Policiesthat PromotedHumanCapitalFormation

f Higher shares of national income devoted to education cannot

explain the larger accumulation ofhuman capitl in the HwAES.In both

;1960
and 1989, public expenditure on education as a percenrage of GNP

Kong2.8
2.0 p.oE
Korc
]46
-2.8
34
'.97~
pV W -

;:,--Smngipoin
?g1ayi
T

23.",
2z3

was not much higher in EastAsia than elsewhere(seetable 5.3). In 1960

the share was 2.2 percent for all developingeconomies,2.4 percent for
Sub-Saharan Africa, and 2.5 percent for East Askia During the three

'.3]9.2.

J.'x',tfg>R;

z lndocs

decadesthat followed, the governmentsof East Asia markedlyincreased

0.9

ovr

*~~~~~~~'

*;*ZAvctnX

;.sc

2i.5{

,.1

Si

---

PakfisraaV$1'I
-- ttdc4cs:oPcdr
;
;

-6'

d0Zn~

0- - ~

SaharanAfrica, 4.1 percent, was higher than the East Asian share, 3.7

-l

percent, which barely exceeded the averge share for al developing

economies,3.6 percent8
3.1S'St;

.7.'.

-. Q:-

wer initial conditions, for example the colonial legarcy,decisive.

\

9'n

5|;

>Nor

econom

oFcDnms'
j^['.''Lli{,.rM<developing

-ourc
esdaral

fl

V's e Korea did have much hier

W

41/

wa1

.-

...

enrollment mtes in 1950

thn

most

,ies
subsequent increases in primary and seconday

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

;'
Ni

t>

secondary enrollment

rate. By contrast, Pakistan's enrolment

rates were

52 percent at the primary level and 19 percent at the secondarylevel.

What proportionof thesegapsis due to initialconditions?At the pri-

.98
--Ef g

in Sub

;t

Et

governments
'
'_t_AX'l
in*'ii;iother developing regions. In 1989 the

.-- J 26{bSs
1

nts- ogVIg
f! Eas.............t2T-J

share
ntheof national output they invested in formal educaion, but so

;.

;;rg
sa
1;9

fiuly

-..

-.

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 secondaryschooling, Pakistan'senrollment rate in 1950 wasactuallyhigher than Indonesias; 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 expenditure on educationallocated to basic education has been consistently
higher in EastAsia than clsewherm:Koreaand Venezuelaprovide an extreme 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 education 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

~~~~~~~~~*

s -- g.

; r ' ; - on

~-

-*

* t--

... i

--.

--

; L;;te>ni8s.
on'! .- .t

- '
cn.

eoncain
as

..

*W-

2-~.

;E~ A:LdcaoonzyGW
Singapore
5.0

-.Taln
, eniag
eu4oi
.t3.ofthe
.

. t-.-

~tini985-8
..-

.,Lltf

.0

10.3,

0f

14.6~

749

Siid,cinollnt
bsc ducanon
i:980. .
30.7:
.
4
-

;E-310
.2' a9d4.3itwfl95.46,iiuhatrthe

e-3

buidgealarcmbic?Iuconwns_81
2..10

dc: ng

St.adalu`p*
e
P-d

'

o 32,N-

't

-1 '-

",

..

-'~~~~~~~I"
7
t
L' a6E
-

..

72..

Mn
. ,9;
iKoeaRepo
0
1vIalay~~~~~ia
79

.4.

cio*

;- educanonsit

-Indones23:

'F.

-atv.Sosc

it
--

-c-

per

m 9S48; and80,

Mg(*

.8.

'I

--

W5Ofli

Qu-.

'99.

in Korea (3.0). After subtracting the share going to higher education,

however,public expenditur availablefor basiceducation as a percentage 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 reflected 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 demand 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 middleincome economiesgovernmentsubsidiesof universityeducation are nor
relatedto need, implyingthat they benefit fimilieswi 'hrelativelyhigh incomesthat could aflordto pay feescloserto the actual costof schooling.
At the same time, in many economies,Braziland Kenya being notable 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
Vocational
Twining.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. Highlevelskills are essentialfor such manufacturing-relatedaaiviries as mnanagement 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

-Wl

UsN

Ed
n :p quesm--lSZ

Mud
gg

FO.R. k-Ai

31

ou

to

.,H,

PC-=
Pff

citi0ii"bf-

propoipomp
Y,
$61

-5-i
-r. n.
ZOM
ulca'don- ud et,'
00
pq*"
90
r
w=ntpmbamrucauonz,-,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

j'an
jopPrO
7

o non,

liEfic,ar-kniv

iham
P C

ZhQ?
:ax

gr

.11;2pcm
"ea

):PM

C=O

1987'
;",O

m 's,

caiiNwduj

Li

&-Co

zm
owcvcr
Lti

SCIOLOO

201

Few studies have been made of training in the HPAESor ocher developing economies, in part because of the difficulty of defining tamining.
One lage survey of 48,00 0 manufliLccring firms in Taiwan, China, conduded 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 Taiwan, 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.Enterprise and preemployment training have produced social rates of reTur
of 20 percent in Maaysia. In-plant taining for welders in Korean shipbuilding had a socal rate of return of 28 percent, higber than in nonfirm. 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 compleAt the same dime, businesses must stand
menary privare invesctnents.
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 augmented by emnomic growth and consequent job creation and are therefore higher in the HPAEs.Moreover, the HPAES'extensiveparticipation in
international markets aided the success of their training pmgrams. Intremational 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 educatdon 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 primaryeducation 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. Government policy also encouraged (and sometmes compelled) increased savings 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 measures designed specifically to boost savings rates. Thcse range from policy fundamentals (such as rguatory supevision of bnks, winch
203

0j

IRACLE

addressan absenceof depositinsurancemarkers),to targetedinterventions (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-savings
relationshipin a broadcrosssectionofeconomies
indicatethat whileincomeand savingsgrowthare highlycorrelated,incomesoften 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.

relationship
We haveattemptedour own testsof the savings-growth
for the HPAEs
(describedin greaterdetailin aDpendix5.1 and its correspondingtableA5.2).Incomegrowthhas beena remarkablygoodpredictorof 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 significant fctor in the speccacularrise in savingsrates in the 1970s and
1980s,consistentwiththe viewthat demographicfiactorsand, to a lesser
extent, the policiesof the CentralProvidentFund, a forcedsavingsprogramn,
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 consumption 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 popu204

STRATEGIES

FOR

At

lation agesand the old age dependencyratio rises,an effectseen in studies of European economies and Japan (see Blumenthal 1970; Horioka
1990).
For the HPAES other than Japan, there is litle evidence of the relationship 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 attributable 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 increasesavings. 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 management of macroeconomic policy and the making of budgetary
choices.
TheRoleofRealInterest
Rates.Foremost amongmacroeconomicfmndamentals was the effort to control inflation discussedin chapter 3. Because high inflation rates also tend to be volatile, real interest rates are
often negativeand unpredictable, discouragng domestic financial savings. Moreover,since unanticipared high inflation erxdes the ral value
offinancial assets,volatilityof inflation ,increases
the riskassociatedwitth
holding,them. Conversely,low to moderate inflation, partculady astable rates, encouragesfinancial savings.Thus the HPAEs' generally superior record at avoiding long-term fiscaldeficitsand restaiing inflation
gavepeople the-confidenceto save (Neal 1990).
Table 5.5 documents average real interest rats in the uAEscompared with other low- and middle-incomeeconomies.Real interest rares
on depositsin the HPAEswere either positiveor nildly negative.In con2.05

MC LE

Table5.5 Average
RealIntkeresRatesonDeposits,
Selected
Economies
-

*.

MJPA&
AIcegtsx

r~~~~~~~~~~~~~~nwiu&wrd

~~~~~~1973-2.

'

* Tan.Iiq,~~~~~d

1
-

Taiwan C.
ThIml

~~~~~~~2.43
.8
4.4197

1177-got

~~~~1976-92

Naqj.dab
NcpI

04

i
'-

.'3

55

~~~~~J314

1935-91

Un4m7

EWF

-.

~~~~~~1976-9

-.

Am.-

'.

'*aIa~~~~~~~~17

-63m-

83
-~~~~~~~~~~~521

144

-243
. -:

:I!-:

;w

334~~19

-. 21033j
9

12793-q

Zsra ore

G2aeret

mAi

average

~~~~~~~~~~LA9
Ism,

~~~1976-1-5
At.tqc

94.9

~~~~

3976-9'

Taft,1

2o6~~~~~~~~~~-l

.2.47

1791443$3'.4

M-ils

uas

7*

~-3.6

117449

177

19,7-91
94-3

',

AS3

1973-90

* MmhI&

-.

-tJ

1Sgonsshw
-5an ecnmc*it1agIeadertso
ae

er

1.

13.1

nZaba.Ra

intercst
109

ra4e

prciinTASseS

weeas'
3

an

evlOS
3

as

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 percent 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 stabilizing 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 addressthe 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 increaseconsumption (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 expenditures is more effectivethan raising taxation. For a sample of thirteen

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 governments 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 ecxceptions, faling in the middle of the distribution. In Malaysiathe high
levelsof public consumption reflectedthe explicit redistribucionalobjectivesof 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 dedined from an averageof 10.4 percent of GDP in 1980-83 to 1.4 percent 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

-~~~~~~~~~~~~~~~~~~~-

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

;
i4o6

J-apa,,*
')0955-70
z |

'

i . R:|?8
~~191480
6

0 17.2t^6.2

4.6

-19714=0
0

312
.2;X

f.;::

-198190.

'2011
0f:

103

187';
15
191i

Singapore.
*.

-4;---o
1981-90'b
1980-B5

1-ft
.--

14.3

-j198

.3-. -- ->;-i24.0
;:'

14

8.6.

.-

Oird deSpigeconemi
-.
-Argendna~S.Kw'lii
-- ;--.1980-85* .....
.

s_
-'
_

,t''-'.''

. 5

I4.2

-19804871,

1980-8.
:: 985647
:
j-

235
1980-84

3.9

*'~~~~ 1980-84~~
*19858

1985~~~~-187
Csti Rica198-8

-~~~

P'ia9i5-8704

.2IO:
~~

.s

'1503

'.--'-,0.;;'--.

2~~~~~~1.6
1364

7.9

06

7.
7.

161

61",

'Philippines~~~~~~~~~~~~~~~~~~J
i.o,9.O
I9
.4-.
,.

,.,

06

1690

1.
97-

'.=-'

I83

STPTRATEGIESFORJLAiPPD&

Coorditon Problems
thatHinder
Savings
Savringsare afficted by severaltypes.of coordinationproblemsinducedbyfeaturesof financialmarkets.Below,webrieflydescribeeachof
thesecoordinationproblemsand the policyresponsesof HPAEgovern* ---ments+The rangeofthe HPAES'
policrresponsesto theses-rual or incipmarker fhiluresrange fiom standard remediesthat the HPAEsmay
...haveimplemented better, such as the prudential regulationof banks, to
-ient

more acavLst interventions, such as-limitng competition or regulating

tinerestrare spreads.
Abse--.,nf
Deposit-Insurance
People will only deposit funds with a

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 Indeed,-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-moresuspec 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 institudons, 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.
:High
rnsacti.CostsPrivate banks seek large-depositsand those tharare 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 lewlsThe: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 increasesavings.

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 depositors from bank defaults. In general, the developing HPAEs have performed 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 institutions.1 2 Sometimes this has been done at the expenseof long-nm
efficiency.
When necessary,all EPAR govenments have bailed out troubled financial 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.Indeed, 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 estate speculation boom in Hong Kongi Japan, Korea, Malaysia, and
Thailand, havegenerallybeen brief and modest, particularlyin comparison with the protracted and disruptivebanking crisesin Bolivia,Chile,
Ghana, the Philippines,and Turkey during the same period.
toProaneBank
Solvency.Central banks and departmentsof
Regu;llions
appearto have been generallymoresuccessfulin su'financein the HPAEs
pervising commercal banks, which have reported a rdatively low proportion 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,nonthan in many
performingloans have been lessof a problem in the HPAEs
developingeconomics.For example, one study of twelveLatin American economiesfound that nonperformingloans accountedfor nearly 14
percent of total loans in 1987, while the proportion of bad loans surpassed 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 apparent in the easewith which most of them adopted the international
capital adequacyrequirementsset by the Bank for International Settlements (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 difficulty achieving BIS capital requirement levels because it introduced
proper prudential regulationsrather late.1 3
standar4
By conast, in Lain Americaonly Ecuador adheredto BEIS
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 collateral requir~ments&
To reduce the risk of borrowerdefaults,regulators
-encuragedbanks to requiresubstantialcollateral.But becausethe collatera. 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 collat2.13

LE

eral and sought to reduce loan concentration by limiting lending to related parties.Enforcemnenr
has been difficult,however,becausecompanies
are doselyheldland dlisdosurerulesare geCnerally
weak.Moreoverinm
severl 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 traditional 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 interaction between city bankcsand the supervisoryauthorities, coupledwith
its tight control of branch licensingi to encourage prudent behavior.
Malaysia'sbank supervisorsoperate in a similar flhshion.
Coping
witfDistessedFinancial
Iustitutioas.
In spite of generaly good
pmudentialregulation, there have been financial crises in Eas Asia
brought about by a mix of macroeconomicshocks,insufficientprudential regulation,and seculative lending (seetable A5.5, at the endtof thechapter). However,goverrnmentsrapid and effTctivcresponsehas been
able to maintain confidenceby acting as implicit insurerwithout excessivefiscalcosts. Rapid growth and spiraling real estate priceshave often
tempted financial instittions into property and sometimesstock speculation. 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 somewhatdifkirenrreasons, among them the heavyindebtednessof firmnsthat had been beneficiariesof 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 KoreaandtTabWn, 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.
Puntein ofBanksfromCompedfioo.
Nearlyallgovernments regulatethe

acation of new financialinstitutionsand the entryof foreignfinancial

institutionsto try to ensurethar newcomersare solventand willremain
so. But manyEast Asiangovernmentshavegonefiirrher,limiting competition and protecting existingbanks both fiom foreign banks and
from new domesticcompetitors.With the notableexceptionsof Hong
Kong,Indonesia,and Singaporewhichwelcomethe openingofnew financialinsritutionsprovidedthey meet capializationand other regulatory requirements,most HPAEs
have had rigid restricions on entry.
When expansionhas been pennitted,it has mostlybeen through the licensingof new branchesof exsting banks.Despitelimitson enny, rhe
bankingsectorsin the HPAEs
areno moreconcentratedthan in otier developingeconomiesor OECDeconomies(seetableA5.6, at the end of
the chapter).This means that bankingsysremsdifferlitde in structure
among economies,although the onrginsof their relativelyhigh internationaldegreeof concentrationmaydiffer,as we shalldiscussbelow.
Why were entry restrictionsusedrather than high net worth standards to encourageprudent behavior?Entry restrictionsgivegovernmentsa morediscretio,-. y tool to influencethe behaviorof banksthan
net worth requiremnentr
' tovernmentsmayexercisediscretionin deciding 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 increasedfinancial savings?The answeris ambiguous.Larger, more profitable
bankshavemoreto loseand so a greaterincentiveto maintaintheirrep..

.~~~~~~~~~~~~~~~~~~1

utations by prudent lendling. Moreover, restricting competition mncreasesprofits, and higher profits increasethe financialstrengthiof the
banking system (provided banks do not distribute profits to shareholders).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,preventing monopolistic practices that squeezedlients, reduce efficiency
and innovation,and hence, lowersavings.'
Therefore,whethe-rcontrolling competition results in &avorablcor
unfaivorable
outcomesfor savingsdepe-ndson whetherthe increasedsolvency comes at the expenseof efficiency.The evidenceamong developed economiessupports the view that financial systems with fewer,
larger banks outperform systems with many smaler banks (Vitrs
econoes financialmarketsare domi1991). But in many developening
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,
of the presare regardedas efficient,despite enty restictions,
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 substantialy lower than othecrdevelopig economics notably Liberiof Sri
Lanka, and Turkeysand selectedeconomies in Latin America (Hanson
Be
p
rtn
cot(asofprcetg
tt t)of
- = ner ~~~~~~~~~~~~~~~~~spo
and Rocha 1986). Conversely,studiesof the bankingsystemin Indonesia and Koreafind that protecion has led to lessianovautvepracbtces(nternal World Bank reports).
Thus the evidencedoes not condusivelysuprt the efficacyopl
acesto imitentry.CWhereted,encies toward large nancial institutions
z16.

STRATEGIES

are strong, govrnment

FOR

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 consequendy the spreadsof financialinstitutions.If spreadsare controlled,
bankswillearn limitedrentsdespiceprotectionfromcompetiion. In genaal, 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 comparable 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
grenerated
some rentsfor banks,a combinationofcompettion among incumbents,regulationof spreads,and evenntalliberalizationof entry reFwue 5.3 Spreads in Financiallnternediatioi, 1979-89
P-cntge

pentspread

14~~~~~~~~~~~~~

' ---

. \.
- 0-

;
,

'

fl---l

~~

---

-/-

~~~~~~~~~~i

,-m'-.

.-

1-

;(t,~ . K

0~~

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

stricdions
madebanksin East
Asiarlaivelyefficientcomparedwidtthose
in otherlow-and middle-income
economies.
Regulatedspreadsare not alwaysbinding,however,becausebanks
can circumventthem by chargingfeesor requiringcompensatingbalances.Banksin Japan and Korea,for example,chargedcompensating
balances.Thereis someevidencethat thesepracticesweresufficientto
circumventthe regulationsin Japan (Ueda 1992).Intermediationincomeofcommercial
banksin Koreahasbeenhigherthanthesimpledifferencebetweenlendingand depositrates.In contrast,regulationson
dcpositand lendingraceshavebeen bindingin the state-ownedcom-mercialbanksof Indonesiaand Taiwan,China. In Thailand,commercial bank lendingratestended to bunch aroundthe ceilingrate, and
interestratespreadsincreasedslighdyafterthe mid-1980sdueto a drop
in depositratesand the relatively
unchangedlendingrateceilings(aninternalWorldBankreport).
prudentialsuCreatng
PostalSaiugslnsoboansIn additionto effective
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
the region's
firstpostalsavin progamin 1875,with
the explicitgoalof fosteringsavingsof rual dwelersand peoplewithlow
to moderateincomesin thecitiesand towns.Undlthen,suchpeoplewere
ssten, whichlackedruralnetworks
effectively
exdudedfromthe financial
or
smalldepositors
byrequiringhighminimumbalances
anddiscouraged
payingverylowinterestraceson smalldeposits.TheJapanesegovernment
andmade
households
heavily
promotedpostalsavingamonglow-income
the interestincomeon smallpostalsavin depositstaxfree.
Similarinstitutionswith thesamesavingsmobilizationgoalswereestablishedin Korea,Malaysia,Singapore,and Taiwan,China. Like
Japan,both Koreaand Taiwan,China,havegrantedtx- exemptstatus
to the interestincomefrompostalsavingsduring longperiods.In Taiwan,China,postalsavingsofficesaccountfor aboura thirdof allfinancialinstitutionoffices,and the postalsavingsservicehaslongerbusiness
In Malaysia
hoursthanotherfinancialinstitutions(Shea,fbrchconiing).
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 deposittaking branch network.
Postalsavingssystemscan be an effectiveway of mobilizinghousehold savings,provided that governmentstake care thar the cost of administerng the systemsdoes nor outstrip the benefits.Geneally postal
savingssystemspiggybackon the mail-deliveryinfrastrucumre,
thus minining overheadand fixedcosts. In the earlyyears in Japan, the system
often operated out of the houses of wealthylandownersthat were alreadyservingas postalbranches.As postalsavingsexpandedand the administrative 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 consumption. Becausethese measures consttute more active interventions in
markers,their efficacy,even in East Asia,is open to question. Restrictons on consumer choices, induding the basic consumption-savings
decision, have welfare costs. Moreover, similar efforts in other
economieshave been spectacularfailures.Examplesindude the Widespreaddeprivationand 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 subsdtuteforvoluntary 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 GDPduring-the 1970s and 1980s(MonetaryAuthority of Singapore 1991).
.219

jILA~~~CLE

Thelackofconsumeraedit to purase housing,constmerdurables,

savings
household
andotherconsumeritemsmayhaveinducedincreased
in somcEastAsianeconomies.Bankregulatorsin Japan,Korea,and Taiwan, China, rsriaetd credit availablefor consumerdurablespurchases.

Maki(1993)offersevidencethat the rapidincrasein savingsin Japan

dumables,
afterWorldWarII wasdrivenby theneedto acquireconsumer
and he showsthesamepatternexistedin Koreaand Taiwan,China.As
inacased,savdumbles
incomesandthedemandfor consumcr
household
ingsas a proportionof incomeroscrapidly.Oncetheexcessdemandfor
andevendeclined.
wasmemsavingsratessmbilized
consumerdumbles

Savings?
HavePordesIncreased
savings?
werepoliciesin theHPAEsaimedar increasing
Howeffective
accountfor
growthconditionsand goodpolicyfundamentals
Favorable
Maoeconomicstamostof East Asiassuperiorsavingsperformance.
bilitycombinedwithsolidprudenrialregulationof financialinstitutions
createdan environmentin which savingsweresecureand retainedtheir
value.Limitsto competitionamong banks and the willinpess of govenments to rescuefinancialinstitutionsin troublehad costs-reduced

in theformercaseandfiscalcostsin thelatter-but thesewere

efficiency
for bytheincreasein savings
moderateandmayhavebeencompensated
hadthe virtue
bypublicconfidencein banks.Postalsavings
engendered
for smallsavers.Govinstruments
of beingbothsecureand accessible
and
ernmentsin Japan,Korea,and Taiwan,China,aswel asMalaysia
tappinga nontradiSingapore,operatedtheseinstimtionsefficiently,
tive coss
tionalsourceof financialsavingsat rdativelylowadminisa
consumersto savemorethantheywould
Havemeasurethatrequired
Theansweris ambiguous.
economicwvelfare?
havevoluntarilyincreased
creditraisedsavingsratestempoarilyinJapan
Restictionson consumer
but at a cost in termsof consumerwdfar Forcedsavingsthrough pensavingratesin
sionschemeshavenot beena majorficaorin increasing
arean imporbutpublicsavings
Eassia, exceptperhapsin Singapore,
Tojusr insavings.
sourceof highaggregare
tantand complementary
trening in households'savingsdeasions,socialreturnson the incread
savingsmust exceedthe privatcoppomruitycos detenined by the rate

Thiswouldbe thecase,forexample,iftotal ficofreturnon investment.

componentof
growthwerepositiveand a significant
tor productivity
orof
overallgrowth,eitherasa onsquenceofdynamicscaleeconomies
220

STRATEGIES

cxtnalities associatedwith investment.The social return also must acceedthe private returnto jusfyr interveningin households'savingsdecisions. 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,however,evidencesuggeststhat aggregae total fctor productivitygrowth is
negativeand rates of return to capitalare low (Young1992). Singapore
may havecompelledits consumersto savetoo much.

Rates
Explaining
EastAsia'sHighInvestment

JNE
OF THE MOST REMARKABLEASPECTS OF THE HPAEs IS

their unusuallyhigh rates of privateinvestment.Privateinvestaveragedabout 7 percentagepoints of GDP

ment in the IIPAEs
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 economic growth. Growth,savings,and investmentinteracted in a viruous
cirde as high investment initiallyspured growth, which resultedin increasedsavingsto support continued high rates of invesnnent. In additon, 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 overlooked. 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 important 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

FOR RAPD
X

i.*

CLE

matesthat infrastuctureinvestmenthas a high payofffor economic

havcbccnhighlycomplenenaryto private
growth.Theseinvestments
manu&cauring
activities.
in export-oriented
investment,
especially
for
development
Table5.7showssomeindicationof infrastructural
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 number of telephones per person roughly quadmrpled,while Brazil,Mauri-

Table5.7 Direct Provisionof infrastructure,SelectedYears

-- ta
;-Wr

FB!N
cenw
f.
Paetge of
roads

e,

w.c4

odaivr
telhone'

padca

Ec; rsonomy^
0-i- 1971- *1986

_.1975 -1985: g

RPAFs
-

.
100Hong-Kongs
^^.
coe.Rep.
-14 of
. $a1
'
.yuia
. 85
S ;-in

80
,'-. i

'= *r,=,

160.6-18.3
100
62.2
. .....
77
... . .
.2 542
332 . 5621--

Thailand 67
'

49:C-'

- : ; ....

-:11.6

Gn

3182.313t3;

396

--18.2,

,...;

12.4.

34W

muridu

174.2

12

5
6

13

7.9

~~48.0-

24

i95.-- -;309.0,

:..........

11'

98.6:

..528:

79.7

209

89.5

'352,

1 .7

21
-

167

'-20

935
3

,60.0

-hilippine. - - - :--

12 5

..
1P7o-mlj

66

FL

=2.2

16

India

Si
tnlt

242

2818

95 8

.0.0

-. C'Ivoi-e

'205

,~~~~~~~~~~~~~~~~~~~~--8
-:351'7
<

~~~->-:

M'-'co

4025

7~~~~~~~~~~~~~~~!

KU

W--

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 policiesdesignedspecificaly to address coordination problems that may hinder investment.
These problemsare derivedfiom shortcomingsin markets for information and in marketsfor sharing risk both types of problems exist in all
economies,but the problemsand the consequencestend to be worse in
informadevelopingeconomies.For example,such developing-economy
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 Similarly, 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 transformadons inherent in development,such asthe creationofnew firms and new
industriesand the absorptionof new technologies,heighten uncertainty.
HPAEresponsesto theseproblemscan be
Polciesto PmouteInmeshment
grouped into three broad categoriesFLrst,as theseeconomieshave matured,policynmakers
havegraduallyincreasedtheiratrteion to the creation
and improvenentof bond and equitymarkets.Hownrer,whilethesemarketsare 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 providefinancig to agriculmreand small-and medium-sizeenterprises.These cforts-appear,on balance,mthavebeen moresucccssfi4in the HPAEsthan in
otherdevelopingeconomies,for reasonswe shal exploreFmally,HPAE
governments have encourged investmentwitha wide arrayof other mechanismsdesignedto inacase the atrcdveness of privateinvestment..
The remainder of this chapter describesspecificpolicies in each of
these thiee categories-supporc for bond and equity markers, establish223

ment of decvelopment
banks,and mechanismsto increasetheattraativeness 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.Severalgovermecnts, 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 effortsto increasedomesticinvestmentby restrictingcapitalflows,Finally,
severalHPAEs,
againparticularlythe northierntier but alsoMalaysia,subsidizedinvestmentfinancingbyslightlysuppressingthe interestrateson
savingsdepo'sitsand corporateborrowing.This last mechanism,which
we term mild financialrepression,playsa k-eyrolein capitalallocaidon,
the subjectof the next chapter.
Bendamd
EqLitMarkets,
Developed
aftertheTakeoff.
Despitethe phenomenalgrowthofAsianco: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 example,the long-ternm
creditbanks govermnment-protected
monopolyon
the issueof long-ternm
debenturesdiscouragedthe developmentof bondi
8
markets.' TeTa
governmentcreatedthe samemonopolyfor its private long-termcredit banks.In addition,tb"~lack of governmentsecurities in most HPAEs,
a direct outgrowth of their successat minimizing
fiscaldeficits,has deprivedembryonicbond marketsof the most widely
utilize-drisk-freebenchmarkrate. This has impededcorporatebondtismsues,sincemarketsfind it difficultto determinesimultaneouslythe riskfree,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
Malajiula.
I

126-Si

Z'~~~.6

$245

$61

36.125

423

137

'I 0,W~3N

~~ 132-W

293

LI14

14

~~~~~311

.-

IjaV

~~3a~1(tnpI2p.frmr&uu)..'

-3.

971-1113

t1
26.0

1564-US'
-~~~~
4 1m971-$

.0

Ca.uu 3 .I97-Z5~~~~~

.7

'23j.-

47

80.

I L-)

7.7

I&U

LO

L32

--

'0i

10d

jan23.2.&)
!P
LuSti.
v~74;-

1.

23

'

T..wi

b.CaWuW.K
m.a1dua~~~~~~~~~u~~~y.,~~~
_:; immu'

Wu

w0.A,i~

Iua

-VNIIA

hl.1z9h2.~
a
hitoicAym

RamilkWm

been

B.ubumiawamuiEsu

small.Makfhzds

appeamrs

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 mrarks'19 Recognizing-the value -of a healthy bondmarket, Japan has
*eased lalwsthar discriminared agamnstcorporate bonds. Similarly-Thailand, a.nd-rec'endy.Indonesia. hav strengthened rating and reg ulatory
agendes, -while-Hong ,Kong has refornictl its regulatory-structures.:
rating-agency.Thus,_whlile.
Malaysiahas
introduced acrotebond
.

'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: development 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 capital formation in either developing or industrial economies (Stiglitz
1993a). Even so, as the East Asian economies have matred, govemments 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 requirements 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 capitalexcept 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 markets have been largely successfUl. The growth of stock marker capitalization 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 phlenomenon, and its timing-beginning long after most HPAEs'economic takeof3swere 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). Thcy3addressed the need for long-term credit for industry by creating development 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

'

velopment Bank ODB) and the private Industrial Bank of Japanaccounted for about two-thirds of loans outsmnding for equipment investment 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. Sarting 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 banking-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 development 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 capability oFfinancialinstimutionsto screen and monitor projects.
Development banks have performed much better in the HPAEs, especially those banks in the northern tier economies concentrating on industrial 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 commercial criteria in selecting and monitoring projects and firms, even
- bwihin
the constraints set by government prioriiy activities. Development 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 another mechanism when in the early 1960s it required borrowersreceiving fumdsfrom the U.S. Agency for International Development to put
-upmathing funds and made lending to a nonperformer a criminal offensei 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 chairmen, 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&cssionalismand 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,establishingdevelopment 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 investment 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 discussion of foiurhighly specificmechanisms usedtin some of the HrWAEs.
to encourageinvestment. low relativeprices for capitalgoods, risk slaring, 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

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 revenue because of extensive exemptions. Income from dividends and interest were tax-exempt until 1981, as were capital gains until 1989.
Corporate tax incentives have also been extensive. Research and development expendituresare exduded from the tax base. New industries receive 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 deprecKorea has
ation, and special exemptions. In contrast with the other HPAEs,
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 industriilization 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 petroleum 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 corporate-income taxes, introduced value-added and urban real estate taxes,
229

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

;Wni XV

x.2'.fCTL
r

e.

Ei

~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~-

q'

adeimntd

obse

. -

fill

Thiln

hii~~~~

-~~pa

-~~~~4

'

-..

clisroriions

fzsoeo

H2AEs.
Business
and incorne

and

pe.n.men.I

-ae1nealwrkda
cetie

apert

conrs

wihtora

aedoemr

a eeroesysc
mx rerenuesP~~~~~~~~~~~~~~~~~~~~~
_,

increasing

otcmpe

xEsn

amtl

ode h

-,. .,. .

: S;

5.o:

oermn

f,

.
,-

:aea.s

iuiesnaX

*~~~~~~~~~~~~~~~~~~~~~~~~er~-n
t.,

frts

rai capci haecsltti

oo m
s, n Taiwan,Chinaniheri
in
itl,
inceninresm hiIn

-~~~~~~~~~~~~~~~~~~~~~~~~~~k
y
.. compl .n
-:f

.~
~ ~ ~~~~~~~~~~~~~~~~~~11ta-t
_ ,
.
.
i:~~~~~-2i
:xI
raS:
ac.. ;.'-V
~ f.eos~av-~~t~
~
oX-, o1anes--

mn

~~~~tahnueabwidrgary

Crmpevi aindii~,lackFTof
b

~.

-~~~~~~A

n0tvsNewon

in rKlucing

: .

f - -

.n

o,
-

0;

:-f

p-:

FOIR-

STRATEGIES

4e~~

17:&tu~t
r,

,*.

4.

Z>-

- .
!14r

fr^Mstoa

{g'

.omorumavntaaeealorne
hstrowCr,,>.mdlnon;

ow

ts)boni.tacnnve~idrssi. rte

xbat'flpC

rte-

'lmiberoniboanM

baafpmeERm-@edlafsicrtoqsnutoC*Ddae
~ 0,fl
GDP~Its
'tl
f1
49.W4.,.
5
!5j7crceXu
alyEboult
&rntlolsescllubTncd
sFMi
PgrOwd ~- S16

ite:
n~
at
,J
=lMQl:}enn,>zem
- 'zi*:<firtl*8if
rmi4mlulceSeancQgi
4$i.<v:eiopmencfI
e ^D
3
4
t.ct
jO?9S
onoImc,3n&213
f,!ri<
-.-.{S

.~~~.rDooe>Doeiln

*^

5brd

gu

c-..s%

mictaves:ant

4nordi#Pnd

w*kw4ti4<_4$SF->
aji&ht

eau:tvrgone:revcnuc.

N~S

de2Tmcnc
-N

ooteono

*u;lro:et

~~~~~~~~~~

tZib1c$varro~~~~~~adn

Did the Hr-s'

ra-I

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 aggregate welfare But the administrative burden associated withi a compliCared 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

ceand
pionolvng
fivestment
i equipmmen,rather than from attempting to
tfine-mneincentive strucurecsto private investors through the tax system.
7231

LowRelative
Pties forCapital
Goods.Unlikemostrocher
low-and middleincome economies,the H-iPAEs
wereable to hold down the relativeprices
of invertment goods, especiallyequipment, during the 1980s (Bhiartacharya and Page 1993). Figure 5.4 shows the relativeprice of investment 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 increasedabout 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'mnvestment goods remained lower than in other developingI economiues
:i~~~~~~~~~~~~~~~~~~~~
throughout the global economicexpansionof the 1970s and the adjustment 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
only began
to rise in'1985, when
the HiPAEeconomies
letog
od rtoo
h necetdfao
oteGrttso)fo
werewell on the way to recovery.By 1989, relativeprices betweenHPAES
fr hih onisenrrel riat iarsten
dt0ar
-foerecnoie
and ocher economies were again aligned. Thus in these economies de-.
h middle-ncmconcs
rieo tenetmn
od-n
104lbe
awoe
NowLMeafe oregopa
ices fo=aia
od low-niems
and
Figure5.4 InvestmentDeflator/GDP
Defator
-

ta~~~~~~~~Suceay
PeTrandPand
Index1970
_

Mad93).a (1992)

hwv

daieprc

f net

120 -

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

- ;:

122

e.peidi2h328sScnd

232~~~~~~~~~1

economnies

fe17sand
eaiv

rc

f neimn

STRATEGIES

FOR

RAPI

cdiningreal prices of investment goods smoothed the impact of adjustment on investment.2 4

Rising investment goods prices affect the relationship between investment effort and investment outcomes. As the relative price of investment goods rises (over dime or across economies), more nominal
investment expenditure is needed to achieve the same volume of physical investment. For the HPAES and the economies of the non-HPAEsample, investment efftortand outcomes coincided quite dosely during the
1970s when the relative price of investment goods was essendiallysimilar. 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 investment 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 sharing 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 mitigated 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 -:
0

A CLX
_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 industries. The cartel allowsfirms to share losses.They may also be eligible 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 mining, 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 incentive 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 industries. Interacton among firms and between firms and MITIare a means
to prevent free riding and to monitor performance.Thus the organization 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 mining. 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 financial 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 fiequendy signals areas of government commitment, providing an additional 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 Industrial 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 private 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 monitoring by the private and public sectors. Severalinstiutions, induding development 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 insttutions. 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 increasedomestic 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 accumuldadion?The obvious fuilure of such restrictions in economicesprone to capit-alflight suggests that capital is too fluiidto be retained wvhendomestic
conditions are inimical to savings and investmenc. Therefore the simplest explanation for the seemningsuccess of these restrictions in Northeast Aiistathecnmsoffered
returns adequate to retain capita
even without restrictions. Converely, for reasons we discussed in chapter 4, these particular governments have been unusually successful in
eliciting the cooperation of economic elites, t-he very groups and individuals who might otherwise shiip capital offihore.
Thus, for certain periods in these dthre economies, government restricdions on outward capital flows may have contributed marginally to domesnc 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 underground or even provoking capital flight, would rank among the comparatively 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, restrictions on capital exports inhibit efficiencyand must be dismantled.
In aniy event, the success of the more open BPAEsundermines any
suggestion
that
capital flow restrictions are somehow necessary for savings 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 nature 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 opnnes (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 liberalized 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 necessary 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 relativelylittle interference,sudden deregulationmay be the only deregulation 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
Repression
ofInteiest Rates.The finalmechanismwe discussthe repressionof interest rates on corporate borrowing-is perhaps the
most controversial.Financial repressionis often used to describe the situation when interest rates are at negativereal levels.A more precise definition, 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, HPAEgovernments 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 moring funds abroad, they have sometimes had the latitude to maintain deposit 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 govermmentregulated nominal intcrest rates. This type of financial repression has
often resulted in severely negative real interest rates, for example, in Argentina between 1972 and 1985, in Indonesia between 1973 and 1980,
and in the Philippines between 1970 and 1985. Cross-economy regression analyses supporc a positive association between real interest rates
and the rate of growth of economic output; that is, large negative real interest rates result in lower growth (Gelb 1989).
237

AR
CL

e
dn
d&rgdl
~

guhuS

k4
U2

T heHofNavust
assessme
suggests'rle'op
5 fnanr

meore,s

L4i

rs
$

nhgh9

sfrgsures to

flV.w%

1"

.a

~4W

~oi'nrArf

iJ*&V

beresoom s codnsitn
se fzinanchial

Hoevr,83acile
elctdco&nomic naecitlrsea.

time or notherheldrntcrestirtes4below
nnthelsHacShaeveinog used ancmia

hl

makt-dZaringlcv

citen
repressionasgrea

and

de-nia

rsourcessfto finnc the bdgvlpento

liiberatepolic fovrtime tosecurentl
greoevriaca
resourcesntoselectedsaig:dcthAineconomicsscos
branksanr
afterany
extenal
bepenssedion,selectd peraniodsrepusually
pression hasca
so
on
shockteto treAsansemresucles?oadete iigidsra
markeostclaringlEeveSmwhil
raespbelow
time orfanotherahelfinterestl
find
pigrtso
aggregateintmtand n hHancial.
achitevighgdn
nonetUyhaels
banksgancoimeodpositinofordsavns diaidltheoAsianleconomiesrowes

23-8

STRArEGIE
=t~~~~~~~~~~~~~NT
.

37

RAIDXO
R0

=~~~~~~~~~~~~~~~~~~~~t

S T~~~~~~~~~~~~~~~
RAE
;7;;

SF

X.
2

-; . .

nr+ridd
rd4
-uMffiedylwn
.iis

a4the'interestr

s
- rpn
9SiytoiBhves
overrnment
hogusaneo ser,Un
Cld

ie

oRnted,thes

y.$0

ulbCC&
-:cooiesinmthreeaimpom.nvhent

rElativdy
moderate arldw
in
pen;istndy

...

subidy
-:

sh

mchiaseher
aatet

r poirsii bostp

cij
2

dshedrefeeshiond
-lir`dexcepions

di

notcrtoet

as.

Iit

of euryn

ubkflt&7Q

wasmo

tablea5.5
In

i3rprsso

*r

s-

d &MCM-.-~~~~~~~~~~~~~~~~~~~~.
iid

real.ninterestrat

.-

mrgina
e
ipnvacerafinnci
12

has
r:Fa

t
ODW.nm_xc0xnVcbUarQto.%cn
n
the
cc, .' t ;:l
t3.es.

* - -ment.by reducUingborrowig

pmasewoduc-s
andservces'inuluhdprvate
b

iid it

ae

in a enon h&gbvn
merntnenrghn

non

ieod

A.....

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

::: ofth

flfiC0

duringn

--.g

,J

to2E'eooctaefprosbpo

'qiymru

- -FandiadorpressincinthuHAEghs
s
iffered froimlatStin ro.swd
o
nthernasrxcs
from
ousehlds o frms inceaesth crprates proitbiit1o
iinvest-.trq

i4owe
=--

the euityoffins,
nablig thm toinvet inriskier-projetsnd bo
Uo
ings is,'more
sufficiently
low
and
t
corporate
sector has a highe
financethese
to
projec.' Thi
ma hvepadyofe
te marginal
eanes
.~~.+*~a.-,..-~ off'
-..

mareooi

an

was

conseque
*1-ai

urintede

no

idingfirns.anatrnatived toa tequt fiainngesrt(Sdlasiity

of993a)ldsav
Fginancufiadpendyion
ind theHopAESeseodiardfo
has
thathin mostgothe
penois inythre
nporthan
th ways.hoFirstao,thedereenofrepressayincrwase
~~ ~
aregav1
oeate ansmnd,
wirthaewrsore,t-lied
fcdeptiontrs,tio retcsultid
inopeusistenlydnegtoeirealinterestmes (see tablrae
5.5).tbiSecyofnd,estwe
-

--

..-

havteshowninychapktedri3g
rheprsion wasonderctakenoifapenvironmentpro
oFmnancninic
rpestainintye
ecoomis

HandEws
nohes unintened
rmta conmseqecofhe
W27S.rlrt,
i

thee

mpotanhe

legee

rpresio3Wa

N.

-of

L E .

0~~~n-t

,~~~~~~~~~~~~~~~~~~~~~~~~
~

1p

-,w

~~~~~~~~~~~~~~.
:

.-

':~~~~~~~~9O,9

.2

,-'''

;':~~~~~~~~~~~a

~unosyxyi}neetrtSadvabnod
~ xb~

--lasqueezetL+Si
: -

..

FQ15g"fim%1,,.,-1B

I.;-i

e:on

"Minterajelnal=knsa
_>9-=

neFEen^o,rmn
B
. Xtn<sn
or4-mom
oisbly
eottzSnprXdarmem;
s;-

.~~~~~~~~R~p

Siz.EbrincrS

pl,o

in,rainrasne.-dig

0_,ovn

2r-

.:~~~~~~~a

.%>l-

-tuv--tr>>

rv;s.>

r.4 }

~ ~~~ ~~~ ~

~oovr>navnolX
~ ~

.f~~~~~~~~~~~~~~~~~~~~~~

oryaetd::-..
-v>l-t-)

cuuenw,te

mxtl!oruenv
sann=
: f^we*,a
ts^

itrFt..4->><stxts,t<i/gt;

,,>,'gsSw4'St;t

_.I

9WillhOtilL9tlt'iOiEiittCrClltOnnpis-:SIi

~,-;

er..

N...................

rv-S ,

\*.M1 sRonn

.eavinten ded:ir- i-f

-ltiH
.Mi*<g

z:pRrC=sOn,,wo

aculii4tz1-

wt1E

-;Si

-;;f4

cosraire-mti=ur.a:
1uuNvdeLAtsts@
_N

-tU

~~,r-J;.>a)>iJ--.i

b
pem6d
a '<l-'4' '^

P
j>S;<,

2* s- e r., Xs >

I_

c.en

ponsdw7e.cmpabl

aiS^-

' r St-, , *-;

0',''~~~~~~~~~~~~~~.
'V
:

~ - ~ ~~~~~~~~~r
-icmiemltGsaco

I B;be>vE;uldshor-rem;tfinancrqprn=svsitrmalsmm4ia;UkI,
: . ; I g ;,=

E >*-5s-^st2
Srsge

~ ~ ~~ ~ ~
.4
,fF
,4
l- =J oSaN
t:' _fi< .
i

Id_

dA

g4i

ul

.i&l i .-a:- S--k; ii f

t, '-tF H:9 :),.U- ] t !

~~~~~~contos.ce

X$Sssrtmus y>dforeacchan^

;-bod

-. ec:
oemihan a6ouvzRcDrcenq-,.t.s
dw mtnznllts-isl-mmedoucti;.
irg;ont
A*

{--

-Wofnloin

itres-onldt

on+

'
'+!F
S..

os

et r >Et
s:

s--a

{ 4swrs emosd(firsJ.
.lE:reaLtdepoitr atesas iw
> ,",snewep.zSerm
68.;e.
&1t'aifer-

T,1

~S- t',S:-

6ti-;5f

=tr2aonnPnrllgno-vnvscLitg.de6tea
v sr-eecv4ec,.ak-o

; ;

t
,,

r,

<S:t' n

- *t i :

-vat.FSeracser.
} "~~~~~~s
vereLstri

--

Ji

Of.ii.rAdilICDO:4Wai&
&1

-O

Sg.^=^,
wer s>.w--\*1.tPGK>
fNtf<!

mof-

uS.7;1vt
bzmatH.fRftAS;sf-he;Ho
S-,R,;wiftW:2.S'sruXs
.{2;;7t>iv1X;742-r
it^.=x
4.-*l -. -r-9. Pj
. f.mE~~~~~~~~~~~~~~~~~-m
;
"Proodx
tssXadMeneeRnf>Bcue.i
zn-<S'6Stae>-en

~~~~~~~~~~~~~~~~~~a.-::'

A.,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

':

.
| - r

:
-

0 -

1, ' '

';--

'

. - 0 f' -7-'

;.~.

stuleding
:

''
;:

.- '

--. 'X S X".

e~~~rapifinfancial

.0

- '

'-.
.

ers,

*,~
-' ~~~

,,

~ ~~~~~~~~~~~hs
0,'
'

thsaye,

indtwirdl

sp=,d,

hav

cosqu-e
.
:--

to

tleatlo

repression

arue

rhatidie

on

WSuring

finiandial

bankt

ehansth

wercpaed

Nonethesltni

~~~~~~~~~~~~~~~~tde
o.- ~ ~ ~ ~. ~
..~
-f0~
~ ~
-~ ~~
~ ~0 ~ ~ ~ ~~eav
,

rate

households)

.-

Av

in

racsae

pId
into

at
(Wrl

depositorsla

posauedo

dmanct

'

8).e'

frae

efetos'ofmow:ayic

reslt

Bank

is infiin

SevAeralc

1989:

Fr

and
198.

pes

:0:,-':

Svrate

th

e.msancs.-stngin2da.r

-:f 'earesutin ,''in' ne''4rtv

..

dcepoit.

rorld ations).

mess

~ ~ ~~~m
grwt

foth
resuts

raest

thrd

fo

that

borrw

with geneitie.

tha

eguatossquee

:.:

has

:on !'paricuarl

e,'-t'o

mod--,-

d-':

STRATEGIES

Evn widhoutfinancialrepression,banks do not allocateloansto the highest 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 success.Since the valueof subsidizedcapitalwas high, firms perceivedhigh
marginal returns from greater effort and competed accordingly.We explore 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, inflaton rates are far more important in explaininggrowti than real interest
rates. i the subset of economieswith positivereal interest rates, the inwrtestrate is statisticallyinsignificantand actually has a negativeeffect
on gowth when inflation is included.
Similar tescscan be made using time-seriesdata for Korea and Taiwan, China We use the differenceberweenthe curb rate and the institutional 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 extraordinary record of accumulation over the past thirty years?Judgments
are difficult,given our incomplete understanding of the costs and benefits 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

FOR

P.A P,r.

.~~~~~~~~~~dr

'.

ronments,universalprimary (and later secondary)education,sound

and solventfinancialinsctations, secureproperty rights and complementarypublicinvestmentsin infrastructure,and lowrelativepricesof
investmentgoods. Effortsto improvethe institutionalframeworkfor
capitalmarketdevelopmentcamelater in the processand werenot responsible 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 important 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 promoterapidaccumulation havebeen generallyunsuccessful.

Appendix
5.1: Granger
Causality
Testsfor
Savings
RatesandGrowthRates
EXISTENCE

OF A REIATIONSHIP

BETWEEN VARlABLES

proves neither the existenceof causalitynor the direction of in-

fluence.Granger-typetestsareofienusedto test forcausalitybetweentwo 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
k-I

(5.2)

s=

k-

S*-k +
k-I

242

s,+,i

g',- +

grk-I

STRATEGIES

FORItAI

'.

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
c2a(s
1 IA,) < cr (rsA%-gt) where A, = {A*ik<t- 1I denotes the information 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, covariancesmtionarydine series,and that S,and G6preservethe causalityrelationship of s,andg,. Ordinary or seasonaldifferencingisan exampleof such
:ransfonnations.
Causalitycan be defined by using severalrepresentationsof the timeseriesmodel for the independent and dependent variables.The autoreTableA5.1 PublicExpenditure
perStudenton Primary
andSecondaryEducation
(US. dolirs)
-~ ______

,cmy

-.

-i

Hog'Y
'cora,Rep.of-'
1 'aaysi=5.

147

90

Primacy - ;C

180.5

512

32

8.15581

3898

'

810.2*
352.5

t0xnanana
-67i
i49

-8340O
1881

.176

,;j 09699

572.5
314.7
~~~~~~~193'3.

PT

t~ e

-' Seconday

1923.-3

28

1975:18

-.

~~2.
I 778.0.' 1r2
~~ 'Brazil
I

_-'3

India53

~23&

Mexicot.

0j

76.9

1~~~~~~~~~79.1-'
15.4-22.

175

4
-

713
566

2 9

~~~~
9

30.9

~~~~~~~~27
62.3
3.

~~~~~~~~5038
1210

312

`6
'

10.

875'

)3

,..,

fl wwevtAunat.wl
c

tea,

'UNFUd--.

ta.

' '.

243

TableA5.2 Testsof Bivariate

GrangerCausality
between
theReal
*2
S;
S Per
Capit GOP
Growth
RateandtheGmoss
Savings
Rate

~~~~~~~
~

*~~~~~~~

~~~~Sig

Si
0

~~~~~~~~~~~~~~~~~~~~~~~~~~~

jas

1S

.m

S -8

MSN NS~~~~S

3;.KS. .

~~~~~~~~~~~~~~~-

S,

-,*-tX

Kqq,-S--

'

N'S
MS'

Mdapua1-

;-i

Z g,;t-fI..
. ..
g

~~~~~~~

..

SBal

.Si

, -~.

Ng~~~~~~~a

Si

-Si

S .i-41rs
NS

MS

.~~~~~~~S

.~~~~~~~

KS

16
,~
~~~~Si

MS

11

-,

Mg~~~~~~S

'

115_ ; ;.

.2

PM-

S
la
In.i

-S

Si

K~~~~~~~~~~Si,

Ti wuyo-u
"S

~~~ ~

- '

~:

~
.-

~~~~

~2

'

,,-

MS

:7

md &"

244~~~~~~~~~~

~ -~ -;~ ~ ~ ~

luhLgO8

''

KS

gX

STRATEGIES

FO

EGERAYIF

gressiverepresentationwas selectedfbr this study. In this representation

: and by Granger'sdefinition,ifg doesnot causes,, laggedvaluesofgcannor 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).

NoteontheRelationship
5.2:Technical
Appendix
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

(5.3)

changes in growth
monetary

also indude

system

in 1973.

the breakdowvn of the Bretton

ship between economic

Both

Fry- and

positve

and statistcally

5.3 has major shortcomings.

ference between the equilibrium

stitutional

rate).

High

.ff&ets on the economy

moderate

negative

ordinary

least

data Both

dine-sries
significant

relation-

Fr,

it fahils to disinguish

system repression

(the dif-

rate of interest and the controlled

real interest

repression

financial

may have benign

by the equation.
economies

with negatve

5.3,

or positive ef-

These difirences

by spliming the sample in two according

putting

or in-

rates have severe negative

and may have driven the results of equation

fects that are not captred

system repression,

Woods

growth and the real rate of interest

small and large degrees of financial

between

Gedb apply

to pooled cross-economy

studies show a consistendy

But equation

a shift variable to account for structual

rates following

squares (OLS) estimation

plored

equation:

G= consrant +-IRu+

Gelb's equations

whereas

(G) using the foilowing

can be ex-

to degree of financial
real rates of interest

in one group and those with positive real rares in another

Second, financial
terest rates. Other

regression is not the only reason for high negative ineconomic

Many ofthe economies

problems

can afifct interest

rates as well.

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 investmenitopportunities. 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 correlated 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 effects will result in a higher-than-expectedreal interest rate. Thus, twostage 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 studlyestimated equation 5.3 applying a two-stage least squarcs estimation to
both thLecomplete sampleand to the subsampleofeconomies with positive average rates of interest. AvenageGNP growth rates were used as
proxies for the growth rate, an1dreal lending races (from the International Financial Statistics'database) as proxies for real interest rates.
T'he 25L estimation induded the previousyear'sinterest rate as the instrumental 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 variablesfor 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

,

Rea hien

'; aJrnsr&-vru.'

~~.L. rw&ratr''Whole

sample(2S):-

P1=2
92

; ;7s
Ekiuado .Rn
1.

-.

rate
. .;.

- ..

-.
f

,F-

Rea *cur
*b.,,
=
interest rnf

,.-

;
0.148

-000;;
1

<tLLUL
LAJIA
5

tuft*nn.. .

'T,

~~~rat

- -'.'-'

A,

0.277

Subsamplcof&os
o-i;
AX

C5

tri-ot

Equaiontl

-.

>tT-.>g
. .
,

'

-0-:

'

i Equaon&
2

Equarioni)..
0.391~~~0.
;'w-.
0.059
Eqarin3 -- . A389

M6
WattTrto-sSt
aeEa
i
on

Icr............,;><,.

regesons

were

.
.;:-;frA.

on

estimated

>:

<

..-

of w-

hmp b

the nm

suhsampl.ofcaniameoLswason
otlie

swarcw&iBank
dima
2;

-~-::~;0363u-

0 438a

C~1;peratlewd.,'
Ieastst

-~

x0

-: -

:l'

r..

;:. f!:.y*:...............
,.. ...... , .. -

:-

the subsample of economies with positive real interest rates, the coefficdent 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 relationship of interest rates, inflaton, and growth in Korea. A third equa-

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 intion

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 Taiwan, 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
shareof.

GNPper

:: - -..
cap

iib

omm 2

Ec--fonomy,;- .f(perytr) -t
; :pon~f;= =2 1:
I96590

: .HogIong-'Indonesia)

9AEV-Japa

Ysrr

.-

71

4Maaysia"'
=

tq.
-gapo 6.5:
Taiwan,China.

1976

Thailandh
-

1975-76 5.6

4.4

*.

f.OzhtAid

5.2
4317.95

6.1;
.51.2
'

--16.03
E11.1
991982683
6-.-.
-3.68:
8.89

35.0
:49.8-

.I - 00---

.;53.9

f51985
--l985b;.- --0
5
-;;
.48.0'

*-0

8.7
74

;375
453

5.1.
995.

hovsseholds

-47.049.4-'
41.0

19e
g70-71
- 03.7

PhilPEippines000:
1.i3

14.6
8.7

Brad!....

-. taiRica
..

1983' 2.47
626
26.0.
1.4
.:.
'1;:-971.
3;3'..........
54.8
} :t. ~~~~~~~~~-.;
54
8c8--S
......;:--1 -16.61.51-41986" ~..3 33 -.

-Mco
p~.
-I{
.

5.4A
--6.6

.0'-.
1973:3.5
. .0;198T'0.:4G6

R -

pertof

pferentof

7.9
*0-XX1979..
'
8.7
1976.- '5.7

f-

tope
botm20

eholkoouse hold

6.;62 1980'
4.50 0>i1976'
4.r
i1~ ; 1969`

Korea,Rep.of

of

arof:
:top20.

...

;.

3..
3

1972

3i3

-5.6.5

2.8
0.2
- - ^: --:- - -- .

1977 2.9
1I72
;L1-98546w 44< -

V&elzuea

2.0

- '_!.

-10

00

0
1970
-.-

57.7--0.
i-9
:5139

3.0
.-

7197'

-:.

19.9
2.11
179

;.:-54.0
.
.- ....................
{;

;76

0~~~~~~~~~~~0
iBoiswana.
--^
=' -- :i -4- 19856 --`2'.:
Cbce'Ivi-re
05
198- ;8-T50

Gh-.na

',-.

dia

-14-.

1-9

19s8j9'

a.-~~
1 ''i ''

--- 248---:-

~~
'

, 0 -..

. -

~
i"'

'.

~~~t'.nerprcznmoc
q

'

'-a~
'

-..

r.

'-,:

10 5

rt

7 .:

- n s04..

00-

527

'19645
-9 6.7'
1975-76

50

48

7.2

49:4

314....
-.

-s.-u
r'|

bou'e.i"k.c'p.ndz'ut.
'~.D
.o

.-

STRATEGIES

FOR RA:PtDiW

TableA5S5Natuire,Causes,and Resolufionof BankingSystemCrises

Naturnfnere.

.:'Enfisrj

::s

..-- Ninedeposi:-aingc-

T-TnnHon
na

-1982-83

....

b.

- ,

Japan.

: -

--;- :-.

Thegovement rem

and mismanagement.
* andw
wak prudential

liquidated troubled DTCs.

...-

regation.

1rg*itrst

..

.' ;..---

t- res

large-exposureto real
The gormentookoverthe
estatelendingand spillover lagebankcandintrduced ne
effers fiom 1983crii
in
cludingtopexecusince che banks o-;ccL- -sewn&de
t-v
from tceJargeieo rnDTCs

'

aSsubsidiaries.

Bank repor has esrti

(a) Excive

and a

4,

- ;.

'

'~~~I

-'

'

-d,

.i44~~~.;'C~~ijrr.

SI"' ':ononur

X . . 4- ,- ' .

dthecosto,

de&eoftbsie
among commercia bank.;

:i'

..

M
i"

'.

Bank

(b) Deterioration rtemsia-

iiit of

OXi

P-

r exposuredititugh
Ins to
nonbanic

,ic

,.gee'
nieterealsoiinacial dis-

...butOO~~
~ .1C
~

t
dalinstidon,and
---

one deposit-

and4 (of
= 47) i'Iedfe
naiicecompa

1'

lon to be

-k;ng (a) Preoinant causesaThe

goverinrrescud -24
ru
n speculationi,
isletcooperatives
and con-~
n`32(if35o)
r. :
d
sldar
In addituo;n4! (o38) bnsstOCks
iic
npM'.Te
eta

-.

~~~cooperatie
ini8cuec

88.rins:

tO

ban-sNo-~

transu

-a-lureof

'1985-88

4.

rnergerioFweaker

banksbwith hthier ones,

r
o b
provide.-

epdeclinein

ral esae prices

Ina-ecjuare
-nc

'

were:.s;rale to
..

ban.kThose no: taken ovcr

Encourage

--- realestatelending,corn-1
prisi about 90 pcentofthe bid loansofd

systrem:ase,.S. .n

eofaount.to'be wrtten off2sabout 1 1-5 percent

banking systema.

nrial
me.

by thgovernrmenrreaivedcrdt
from toher
com rcialnk

mated the sizeof problun loans

of the top 21 banks o be.
betwc 3.5-4.8
_
percentof
bki

reg-

ulatoryandaudiuing
system and

tr

ACenini

; -l991
11;

m or:.

Eight banks, includinga maJor

-international ankwere in.
'fnancial difficultie
.

--

Hihinternational

g;or&.,
inhteiiga

[7stwi

or re

(a) rg exposurc
toreal
eestateiled
lending f-raud

t*

1{o4 Kong,
:1;8-86 ::

Nature ofbiot

Causes

.n7.'

i..,u..

,.-

n.r-'$20

4..h,4.to

s,.

J8'4r y0'"C-:
..

~~~

OUr1Uust:-&rnpan
_S
n

~~..
.-...

Z L'ooeratLVeS

ae

,L'

-J,

.luo
trbek~
.
'S44.

(&4kivirnnre
-. '-

opt
-

,,..

,-4

Z9

249

LE

TableASS (coawiaWed)
Economy

NantresJiitttn.iI

ThailandL-1983-87

Thegovernment'st
ofbout of 50 financecompanies
w..as
estiinatedtUS$190 milion or 0.48percent ofGNP.

- 0f:-; ---ii -

0 .X

-*--

(a) Fraudand_speculation on real,estareand

exchagerae
c--ansac
ons.

t 3 u;9 ; - - ; -

45
.an

() EHigh
concenaon
unscuredinsider
loans.

financial-difficulties
in

(c Hit

*a
000t9B6-;87
1 - '' .< < ' ->>
g 50t,0 -0"~

~-

- 0- - 0- -; -:0~;
-.-Mat-uri-y

-t soldthmto ncwmvsots
-

ands
tinvestors
i
n o
banks)
It bouiht someshares ofrou-

bledbank
-

.6

'

-'

--

ksocaeda-Ioarid
to_provid emnergec loans to
rubd
-TheRind
would be finlancedby contribui-:
tions frdmn
commercial-banks.Trobe
-m-00-orriom
t

of

Matitymismatchesof
.Troubled
assetsand liabilitiesas
liqudat
~~~~~~~~~~~~~interestrates
woscandi

- -

The governmentliqtuidaed 24
financecompanis and-meriged
another 9, andthe Centra-:
Bank cookoverthe other 17
:

of

te

" - - inestrae

Gl' ;;' :- 'fChide,'

X 7';EntLrc mortgage bankingsys0-X-1976
te0 - was
- - insolvent.n

mowwcre'

~~~~~~~~~~~~inadequate
Capital
~~~~~~~~~~~~~requirements-,.

*
-*:

Five
0 ,commercialbanks
accounting for 24 percent ofcommercia bank assct were in
-.

fbaib ut

Aauee

-Chile,Severn!financialinsdiutons (or -a) Severe.r

i
' :.':198b-83. :-; t: :iabout 60pcrcentofrhefinan-,
,'1981driasi
case
- .0ni'sln
.-..
;.:al
1porolio) m nterest rts,adX
w insolvencii-n1988;ceniral
declinei termsof-

-5banlholdingsofbadconcrpe

cost

est

&iiloutrct
dat
-was
US$6bil-

lion, or 26.2 percent of GDP.r

7 .

~~~~~~~~~C
ofthe banking

'Ghana

Thlicnetwot

-1988

system,-lose
waxnegativsiat

1hr

ec

e rate.-,

schemetomee-rcgnkabik
p
'
ltoadsupervisowo
and r2Mpaniisider
and speculativelendling

tesd

interest rate sub

er4utpwp
o ea

raie

nc

--

,.-,,

andahppor-,

(a) Collapse1in hepic;of -Government-

_'50

...... I.

'''t
-. ,';30>-.Maa

'

'~tew.

wie oOamoans. resieur

n,' capta

~oi.
ipuin ~A enig

Encourage drpe:and

L....

0,

'rionirononiperfrrnng2ans
o

ng

heavylosses.~
~

Bankalsoreqs--

-.Hug foreinecag

*c&rof restructuriii as 7
US-$300 millionornealy 6

tradedueto:ihecol-.

n1981. thegoverrncmntbqu
dat ed:8insticunon&sIn:1983:it
rook ovceS(liquida*ng-3nd
rccapitzng5).n The:Ceir

coppero
(

. .0 !, _,'

-.

.-

STRATEGIES

Naftre ofLalikt

~~Nature
offlwwal

Economy
-1-;
Eco ov

dt-ress
.~*-PhiUippinas,---- :. Amongsmaller finanacalinsti--

cause:
(a).Inrcrascsin oil prices

i- 5;- :1981-87

wituuons,comprising7.6 percent

.Ilb)

more than two thirds

i of the.
asset.s .-.ofde|itnolney ban.

-- ;;:-.By-1986,
centl b:.lcassis--e.- :;: -- -- :-.tance to financialnstirudons
am9-:outto 19.1 bmiionpesos
; . E .!.
y:.Sor 3 percent-ofGDP.i

j .--i4,
r-

i;Turkey.-

:-

- .'.198245-8
5
-'

piud al rigon mMnaeent.

Five private
and--:suevso,
.
bak reaie
under.........
direct..
c tral-banlk
!ene'X
~~6-.~
supervision;:
t;meat

Thegovernmentrecapitalized-

the

the--a
banuks an'd isistedonthe-ir
op-opucrurt,
erriaires

ofsal brokers,=

collapsc

5. .-.- 8-

-adfeba
weres

re+ xc=ured.
'Of

superision.----

:- ~

dara

appodirel 4,500

X i57s9-89;
;

'

.-

(a)Predominat &wewa

Q
- '
'D?.

i
.
''

..

4..~.,

themers'nd.rnew'ana2ei-

nce :Corporationl riquidated~

banks(a d od

pru

x!srctn&-

mcesineetrob
'to
poo7r-q,ueprtc~
alr ecerpaid t depostors,.

S&Ls)

f, . .
'.

Simia approachtora oth

in
S&Lgoemn-e
FedraDeoi
Isrc-).

rea esat

1989
_-~0-- (o
ne1~fift oif~
~w over60
were solvent
-'.Toed

-!;4
-.

of

~~~~~
(~~~~~~iarioi1987.sB
KiCwewn198
(19)lindz
,~ i:
_lrg ,b=., 6
.VV~'p

er

...e0

liaosoS

-- A.

L a

,,

.4-0

'

~US-$13 bilion

ibi-i,sl9

'

4A
'

'f*_

r22

'

acquiredw

(96;Seg(18A9)Sii&ae
CO'.rbn

costs'OLcLa.ofA.Lupec,rtsoFGN
t 1~~~~~~~~~~~~~~~~~~~~
!ei2 ge ioanC.Q'I otspJ
additionlg'
up

~~~~
inic
vi
aennas
supervision;
let

__

(b Dergultion ofS

'

S&._ .- to le_h
'
-

.-

r. noe

~il~jii15o
~S'

'-

on its prudentialregulaons.

Sincel9i
epe ~ ( ~ ~ ~ ~~"rceve
Berien 1
D988, about
1,5O6
..f :-'.:(ofES,147)
S&Jwre
2 l
98Ireiducedrhev
~~
aluc -;ur;;IeC&porons.ised
al crisis
uptea wid-nT
:i

sp

~~ ~~

.-

~~~~~-:.M'
fiwxra.ouc'acdequateat

-'1.

175 liqudated.
n
-:ofanks'aserand
.21982-87.
bank,
spcltv
A
meried1,575
'andfi eriarporen~
b1~!
upcr
;N ; E >- -'-troubl
in 1987.tec en 1982 - loans:;;
- and1988,
e
ou 424banks

i.

liquidatedand depositorspaid

Afinancialcrsis
Aeruptedwith: ~Inadequate prudential

'-:',.

.>,Uwe Strs .

or rescue
Insolventisttutions weas

off by ie deposit insurance

deterioration'ian cc
he
corporatio.
tenmsoftradc.
-banks.
Th
2 publicbankswere.
High concentration oforestu-c-tr:
in 1986, and their.
lending to relate par,
adporios wcr put und raa
tienes
and inadequate :
epate agencyunder over

'

private banks, accountingF

nd internationalinteresi rats, coupledwith

o- deposit money bank assets,w

182
iilai In
a
In,a~~~~~~~~ddizio
erment intcrvened
in thc two
largestp bli
f

-.

4>

FOR RAPI

,_

.A.,

-:'

,s, -t rb -7 >'thertJ5Ss
of

~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~

&rp

_.e

.'

,.

Tnw

.;g

TableA5.6Comparison
ofBankingSystem
ConcetainRto

* g f

~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

-tuSI

~~~193

Hmlshnet
Indaish
lapi
'
KcrsRep.J

75

28-54

_d qil n

15 94

: >;

0,0.1959
.00

77.7

59

2
41

40
50

2:
60

~~~~~~~~~~~~~hil.,nd
.
359f8
9
56
Pidais
~~~~19116-U
hipln

tips-Y;an

.2

690.10
37

1.-

56
6

'51
61
7.

DI OP
0i-.,

66

72

t,

3986-U

t<N-x-~~~~~~~~~~Dhlds)
CA"ac
'.19156-111-

O.26

40

CLOW)

74w
20..

(UGh

90

0.50

52

X=atbr i

.Camdz,
Pane
*.
.

993

.. ,

~ ~ ~ ~ ~ ~

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

~~~
~
;

~~~~~~

W..~~~~~~~~~~~~~~~~~~~~~~~~
cs~~~~~~~~~~~~~~I

-.-

~~~~ ~ ~ ~ ~ ~.91r ~ ~ ~ ~

.20.

Iy

0040

41

6
69,

.43

LU

431a

as

71

:2

,.M.Ifbbrh

mAh6hIfl6&lkj

67X,

4:02

**

.7

d k=ni htl

klptn
3

76

.-

3.8

.r

cineanmtflrKd
ktmdinfrask~.~ah.k
indah SmwcnSm
d
Iph.Pn
ISbehbrasO
tnzh,S),,
baptvv
I
[naMti cm(x9kzphw,JnJ wi
3,nishll

u~~~~~~~~~~~Ti,;

*~~~~
..

5~~~4

. tlkHcdnMiI main&Inedbna~iuIl.,w
Sahi

DdU n(Ihhkaa&
higuhie Saif
wn(5

~~~1383
..
~~4enua1abJc

S'ahgznbnd

~~~ ~~.C. ~

' ~50

1359U

5.

2989
1989

6CM.

! .9!i

l O s-u

ses

-.

ka

mi

U.f:t

- -

-.

FOR0 RA,_s
STRAT~~~~~~~~~~~~~~~~~~EGIES
F ST1_s;RAT
~~~~~~~~~~*

Banks,SelectedHPAErs
-.TableA5.7 InterestRateSpreadsof Commercial
@
S
;.^r
>
s;=
r.z<-i
=
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~..
........
P > la.; s;] s r _ps

tS

i ;J1980
*',i-i#t..
-3, 21. ..
;-;
';,'1;';
'.........
l'-'
'....................

r19909' I--1

Silg9O8|e
0 1' r

to

-7.0
-- 0S@03e2

. r ."5

,s;
_L _Na

;~~~~~~~~~~~~~~~
X .. @V

ac:

tclHai-ners

or:ptnsn<SiXpn:-il-cl

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

-1aitltwo

- . X .e.

>nfre.

<,>3.g,c<anddwoslE.!
kV'N

-tasta;6OrM-.7;Ft..f

we;ostffnsfrIhln,&-

tiluo,,.ii.rn=ii

.- ,-~~~~~~~~~~

-::~~geed

lGo e s.v

x,,ua,o.lpoxasedr

fl,c.,.

.l

.te

-r7- 4 . .- -.

:-

s4

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

i
.

;;

.;r4,4

:!i

. -IR_AL
E5

EE

TableA5.8 Outimes
of MajorPolicy
Implem on inancing
Institutions
-Aals a

(Af it

-~~~~~~~

-PtN009 '

Bath(

0rWSM7mIlm,
(

M3 manC4
- (GO~~~~~mt)
%nr .

;f0Q00Q.--

-.

-. .

< . ;.

U..e

--

bafowa
nd.Acwith bath.

De

f-

terstssX
n1ef

'1135~~~~~~~~~~~~~~~~~~~~~~~~~ii

Ewe uneur

Das.

'-

hAS?bik

r6'w-S>

'f,

05,
C

lat
l

,x

'MWkm~
;

**

(US$3LEADe38)
H
- 351 ,Wtau
,

,.

ra

DcbuaP.aceD,adepeet. BakefMaLquh

nrlmam4ea

n . n- ,---s

iicrs)

821sata

.huu

~~ ~ ~

-m'.
''

hp

0;..

Iaaswmi-aawha-uda
h-

-to

mmdmmBan
Mas.s-~

beta-1usdm
b*r

e
-

inpesmu,

haEraw$

Fw*

'tn

i-:)k

W"

'

.-

rtu7

(US$54

r"T"'ad

5, p=^

UmmlUll 23

MO

-u

*t

we-'

b;;-l -l,e
A,
t ;r-{lwace
.-

*t2
tss9,~lnsgTe

.*, raEd

.0,--;0Dl

~tirl;.u;
-

17a= l,

Pksci.fwpa&M.y-

.4n

LnU1S

wi4

nnii7.5pantna..d

..

pre:,:

..................
;
M

~~~~mylwdncg

15*41=

s'.-

iu

(383fpm
f6ai,na

Saa,Dwqscfa

Ouse

aena,iacpehfehtd

cMa&aashqm

na

.' "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~w

(U

-.

-'ps,
1
1-

F.

:w

aucef,u

..

DZI~hhy

.!

~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i.anuctl&nl

_~

...

254
-. -

.. -

.e

. .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

ST RATE GI ES

notAum
bnal.pnwant nkdkdw
Phllpphim(inw) .

FORRWD

56Na'.".mnouvkt
aayTciamr

-17MpcLIpa:Dad.xemowwt

UISI

Caa,g--st

WUDNW

salmawn4 younfixwmldtus
Hai,

Ealcmwmaaima

25dluftolankga.safaw,

fhlsi5p).Amam1ka(vn$

Ml~~~~~~~~~~~~~naCr)

-*.

-.

~~~~~~~~~~~~~~~~~~~~~~agmriaa
ftfiwm.Fwre io
.

taandeatnamI

&Wtuat.,.

'ra'.

Cia...
lwx.~

L U.Jm..

...

flakdC.ma,wdcttaaa.
..
I

Indian]

Psunatatd

-. -

~~~~~21.1 raiKonbAixtaIwna
~(USSLIS
ils.)

Rime.

Capstans: Jlbadand

~~~~~Damu-lag
rumtht ptwmsnt.

..

M.&Sns-toharm

Ova.sbmmPt-gbaii.ouwbg
fa
pnaar'AaSlak

hAlfpnaee

- -isllh-

spn

s.leantam line.
fiwrdxsA aee nmt&

nxd.dbamd mAman

~45JW
KWon

Back fhcAgcdswc
(MAO

54 pm--

Drpsu

(W1i5tPSmaxlonwund-

PiLbSzaa
aghhinAxn

0(W)
G Na'orSbn
Haaxid2781
.easmatn

snaenntw

ix ai

t 13tmx

aotnaan

=it

dndaif.d,136

(n

bnvarFan&nsn
lelmnaina Rantm (199011. nc.Ceainullash ripen Exap nImpaMllk a Japa bmtlwa
kandblnkLTuixwaLnz
h
Eaptut-ap Bank f,V bases
mixs
dhcDdJchi KangacDank.lad. Malayt, teat] banksa.iaa.Dn
fllnAadRps
33Twhfpa.Ca
ai
u.uo
Iqi
Bthtapnbahndoknet
Japa Cntlnmmod
Kot.spaeqtaa(nip
am
TDank 52ldr.
w
mi lee
N.re
hl~C"lbakum
T
a

Novtes
1. See chapter 6 below for a mor'ederaileddiscussion
of the relationship between educational expansion, the
* growth of labor demand, and the educationalstructure of
wages.
* 2. From aross-economny
school expenditureequations
esinared by Schultz (1988), the elasticityof enrollments
with respect to the relativeprice of teachers is -0.80 and
--0.70 at -theprimary and secondarylevels.

3. Of courseGNP per capitais much higher in EastAsia

than in Sub-SabaranAfrica.Becauseof the bias that results from using exchangerates rather than measuresof
purchasingpower parity, we do nor attempt to weigh the
effect on costsof higher GNP per capita against the effect
of lower teachers wagesrelativeto GNP~per capita.
4. the relationshipbetweenthe abundance ofthumnan
capital and reachierpay is not firm. In Korea,~where

CRACL
E

human capital is most abundant, the earningsof primary

teachersare roughly 5 times per capitaGNP.However,because pupil-rcacherratios are higher in Korea, operating
costsasa perccntageofper capitaGNPare, at 16.5percent,
not much higher than in Indonesia and Malaysia,suggestingan important role for policy both in determining
teachers' vage and in controlling costs. Sec Tan and
Mingat (1992), pp. 28-34.

5. Implicit in the calculationsis the assumption that

expenditureper pupil would remain the same even if the
growth of the school-agepopulation was faster.
6. There may also be problemson the supplyside-inadequatecapital marketsor entrepreneuri deficiencesthat limit the private provision of educational services,
Psacharopoulos,Jimenez, and Tan (1986) present estimnatesof private and social retrns from education.See
Psacharopoulos(1993)for morerecenrestimares.Behrmn
(1993) citesthe manycritcisms of hese estates, indudingtheir failurcto acoountforabiityand forqualiy ofCeducation. Neither he nor others, however, question the
existenceof some gap betweenprivateand socialreturns.
7. In fact, it is difficultto distinguishcmpiricallyb-i
tween the e1*lrEsof nor being able to borrowand the effectsof poor inbrmnationon the demand for schooling8. Governmentexpenditureon education,exprssed as
a percentageof GrNP,was used as an explanatoryvariable

1O;Given the smaller size of the basiceducation age

cohort in East Asia than in Latin Amcrica (see bdow in
chetext), this difFerenceunderestimatesthe gap between
the nworegionsin the strength of dte public sectors commitment to basiceducation.
11. Howevcr,since the mid-1980s, public savingsin
Argentina, Brazil, Chile, Colombia, Costa Rica (since
1983), and Mcxico (in 1987) have incrcasedin response
to fiscaladjusrments,leadingnationalsavingsto rise (an
intcrnal World Bank report).
12. In Japan, by managingexcesscompetition, industrial policymay havealso assistedir the stabilityof financialinstiutions. Japan's recessioncartels,fir example,may
have avoided large and systematic bankruptcies,which
surely helpedthe solvencyof financialinstitutions.
13. The danger of casing intcrest ratesand entry controis without simultancously strengthening prudential
regulationsand capital adequacy requirementshas been
evidenrin Indonesia, where low capital requirementsfor
entry and poor enforcementof prudential regulationsled
to mounting concern about insolvenrbanks.
14. In severaleconomies, the risk was aggravatedby
the banks' tendency to speculatein securities.TableA5.4
shows that the major source of financial disruptions in
Hong Kong,Malaysia,Thailand, and, recently,Japan was
real estateand securitiesspeculation.

in a cross-economy regreSSionin which expected years of

15. Stiglirz(1993a) notes that in economiesin which

governments do nor take an active pro-competitive
stance,there are a limited number of firmsin the banking
sector.Furthermore, evenwhen there are a large number
of banls, e&ctive competirion (if viewedfrom the perspectiveof the receiverof the financialservice)is likelyto
9. In Koreaand Taiwan, China, the share of public exbe limited,This is becauseinformationpertinent to a sinpenduimre
on educationallocatedto higher education has
gle borrower is specialized,and each of tie major categories of loans represents a diffcrent product markerk
increasedduring the lastdecadeor so for two reasons.On
the one hand; universal and near universal enrollmrient There are differencesin the kind of infbrmation that is
rateshavebeen achievedat the primaryand siondary levneeded across product categoriesand few economies of
scope across the categories. Patterns of specialization
- , respeccively.On the other hand, the increasehas been
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 availthe demand for engineersand other skilledworkers.
able measurements suggest that infiastructure in the
schoolingof the school-agecohort (essentiallyan aggregateof enrollment rates) is the independentvariable.For
a sample of fifteenAsian and Latin Americaneconomies,
the expenditure variablewas insignificant.See Tan and
Mingat (1992).

-256

STRATEGIES

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.
HPAEs

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 largecorporations (an internal World Bank report). Many of the corporate 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 example, Eximbank, the Small BusinessFinance Corporanon, and private long-term credit banks).
22. While the experince offEastAsianeconomicswith
developmentbankshas been generallypositive,there also
have been failed development banks. In Japan (Reconstruction FinanceBank in 1952) and Thailand(Industrial
Bank in 1959), insolventdevelopmentbankswere closed.
Development banks lending to srnall and medium-size

-.

FOR

R_

enterprisesin Indonesiaand Malaysiahave reported difficulies 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
1993). The Dethan private commercialbanks ODaIJERI
velopment Bank of Singapore has evolvcd inro Singapore's largest private commercial bank, largely on the
basis of its excellent financial performance (Gordon
1983). Yaron(1992)suggeststhat traditional financialindicators 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 between 1982and 1985. Korea continued to show a declining 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 similarresult for the relativeprice of equipment (to the GDP deflator). 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 significantly belowtheir predictedvalues.Real equipmentprices
are unusuallylow in theseeconomies.

.~~~~~~~~~~~~~~~~~~5

0;

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 capital formation are not a guarantee of economic success,as the
modem economic history of the former Soviet Union and Eastem 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 international trade. It then turns to industrial policy-deliberate, governmentwas an
sponsored interventions to alter industrial structurewhich
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 bestpractice technology.

ExplainingEastAsia'sEfficientResourceUse

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
the Republic of Korea, Hong Kong, Malaysia. and Taisome-japan,
-wan, China-had among the highest rates of TFPgrowth in the world
259

_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 developing and industrial.
-How have the East Asian economies succeeded in using resources efficiently? Good fundamentalsagain tell much of the story. Price distordions, 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 regions. Whle capital markets have been frequently controlled, resrictions on deposit and lending raceshave generally distorted markets less
than in most other low- and middle-income economies (as we demonstrated in chapter 5). Government efforts to direct credit have generally
been undertaken within the framework of moderate interest rare subsidies, and financal institutions have been given the final decision on the
creditworthiness of investment projects. This has meant that loans supplied through directed credit have been monitored and generally have
been repaid. The relative prices of capital and labor have broadly reflccred 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 incentives 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 intermediatc and capital goods used in export production. As a result, domestic
prices for traded goods in the HPAEsare more doseJy aligned with international prices than in many other developing economies.
Apart from these findamnentals, however, most IIPAEsseletvely intervened in financial markets and used a combination of policies to promote 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 because 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 relationship 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 strategyemployed by all eight of these econormieswas responsiblefor their superior productivity performance.
Manufacturet exports acceleratedthe acquisition and mastery of international 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 orientation and high human capitl formed a virtuous circle;exports raisedthe
returns from education,and educationiraisedthe retuns from exporting.

UsUsng
the Market LaborMarketsin EastAsia
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 Philippines, the republicsof the former SovietUnion, and Sri Lanka areprominent 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 investment in education and training is as unportantto growdhas the accumulation 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 investmentdecisionsin 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, flexible4
and responsiveto changingconditionsso asto ensure
thatworkersare employed
in jobsin whichtheirskillsaremost productively 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 interactionsbetweenthesetwo 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 EastAsiacomparedwith 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

RESO5 RC

Figure6.1 Increase
hi RealEarnings

Japan
Rep.of Komea

;-

__

Tawan,nChina

7=

HongKong|
Singapore

Indonesia

- -

Thailand

LatinAmneuca
andCertheanF-

mi

SubSahIranAbica
d980-90
-'AI

1SrO.:It

-t

l
I0

-Z

2
3
4
5
6
7
8
IncreaseInrealeamings
(p'e.nt)

10

11

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 nineteenth century, the populations of the currendy high-income economies
grew, on average, ar an annual race of only 0.8 percent.3 Today, the population 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 highincome economies in the nineteenth century. In Sub-Saharan Africa the
rate of labor force growth continues to increase. In Est Asia, despite increases 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.

H; , s;Indonesia: . -~-;
Rep.of.

2.5
-2.4
2.7
2.9
1.9

t'rKc,

Malaysia

..Sinore

Thailand

-;

2.5-

thAsria
L4 in Am6icai

1985-2000

J--.90-85

bhea-n

ST--S
m2ranAfr.
.

.-

1.8
2.2
1.-9
2:6
0.8
1.7
0-

2.2

2.2

2.8

2.6

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 services have tended to be substantiallyhigher in East Asia than in SubSaharan Africa, Laiin America, or South Asia. This is not because
gowth of employment has been unusuallyslow esewher Widt the exception of the 1980s, during which output and employmentstgnared
because of the macroeconomicadjustments made necessaryby debtserviceproblems, the pace of industrial sector employment growth in
Sub-SaharanAfrica, Latin Arnerica,and South Asiawas more than double the pace in the currentlyhigh-income economies during the perod
1880-1900 (see table A6.4, at the end of the chapter). Rather, the interregional differcnce reflects the unusually rapid growth of employment 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 percent during the same period (seetable 6.2). The pace of change in the
occupational structure of employment is lower in other developing regions. 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

Kora,JRp of

1980
29.0:
.1981; 292
0 S-; 198200:
f - . 30.6: 0 :

30.8
.-- 0

1983
-1984
1 l985*. :-1986.
- 1987 2
-1 98 8:. . . :
,1l989~0 0

34.6

42

31.8
32.7
33.1
33.1
33.7
34.2
34.3;
35.1

45.6
-

379
-

- 35A4~> V- -? 36.2

;1990_

*/

34.3.3
34.1

Tariano Chbina

-42.8
.
*;:
-;-44.2

32.1
32
34.3ta;;

Shngapore

46.5
- |--.46.5
46.247.0
.-X

46.1

-t-46.8
60.0

37.2
X

38.5
39.8
-

r'-_Notvanilablc;

aToml
employmentforHoneKongindudesLinmployed
whohavepreviously
bald
KIidudes serviceworke
:Somiwfe::ields
(1993).

-X

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 manufactured 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 important to consider the extent to which the path ofgrowth of output has
been more labor demanding in East Asia than in other developing regions. 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 interactions between employees and employers. The labor market thus bears
the responsibility for melding the changing needs of the economys pro265

iR_ACLCE

ductiveapparatuswith the changingskillsand preferencesof laborforce

participants(Berryand Sabot 1981).In the courseof economicdevelopment,workers-most of whom are followingeducational,training,
and vocationalpaths differentfrom precedinggenerations-tend to
shift toward more dynamicsectors,from self-employmenr
to wage5
occupations.An efficient,flexemployment,and towardhigher-skilled
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 temptation to intervene in the labor market to counter outcomesunpalatable

in the shortrun or to particulargroups(forexample,wagesbelowwhere

workersbelievedtheyoughtto be, educatedunemployment,or the employment of educated workersin lower-leveloccupations).A relatively

high levelof efficiencyin the allocationof laborwasachievedby allowing wagesand employmentto be determinedlargelyby the interaction
of thosesupplyingand those demandinglaborservices,ratherthan by
governmentlegislation,publicsectorleadership,or unionpressure
Perform?
HowWePl
DidEastAsianLaborMarkets
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 anyparticularsector (Fields1992).In economieswith highlysegmentedlabormarkets,
the oppositeis the case.
Labor
Markets.Bynot allowingthe price of laborin
EastAsia's
Flexible
somesectorsto risewellabovewhat workerscouldearnelsewherein the
avoidedthe creationof a high-wagelaborelite.
economy,most HiPAEs
For example,thereis evidencefor Taiwan,China,of a remarkablyintegrated labor market; there is no significantcorrelationbetweenthe
growthrateof earningsand the growthofoutputwithinsectors.Manu&acuringwagesfor unskilledlaborare only20 percenthigherthan agriLulturalwages.By contrast, in Colombiaand Jamaica,where labor
markersare highlysegmented,workerswith the sameievel-ofskillearn
wagesby 150 percent
nonagriculturalwagesthat exceedagricultmral
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 unskilledwages.Rather the growthof demandfor unskilledlabor, in combination 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 segmented labor market may provide an incentive to workers in low-income 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 privatesector. The median share of the public sector in the increasein total
0
employment was between71 and 87 percent.'
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 medi-anshare for other developingeconomies.The limited growth ofpublic 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 inputs from public enterprises lobbied for low prices. Moreover,governmentsstrivingfor 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

1Figum
6.2 Industrial
WageGaps
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~t
'
.

..

~~~~~~~~~~~~~~

'

Wage a Taiwan,Ch._a
of 1955 NTS)
Meanmonthlyearings (thouK;mds

T-

....

:~2~

'

to-I

O.

-l

rea walp

F4

IF=

~~~~~~0.5-

--

I|

X t

'

]I

I .

UnakflledlIndustulal
realwags4

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

10

1963

_________

1965

1967

1969

1971

*I
-.-

1973

Unskilld

odustrialreal wa

Average Industrbl reawuge

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
__

&onorny/~on
Peno
LnAmericaazndC"bantbe:

CostaRica
Panama *':,
Peru
Trinidad`
Venezuela
'

Middle ase and Notnfipc

Egjp ,
.

:1970-84
;1967-82

1.4
:6
.

~7.3

Gh~~~~ina

~~~196246.

-1966-080
K

India

6.1'
7.2

Taiwan,China

.:

'

-196.5-85
~~~~ 1963-83

I'

Unwciglired.mC"ean"

p b tsector In
inrase p/otl
wageemplyment

00.

0.3

92

2.

45

37

27

103

28353
061.4
12
34

,195
'.

ahr,pbicaresons
employmentEwswns

.
1.619

67

7
09393

23-

665501

6.

..

33,

.~~~~~~~~~~~~~~~~~~1

bl'emlyetwsntarsostolbyninud

publennat0.

__

Toal

--3.8'
-r0943

~~~197183"

~~Thailand

__

rkwre

23-.

.~~~~~~1960-80 42213.2
8.0

SnI nka

__

-5.93-4.
64.2.

196381

LZambia

-c

~~196048-734

Tanzania

__

,.,18

6.61-

tA

,Keny

~~~~~51

1~~966-76f

n*nj.I

__

Public

~~~19733
1~~973-43.
:1963-82
~~~~1970-44

__
__

by2.la0or

tolbyn
nue

mar2-

bynlaborcma-

ployment opporrtunitiesfor the historicallydisadvantagedethnic Malay

Majority.The Bumriputeraprogram can be seen as an aggressiveaflirmatiye action policythat resulted in considerablepublic surplus labor (Freeman 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 dynainics in Eas Asia contributed to superior labor marker perforrnance.
But some HPAEgoverniA.."ts also intervened in labor markets, primanily 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 contributed to the explanation of why, in contrast to odchr regions, wages
were not maintained above market-dclaring levels in E-astAsia. An alternative, 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, espediaily 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 elsewhere. 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 compensation 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 market can rake two forms: dlirec restraint on wage levels or growth (wage
represion) or restraint on organized labor (labor repression) with indirect 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, leadiing 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-owever, 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 aggressive, 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 saying, "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 investrnent
and aggravate the unemployment problem" (reprinted in
Fields 1992).
* .
EIfwagesfall well below market-dcearing levels and employers are prevented 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 acplicit 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 interventionist 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 aggressivelyrepress 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).
to Growth?
FowDidEfficientLaborMarketsCoutnIbute
The benefitsof maintainingwagesat or marginallybelow,rather han
levelswere numerous and substantial.13 FiSt,
well above,marker-cleawing
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 investment, 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 intensity,this was particularlyimportant, given that government interventions 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 producdivepotential 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 prevailing (relatively)high public secrorwage to do jobs in which their marginal 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 conscquent mcreasein the educationallevelin those occupationswould have
contributed to an increasein labor producdviry.
Becausewageswere responsiveto changesin the demand for labor,adJustent 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 macroeconomic environment, then either the quantity of labor must adjust or
the ratesof output growth or inflation are liely to. In relaton to fluctuauionsin 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 finance 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 documenting the highirare of financial deepening in the H-PAEs.
We then
describethe nature and magnitude of directed-creditprograms and assess their performance. We show that, where these programs appear to
have had the most beneficial results, they supplemented fundamentally sound capita market policie-swith improved selection and monitoring. 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.

EastAsia'ls
RecordofGrowing
FinancialDepth
The growth of financial assets in East Asia has been ve-ryrapidfaisrerthan 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 mcome,

and theseratioshaveinacrasedmoreslowlythan Korea'sincome

growth would predicc. The largest incremental change was evident

after the inte-restrate adjustments in 1965. However,M2 fails to capture 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
OECD
Economies

~~~~~~~~~

.~~~

East
Etropeand
LaiMaidde
HAmaEs

190:7

Sub-Saharan
Afrks

OECD
economiesI_________________
0

20

40
60
M2/GDP (percent)

100

so

latin America includes Argendina Barbados. Bolivia, Brazil. Ch_c. Columbia, Costa

Not

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

and Switzerland.
Finland,lGermany,Ireland,Netherlands,Norway,Spain,Swecden
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 maceasinly 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 better selectionand monitoring of projectsis a key function. Imnperfecrinformation 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 monitoring their performance. But credit markets are not like auction markets 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 interest 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 standmgin order to renew loans.
Financal institutions evolvein responseto these market limitations.
For example,bank-dient arrangementsrespond to incomplete in$onmadon through relationship banking. Loan contacts are characterizedby
more than lust interest rates. The namre and amount of collateral become key. Yet despite efficiency-enhancingmarker responses, capital
markerSremain imperfct. Market power by banks may raise inrermediation 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 involvement are thereforecommon: first, efforts to reducemisallocation of resourcesby 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

ReducingBadImuestmentChoices.
Portfolio requirements,

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 underfunded 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 speculative 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 agriculture and limit lending to other activities, induding real
,-re. and
securities.
Most BAEs were assisied in their efforts to influence lending decisions 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 allocating s;-vingsheld by public finanicial institutions. The financial insttutions 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 FILaabsorbed 20-40 percent of household savings during the rapid growth period. 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

Wi.h
IM

or
Wi.h
iteI
pus ed singspwvidearinsturance
Economy/year
insa*adem
fiends
~~-

.:mmcw

XjiOQ~~~~~~et-r
.,1971-76

>ing-sy

,91960
81-8 5
-

-10-.

-.
N'

N-.

;-

1981-85

1971-75'

NA
8.79.9a
NA.
5F-A

~~~~~~

_1971-75-

197

1976'-80

-.

44308
14t3}
36

'.0

.19 0

---

.2

Thud

7 1.

7 1. 7

7M

61.3

'

4.

NA.

Noa

NA61
~
~~2.280.7

N'

461.3

151

855
4
22-.3;-':r.
_

NA.

24.7

02.

.~~~~~~~~~~8
..

~~ ~ ~ 22,N.37
.

198145:~ ~
~ ~ efiieg
1906-89
~
~
~ 15 itede

.~~~~~~~~~2
.

.3-

: .

108,

rid aem
n
s~Pecielixzdi
banks'

:0;

-,

NA

NA

-6

0-~~~~I

dcigtevlrl9o
57rsueealoainb

30cora4 prdn

beavo

netel

an3nhne

1ovnq9

: -0 - coprateos.
; ;. - - - =.: -ndesprad
- --~~~~~~~~~~I
stradbtutoryboardsinand goernment
Kreauete

Mhalaysa and Thailand hyagven

guided theisaig.Tate-ownedcomerieal
banks ofinanen priority projmbiiets.
cilsngsb
rmougi

n~ ~~

2t78

-:

~~~~sra

duvencthoeef
bewe

fcmptto banks and

mn
borwn.n
edn

aes

inceas
ihsrasrdc

thecen
averageins.h

USING. RESOI1

With pvernsment or

Vit? pensoi

~~~~~~postalsavings
~provident awntrance

**

4 J pbi
commercdialad

*SingPapre

1971-75
1976409.
19814~5
1986-90

2.1-1.0-

~~~~~12.7

1971-751976-80
~~~~~-13.81
t18-5

-.

-:1986-90

10.8
92N.53.5
19.4

21.6
30.0
34.6
34.9-

-23.7

39.4
456'
47.6

A67978.7
NA.

63.6

NA

52.4

77.4
727
71.8-

.Thailand

-1971-75
1976-80
* -1981-85
*1986:190

.12.7
10.

N.1.1
NA.

8.2

NA1.

7.6

NA-

-13.8
13.7

.
-

2.2

is fbr 1978-80.
a.Th fguxre P6rpotd
c Te figure-for-Taiwan,
China, are bed on the depositsof all COMmercial
banksand'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 imperfrct 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 declined 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 ornm279

~~~~~~~A
LE-

petition renrs. HPAEregulatorsdid not alwaysget it right, however.It is

notable that, in Indonesiaand Thailand, regulatedintermediation marginswere substantiallyhigher than those afterliberalizarionof entiy and
interest rates.
UsingDirectedCredit

Like many governments,those in East Asiahave gone beyond the indirecr guidancedescribedabove to target credit direcdy at priority acrivities.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 standards to the selectionand imlplementationof projecrs.
Criteria
forSeecting
Dreeed-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 accomplishsocial 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 associated 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
Japan
-22.
.SE5.9uS:
sare f ne'.7--

I96'':.iV

Republic~~~~~~~~6
fKa:reat
k'onsas a'
o
sa
or
,

'Pa/C

. 1 lons
Pol.

4965.
09

~~~~~~~~~*
;,..e i

..

1955

J3~~~.2.

32.

~~,1970 ~~~~~~9.8
.,W

~
1975

............

'

t; .. ',

''

..

-13.8

47.41
................

I'

W
106'
:1.75.390
*~~~
PoS loan indude loans0
udrJ
t
some kind oftexii ;rrna
;&I prg;m
1980 1
141
,~~~w'.u 176,
.D,>,,0inuerestzat
eandaada8llt
or suppor;tend
b'' ch Cenra Bn's auomr
8:12'.~~~~
Srnc:JpaCiDqE
;Wr;-;rfX(

19).0 i-r
uS-.rn;n-oth:

Table6. loanspertainwi Lows Adeounder

c-unts -ob

7,

.,1'

tht,

(192.

8- :,

byiavemmentad

'

lon eted&emnali
h au

59.78"

term

19.4

ofal lan

Imn'Policgram(FIlP) Outstanding
Rloancre
cd

',

and d-

f2 finanKial

,_,_
iwo
_

rh6~~~~~~~~~dveometinttuin
~2U'Foreig
'rmd.nn

va7&bni

banI d

oF

:;rl)w

* fi:'ir
-',ey b- ',
j'Ban
privasector madeby &- '::'; of ICor&de-'o'

o icudkeyal-lansfo foa- i n seb,-dt

.-

"11.~~~~~~~~~~~J771:

w:

'nicws&tdnglar

,;;

-:

a d loa

;te

of Trkeyand
exio. Lans y te bakng systemto public enterprises
in Turkey'in 1975were 28 percent of bank assets,and to the public administration, 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
oInS-goes
1
9t* Malaysia,and
well economicallyor fina-ncially.
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 broadest functional-targetingof credit in the HPAEshas been to exporte-rs.All

Easturkeanecnmisexcept
Hong
and Sigpem
subsidized
-f creit
-- Sngaore
o steKong
avedircte
nterriss,
ut hehave
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 axportfinancingone of theirfewreadysourcesof fnral finaneow.
Box6.1
rAa, Maltznus
onpagd284)

.AC

SKAN MIRACLE

in EastAsia
Table6.6 Characteristics
of the ExportCreditSchemes
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,Irediscountcd 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 exporrrelated 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 postshipment finance.
Provides export credit insurance, but this has been used sparingly (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 shipment 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 discount 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

-Cref'itWork.sBox61 Directed
as aSupplemen,
,Bed

MANY ECONOMIES

CREDIT TOE-:

HAVEmID TO USEDIRECD

fewhavesucceeded.The
smalland mediuni-sizeen erriss;,
prounsusinuiable
-contrast
betwcen an unsuccessfiullandulumately
-a .., =;.tE,.,;-n';
in K
,..
i..
gram in J-Idonesiaand a successhflprogram in Koreatshowswhy.
f--'
.- In Indoncsiawhich itdl 'th_early.198 s hada heavilyreprcssed
financialsystemsimilartothatof' niany,developingeconomies,heav-,
'..ysubsidizedcredit dixectcdto enterpriseir. pnvdmnent-d&ignared
wre divcn-bypoliticalf
- categonesoften went awry.Cr&dizallocautions
=aswellas conomicconsiderations.Repaymentrateswereas iow as 53:
percent imone program supporredby'the World BanL Meanwhile,.
the politcal
medum-ze ; firmsi that lacked
porentialy aedrcwdrth
-,P.
q.R. d-,--,!j,:V
d R - ' ;,m a um'
..
-:influenceofireenterprises and the speci.alcret access cnjoyedby
- counge

.. smallfs werefirzen-'ourtofh finanial systeCM..-the govenimentbegan.o liberal-'

PartyPto re=edj thesewproblemts
"'-izz he fiianial systmmin te -early1980s, permitringa rpid increase
-d. dimann
intert
'iralizin
in thenuiiiber of'banls,
programs.Aa result,medium-:sizefirms'access to iii'
cdirccted-credit
.nance hasevidentlyimproved,althoughte situation mayhaveworsl-that
lack-ur established
for vey small firms
died somwar
:commercialtack record.
.; In- contrast,
ectcaei
-or small.. nd
-M., t-progrmn
- erMe.. -. ~h;--.;.S
: . .= -Korea's
a
medium-size. enterprises has beencompaiblefromthe outsetwlm'-aks; whlchflave
disci&pnec-miarket-responsiveFminal:isysen.B
ialarge degreof autonomyin:choossingborrowers,,havepref-r-ed.
''..had
i 'estiblisheduack'record:of 122firms su2rvd, only20
frinwith
''
-.,
inaik&trcsouraesduring rhei.start-Tuhje Kiorean.:.,drew.onKflnand
Tosector
program raiieed hrms withournor acesothe banking
sMepapMrv idcdfoSorldng,aswdl
keent&rpdsesgoi
5
eexp,r&avafable
on pariutvesm nt finance,\xth workingcirzal far
ii Kbrea hare beeAin.-excess-..
-libenltem&,Repayment.rates
.:cuailawly
npn;rnr

Perhaps

..

-. '.

most important-todmesuccess&4I{eKoreanPprogram,Jn,:.

subsidieshaveben
-. teresr.iates

thoseTin
wtcompason
wih

modest-

h'hdy;epies-ed fiand other developzg-e'onoiiest

.:I.
n*edonesia
KoreanD
perveitnrern
gov.emnenr's
Despite
the
nancal>systems
-a..-.,;.Fs.,,.
,.,{,c lx,
ysr
=n iumsz
dr:f4o'r.srn4all
dleedcr
2n
ti7;nfiacamar

rt

r,0I.

.-

,.!],

_',pi,'

its;'-

-4h

_..

Prlss-

0--

2:rCm

SU-

c.

'

RESOURCS

*II.MIRACLE

enterprisesin
contrastsdirected-credir
effortsforsmalland medium-size
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 commercialbanksmonitoredperformance,sincetheyonly collecredrepaymentson realizedexport 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 required banksto extenda proportionof their loansspecificallyto small
and mnediun-scale
enterprises.Smalland medium-sizeenterpriselending throughspecializeddevelopmentbanksis a growingproportionof
Japan'sdirectedcredit There are also attemptsat facilitaringcrecliaccess of technology-inrensive
producers,especiallythrough Taiwan,
China's,Bankof Communications.
Targeting
socialobjecaieL.
SomeIuAEs haveuseddirected-creditprogramsto achievesocialobjectives.
As we noted in chapter5, Japanhas
directed credit to facilitarethe adjustment of decliningindustries.
Malaysiaand Indonesiahavedirectedcreditto promoteincomeredistributionamongethnicgroups.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 directedcreditprogramswerefinanced.The creditprogramsof Japanand Taiwan, 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 HMc
periodin Korea(1973-81),the more
prudent macroeconomicand interestrare policiesof all HPAEs meant
that realinterest rates to borrowersof directed credit were generallypositive, if not at market-clearinglevels.'3
Measuring effectivesubsidiesto borrowersacross economies is difficult,since it requiresestimates of equilibrium interest rates for eachtype
of credit in each economy. Table 6.7 contrasts real interest rates on directed credit for the HPAES with a sample of other low- and middle7,84

USING

RESOURCES*'EflI

Table6.7 RealInterestRatesonDirectedCredit,
SelectedHPAEsandOiler Developing
Economies
(percent)
Economy

Direcedit

FIPAESI
Indonesia, 1981-83, liquidity credits
Japan, 1951-60
Korea,Rep. of, 197040, industry
cxporrs
Taiwan,China, 1980-89, industry
1984-85, exportsQOterdevelopingeconomies
Brazil, 1987

3.1-4.6
2.9
2.9
4.6
4.6

-23.5

Colombia, 1981-87, industry

India, 1992
Mexico, 198748
Turkey, 1981, indusay
1980-89, cxports

-1.7-4.0
0.5-3.0
-2.7
-6.7
1.9-3.9
1.5

Nonireted credt

1.5

13.5

-2.5-4.0
-24.0

7.0
6.0
13.9
13.0

-4.0-15.0
-14.0

,,- _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 exampie, in the early 1980s, roughly tiree-fourths of loans from Turk-ey'sfinancial 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 average 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 similar: interest rates are gradually deregulated
1 directed credit is reduced,
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 financial markets. The deregulation of entry has been much more limited, however, except in Indonesia. This process in the East Asian
economies has been gradual and was meant to minimizedisruption in
the manufacturingsector.
In contrast
forSelection
andMonitomng.
Institutional
Mechanisms

romost

other economies that have rried but fiiled to use directed credit, Japan
and Korea appear to have strong institutional capacity for project design, appraisal,and monitoring. Becauseof this, direccedcredit has generally gone to projects that are creditworthy and viable. More
importan-and in sharp contrast to many developing economiesfunds have usually been utilized for the purpose for which they were allocated. 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, scrutinizing projecr proposalsand the firms involved to ascertain their abiliry 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 resuited in very low loan losses.The Japan Development Bank, the main
government lender,experiencedwrite-offsof a mere 0.09 percent of average 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 performance may be partly explainedby its greater reliance on collateral
security, particularly equipment purchased with the loans. Fear of losing collateralized equipment gave borrowers a strong incentive to repay
their loans.
In Korea, the government, banks, and industrial firms worked together and shared responsibility for the success of directed-credit pro-

286

USING

RESOURCESO

jects. Various institutional arrangements such as monthly axport protrends, discussed

motion meetings and monthly briefingson ecoonomic
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 continual and pragmatic reviewand adjustment of policy implementation
reduced, but did not alwayseliminate, the moral-hazard costs of government intervention.

Other HPAEs fiallalonga spectrum betweenTaiwan, China, whichhas

utilized mechanismssimilarto Japan and Koreaand has had similarsuccess,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 noneconomic considerations. Projects were designed with conflicting objectives and lacked proper evaluation. Moreover, the very large rents that
could be obtained from subsidizedcredit werea strong incentiveto corrupt 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 government-monopoly negotiationsand were prone to poor appraisaland disbursements without proper documentation. Credit allocation to large
and politically powerful firms or state-owned enterpriseswere characterized 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 actor explainingthe contribution that capital marketshave made to the 6fficient 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 contributions of East Asia's financial policies to growth. Recent crosseconomy 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

LqR"A`ThMIRACLE

1993). Moreover, the level of financial development imporEandypredictseconomic 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 successwas 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!spolicybased lending in the immediate postvar period (DB/JERI 1993). The
study found thar interest rate subsidieswere no more than 2-3 percentagepoints 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 economic plans and directed credit; and respect for the managerial independence 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, particularlymanuFacturedexports (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 mid1980s, 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. Directed credits burdened banks widt nonperforming loans, interfered
with the financialsystem'sefficient operation, hindered the growth of financial savings, and sometimes threatened macroeconomic stability.
Koreawas able to weather thesedifficultiespartly becauseits closed capital 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 either fimneled into directed credit or went to bail out troubled enterprises 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 repression and directed credits are now less viable options for developing
economies.

22

To argue tlat directed-credit programs in general reached their target 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 promoted investment and did not simply displace investment from other
sources, and the social returns on the incremental investment exceeded 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 programs in general machinery, electrical madcinery, and precision instruments 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 significant impact on both gross and net investment; the three-year cumulative 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 borrowers 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 receiving credit had researchand developmentinvestmentshares in output 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 effects 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 investment during the 1960s, when JDB was the major policy-based institution 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:
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

* JDB
*

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 insurance mechanism. By selecting among borrowers, the JDB encouraged
complementarylendingfrom private financialinstitutions with a consequent 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 affiliated 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 additional 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 returns to scaleand that produced basicinputs for other parts of the economy. 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 dissimilar 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 desirability 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 argued characterizeseconomic policymakingthroughout the HPAEs, abandoned 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 providing credit for the private sector meant that, in gcneral, loss-making public enterprises were not priority lending targets, unlike in other
developing econonies.
Z91

ANL~MIRACLE

Using
the International
Market:
TradeandIndustrialPolicy

HILE THE LINK BETWEEN INTERNATIONAL TRADE AND

economic growth is widely accepted,the precisenature of

the relationshipis controversial.Earlywriting on trade policy and growth stressed the benefitsof neutral incentivesbetween production 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 importsubstituting industrialization.4 Recent theoretical work has argued
that, where market power, economies of scale,learning, or externalities
are significant,departures from neutral incentiveregimes (lowand uniform 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 departures 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 success.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 instruments, are motivatedby the belief that shifting industrial structures toward 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
industriessuchasaircraft 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 increasingproductiviry.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 international best-practicetechnology,thus boosting productivity and output growth.

MarketFailures,Trade,and Industrial
Policies
Why should governments inreriere with the level playing field created by the international market? Our ditcussionof coordination failures 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 upstream and downstreamindustries.The larger the indivisibilitiesand returns from scale,dhemorelikely that private initiativeswill be absent.
a Straegic negotiations.In negotiatons with other economiesand foreign companies, governmentscan alter the nature of the marker environment by intervention. The outcome of any bargaining problem
293

RACR"MM!IRA
CL r

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 attributable 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 between 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 incompleteappropriability of knowledge,individual firms may spend less
on obtaining production knowledge than is socially optimal.3 9 Externalitiesdue 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 arguments supporting the protection of so-calledinfant industries. Protecdon 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 capitaland knowledge-intensiveindustries, they advocate a rapid shift in
industrial structure toward these activities,evenif they are not intemationally competitive at present levels of scale, knowledge, and fiactor
prices. In this view, the short-mn aflocativecosts of establishingintemationally uncompetitive industries will be outweighed by the long-run
benefits of rapid productivity change in the promoted and linked sec294

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 efficiencygeneratingbenefits of international competition through mixed trade
regimes:they granted exporters duty-free imports oFcapital and intcrmediate 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 production offset profits in the protected market, while competition in the international market ensured that the firm would not suffer from loss of
have reduced their proteccost discipline. More recently,all the HPIAEs
dionof import-substituting industries.
Belowwe describe the evolvingpatterns of protection in the HPAEs,
with the exceptions of Hong Kong and Singapore, which adopted essentially 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 protection (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 research indicating that the Japanese atuthorities engaged in an intensive
effort to develop this sector. Eiu levels in the machinery sector were reduced during the 1970s, only after it was evident from export performance 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 remained 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 gov295

RACLE

Table6.8 EffectiveRatesof Protection

in Japan
(percent)
By type afgoork
1963

-A

19 8

_-NRP

NARP

ERP

3.1

-P.
0.8
29.6

3.9
15.2

0.9
22.3

-2.3
15.9

-3.0
7.1

Intermediate producergrods

12.3

28.0

14.1

21.7

15.7

7.6

Finalproducergoods

15.9
21.6

32.3
44.6-

16.9
23.6

23.2
.35.8 .

15.4
23.0

6.3
12.2

Type

Rawmatcrials
Producergoods

-13.7

Consumer
goods

1968

B yindry'.r
Iad*m 1

:*1963

Manufacturin
Tattiles

32.3
54.3

Spinning

27.1

Weaving

-44.6

Products

1968

1972

24.2
282

.18.6

14.4

12.5

I.0

30.5

15.5

72.8

32.8

22.4

Woodproducts

14.0

25.6

16.1

Pperandpulp

9.7-16.7

18.0

11.0

Publishing
Learherand rubberproducts
Chernicals
Perroleumand coil products

30.9
33.4
19.5

1.0
21.8
17.7
14.5

-0.9
12.3
8.8
7.1

Nonmeallic minerals

22.2

15.7

8.1

Iron andsterl
Nonrfrous metals
Metalproducts
Machinery
Genemal
machinery
Electricalmachinery
TransportequipmentPrecisioninstruments

30.1
30.4
13.8
36.7
23.0
30.9
61-5
34.9

30.0.
34.1
-19.9
20.0
14.5
16.5

17.1

22.1

9.9
7.7.
8.7
-5.4-

31.0

9.2-

22.9

IG.4

a. NRP=Nonial RateofProtcdon.
b.-ERP=Efficdvc
RarcofProtecion.
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 industrialpromotion, the HCI drive,and the subsequent liberlization. The 1966 figures show a relativelyprotectionist stance that becomessomewhatmorc
2.96

USING

RESOURC

SE E

Table6.9 OverallDegrees
of ImportLiberalization
byMajorIndustry
forthe
Republicof Korea,Selected
Years(ConsoEdating
BothQRsandTariffs)
(percent)
Mgjor indunrLy

Printary
Industry
Food,beverags,
and tobacco
- III. -Textiles,clothing,and lcathcrproducts
IV. Woodand its produas
V. Paper,paperproducts,andprinting
and publishing
VI. Chernicals,
perroleurm,
coal,rubber,
and plasticproducts
VII. Nonmetallicmineralproducts
VIII.Basicmetalproducts
IDC Meralprotucts, machinery,
and
eqwupment
X. Othermanufiacturing
Allmanufacturing(II-X)
Lighrindustry(II-VX)
Heavyand chemicalindustry(VI-DC)
All-industry
avcrage(I-X)

966

1970

1975

1980

1983

1985

30.0
34.7
24.2

5.5
56420
32A
.36.6
64.0

55.1
38.4
43.8
69.1

58.8
49.6
74.0
76.5

65.3
49.6
74.3
82.7

71.2
64.1
87.8
92.4

39.5

54.7

54.5

78.4

88.8

90.8

47.9
41i.
44.7

57.1
77.0
73.3

58.4

65.9

74.8

89.1
86.2

75.2
89.2
90.1

80.0

76.8

91.7
92.0

41.0
31.8
37.5
33.7
44.5
39.6

59.3
39.5
47.7
38.2
62.5
50.8

55.4
38.7
52.6
43.7
61.2
52.3

63.6
58.2
66.8
62.3
70.5
65.6

69.7
65.4
71.3
63.8
76.5
70.4

77.8
76.1
79.8
76.8
81.7
78.5

I.

* II.

Near:This tablegivcs
thedegreesof importliberalizationbymajor indusuy theseare obcainedby averngig scrmal damweightedby

productonforrespcriveyer
thecurrent
pricevalucofdom:sric
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 variety 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 extent of intervention by authorities of Taiwan, China, in product markets. 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, Indonesia, Malaysia, and Thailand all had import-substitution regimes that
while modest by international standards, nevertheless favored production 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, protection 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 effective 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 exporters in these economies partially offsets the structure of protection.
The Closeness
of EastAsian DomesticPricesto IntemationalPrices
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 bodies of evidence lead us to this conclusion. First, nominal tariff rates adEffeciveRatesof Protecticn,
Table6.10 Selected
Malaysia,
Selected Years
(percent)
Sector

Foodprnducts
Bev:rages
Tobaco
Textiles
Apparel
Wood products
Furniure
Paper and paper products
Printing and publishing
Industrial chemicals
1Peroleumrefineries
Rubber products
Plasticproducts
Iron and sted
Fabricatedmetal products
Nonelectricalmachinery
Electricalmachinery
Transport equipmenr
-Not

1969

1979

1987

83

88
38
58
45
38
84
66
26
32
22
129
312
63
26.
89
4
59

-22
-26
15
6
82
43
29
-9
12
5
14
163
289
30
19
12
65

20
125
-

400
33
40
140
230
60
265
84
40
1,600
130
185

available.

Note Theeff&cive
mceof prorectionis thepercentgebywhichvalueaddedat

domestic
pnicesexcereds
valucaddcdatworldprics.
SoweSalleh,Yeah.andMeymnathan
(1993).-

298

USING

RESOURCE-SC

Table6.11 SelectedEffecveRatesof Protection,

Years
Indonesia,
Selected
percent)

Seaor

1975

1987

1990

Food, beverages,andtobacco

336.2

122

124

Textiles,cloth,and foonvear

35
33
20

231.8

102

Wood products

-12

25

Paperandpaperproducts
Printingand publishing
Industrialchemicals
Petroleumrefineries
Rubberand plasticproducts
Ironaandstcel

873

31

13

14

-1

-I

426
18.2

57
13

48
10

Fabricatedmeaulproducts

*Nonelectrical
machinery

Electricalmachinery

28.4

-1

63.6

60.2

53.6

74.1

68

59

Transport equipment

Non-oil
:
-Cement
Manu&crmnrng
Total
(eclduding
oilsector)
-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
Secor
Averages
Allseccors
Exdudingfood,beverages,
andtobacco

87.2

18.6
45.9

70.2

90.3

44.2
-

April
1985

1978

Textiklprmducts
Woodproducts
Paperand pulp
Chemicalproducts
Rubberproducts
Othernonmetalproducts
Metl products
Machinery
Consumergoodsandmotorvehicles

Nov.
1984

Oct
13984

1974

Manfacmuring
Leacherppmducs

Marc/i
1983

1971

7.4
248.5

151.3
60.7
39.5
54.3
48.7
99.8

1981

633
14.1
34.9

67.4
144A
199.7
53.7
43.0
52.9
46.0
102.7
62.0
23.6
33A

65.3
127.8
199.0
55.0
43.0
51.9
46.0
102.3
61.9
23.6
33.5

57.0
108.0
167.0
49.7
39.0
46.6
39.3
89.8
53.9
21.0
30.6

66.3
118.4
152.7

62.0
53.5
44.5
42.0
108.5
70.9
29.3
45.6

Nor available.

Noz The effectivemrre

ofprmtectonis the percentage
bywhichvalueaddedat domesticpricesexces valueaddedatworldprices
SwxrcBrimble(1993).

Table6.13 AverageTariffsandPara-tariffs,
byRegion1985
Tranip
para-ta

Ceatral
Caribban America

Soudt

North Saharan West

America Africa

Otler

DeveAlpg

All

Africa

Asia

Asa

PEAt'

aglow

Tarffi

Economies
unweighlted

16

23

34

29

32

36

19

26

-import-weightedb 17

24

38

30

35

22

17

24

18.

65

46

36

34

42

24

34

import-weightedb 17

66

51

39

36

25

20

30

Economies

Tariffplus

pars-tar:ffi
.

Economics
unweighted"
Economics

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 internationalcomparisons
of 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 axpecred, Hong Kong, Malaysia,and Singapore rank in the top nvo
decilesof the index, consistentwith evidenceon their stmcruresof effectiveprotection.Somewhatsurprisingly,Thailand, with higher effective 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 systematicallyin 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

LatinAmeadca
andCadbbman

MldcileEast

SabSaharan
Africa
0

Svnnre:Dollar(1990).
301

.SIAN'-MIRACLE

of internationaltechnologytransferopen at all times,

severalchannels
even thoughsome, such as directforeigninvesunent(DFi), wererestrictedor dosed for varyingperiods.In contrastto mosteconomies
strategies,evenwhcnprotecindustrialization
withimporc-substituting
dionwaspracticedwith respectto the domesticmarker,the searchfor
and absorptionof foreigntechnology
wasencouraged iWhile
Japanand
Koreaset obstadesto DPI,theywerehospitableto licensing,though
evenheretheJapaneseattemptedto ensurethat theydid not incurexto directforeigninwelcorming
cesscosts.Singaporewasexceptionally
vesunent,and a majortask of tie economicplanningagencywas to
locateappropriateforeigninvestors.
Malaysiahasaggressively
soughtexport-orientedDFI, particularlyfromJapan(see box6.2). Noneof the
of localbuyingofficesby interHPAEs washostileto the establishment
nadionalpurchasers,
an importantsourceof productionand marketng
knowledge.This selectively
pennissiveattitude towardthe acquisition

to MablysiaTechnology
BuigsExport
I stEn
Box
6.2.ri-Forign
a
:
,
.
:;W.6.:--,
chemical2
toward eecrcal- and electronic;priductsC

.INmTHE 1970SAND EARLY 1980S DIRECr FOREIGN

Gonenbbe
investment
Was weloMe in Malaysia.

htanaur
anddecliniginvesmt broun
.,pC
.

. . for. -xpos
. :

. PO..

The-benefits toMaysia of DF havibekensubgenrang

reign echange
stnal par
that
er, somecitics -argue
a
ploymenr.-.Howe

-. e

;-Dlocal
partcpaton incorporateowner. Then,
-receson of 1985-86 largefiscalantrdedefi

spuh
for
,

. o

melproucs, and

rduc

:
pelm. Moreoftheinvtents were-dedicated:before.
t- exorsiv

policy
-encourd foregn investorye the eec
ing
a of i
i
was.inhibitedby-

Z .-

eloc
.

no-my
V
.

... .

..

I.

atogas i neighig&
t--- cenives
ax andrladdomesticquity strast,not
cpits
of
DETwoimportntreasonsfirtatare,
potentialinvesto,
bparigpaiong
requzrcnmcis-or
:#ny g of the investments LIn
tecleogy and-e-portan& employment thec young
MC
Wr-.-an
~~-.,,CO~31
MU

opporrumnes.

The newflcdilihad a draiaic

confo

averagd
U US$300millionayearin198385,jmpe

eaiescaz&cl ofmangainet cac

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Malayuiidd
lbor,and
h-qualiysupplieratepadiyIskilld

foreig inveset

pj

cts
thmselves

-h

to blng. no-hoywin iec

t "became rnbuied
d-- ! - f-mc
1985tn tirie
mrcnsmiocntrm
moredynaic.
n>198a qU2rerFEre,T og

to US$2

bbi981

-''. Som-L.@

tii&&ibJanii

nare frriiJapzi>fllo~1..arwah,Chi&anc
y

302

-.

hd'bcen_:W
when thegreatest soureof-D.El
gaporet
afraed
nanfrlly tiicafti&and

than$0 ilo

&crrinMlysase9io~
u
i;L
bfisb&ffrtedrd

tsi
hertcnooyr

-k
kceepu

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 emphasize 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 permitted in inward-oriented economnies,it was not viewed as providing access
to international best practL e but rathieras a source of additcional domestic production. Thus some economies in Latin America that were hiostile 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 multinational corporations locating in the latter could only do so to export,

b~

--

~
-

er

~~

., --

actmis
M2)Sa.
.o

eprs

i-h&

t'u

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$3,''iniIl souibfware:ztune of 85,000 inelectronics alone in the late 1980s
cetr ocreate4applications6r,UN-aeworkism
MAiyaysihas found :that,-ainetments mature- sin>a
trce
oMl
yiabth&r'arts
adluand ndrginl'ikae
aefrtifed
ml-.
sotaeegnes&roe-fifth
of Germany4snre
nainl hat' are fe' huh
fa "foodoiose" *.calledii its Mlyineprst_hpstu;ci;.
stay, and the~poiiplovers
~created-ysuccesfu
asmlline
iinArzo'na.'
vntures encouragemore.investment.Fo xapc
Mlasa
spanproduce:a:-rang-fpout*-~
Si cth 90,DI1
nd
aysi '&
rom coiorTy<~,xado6Lassett-pyepl;;ar
ioatD5asters
cx
word'sithirdargest,--producer f. semic-bondiu"cto&compdze"eiiheras. Mcomom
istat scarte&
.n.flt.2
-garnering
almostbillion.
US$' 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:
i--,

Tmrld'
I

li-

Ma

Unfiterie-s).

Birocincmais

a&okp

to _seinc5ndu&rois--vestmentsin dectroAicshave

r~~eincohduccbrPa4s.A
~uppljilocal
Motbrolaind
~llo(r&
Hiachilii~e
tSei~a&has

os

oief die ornporflezfbr:itsnew clri

plntffaiya,jn

th&tat&-1980 Matus

i%5Pi~&nr4'
~conchcioner

`ia's

rei:ywor1d=
m rke Sr-roomtair,-'n

op.(-1991)

303

MiR

LN.MIACE

given the small internal market. Korea and Taiwan, China, with larger
domesticmarkets, created a similarsituation by encourging DFI in production 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 govermnent efforts to alter industrial structure to promote productivitybased 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 systematic but were nonetheless widespreacL Industrial policy in Singapore was
more functionally directed at the rapid upgrading of technology by direct foreign investors, regardless of type of output. Indonesia, Malaysia,
and Thailand have all used industrial policies but much less systemaucaly 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 industrial subsectors: metal products, dectronics, and machinery; and texies
and garments. Table 6.14 shows the share of value added in nine Internatonal Standard Industrial Classification ('sic) subsectors as a percentage 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 garments 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

ValueAdded
Table6.14 CurrentPriceValueAddedasa Shareof Manufacturing
Eismjf

-m
hwo~~~~~~'"

w-d

TM"&

ad...d

ad

X.-

~a

"ad

;dr

pA'

HPA&
'

.I

3973

0.05

0-.

3573
3511

0.-16

L15
.4
0.35

0333
all

3053

IL"9

0.36
am

oi3

MP

a"
.2

-Y196

0.13

ISO538

0.39
0.35Om0

Ut

0.01

0.

"I
0.01

0.36
0.35

00)
0.05

DA3
an0

0.0
0.34

0.36i

035

0.12

ON2

os 1

0.46

013
Alli

013
0am

Wu0
.0

3.00
La

e5

-0.6

I.3l1

gmJZ

0.12

0.10

WI

33

gm

0.05

0.21
.3

0.4

00Ms
as6

DA.
.03

IW
0.59

0.02

IL"
a

1960

0.36

0.7

all

0.27

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

0.33

1.04

0.16

-D0

0.32

.Ot

3936

0.42

0.03

0.103

0.9

DA0B

0.0

0.1I

0.01

1.03

0IL"

L137

02
0-

-O0

1A

3970

0.32

0.23

3101

0.A5

0.12-

0.2

O.01

0.12

0.5

001

004

WO3
0L4

04

1916

05

0.10

0.9

01

.03

1970

0.24

034

0.0

0.0

0.0

0.04

Om.02

014

14"

Iwo

0-17

0.33

0.5

3965

0.24

-04Z

0.19

OA

0.12

am

0.0

0.24

0.0m

0.011

.3 01 -L30

W0
0.36

0.03
0,4

.06

I0D

020
.27

0.0

30IJ

4.36
0.23

0.7

0
am0

Lot

00e4
a3

0to
GAe

.314
ILl5

0.l
01

30
.00

0.35

0AS

0.24

01e

0.0

1.03

0.37

0.33I

41.1

0.5

amo

3.0

g 3970
1916

0.37
Gm2

0.12
0.33

0.3

1970
19[g

o0e5
0.30

0-5

013
103

0.06.C0

0617
01.15

I=7

0.31

0.8

D101

0.0

*1,34

0.~~

01

~~28
AM3
OA05
~ ~~~~~

0.06

'-56

(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 minerals, two sectors often assciated with indusErial modernizadon, ameno
o.ndards.
-largein the HPAES by interatdonal s

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 accounted for a substantial raction of the growth of constnt price value
added in manuFacturing. Table 6.16 shows a varety of-esmates oflong305?

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 attributable to cyclical phenomena or growing capacity utilization of initial
large investments.3 6
Sbtres of SelectivePrmnonion.
Here we describe the strategies of selective 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 exhibited 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

ir..

ITC
&nwqd#

hnnwj.

rqi"

a&d

&tS..

gadbe

a"

P4-

_d

ukA

ad

p-AKw

jminaAw

ruUvr

e12

.89

0.53

0,7

e.14

0.43

IMF95

036

065

Dh67

0.14

1973

1.63
0.75'

[.7
03'

0h30
1.42

O3
IAO
0.6

0.54

0.20
095

15SC
19

on

9s

_,~~~~~~~~~~~~~~~7 :-5

1329
O.98

2.412

36

Uhr

13a52

ns

a,66

0.73
0.67

0.63
0rt

TsIt
m

sad

0nay
ter

nuMb

uhmn6wf

1973
9ISBt

lo
10-3

;,.daV

N..Mlfk

0.37

0.17
0.00
Z.62
I.Woil

2.6

0o35

1 2s

11.6

0.3

1.2

0.44
a.ll

0."
0.01

Oda
0.57

L69

0,14

423W7
e0

'

Kar. Rxord
1563
19

0.99
D35
0.

17
2:4

2.0
o.n1

1.04
0.64

0.94
0.56

0:73
02

0.52
1.13

2.0
276

0.20
0.32

1.04
1.26

155
19B1 -

0.J5
0.94

025
0e5

1.55
33

O."
0.7

G."
C."9

0o.e5
0.70
Jo57

@0A

1.03
'.44

1.43
0.07

0D4
OS7

1973
t9S9

0.31
0@36

2.14
11.32

2.40
0.31

0.49
0.47

L09

0.75
0o27

1.53
oa

1236
50

0.13

t.7

0.34

IAI
to3J

1968
19: 6

0.!0
.34

121
3.33

4.35
1.85

al
071

0.3
0.65

1.9
135

02
033

3.16
1.32

0.09
022

0.95
3.s

6.00
2.39

0.54
(L

233
1.02

033
0.39

0.37
0.35

1so
3,74
0.63

L51

0.32
D.09

1.10
052

SeM

1I56
19S6

0.50
0.56

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

193
19t0

1.11
1.1

3.04
3.M6

1.21
I1'[94

1.07

oi9

-. 67

0.73
0.43

L63
436

2.65
D.8S

0.24
4.03

121
1.1

0.45
4

0.41

-76

0.84

37

am64

Chile

1915
1954

-306

?dkcJ

:
_

uhiew

37
1.57

90711C

'

aku*t-l.t..p..

0.15

15
t

a.70

1.04

ftrn

am

0.0
e

_32
l

US ING RESOURSCOU

Table6.16 Long-Tenn
TFPGrowthRatesbySector
TFPgrowtldrare
Taiwan,

Sector
FoodBcverages
Tobacco
Apparel
Tcxctiles
Lcather
Shoes
Wood

Japan,
1960-79
-1.76
0.0

Kirea
7.30
7.90

C-/ina,
1966-6
2.0'

13.40

10.70
12.60
-

1.98
0.47
1.03
1.03

10.5
7.6
_
o3b

9.40

2.81

Furniture
Paper
Printing
Chemicals

12.10
8.20
10.70
13.10

1.74
1.44
-0.18
3.36

Pecroleum

-0.30

_3 55d

O.Ob

Rubber
Nonmetallicminerals
Basicmetals
Ironand-steel
Metalproducts
Nonelecricalmachinery
Elecncalmachinezy
Electrical
equipment
Transportequiptnent

11.40
2.80
3.70
7.60
.00
10.70
11.20

1.02
134
3.41
230
537
432

63b
2.4
7.2

2.3c
3.3

4A
67'
7.1
2.7
11.0

Prccision instrumencts

Plastic products
Oilhermanufacturing
Average

o.92f:

0.0

-1.76
1.2

4.6

7.50
8.8

-Not availablc.
a. Foodmndbevrages.

b. Rubber,petroleum,andwoodproducts.
c Paperandpaperproducts.
d. peuroluemref and coaLc. All machincry.

f Plasdc..
SowrePack(1993b).

achievement

of international

respect3

In later years, industrial policy

had a narrower technological focus, for example, promoting establishment of the technological base for very large-scale integrated circuits
(Borrus, Tyson, and Zysman 1986). Among the sectors carefyl 7 nurtured were steel, automobiles, textiles, shipbuilding, and aluminum refining 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
and Yasuba 1985). The government di(Yamamura1986; Yamnamura
rected credit to large firms and protected the domestic market to help
allocated
the firms realizestatic and dynamic economies of scale.mmTt
foreign exchangeto favoredlargefirmsand tried to ensurethat technology 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 effecto help firrns to export successfully.
tive rates of protection across the manufacturing sector was 22 percent? 9 Such rates may have been responsiblefor the very low level of
manufactured imports, an issuewidelydiscussedin the scholarlyliterature (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 deficiendes, usingexport performanceto determinewhether firmsdeservedcontinued 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,electronics,and industrialchemicals.The motivationfbr theseappearsto have
been both strategic-to increasedefensecapability-and economic-to
shift to capital-and technology-intensivesectorsin anticipationof a loss
4t The costs of the
of competitive advantagein labor-intensivesectors.
Ha: drive are still not filly known, but theywere high (seebox 6.3). Although there are no recent esrimatesof Korean rates of effectiveprotection, there is considerableanecdotal information suggestingthat the
governmentaffordedthesesectorsrelativelyhigh protection.
Taiwan,GCina.The intervention of Taiwan, China, in manufacturing has been similarto Japan'sand Korea's,though less important quantiratively.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 standard 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 difficult to discern a pattern of economic incentives. Wade (1990) notes
that a guiding principle in the 1960s may have been the existence of
. ........~~~~..
~

Box63.31TeHCIDrive:Costsof
lidtervention
THE-KOREAN C;OVERMENT'S INTERVENTION

sidered the lower bound, however. Those industries

.policy,most notably ihe HCIpromotion, has often

. been
evaluateCd
from thc.perspectiveof the successor-

whose credit was squeezedwere,frced to borrow at

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,

woud -betoestirmate the fisal/financial cost assoWo

the Korean goverment opted to bail our.the.

strugglingheavy machiney; shipbuilding, overseas

r CaCed
wit intervenCiin. Govcrnment intervention in-:X-.- ',.cureddiecthefosts thform ofsubsidies to strateic
sectors through-policy loans andl tax-exemptions, esecialfr during the..1973-79 periodof.HCI .promoMcQf
nflosPA
tidon. Intervention
rred idirici c
t

construcrioh, and shipping indusries in' th mid1980s, nonperfbrming lo

of conmrcial
banks
accumud
rapidly and accordny
bank profi;abi.y seriously detoed
During 198-7,
fr

in:iec

exaanplIe, the share of nonperfmii

lonsI
to
-rmn .of accuimiated nonperorming
loans
assets readied
10 percent. Between -1985 and:
resulting port:bli difficties'ofcommeicial
banks.:
1988, 78 coqorations were "rationa
DrC
ThepornmoUooftea
scows
spporteda
ti
process,
rn
b. abroaod ran.oft fscalae,dFCa.ncirwa.i'srumets.
iwntie-o.
of principal aonc amounted
iy~-oroid-='rangds
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-

eInost

amounwird roperceatord2tc
-romotrionpmo.n
aton,
Upfipjwo
100pr.fr:hi
percht-irfor the` ew
yas
;'-rate aboutatwo-tl

beilijoeawon q-;~

rationaiefis
ations
ea
ad 50

to Cen
.. balaces

-provi

'

1977-a

piceit
_;-

4 tGNP'(6kIiby

issW_-_.

bankIoSii4- Ths
L-:-,

75io

aitg

an

eshdmshoulbe
V-

It...

hash:urtbthe

rf

nt of.
l ugb'p

5qt;ra;n it has taken,mre tan-a

.idede
toawith
.sdircefi-

ind .th costs

i
n

form of accuulated.nonpdomn'on

'squed

ba

pr

tKore
governt
oubida

to

an
C45
93t

aonce
med
pl
n6perceinnt
domesais,eilo
stlact

HdVseccoI'in. more than40 .

wroe ati5&

Afrgen&al

encesd uring the traumatic economic adj(nt

1l979-8; the direct costs ofS
s o(3t
sixt

<collecned)>a.~~~~~~~~~1ost
of fi

;-XstedT
ha..spredomnant
-.
?-.-,.&asupoirH
U-wa:pr&etl
pOliclon
1
dirtedtokey-industes.
snp
n977r
percent
wfahe:oiidome
tioflthe
T&
kaMng
rhplicinin&i&~~~~P
ib
.nanci I
r iii dHCIi uppsorCofInsfpoic
theiic

,.sectr: I.l

Bakdiscoun-rswhich
; .0
overnmeab r

eK

owerfor
strategic IndustrIs-,
drive in.197 alone,abou& 82
e vbenuo
percenmeofrtot
won

adt

of tot

&

separate.; Da&k-

ad

nerest<

troughicbuaet

atyetto bedeermicos:ts.-:.

(1993 NM-stB.

ank

con-

orin:.gounentannoLnce~a-memura-;
;.-...;

to.z adle.t tc

(1987) LKep

Pew

.pan---- -

309

?SdRA
C

EMIRACLE

gaps in the domestic input-outrput table that revealed potential areas of

import substitution. The autobiography of a major architect of the island'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 government may be discerned directly from two infrastructural effiorts,
namely, the Hsinchu Scicnce Park and the Industrial Technology Research Institure. Both were major investmentsundertaken by the government 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 emulate the heavyindustrializationeffortsof Japan and Korea,which Nvere
regardedas successes.The governmentcreatedthe Heavy Industies Corporation of Malaysia(HICOM) to be the vehide for the heavyindustrializaton 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 companies, employinga totai of 4,350 workers,involved in steel, cement,
motor vehicle,and motorcycleengine manufacruring.The highest profile 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 maaoeconornic 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 develonment 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.

many so-lled

-THE PUBLIC SECIOR IN MAIAYSIA WAS RELATIVE-

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 percentofitsouri,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

cEdfictraa&la
mngs of trade threat-

-w:er privatized'showsconsistentgains in soial we0

p'ublic enterpnsestached 3.7 p ercentofGNP.Asris-.-

,,

Dipg

fiA:il

-ened -macroeconomic

stability

achieved A derailed analysisof the compauies that

sthe govemenc

fare. As two cases belowillte,

begat
pdvaize in 1983.iOfm
orerhan 800 state
fiMrs existing Mi th mid9Os, about a hundred had,
been
by'WO.Maire
s
conmecial
.manuring
oper
s,butsome
were quite big,.

incl=du,ding
an--airline; a shipping' company,, and a
telecc}tnmuincatiotis
fim,
*
there
have
beeinglitches.-For
offeingshave

been

heavily

; the:government
in
that
Imps

sing

as much

instance,

wa no
gc in
an
Eve so, managment
gnd

public

esscbed,:suggt

upde.*ied.i'

as US$150

investrment

sha'res
s

million

i,

per-

sl

decisions

to. raise domescconsumers,

of

while ir contintue to
Sect or r-femain
;
X sL.,ibni,

more.than

,, d .

--e.;
;;

H-tover

theaiSaxontrall PRO-

company

ON,taornil
t.

y__Fxpcrcnxe

Ke,'SczE,,edr

,h&,,,6nrinf

eII&'
'ubl-PiIcsc~
r
phec

*as-

no,

ra

b
becamec
profitablta-paying
ALA a~
l'i2bu
',Rb-,cn
:.
even,

Co.tainer Temindl (z)..

'In 1985,90f Kr was sold to de-izep
-

;piblic

-o -tickets

ile

har.

dlS.
.

. to.,t the
;,

pia

sales. Ith.

~~rodxicri
tvi&de
6&
tiea
cotrings

:2 ~flrn ihake

permission'

the farie

hysia.
s -meant
thoughte
ublic secor retained control, the
indmcude'l ashosr"of
'
;.rm
that took over, many repres'nnt''
Bumi
-p
d h c.thh'
ontro
srests,
-'
ireincentvet
to ear Pr6fit.nK

ounigh

:schems
tovrpro

--.

and eventulywon

less, rationing

Ii.-mg

t or de 'jur autl{orit'"
de t
contrl
over

fares. Whie

ur big companies.A-lso,morc atenon to promorw r Moreoe,

h wr',p
monopolisriasitu?aribnsvol r;
'in
18sir
, b- mg5mpedrion
u
"O
';L the early
Mocvr-1ic
Dnont-.
Finally,priver&.to-h
_a;
17Y corporation?
;'on-as proceededveslowl;
s day, -ie Malysia

changes m incen-

tives and a broader role for priate sector interests

have 'frequently been more imporrant.-lthan
the
change in. oners
pe
.
M
A&r?Systems Berhad (MM).- -Wlhi48
.percent of ASwas sold to the publicin 1985,ther

ui

s a luc

-- 1,ing container.

inWCOM. E
ienad
Even
ref.ier

.nb-e'rir

the-fi-ms&vi

s.s

sub

m mnagrsinat-nyaestanisinad
taonintd

kspoldecrk&4he~~
mor' toit~z

r''

wa tha

traicinand

ofdie

nopbzcaf

reion
nil

. ah

tanrlyimprove

iworkeri,productv--

shamtrome.' w,

Sw,,t;

~ Slkii(199);

es,

out

'-;

(iu)

od Bn

(1992);

(.1988)"'-

3"I

-C

_AIRACLE

Japan, Korea,Malaysia,and Taiwan, China. Thailand's Boardof Investment (BO})has attempred to promote variousactivitiesviewedas having
the potential for technologicallearning, but, as we pointed out in chapter 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 technologyindustries. 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 technology industries-Indonesia's attempt to create an aircraft industry-is
describedin box 6.5.

DidIndushialPolicyIncreaseProductiv?
Have atrempts [o alter industrialstructure helped accelerateproductivity change?We attempt to answerrhis question by addressingrwo related issuesin appendix 6.1:
* Firsr, did industrial policy alter the sectoral configuration of industries in ways that we would not predict based on facror intensities 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 proctucivity change in industry are low overall or in promoted
sectors,prima facie, industrialpolicydid not meet its productivityenhancing 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 efficacyof government effortsto promote or discouragespecificsectors. In
Korea, for example, despite the government'sextensiveefforts to speed
the privatesector'sshift from labor-intensiveto capitml-and technologyintensive industries, the relativelylabor-intensivetetiles and garments
312

USING

RESOUCES

sectorwasnearlythreetimesbiggerthanintemationalnormsprediaed
in 1988, a substantialincreaserelativeto internationalnorms from
1968.Duringthe sameperiod,Koreamerelymaintainedthe internationalnormin chemicals,a heaviLy
promotedsecror,whileocherheavily 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 springboard toiinternation2lsales.Te strategy
is 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,
.beensold to Indonesianairlini<the miiaiyrhe pmoli and th&
stame-ruoi in
Foreignsalesareuniikelyto imptrove.Whileotheraircaaftaimanu&u--r&sJcihaive
been
zibusy,foginhnological
ianc&,to%Iinkrhe
-Cl

-timefromdesign
compleonoplucon, i is-

iesil

alone. Its majorpwd@ forthisdecadan dendeily

designed fifyaurboproppi
.nor.
rs
fr el
u.
igXto~ gDIt

-1995,f ieryearseimdesignicoxpiieon Byh

..for slmalarar f&alr~adywedwill
havexseral hern en
t -ranrsallpncing their uairdfrCaggressively.
rr cn ill affztpnC
arompyr
srbu195'hle,
m ostsrc:,o4i:catethat IPTN-was.nor
pfiptirabWlewhen
capitaLcoAstsarleen into acCOUUL 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 sector 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 industrialization, the share of labor-intensive manufactures in toral manufacturing 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 intensity at the beginning of the period of active industrial policy will
predict the pattern of industrial growth. If industrial policy to promote capital- and knowledge-intensivesectors was important in determining 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 twodigir isic classificationsof industrial sectors (seetableA6.2 in appendix
6.1). Our efforr to differendare berween a comparative-advantagebased evolution of industrial structure versus one characterizedby significant intervention is notable mainly for a number of negativeresults.
In particular, in Korea-the economy for which significant intervention is best documented-during 1973480 the most rapid growdt in
seac 'I shares of value added occurred in lower-wageor lower valueaddeC_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 encourage increasingly capital-intensive development. But, given the
3"4

USING

RESOURCES

rapidgrowthin the capital-laborratioin Singapore,this resultalsoconformsto 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 neoclassicalexpectations;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,
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 haveunusually 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 consistent with one another. Most striking are the high values ofTFrPchange in
most sectors by intemrational stndards (Nishimizu and Page 1991). Although 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 decreasing, was characterized by considerably higher-than-average TFPgrowth
during this same period.
The government in Taiwan, China, did not attemnptto influence sectoral evolution as strongly as the government of Korea. Nevertheless,
there was more than a small efforr devoted to encouraging specific sectors, particularly those viewed as either capital- or technology-intensive.
rates

315

STN

MI R A CL E

The threesectorsthat exhibitedthe greatestexpansionin the shareof

valueadded-apparel, electricalequipment,and metalproducs-are
wages,capital-laborrarios,and labor
by below-average
allcharacterized
product.Moreover,there is no statisticalrelationshipbetweenwageor
capitalinrensityand productivitychangeat the sectorallevel.In fact,the
highestsectoralratesofTFPchangearerecordedin textilesand apparel.
RecentsectoralTFPgrowthraresfor industryin Malaysiafrom 1973
cheTr growdtrates
to 1989showasimilarpatternofgreatvariabilicyin
machineryand paper
in promoredsectors,from high in nonelectrical
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
changeisweak InJapanthereissomesupsecrorswithhighproductivity
while
port for theassertionthatTFPgrowthwashigherin selectedseccors,
in Koreaand Taiwan,China,activitiesthatwerenot promoted(forcxample,textilcs)had TFPperformanceas impressive
as those that were.
arremptsto determinewhedterhighratesof TFpgrowthcomMoreover,
binedwith rapidgrowthof promotedsectorscan plausiblyexplainthe
veryhighoverallratesof TFPchangein manufacuingyieldmosdynegativeresults(seeappendi 6.1).The mainreasonsfor manufactrings
successinJapan,Korea,andTaiwan,China,layin the highgeneralrates
nonpromotedsectors
ofT"Pgrowth,includingthosein labor-intensive,

Increased
Productivy
Exports
HowManufactured
OES THIS MEAN THAT PUBUICPOLICYHAD NO ROLEIN THE

We berapidratesofTFP 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 "productivitybasedcatching up." We believethat rapid growth of cxportsi a resultof
the export-pushpoliciesof the HwAEs,
combinedwith the superiorperand allocatinghuman capital,
formanceof rheseeconomiesin creatring
provided the means bv which they attained high rates of productvitybased 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 exports 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 exports 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-leveladoption, adaptation, and masteryof technology.Thus, exports and human
capital interact to provide a particularly rapid phase of productivitybased catching up.
Of course, it is possiblethat the move to a higher production finction occurred beforethe growth in exports-that TFrgrowth causedexport growth rather than the reverse.41 But even if expors began on the
basis of productivirychangedue to such domesticeffortsas plant reorganization, the cumulative magnitude of productivity growth over
nany yearsis most unlikelyto have been a resultof purely domesticefforts.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 technology 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 knowledge is simply not sold, because its owners fear that licensingor direcc
foreign investment will eventually Ieakthe knowledgeto future competitors. 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 economy 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 exports 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 embedded 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 performancefrequendy 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 developing 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 diffusion, signicant knowledge ranfers are inevitable.6

* Technology
licensing.licensing exsting technologies,both rights to
proprietaryequipment and derailsabout production pocesses,offersdevelopingeconomiessubstantialopportunitiesfor 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 licensorsin the OECD economies,havebecomewaryof helpingpotentialcompetitors, even if contractspredude exports to ochereconomiesfor the
duration of the license.Particularlyin R&D-intensivesectorssuch as
chemicals,machinery,and electronics,firmsare increasinglyunwillingto
they believeroyaltiesprovidean inadequatereturnfor
licensetechnology-,
This creates
actionsthat mayimpairtheirownlong-termcompetiuveness.
an irnperfector nonexistentmarketfor criticaltechnologytransfers.
Exporting economieshave an advantagein coping with this situation. 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.
technology.
The freeor inecpensivetrans* Trasfir of nonpmprietary
fer of nonproprietry knowledgeis easierif an economybeginsits industrialization 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 industrialization in labor-intensiveand low-technologymanufacturing. The
equipmentwas readilypurchased,and the production knowledgewas
availale at low cost from engineeringpublications,trade literature,and
4 7 The simpler equipment was conduciveto
independent consultants.
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 cx319

-ALRACLE

portscan be quiteimportant.This hasbeenshownbyWesrphal,Rhee,

and Pursell(1985)in Koreaand independendycorroboratedin Taiwan,
Gee 1989).Buyers
China,bya numberof researchers(see,for examnple,
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 publicdomainor arenot closelyguarded.
a Knowledgefromreturning naiondls. Drawn partly by the high wages

madepossibleby exports,many residentsof Koreaand Taiwan,China,

trainedabroad,particularlyin newsectorssuchaselectronicsand computing, have returned home to work. Many retuming nationals have
receivededucationin OECD economiesand then workledfor OECD48 Their return hasprovidedsignificanttransferof besreconomyfirms.
nationalsaccountfor
For example,foreign-educated
practicemnethods.
al/the postgraduatesemployedin the industryof Taiwan,China (Pack
1993a).This source of knowledgeof internationalbest practicesbecomesmoreimportantaschangingfactorpricesdictarea shift to more
sectorsin whichhigher-levelskillsare
capital-and technology-intensive
neededto unlockknowledgethat maybe embodiedin patents,licenses,
or the useofspecializednontradedequipment.
In both KoreaandTaiwan,China,a considerable
* Domesticrseanrck.
proportionof R&D has beendevotedto improvingexportsand reducing productioncosrs.There is no evidencethat the gap betweenR&D
econoneedsfrequentlyseenin the import-substituting
and commercial
mieshasbeena problem.
CapitalIncreaseProductivity
HowExportsand Human
technology
and rapidfiomnarion
of
Accessto intemationalbest-practice
humancapimlsupplement
and reinforce
oneanothecIt isdoubfl that tie
useofforeignknowledgeand imported
capitalwithout highlysilled domesticengieers and workers.Convey,
HPAEscouldchave
madeasprductv

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'
rapidexportgrowthhasoftengeneratedpositiveinteractionbetweenhumancapital,physicalcapital,and knowledge.
The externalitiesgeneratedby 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,lowdomesticdisrortions, and exportincentivesall interactedto promote high ratesof technologicalupgradingand 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 basiccrosseconomyregression:the index of openness(as describedabove,the degree 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 remainglobally 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
5 2 We chooseas our measureof export performance
grewmore rapidly.
two indicators-the averageshare of manufacturedexportsin total exports 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 manufactured exportsin totl exportsis introduced togetherwith the openness
index, only the latter is significant.Conversely,when the share of manufactured 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 reasoning concerningexport externalities.54
Estimates of the sources of total factor productvity growd (described 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
(0.1650)

-0.0124
(0.0083)

-0.2324
(0.1550)

-0.0055
(0.0085)

-0.3943 (0.1748)

GDP relativeto U.S., 196O

-0.1033
(0.0268)

-0.0459
(0.0108)

-0.0837*
(0.0256)

-0.0381 "
(0.0123)

-0.0892
(0.0280)

Primary enrollment, 1960

0.0259

0.0216'

0.0221

0.023!W

(0.0081)

(0.0064)

(0.0079)

(0.0069)

(0.0085)

Secondaryenrollmcnt, 1960

0.0317
(0.0203)

0.0211
(0.0134)

0.0218
(0.0210)

0.0206
(0.0148)

0.0156
(0.0210)

Growth of population, 1960-85

0.1322
(0.2699)

02693
(0.2346)

0.2054
(02806)

0.1856
(02376)

0.1222
(0.2870)

AverageinvesrmenrlGDr.1960-85

0.0568
(0.0259)

0.0659
(0.0232)

0.0625 *
(0.0291)

0.0444
(0.0242)

0.0436
(0.0276)

Dollar openness index

0.0042 *
(0.0017)

Averagemanufacturedexpotslrotal
exports, 1960-85

0.0210**

0.0023
(0.0016)
0.0002
(0.0001)

0.0003
(0.0001)
0.0011
(0-0003)

0.001a
oi
(0.0004)

0.5217.

0.3530

03938

0.0132

0.0147

0.0155

Averac manufacturedexportslGDP.
1965-85
AdjustedR2

0.3947

0.4912

RMSE

0.0159

0.0124

0.0040
(0.0018)

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 contribution of the variablestaken together to explainingthe variationin TIP
growth rates,however,it is statisticallypositive.We conclude chatexport

of TotalFactorProductivity
Growth,1960489
Table6.18 Determinants
(dependent variable: rate ofgrowtlh of real GDPper capita, 1960-89)
67

Numdrmrofobservatons:
Intercept

-64.9123
(14.0585)

*GDPrelatve ro US., 1960

Educationalattainmcnt, 1960

Dollar openness index

-72.0692*u

-70.8604

(14.8657)

(15.6104)

(15.4608)

-5.5757 **

-2.3509
(2.3225)

-2.1562

(2.0637)

0.1471
(0.0874)

0.0680
(0.1082)

0.1574
(0.1064)

0.0738
(0.1207)

0.7154 IA
(0.1508)

0.7225

0.7134
(0.1558)

0.0314

(0.0066)
Interaction ten: Educarionalattainment 1960 times
manufacturedexports/totalexports, 1960-85 .

-71.1186

(1.9771)

(0.1417).

Average manufaccured expors/GDP, 1965-85

0.0159

0.0032
(0.0026)
J.0686
(0.0966)

- .. . . m:

;-

(0.1574)

(2.3008)

(0.0142)

0.02

Interaction term: Educationalattainment 1960 times

manufactured cexpors(GDP,-1965-85
AdjursedR2

51

-4.8047

O.6493

Average manufactured cxports/mtoal exports,1960-85

51

67

0.6333

0.6376

0.0284....................,

0.0625
(0.0269)

. 0.0284

0.4507

0.4628

(0.0201)

** Smzisically
significant
at the 0.01level.
SmziricaHllyiignificantat
the0.05 leveL.
XNoeCC
dicmt
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 manufacturedexport orientation and high labor force skills interact to faclitate the acquisition and masteryof technologywith attendant spillovers.
The evidence from our cross-economyestimates is supported by a
number of recentmicroeconomicstudiesthat attempt to test the link betweenexportsand productivitygrowth. Packand Page(1993)present evidence from Korea and Taiwan, China, that ar the sectoral level rapid
export growth is correlatedwith the pattern of productivitychange; axporting sectorshave higherseoral ratesof rF growth.Wei (1993)uses
city data from China and finds a statisticallysignificantrelationshipbetween 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 neutrality 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 manufacturing 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 understanding establishedby government statements and actons in Japan,
Korea,andt aiwan, China, that exportingwasthestandard by whichall
economic actity would be judged, even firms benefitingfrom higherthan-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 pro3274

USING

RESOURCES

ducersin the SoutheastAsiannewlyindustrializingeconomies.Governments were credibly committed to export competition. Exports were
important becausethey ensured that, giventhe HPAEs'high human capital base, productivitygrowthwould be facilitatedby the improvedability to tap international knowledge.
*..

What are the m2in faors that contributed to the HPAEs' superior allocationof physicaland human capital to high-yieldinginvestmentsand
their ability to catch up technologically?Mainly,the answerslie in fimdamentally sound, market-oriented policies. Labor markets were allowed to work Financial markets, although subject to more selective
interventions to allocatecredit, generallyhad low distortions and limited subsidiescomparedwith other developingeconomies.Import substitudiorn,although an earlyobjective of public policy in all the HPAS
exceptHong Kong, was quicklyaccompanied by the promotion of exports 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 foundadon 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 establishingautomaticaccessto creditfiorexporters;this simplecontest wasan important component of their export-push strategies. Other directed-credit
programswere lesssuccessfulWhere rheyappear to have resultedin allocationof 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 exportsall combined to allow rapid growti of manufactured exports and,
through it, rapid technologicalcatchingup.

Appendix
6.1: Testing
theImpactof Industrial
PolicyonProductivity
Change

OVERNMENTPOLICIESTHAT CORRECT MARKET FAILURES

should give rise to growing total facror productivity in the entire industrial sector,though the resulting pattern ofrFrpgrowth
across subsectors is not easily predictable. Knowledge-basedimperfections are likely to yield TFP gains, primarily in the subsecaors in which
the intervention occurs. Subsidization of knowledge acquisition in textile production will mainly benefit other textile firmnsas knowledge seeps
out to them or trainecdengineers go to other textile firms. Some knowledge 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, empirical 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 implausible. 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 classification Uaffe 1986). The nonsubstantiated claim that the spread of external326

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 manufacturing 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 descriptive 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 weightings 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 represents confbrmance t the international norm, values less than 1 indicate
that the sector is smaller than predicted, and values greater than I indicate that the sector is larger than predicted.
In general, the results conform to our expectations of these economies. 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 predict 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
that used indusparel. Only the former sector was promorecdin HPAEs
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 conventional competitive advantagebut that the governmentcreated competitiveness in new sectors. Sectoral growth was stimulated in capital-or
technology-intensivesectors rather than unskilledlabor-intensive sectors that were more "natural" fur economies such as Koreaor Taiwan,
China, in the 1960s. Moreover,it is implicitly assumed but not emphaComposition
on
TableA6.1 EffectofSectoral
Manufacturing-Wide
Growth
ofTFP
TFPgrowh,

Economy

actua vae-ad

:TFPSrow*,,

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

58
result of static and dynamic scaleeconomies.
If industrial policy was important in determining the rate of industrial productivity growih in the HPAES, it should be the case that
a. Sectorswere begun or grew more rapidly in terms of factor commitment 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 consistent with growth accordingto comparativeadvantage.
b. Favoredsectorshad higher levelsof TFp,Af> A,,,l. Evenif (a) can
be demonstrated, it is necessaryto show that the governmentcorrectly forecastfitture growth of TFPin sectorsor thatit createddifferential rates of TR) growthresultingin * >A*t
c. There was a quantitativelyimportant introduction of new, modern 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 internationaltradeRicardian, 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 depends on one's preferredtheory of internationaltrade- Nevertheless,a
weakertest can be applied. If the evolutionof the secaoratstucmre of industry in the HPAEscan be attributedto comparativeadvantage,it is difficult 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 comparativead329

;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 content of exportsand imports, not directlyabout Ehe structure of rotal production. 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.

Predicting
the Pattern of Sectoral
Growth
The newerliteraturesettingout the meritsofinterventionhas not rested
whetherthe sectoralevolutionof successfilAsian economieshas differed
fiom what would have occurredhad therebeen neutralpolicies.Nevertheless,showingthat many instrumenms
wereselectivelyemployedis not the
sameas provingdat they had a significantquantitativeimpact on the sectoral structureof production. Even'whereeffectiverates of protectiondifferamongsectors,the impact on sectoralstructureisnot easilypredictablc.
In generalequilibrium,the movementof resourceshas no monotonicrelation 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 secror 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 period, 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 possibility of divergence. Governments also may have employed one or the
other as indicators in targeting sectors.Wageper worker may be translated 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 incentives.The static Hecksher-Ohlin model predictsthat for low-income
economies sectorsthat exhibit low physicaland human capital-laborratios 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 emphasizing increasing returns subordinate differenccsin factor endowments.
they are more likelyto describe intra-industry patterns of tradc among
industrial economiesthan the evolutionof the sectoralstructure of production of developingeconomies whoseproduction structure islikelyto
reflect differencesin factor proportions.
If interventionhas been of great importance in promoting the expansion 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 , supporting 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-addedvariables in Koreaare negative.For the period 1973-80, the one characterized 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
Shareof Sectors
Changein Value-Added
Economy

Period

IHengKong

Japan

WE

1973-80

1980-88
1973-88

+
+

+
+

+
-

+
+

Singiporre

Taiwan, China

VB

1953-63
+

4-

+
+

1973-80

1980-S8
1973-88

---

1969-73
1980-89

-;

1973-89

-+-

1966-86

(end) of

heperiod. VB

1963-73
1973-80
1980-89

Korea,
Rep.of

VE

Wu

1968-73

+ and - signs re sign of cDdenciLent

-

'

Signficncatthe0.10IeveL

Significnurartie 0.05 klveL

Significantat tie 0.01 leveL

NbOwtWSmZa Wage per emPIoyeeac the b%eing

= 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 growing. 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 sectors despite the promotion of ostensibly capital-intensive suzbsectorsis
that the major branches that grew most rapidly in these years, and indeed over the entire period 1968-88, were machinery, electrical machinery, and metals-all sectors exhibiting somewhat less-than-average
labor productivity at both the beginning and end of the period- Conversely,some of the declining sectors induding iron and steel, industrial
332

USING

RESOURCES:E'FPtC

chemicals,and paper and paper producrs, were sectors with greaterthan-averageu and va1, A more completeexplanationof the contnuing growthof labor-intensivesectorsisthat they experiencedmore rapid
growthin TFr, and the cost-reducingeffectof productivitychangestimulated 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 laborintensivesectorsmaintained their relativeposition or grew, (b) many of
the promoted sectors themselveswere not capital intensiveat the tWodigit level;and (c) subsidiesto the cost of capital may nor have been sufficient to overcome the ongoing differentialsin TFP growth among
sectors.Moving away from the anecdotesabout individualchaebol,the
quantitative importance of governmentintervention to alter the structure 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 sufficiendyadvaxced
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 permit, 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 sectoral evolution of production.
HongKongandSingapore.The cwo island economies revealdiffcrent
patterns of development. In Hong Kong there is no relationship between 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 structure. There is simply very litde variance to be explned in the valueadded 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 Singapore, 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 insignificant.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 cdifferentiate

besweena comparative-advantagebased evolution of industrial structure versus one resulting fiom intervention 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 neoclassicalexpectations. 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 governmenr policy that industrial growth emphasizedcapital- and knowledgeintensivesectors. The sectorsthat grew most exhibitedhigher wagesand
labor productivity than the laggards.There is no evidence of a factorproportions-based expansion. In contrast, the evolution of sectors in
Japan and Hong Kong does not permit a simple view of its determinants. This is nor surprising for Japan given that HOS is designed to predict 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 parterns are more likelyto have been dominated by technologicaldevelopment and scale economies. The failure of Hong Kong to conform to
334

USING

RESOURCfSts$r.

neoclassical
comparative
advantageis at firstglancesurprising,given hie
enormousimportanceof its trade. Therewas, however,relativelylitde
variationin the dependentvariable,the changein the sectoralsharesof
valueadded.

Growth
Patternsof SectoralProducfivit
This sectionconsiderstwo issues,namely,the patternofsectoralproductivity growth and the impact of the changing composition of productionon 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
sectoralTFPgrowthratesare
availablefor Japanand Korea(surprisingly,
nor availablefor Hong Kongor Singapore),but theseare two of the
three economiesin which activistindustrialpolicieswere employed
cardtingup.
mostconsistendyto achieveproductivity-based
Korea.
Calculationsof A*ihavebeen carriedout for a varictyof periodsby Dollarand Sokoloff(1990)and Lim (1991).What is mnost
strikingis the high absolutevaluesof.At i in mostsectors.Fromthesestudies
a numberof patternscan be identifiedthat are broadlyconsistentwith
one another.Thereis no simplecorrelationbetweenpromotedsectors
andTFPgrowth.Wasselectivepromotionimportantin the sensethat resourceswere shiftedtowardsectorsin which TFP growthwas (exogenously)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,utilizingthe A> valuescalculatedby Lin. We then recalculatethe valueadded weightsassumingthat the sectorsin isic 38 had been at their
predictedvalue,VP,on the basisof intemationalnormsasa shareof sectorwidevalue-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 percent 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. Indeed, a verv good case can be made that many of the major policiesoccurred 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 dedined 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 slowerwithout 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 machinery sector, the externalitiesshoulldhaviebeen revealedin its own TFP
value.
336

USING RESOURCES. -FFYCZR

Appendix
6.2: Testsof the Relationship
between
TFPChange
andTradePolicies

ECONOMY-SPECIFICTFP GROWTH RATESCALCULATEDIN

HE

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

- inrn

rwktivm US. I960

*
.:wwPu7
iu.

.L7120(0.)1483

lnnlem
I,%0

2.

SalmlSycndlmr9G.

-9)

51

67

*.634 *419
(034n7

.2.40r
OM095)

.5.0403
(2.3467)

.4.4383
(. 2599

2-2-1
0.6722)

2.0941(44757)

2-6731
(0.6646)

2A.1(0.65939

I.6WS
(5.5121)

1.6382
(L25)

(O.59

390430
11.1417)

2.604 *
(1.1406)

3.069*
(1.5790)

3.443'
(I.6325)

015543
(LI441)

L52Z
(0.150

a4u6
(MG02M5)

.356
(e.ecs

1.1991
(1.045

(0o1073

-.

HiAtt
-

r,-

M)uud IV

LOW60
(2.55)55

047!50

(0.13Z)

(0-23712

025D0
p(0593

0.0255
(a.00W

E0L=75
(0.0073)

at7G
(04261)

55
(0Q4531)

CO50

-0

Q4457

0UA

.604

OD354
(0D33)

%0641
(.U0246)

M0
54),
CAW67 -

aAIeSs7
4817.
(0.50

e70

03673

Mba. Sma_rd effrits

KNG,'tr.
N -- >-- ff-

0R5579J

(24.A871.

-53.0356
(t4.5

.33W
(ld0219)

2399
(2.0633)

004
-

*-5L119I

.*t3538
(2067)

t4943

mliauraape&.n
apomn.
I9GWS

5i

4.584
(0A3041

*5F311J*
(189294)

3S53
(0.56501 (

AmW

53

-I .376(0.437)

49925 (0.9791

Dallwqrnamkrla

*'

69

St

(035712

:(0.573C

*-

4&5615(135162

50A684
23.W095)

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)
ofmovingfromlowerto highertechnogainsrealizedasa consequence
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 externalities 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-enhancingexternalities through the better ability to use and mastertechnology.These externalitiesmaynot be capturedby rhe human capitalvariableand would
then be included in the estimatesof TFP change.We attempt a similar
rest for secondary education.
for
Openness to the world economyis another imporcantcan.dlidate
explainingrapid TFPgrowth. The theoreticalcasefor this viewrests not
only on allocativeefficiencybur also onlextemalitiesassoiated with
tradingactivitiesand on "X-efficiencyfgainsfrom creatinga morecompetitive 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 international relatrive
prices.
TableA6.3 presentsour restsof the relationshipberweenexportsand
productivitychange.We chooseas our measureof export performnce
two indicators-the averageshare of manufacturedexportsin toWtle62 When the share
ports and the share of manufacturedexportsin GDP.
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
Industrialeconomie?

Atcunm

1880-1990

Indury

Serikes

Thoe

0.4

2.1

2.1

1.2

1.1

3.8

3.9

2.0

DDcvropingeconomicsb

1960-70

a. Developing
economics
includethecountiesofAricawiththe exception
of 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
m

a~~~~J

bawwnjtunkt ?tlautwrr
'
nrmaBen Us tanRg Uqunc d
by do m. neei,i-g

bush sad

inyharlw.
ISMNWIh'NZd
kshn*qjuio.n
rwniy
timenateAmhek tug Kn
-Mndab,odbsnaid(M.AI-fIAB
dmmcniewairAn

nsw&ayrndeumciwa

swjydryc-nnpa
ben

atn

Theen,ttiian
ints

-- :

: f:

---

'--

, Lnalnbemi::

nmppoadi

-hn --

i-flOAbbidIwbbnS
amaiS

IX6-9& Fad trsrininen

thituNialneutinia tim

ledsnso

ksduapnd

hSI-UniWatSde
umuck
adanhis
d
sibicckExbnGDe,nnilmicce
u-

iltrHn
tngLn
dewlukInde=

FunnUmbw

Oft C:Gti atm

- ltea
Anie

pe of

num.

t
It ini
rels125elmns
panytq.o
- 6..m DTLCsd iumda
ar W mn k
ionm
eymm.totx is ,kni,a d
195* Etsd fdsetl9dwSuikmadl Fume
CRn
Onindelesun. iehuuansiiYwEmk
aa
Cod eFda Edung cmetd puidestpeindn

:CntwmmI
baokcahmn
ntto
,i
1tu-

4apter
in

6ih-k-m
hi

awstdasd

&g.

anchahe
s
bends iuqm.

enag

'

bcnriitss&s.-.

ut
pin

.3hiselnkid

innS dthw

se
epdm

tsaty

S
aujxMEn;MCmmi
'w u
capiaIcstwmrlnsfiS.
'
MJ*b
lksdiJ*l 6anchintnnas.
rkaalidlifirdismacmor
stm puivute
huskid widfir4ig

tcru

-nqtsin'nionvi_en-

SW:Smtuib aetbl"i a] 0
line in na. inplnst nalFuiwtnats-a
eneaominmhemsy
ainmn
tpedenn
dhbsufl4hiWntoteoIbesL
la
ualrB&)inua&cnL
ruw
1Y-21ailnz_
r c
AMS,EstdghAdNWwr-ducmimmi-T
INnamDU
finnkfj,rlnsnntkst5ankmn
(HI) mki&
tn

O
pplic

)etanassp
,dbmk&mxli

,l,ssi

dzctc
-

f&

-fitovr

sakvi

b6iIss Fd6ER.PwdioInwili.
ctadseopintrtCD4)

annl.

otnoMpUifMttI Ad

hEn.r.

i(DtC4md
Qinkiua

dumt. .syaniydsngd*
Usn
RadA Reuian LiCe
Lnb.aid Dlr
At danethire

uw

'wedaspheolaiu. Cmavitccilnp
ahh&C
,, -baksiiar.-i.
u
dw p- l-dia wm &LSwpAOt
qykriogiDI
mi" ,
'
lmorntd1rmllysw
nadce1Jhun
trdnefidlkiusndtelkisteel
- m'- idadwiegnn

dinm.

,;Mstnie pidhoe

enom (B
1a lmd
25d5f-imdtrnim.

IMtlnsnemutermsicdsmanitnl

depCSfbUIadruttnlub
luta~w
Lrmnamo(iutehab.

ITden
dwpntncdadmifillsiA
iumadiaitsdwaWnim
ILkand Uno" ikerdflq'ek.it

/ort..rDatEd mknatu
39Q,
he irans.
fer
liauibsnbe fRn HlU5et to Hls 1in litA
murafltk
t
andscowinnisldu&
y,'% eiW-Balimcomuimnu cirye mebawithdie
sin rdIasEqking,ts
G et
qBiz.Wl.tobmU
mnndma%imaams nid
hT,
dti-nirnma AiedmdnpduWed

o inte
fini ecano.
St
imoatnspipsad

. IZRninddasSihcF'un'au-rmk dsq,. Eri.b

vith

w-binnlnge1yforbamla

;nrk"du

'

bkh

IVA4ime-wgnla

9IS-Tbeerrcrk
cadmftgeablkxd
19P3I:1iedfeecI
odue
aesnu -r pin
Hanl og

tqir
Ddgt-TudSC
sii
bynttnn
kn-

gstdi

!tmoba..It I.L

rihadjrnnnat
TuingC.o

~~~~nsaa
h- (3W0
snunreyA

.fflaw-dd
-In

Tudiduifr
r&d
hum 6W awi
eoniumNMmforfrwva.x. Yeq ib
n&kr inarusul bsdznmgpam.
Aldwien-cking
kn.utim we

19*
6kTn
sop innrvmewidn
5
-i nenuis
dtpma .iln - undo th
dimisonJibe achuop wlln6
A-dairt$EI
byvkita&-s

Jfrg fly

)Isrssien

N.o

dRtel
puMladan
hd
pwrdfBtk.w-i eye
nedmw
Laureleng

~~ ~~~~~dcwys
tna..sTsdahgaingaina

(^~~~~~~~
p:Xd
im. _

aqu
muse

quoky. Pa-

- inaI
- :Eri*ntd
.
: Ninmss
- -- - : :-.-

SwitinamdlEueGCmrm

--

lil

nde

h sh &eupa

plscigsghe,
nw.rWa'

nmh339

339>

MIRACLE

iASJ*

(confinued)

Table A6.5
Fi_nes 1

tflnwu&gku

-fnanq

.f?e.emw

hmaClg7talwdmailunlmeuhaa
(I 91
hig CWaet maeberamtnls
(bean.
'ngleanEcesnlldCa
Ml
eF

e-

ra(nS).1

a'el
amarke

JImeqIad

ruin

Lading maawarmneguled bn llealminks..

Idewldeindroddidiol
Dql dr
I manimu sep. Oqee
weprr&ulyco-IaL
Iw pgalt.
lPf-lSC:drLl
lilea.
-finlrualnefetreadensnl
rm.m Feed bawur
hsr
lotion ofin
munibehinmrbskalkudIl b mr-

Jq_

queamalAa4d dqeie rgetrorminga

nemyskLst lhew
t9t Rea e
Ilntve rum
ncmlad.

~n

Ifl&selandpwLeaand(pualk
I"h ly bn6 pwnal.nLapuend
ceargnadrMh,lmwL
by Ereiga.
wesneJ
TniT
uu hu
IJ
hakbd
boan..
lWtSinI9 NWiedlum
btw6n(Seg)wem endvn aul
BAten
Inf:

te
ro

l esgi
Batik Iti

incs

sudiarlsnlf iea

s thn tint JrpaI nraipal.

elek
bfldJUb-& EarnWerd
Tdkyq DrLlr
trannd
lmae.
hebd
petin
I97Puealing
eqeriancC ietCFi) muSs
mpl6.
flfaRnictiuon.sinurinral
eiclLd n.UhULi
/misarkar(pulickmmdAutiuL
I9eat*Gw nubeaeeet a
u
.n
nrlca andt
ISW&T mmuyiuco

ng

nwaed

fi.rm

SrngpnErd
Feuip atitia
maianheniphntheTLy.Snk
wme
m
ExdaKJ.wafuned

-tan.

p4mt
Iwkueai
Sesehreem
(CC)mkask mmneiL
.trdin

thteUs

mya
andns

IVRf=

M."
_y

,fiecu

flMtAlhyecd cmey.C 6lr

wIdch by IPRI wee ulpalin anall
qrml fnsinguith ldmik bcAl.

tse

is
bue
witi iwn
The MinyfFlhlna
mpmnimry
eauu
eA
U werwik pwns fr. the akae Lwa%
e
SecuidmitedadEhnd LnTreeleandlki
.J.lr
. _J.mlrghti6d
J
ELASp
mpAmdeycaie
eunainenda ituahuks..
son
m
pnd udea
iItStnMU.ddl
iaJ
ndwepirle
dia
Mea
inada
In
Dq*
191: EatbiMlu dthellWd
prcedes
Coonamtidenaeda.ceqnr
I...caaaar.L
cw lon&
tN
Itnh beanit anenimpat herniae law
iftiUenUIStrd.
r
ed
u lanird
ag
I

Sequlds

Edia nse ered. Dan usMes

rf nd intne arn
almond tv psiidpau inh m
Ca,ndial hajneaaMrt
AtktSmckindfis ma wading
ru Minr
rfTey
un.mj
Ms-Tkofiac
5ketmaut.
6at
r
haairnl,n

.Kern

4p.f

1* ll nareuanladlaetpm hrnitinyalaminhinld lrtam nem Rorm

ttacauac rmcatlain1973.
IPRAWlderadeiteecinap and
ne.ro rncdveedulapd
il.l Jdaapreliratth inse
l.
-bell
19-* Minfimumed mem&nn but
incnhAwL Isa
ecago
won
h
nae niuluak hding
mee.tr
n enr CD.
uWhi.Ll3abnUnseA

~~

-~

-r msigid.sunfreinrdepode
=
tnt Mm bekake .ukflntin
b he.deeetw (NR loaig
eiq
Dil insmialen (NOR) imnm niem
dcpdtmc a
MON li

mlnrt

ihmsm.L

gd

1W

6r

maini.

IA9 rlhid

.a

CD,

:.

:r.'o-Sf.th,

Ilsotlrfeaciual eapelisn t
ie td l.
i
pq-e takre
Jilt csneni
cualdieetI.Setupcstmeaed
lieiwg
ifnhtet-Sr
i;1..
s up OTCtae
Ieefarkinu
anM*bUn ieUd ahea my
I bgewen
. PIa
po4u
nm.nkn-aae
banldlagnrld
)mPAwdl kgwidaehID cace?
JW
lamemat lnaancnmpudics ate

dqa.it inasunacule .cs.

InEudAaIo
smt
tt
lRni6ill
J9n1f lv
e-in grcyupebieic-s Iinulcail
ide
d'bg
cheOQeeOf
Madudeald
mt mnd
Eeaneiicn (ONSE ID enmdu Rn
&eraidwa
e
wauphiias asally. Cpa
bnY IWSd Xlirlilnab
rits UlflnS a
i
amkte
peta
nsi
e
cpened
na
lee
i
uhefhbne

scaaelckshanSipuaeauadmaqqwd
licthcc
JNIPouaiaac
fmlap
a

hey.mkannrioait
atcljeeitk
inpdoeiii
ipe

roNaliq_iIFi

ii

r-oel
vnXd

iFa_balloerd M eAs

_:,_

340

340

&i*.h4n

NllFIswrb-UL

dkpoir

A-m

wu: end vmleuneetaStcdad

csneed
t% t4iodd Inla
eptoh
lJuCaeGnrt
cheat
eua-mnani cn?i.eri
IWI-AS:P
nwn,61 Udissad ailowdaeyof
rewdmmicaeidfrnephink.
eagn
lt
fwaegunla (enenere r`fieigs
te
Bn IS is I
btd.ia iama
incndtudi64 in 199li. Deei. p
if cngrefAmuiei6c%ukhleaditccank%
erpsdamenprctr.nbuineband
Noa atw bankas ma any &W
tna.2
qchncAhelFlsmeded
FIksaJ9c.
s
ewiInen

.-

la shiauia
iinpne
mhlhiscynrf inae(MOljwapeniassal
rma pe.tionnsftWBk
pn&i6l g;dd

fahnv

in rej

rnd
a p.*hmfia
arW
r=;o&m

f6rtLirg'

USING

~iapig

1971:hmnaei aianhrfind&pdewicf
nmmzho flmyw,g. hiajlty 6cua
ffkb loUtes aam.csa11icMFumnl
Inkik&p.sitmuinr am pr uawiru

aM.opta

Mi.hudantf194hmoaku
haiLbmab Howemi.dndc
bmil kw Inured aainsO
theyhawha. napmdto insni
w6iiu ne&lernn

fl%-KmJus"s
Li
cIS&E
w nab&hd
NW
50, &Ldqwia p-cer.
wmardubinmnbedq
rmaota
nmnduat

if LRid
hJAiog&t-eonpickalyddsdn
p &hf.
Idga
Jtaincfbhulu prcaci, of
dqiWmnc anvocrUe
i maumpnamu
andaural of
9
buil
jMs--aoVmei.fUanaic
6i.& mnd0TC

1374-KNew Anaidelimuamsn

ngmitoinhclupot..49 prenm

139W8
h A kaC.1'dohia Wiath
gasiddnmenogle
mu' KManddhcfamdanX
AnbudgesZudiddspli
dfitpakiLm
rordineindng Ktndialaw
1398 mALina%ar
Finadi lasituhm.Ar-~cwjw

~~~~~~~bn~nchh

(End.
nhh.rCmbsponlosasd&4
57&MA wnianuadu
Laekbaing

Ewa

.aac ouhAseleiA
CD%
mnd
hr&r
sit
alibe,tindaweL
drpca
aaqumcninamn
Flame.ma
h1 Bwm.l melube.

aua0cmdrm
n
oa sriau
19~~~~~IN-%O,SshSab,rdan
cff-,bca

tifan
manroe

plfthklAngSram

biWngto
1 UL

-)flCrmmrulbmk-alminhunced Bar

phmuLSac
cadlgeuzie

h Sfiogper.l

~~~~~~~~~~~17r0
asa
hraiaglirnbna
aiiua

aWeCD*

ro

frran

t^ii0ilbding
Sk~~ar

CLCruStbaku6,

iuqMaCCaKrhia.
1:&ibdla
c ue,pWcctc
'n..-taupary.
2dIcrjta
/>sindl
3 aflhlsc

LadIng1atehrodcednmanlhe.
199-.(snt ISZ211lutternilaprnn
at liamfln
car-

RESOURCESSO

sliad6

poad rtheakingAadZ5l'ManlFonmanceGimpma
Act
p
wdn f
SoL:i..tm,.lacu
.1M.
lc&g guidolina
andactig awtdiinmsu.darda&afiaaCd,hainnd it

.aitardind rnamisffi.aeWupcidg anLinac

so pwpawapi.Mcnd
il
INtn SitdiEuaqrlam

- dhr~~~~~.M aJeimonqefi

and -gi

)ni cceaDTlcsdbi

vm&g-;ia:-r-

tnO

-w

ii7kCamplcc~fi-Nrad libaclin-

IWI
9Aanmoderelmcnminoal

IAf

uam.Bankouitn

aa J&odeurmia.

ofbanWigumwawh bklbal p-x

J"ifSupwpurS&Exhap

kg fitcennuccd*mow dqaodtand
IcadhguotaJimam Rial

aintuudrlpmz
1
aftoe Ads=ODi. AdLs
I6IxMA a

I=
9laUhlmI
dwSIngar pveI.
a!,1niidloocuiy
E
aFeat

maicsncfainea dut baki mow mimutWit.

BabigAauet
oidha.apiiada
aiuhnerma,tin.d-

etehaFownand watplevalldratmarlod by 197K

nEabat boideg maskts. Hi limema

in shyumgarn to.aADwiplbikano

MqencdcwlS1VpmrCG&1dEn
dbze.pupcdatnimnmsd

Bu ala r1Mnms 10 .liqddkrruaz..h*j

baeulamts
adautraiomtea.opm.perJlomauidlnid

--_iona
tSlgt

had..rbankiang

ai. DAr Miadot-&lai

lnnard
lacun

banlh Fontoeadat,i oaptiax

flflSWtFocf dooic

15M.kEorkWhh MqntuAki&
or

iyaVSaypm

AlonigwmAt
canal bau'ngAocticm. mope-bnou-

and
Mlaa1fi
Stock

buk par-

lanti

c..mnnpecl

in apitban orfbeeW

c L

hniari

mir"uiomefluainidey*riwai
pIMatler4l

inlocalboaokomps.Poretbasik

oatior
Jfmgn t

madlbomgarvhow, nuk(hd
tacnve tehdi,19 poite sula,

re...

1950(Lirbclsd C

aim.

Ckit
at.

194 lAI,Jnaamsaimds.adinisnladugawa mnd:auccl.

as. da
B
ncc.dbaglachain
lylSCosa&lrf
hanktulnuadM
napinlene maewi*inanmnl
bank-adaalnimndinapt. Rrmbrd
delnt an Piamcwrugahndimun
hidinpia

*drpltionsadlarcog.it
iSa
meien

kb;-itpebid.
*lirinn
I.Ab

rofihltDam
reTolwet
awidr uatrumap

.aatnr
*

-aciallraktimcrpoitas

g~~~~maetmaa.ryraL
* Imhbclitttallepa&
nemlbnga
-n aaadalbaukaledimcdepm-o
.bat anepm.Abormiued
.h&dbAlh,v&o.5oa.dii
hl.z

T.iwao19t

IQiEahlsbehrx
Suit.tud.a

t mccoy
nTh.alc
RappmhscMakanr
mlet11CP. CDaamnW.4Atwit
awknntctnwinal. inauacsmm.
1956 Craed .W.&bovebadaig

long-mmlendin hwodam.

N Cepnui
c.rtahn
naaapmnicaanal
InacknEacsof

ioadligrolumbrod.
anl

nof oiunp

Jwifletnnrymecti6pkbocWitnca
,nau.rin. bmI ficoInuSpnnz hid
iatyiuedbyuatybw

73&11J

AttaInhit
1 ban mw
fmnddradcpitbd
d
a
e
ow
I*oaow
r.vlmrrial
aJbn
padpermttdartoiWtid.scutlq6a1aw
.ppmnL Madaadnsmcuicaiomfiurfoaig bade.malm)ined

CdllhOGtoMldkrhgaenIf1
ew .dn
lNtPcekd
1562habonglwaIn

bat11strada

;
uclikehagnharm.aidndunl.

r-

D.SsxhiS ad Exitnu b
Crwnuiona
ucpbcnm.pnvreaepital ni.arktpeaaks.
IWi tfmbl&d aedqmiIa rpnce
copses .
lflnlSboib rh
tlni
ucccmopatuy
Zn6
IN*-JflA.."dclSriabTrmanatdn
I.A. t
acpdwMaaenrmandb.tnauborahmIthan
ainp.
Raiwd banbinglausnD
puohlbktcrdiemn Iiewma
kdaiuAwld

maas

mW

ITTmanrortoa
ltpleamatedo
ate.
f/m"an
gto
purcm fiadand a
ansby~

EnltaqccftluiIp

trdnStds

J^*LdC>

FanceCsNp,
ticrmuAdpmdu

/Is*hrbeglarff@6f:-_IT
-n
/fftSBpS:r3

caabl Lsa
II"urewse
E uubcd d.c trer;Anbc

Rtrctcioa;fr
titmmucaL

Ca. maktow

heiJMA&iur.jpnba.ndioemc
iH mintd-i
machemmanh.wiOgdiwtcndhac
Oay.r Pa
cdt cmibriuinafocf mersan
nd iea
acsaaahimunsdrdlpih
wmaunaal
mi5i
dnfitlcpipntadtc
.na.Xcmain
q-yydm.rllaoap
sIrkn

:flhauzJMvs

baJutk
Ret.

ti.dindopMMimdaanigf

~~~~~~~~~~~InhatWdcCS.aI

~ff

WDmillion bOIL

am padh

clicducaon InF

ISthSaundsandSabaa
-

:~

,9,d.i4

aada

~~~~~mn&,uno
ltmcl napdcnacuaad miumntimarrff

113qsskd.p
startd buioandoin 197tkdua.
pw"-Ph.
-

,,em ledwiard

0am.d

cmasal wnit

--

-*,-*-

5jE
eMRSCINMRALE

Notes
1. See Greenwaldand Stiglimz(1988) on the effecrof
coordinationfiailuresin labor markets.

7. The costs do not derivesolelyfrom the misallocadionof labor impliedby gaps in marginalproductamong
workcrswith thc same human capitalendowments.The
rarelyexceed 1-2 per2. Altematively,
conscantrealwagescouldbe due to a re- costs of such static misallocations
duaion in surpluslabor(Lewis1954; Ranisand Fei 1961). cent of GNP. Dynamiccosts can run much higherwhen,
for example.employmentin enterprisesis excessive(im3. A populationwilldoublein 87 yearswhen it isgrowv- plyinga wedgebetweenthe wageand the marginalproding 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 cur8. A labor market is segmentedwhen some workers
renly high-incomecountriesin the nineteenthcentury.
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
9. Clearly,a high share of the public sector in wage
Thailand.Indonesiaand Malaysiaare forecastto growat
higherratesof 1.6and 2.3 percent,respectively.The share employmentgrowth is not definitiveevidenceof a relativelyhigh share of surpluslabor in the "modemsector.'
of the populationlessthan fourteenyeawold was31 percent, circa 1900, in the currentlyhigh-incomecountries. The publicsectorcould be findingproductivejobs for all
its employees.
In 1989 it was30 percentin EastAsiaand 36 percent in
LatinAmerica;in 1986it was46 percent in Sub-Saharan
Africa.
10. In more than one-third of the economies,private
sctor employrnentactuallydecreased,implying that the
public sector accounted for all the observedincreasein
5. Increasesin the rechnologicalcomplexityof nonagriculturalproductionand in the sizeof establishments total wageemployment.
bring greater divisionof labor and, hence, increasesin
skilldifferentiationat thesame time that greateravailabil11. While there is no doubt about the roleof governmeat in suppressingunions in Koreaand Singapore,in
ity of educationalopportunitiesincreasesthe supply of
Taiwan, China, there is some questionwhethergoverncducatedworkers.
ment action, for example,
prohibitingthe ChineseFeder6. Labor market efficiencyis assessedby comparing ation of Labor from engagingin collectiveaction, was
marginalproduczivitiesin losingand receivingactivities. decisive.Market forces,and in particularthe sizedistribuworkersin Taition of firms,alsohelp explainwhyso fe%w
Evidenceof changesin laborallocationsuggeststhat individualschooseamongalremativeoccupations,industries, wan, China, are union members.See Fields (1992) and
chapter 4.
and geogaphic areasas though they had a fine regardfor
marginalcosts and benefits.Ihough behaviorconsistenrt
12. SeeFields(1992),who noresthat until 1972unions
with rationaldecisionmakingon the part of individualsis
a necessarycondition,ir isnot sufficientto ensureoptimal were freeto bargainoverwages,whichwerelargelydetermined by marketforces,and Freeman(1992),who, refrallocationof labor. It is also necessaryfor workersto be
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
deliberationcouncildiscussedin chapter4-was useddurworkers'associations.
342

USING

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 incompetent in economic affaiirs-vide Burma in East Asia,"
though he recognizesthat this generalizationis not an adequate 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 response to unemployment the creation of public sector
jobs with zero marginalproduct an overlitde more than
a decade reduce a moderatelygrowingeconomy to economic 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 universitygraduates. 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 industryin the late 1970s,ir wasimmediatelychanneledalmost 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 equivalent 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).

RESOURCBS$T:

19. Other evidence on the extent of subsidiescan be

obtained from the reduction in financialburdens to enterprises that borrow from subsidized lending programs. 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 benefits 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 exporters receivedsubstantial interest rate subsidiesduring the HaI drive of the 1970s, but these largesubsidies
havesince declined.
20. King and Levine(1993) also show that high cenrralbank sharesin rotal credirare bad foirgrowth.Hence,
in chapter 3 perthe macroeconomicrestraint disLcussed
mitred rapidexpansionof private crediL
21. When exports were not used as an explicitallocaion criterion, Japan often used inremational prices as a

stndard of good performance. Its industrial rationalizadion 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,machine tools and steel).
22. The meritsof well-managedand fbcuseddirectedcredit programs have long been advocatedbv Japanese
govemment officials.For example,a study by the Overseas EconomicCooperation Fund has argued that government involvement in directing crcdit is warranted
when there is a significantdiscrepancybetween private
and socialbenefits,when the investmenrrisk of particular
projectsis too high, and when informationproblemsdiscourage 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

the maintenanceof macroeconomicstabilityto minimize
distortionsin incentivesand on effrive monitoring to
ensure the timelyrepaymentof loans.

menr, sayingthar if a firm reallywere to be profitablein

the long run, it wouldpay for the firm to encounterlosses
today.But this counteragument runs into problems:it is
basedon the premisethat capital marketsare perfect.

23. Thesestudies are preliminaryoutputs of a World

of Credit
Bank researchprogram on the Effeccivencss
Policiesin EastAsia,which is intendedto provideempirical testsof the effectivenessof directed-creditprograms.

31. Exportpricescan be set in the international(unprotected)markersat long-run marginalcost. Is is possible to show, under certainsimplifyingasswnpdons(see
Spence(1981), dhatoptimal production of the firm entails setting the marginalvalue of output cqual to the
long-runmarginalcost.

24. Among the best known of thesearc Balassaand

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 Kiyono 1988;Komiya1990).Althoughthere isdisagreement
linkagebetweenneutral incentivesand growth.
over the extent of the reductionsin effectiveprotection,
there is agreementthat effectiveratesof protectionwere
25. For an interestingattempt to test one of these
propositions-extenmalitiesaccruing to export-in a
substantialbeforethe early1970s.
gcneralequilibriumframework,see de Melo and Robinson (1992).
33. While effectiverates of prorecrionor subsidycalculationswouldprovidea better measureof thequantita26. Among the earliestformal discussionof market
tive impact of protection, the figures shown furnish a
notional indicator,consistentover time, of the extent of
failuresin developmenrareScdovsky(1953)and Chenery
(1959).A thorough surveyof many of the issuesis given
reductionof nominaltariff ratesand noncariffbarricrs.
by Corden (1974). More recent discussionsincludePack
34. It would be more desirableto compare efnective
and Westphal(1986) and Itoh and others (1988).World
Bank (1992c)providesa recentsurveyof the issuesand an
ratesof protection,sinceit is value-addedpricesthat matapplication to the evaluation of World Bank industrial ter for resourceallocation,but the most recentsystematic
cross-economystudies based on comparabledefinitions
developmentprojects.
udilizeevidencefor only one year in the 1960sand early
preclude the
1970s Thereare fewrecentestimatesto providecompara27. Economic growth will eventuaHly
need for policiesto capmre the producrivitygains from
blefigures.
economiesof eitherscopeor scale.
35. Singaporehas been exduded since its zero duty
rateswillbias the estimatedownwart.
28. For a reviewof many of these, see Stewartand
Ghani (1992).
36. Pack (1988) expresseddoubt abour the robustness
29. In general, real externalitiesprovidea validarguof 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 sociallysuccessfulinperiod 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 capacityutiliza30. More recentdiscussionshavediscountedthis argution eliminatedthe residualfor much of the period in
344

USING

question.Additionalobservationsare now available,and

improved estimaresof the capital stock make it seem
likely that the measuredTFP accurately reflectsunexplainedgrowth.
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,however,arealsothe sectorsdtar Dollarand Sokolofffindexhibited 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 Herrara,forthcoming).
43. Undoubtedly,domesticsldllswereimportantin allowing the technologyobtained from abroad to be utilizedproductively.
44. On the roleof infi-astructure
in Taiwan,China,see
Ranis (1979).

RESOURCES

in Japan, Korca,or Taiwan, China, but it has been very

importanrin otherHPAEs.Evenin Taiwan,China,DFIaccounted for 15 percent of employment growth in
1975-84 and demonstratedthe viabilityof new exports.
Moreover,muchof the investmentwasin relativelysmall
firms in which employeesvvereexposedto new equipment and the relevant production cnginecring.Since
labor turnover in these foreign-invesredfirms was substantial,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 import-substitutionregime,an exmple of the too oftenforgotten point that protecton has both income and
substitutionefects on theeffortexpendedto reducecoscs.
In thefrms considered,the substituion effec dearlyoutweighed the income effect. Nevcrtheless,calculations
based on most of the studies suggestvery low rates of
growthof capitalproductivity,an exceptionbeingthe results for one steelplant analyzedby Dahlnian and Fonseca (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 developing economnies
increasedby two-thirds, yet in most
economiesthis had littleeffectin stimulatinggrowth.Educarionhas a smallereffectwherecomplementazychanges
do not occur,thus, high educationmaybe necessarybur
is hardlysufficient.

45. DFIwasnot a major sourceof invesrmentgrowdt

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 opennessvariableDollar'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 operaring at export prices;in that case,the share of manufacrured 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 hypothesisthat, takenjoindy,the coefficienrsonthethreevariables (education,manufacwured
exporrsftotalexports,and
the interactiveterm)are norsignificantly
differentfromzero
(taken together, the three are significantlydifferent from
zero at the 0.01 level).Likewise,the Ftest rejectsthe hypothesisthat, 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 separatelyinsignificant,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 movement 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 constant 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

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 fundamenals 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 constrain fiiture use of the HPAEs'most successful intervention: the exportpush surategy.

Foundatons
of RapidGrowthGetting
theFundamentals
Right

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 accumulation, allocation, and productivity change. Here we briefly review
some key aspects of these policies. Although implementing these fundamentals 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 necessarypreconditionsfor rapid growth in all eight HPAEs.The keyelement was the management 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.Moreover,budget rules and legislationrequiringcharsovereigndebt be used
for developmentexpenditureshelped to ensure that borrowingfunded
investnent rather than consumption.The successfulrecord of low inflationand sustainedgrowthof the HPAEsreinforcesthe viewthat there
is no substitutefor fiscaldiscipline.
The HPAEs' successin export-drivengrowthwas predicaredon competitiverealexchangerates.Prudent fiscaland monetarypoliciescoupled
with flexibleexchangerate managementwereusedto keep exchangerate
movementsalignedwithdchangingstructuresof trade protectionand inflationdifferentialswith tradingparmers.The HPAEsavoidedstrategiesof
macrocconomicstabilizationthat stressedthe roleof the exchange rtc in
breakinginflationaryexpectations.Rather, exchangerate adjustments
weresupportedby expenditure-reducingfiscalmeasures.While Indonesia,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 enterprisesoften 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 protectedessentially
conservative
macroeconomic
policiesby limitingthe scopefor politiciansand interestgroupsto derail
thosepolicies.
Building
HumanCapital

Wehaveshownthatthebroadbaseofhumancapitalwascriticallyimportantto rapidgrowthin the HPAES. BecausetheHrAEs attaineduniversalprimaryeducationearly,

literacywashighandcognitive
skilllevelswere
substantially
abovethosein otherdeveloping
economies.
Firmstherefore
had aneasiertimeupgradingtheskillsof theirworkers
andmastering
new
technology.
In addtion, rapidhumancapitalaccunularionreducedincomeinequalitybyincreasing
therelativeabundanceofeducatedworker,
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-scale
importersof educationalservices,particularly
in vocationally
and technologically
sophisticateddiscplines.In addition,enterpriseleveltraining,muchof whichwassubsidized,accelerated
technological
acquisition.
The benefitsof dosingthegap betweengirlsand boysin primaryand
secondaryschoolenrollments
havebeensubsmntial,
evenin economies
in whichfemalelaborparticipationratesarelow.Bettereducatedmodtersraisebetrereducatedchildren,complementing
publicinvestments
in
basiceducation.The rapid increasein the educationalattainmentof
womencontributedto Ehe earlydeclinein fertiiy, whichin turn resultedin 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 deposits 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 secure banking systemsthrough a combination of protection and regulation. Protection of establishedbanks induded governmentlimitson the
creation of new domesticbanks, the enuy of foreign banks, and the operation of competing non-bank financial institutions. Regulation included 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 programs as needed, sometimes,as in Japan and Korea, pressuringstronger
banks to absorb weaker banks. In addition, many of the HPAEsencouraged savingsby creatingpostalsavingsplans to lowertransactioncostsof
savingsfor small savers.
Balancingthe advantagesof protection against die need for compettion 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. Governments 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 developing economies,which tend to be subject to extensivegovemment meddling, 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 economicdownturns ratherthan die willful default common in economieswheregovernment officialsare closelyinvolvedin day-to-day management.
Limiting
PriceDistortions
Despite the HIAEs'many market interventions, reiativeprice distortions 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 protection 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 internationl prices than the import-substitutingeconomiesof atin America,
South Asia, or Sub-SaharanAfrica.
AbsorbingForeign
Technology
While none of the HPAEsexcept Hong Kong and Singaporewas entirely open to international trade, al were open to foreign technology.
Japan, Korea,and to a lesserdegreeTaiwan, China, relied heavilyon licensing, 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 bundled with technical, managerial, and sometimeslabor force skills. Licensing 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 technologicalacquisition suggeststhis need not be a seriousbarrierto growth.
The alternativepath of self-reliancehas litde to recommend it. Some
developingeconomies,such as Argentinaand India, have restrictedcap35'

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 limiteddirectand indirecttaxationof agriculturemeantthat ruralincomes
and productivityrosemorerapidlyin EastAsiathan in other regions.
As in othereconomies,agriculturalsectorsin the HPAES
werea 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 investment that led to very high levels of private sector-led growth. HPAE
civil servicesrange oromthe highly meritocratic and insulated bureaucracies 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 technocratic 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`

Second, in the HPAEswith high-quality bureaucracies, rules and procedures 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, discouraged 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 communication with the private sector. Formal deliberation councils esrablished in five of the economies-Hong Kong, Japan, Korea, Malaysia,
and Singapore-included government officials, journalists, labor representatives, 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 economically 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 rentseeking. Information sharing made it more difficult for firms t curry special favors fiom the government and for government officials to grant
special concessions. Thus the deliberation councils helped check opportunistic behavior. The deliberation councils in Japan, Korea, and Singapore also performed an important monitoring function, assessing
performance at the industry or economy leveland administering contests.
Consultative mechanisms in Indonesia and Thailand were less formatl, and their role in facilitating business-government communication
was much more limited. Malaysia's experiment with deliberation councils is quite recent but appears to hold some promise in the area of
macroeconomic management.

Inteirvening
in Markets

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 char353

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 economies, where interventions have not been consistent with macroeconomic 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 therefore 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 developing economies.

PromotingSpecific Industries GenerallyDid Not Wark

Industrial policy narrowly defined-that is, attempts to achieve more
rapid productivity growth by altering industrial structure-was generally not successful. In Japan, Korea, Singapore, and Taiwan, China, promotion 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
comparative
advantage.
marketfoirces
and factor-intensiy-based
It is nortaltogethersurprisingthar industrialpolicyin Japan, Korea,
andTaiwan,China,tendedto be marketconforming.Theseeconomies,
indusalthoughselectivelypromotingcapital-and knowledge-intensive
tries,still aimedto createprofitable,internationallycompetitivefirms.
Taiwan,China,placedgreatemphasison pricecomperitiveness
in evaluatingthe performanceof its largepublic investmentsin intermediate
and capitalgoods.Moreover,in Japan especially,
but alsoin Korea,institutionalmechanismsthat stressedcooperativerelationsand exchange
of information betveen firms and governmentintroduced a largc
amountof marketinformationinto the processof industrialpolicyfbrmulacion.And the yardstick used to evaluateindustrial policies'
success-mainlyexportperformance-provideda markettestof the success 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 promoted industries and the use of export performanceto assesspolicies.

Our testsmayhavefailedto captIe one importantaspectof industrial policyin Japanand Korea:informationexchangein a competitive
framework-in short, contests-may havelessenedinvestmentcoordinationproblemsand resultedin earlierand betterinvestmentsin what
ultimatelybecamemarket-conforming
sectors.But among the HPAEs,
onlyJapanand Koreahad the instiutionalcapacityto organizeand refereethe kind of contestsusedin formulatingand implementingindustrial policies.
Despitetheir potential benefits,collaborativearrangements,especiallywhencombinedwith accessto scarceresourcessuchas foreignexchange 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 repeatedrelationshipsbetweenbusinessand governmentand the useof exports,not
domesticGDPgrowth,as the yardstickby which the successof the industrialstrategywasmeasured.In other HPAES,such elaboratecontests
were not used-for example,in Hong Kong, Singapore,or Taiwan,
China-or wereunsuccessfil,as in the casesof the heavyindustriafization 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 increased 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 undoubtedly discouraged banks fiom taking on riskier, nontraditional
dients.
Interest rate repression was key to the HPAEs'directed-credit programs. 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 consequences elsewhere. WVethen consider whether such polidtes could be effectively adopted in the increasingly open capital markets of most
developing economies.
Interest rate subsidies were smaller in the HPAEs than in other developing 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 increased 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 inefficient 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 interest rates on domestic rates reflecrsa similar trend towardglobal integration. Indonesia,for example,began to open its capital accounts in 1971
and freeddomestic interestrates in 1981. Sincethen, interest ratesin Indonesia have followedinternational rates.
Although capital outflow restrictions are generally ineffective,they
do hinder the movement of some institutional capital, for example,domestic 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 financially than they are conmmercially.
This can be seen in the ratio of
gross capital flows to CDP compared with the ratio of exports-plusimports 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 example, Chile and the Philippines, the removal of capital outflow regulations 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 policymakers is not between financial integration with the global market and
isolation. Rather, it is between two-way integration that permits outflows 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 sectors 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 performance as the main yardstickfor credit allocation.In addition, thesegovernmentswere able during the earlystagesof rapid growth to closetheir
financialmarkets to the outside world. Few developingeconomiestoday
have the institutional resources to consistendy impose performancebased criteria for credit allocation. Moreover,the capital markers of all
economiesare becoming moreand more integraced&
Thus, successfulrepressionof 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 interventions 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 targeted 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 abandoned, 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, despirehighly 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 substantialremainingprotection 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 contestbased incentive strucnures required high governmenr institutional capability. 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 cooperation 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 actions. Interventions could not be made arbitrarily, as these could be appealed at a higher level of government if they interfered with exports.
The more recent export push-efforts of the Southeast Asian newly industrializing economies (NIEs) have relied less on highly specific incentives 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 Indonesia, Malaysia,and Thailand have also been explicitly export oriented. 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 industrial 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. Economic 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 practices by imposing quotas or countervailing duties. In such a trading environment, 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 manufactured exports? Second, will the need for market access constrain the developing 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 export 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 improvemencs in export insitutions and availability of export credit, suggests 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 limits 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 improving market accessfor smalland medium exporters. But, these measures will necd to follow internationally acceptable rules to assure
market access in importing economies, and must be of limited duration. Fiscal discipline, moreover,wiUlrequire that the costs of promotional 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 contrast 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 expand 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 European 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, indcidual developing economies, particularly smaller economies currently
contemplating an export-led expansion, could safely assume that demand 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 unlikelyto approachthis level for many years to come. In 1989, for example, 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 opportunities for South-South trade. As recently as the mid-1980s, importsubstitution 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 becoming 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 circumstances, 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 laborintensivegoods that compete direcdywith the products of other developing economies.
Despite this heightened competition, severaltrendssuggest that expanding global mnarketswill remain open to developing-economymanufactured 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 opportunity 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 import. To compete internationally, exporcersmust meet global standards 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 low362.

POLICIES

s7fFiTr

Pb~~ ~~~~-

AN:

D.

er' he;|G

Access underthe GAiT.

Box7.1 Increased
.'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 current UruguayRound of trade negotiationsis addressing .~~~~~~~o

.
.V.i s is ad ci.r.

..

*Increasing market access,inudingmore-open trade in agriculru.-s texties,

na l rsourcs and e remo or reuc-. tion
ofrmainin tariff and nonrariffbarriers
a Extending the GATTto new areas, principillytiradein services,
standards,and the enforcementof intellectualproperty rights,
and trade-relatedinvestmentrequirementsimposedby govern-

ad

; -rnents
CAT rules, especiallythose coveringantidump.: -Strengthening
ing subsidization,
productstandards, import licensing,"safeguards" (temporary import protection)

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
exports,particularlyin sucb arcasas agriculture
:developing-economy
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.
of intellectal propeAiy
'be:obligedtto ughihen-their7:nforcement
-rights, while easing rade-related investment.reqwremens They'
:couldalso -fi;e.touglie rules of oigm :and anricircumventionmnea6the
-.sures-concerningdumping alhgarions. On the whole, hwe
?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 beffecd 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,prodlosed 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
Community.
Typically
these rules require that products entering
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
formation of the Ecwas an imortant
inipulsc for the
,Area, a 50 percent regional value-added content re. quirement was required for eligibility for regional
Kennedy Round (L.wrence
1993); As Mexico
access. .In the NAFTA,tihere a`re special rules. for automoved into the' NAFTA negotiations, it simukmnemobiles 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
signed agree..- "..'.and
sewn flom &ibric spun rom Noth American.
area wthVenezuelaandCie,andhas.
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
,- .; con.erns
But
-.
about regional .protectionist
-endencurrendy enjoy prefrrentiaJacc
to'the,U.S.
des are easily overstaie. Unlikc-their predessors,m..arkets
under schem such as thc gene

tha
and EC
ed S

m- -.
ten-of prefer
th Cabbe
Ina
th:
en 5 wblcs go beyondliberaizationoftradeto
;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
em.
''erging,
hie' movation -has b'e'a at.re
LrOato*
further.access

and internationl

:opportuities.

wage economies

such as Indonesia

and Thailand,

cated operations

such as design, marketing,

from Hong
utation
rrade.

364

succeed ,in stimilatig.

e
w
.hose
outside.the regions will.n

Kong, Singapore,

for protection,

and Taiwan,

Asia has generated

of teiembcrs,
increased expor
-

and more sophisti-

and finance being provided

China. Thus
particularly

despite its rep-

rapid increases in

POLICIES

AN

RG

liT

intenweuions
Howthe Needfor MarketAccessWillConsbtain
Developingeconomicsseekingto increasetrade and attract investment 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 exemptions from GATT rules (Whalley 1990). By invoking Article XII, to
safeguard their balance of payments, or Artide XVIII, which allowvspromotion of infant industries, developing economies have been able to escape 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 proposals, 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 increasingly 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, interventionist versions of the export-push model could not be implemented within the EC with its stringent rules on stare aid and its
provisions allowing for the fiee movement of capital and for the community to implement competition policies. Moreover, no economy exposed 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 market 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 decline 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.
.*

We began this book with the objectiveof understanding East Asia's

success.We found that the diversityof experience,the varietyof institutions, 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 funcdons 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 theseinterventionscontrbuted 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 vigorouscompetidion.They did this bycreating contestswith exportsas the yardsick of
success,with subsidizedcredit and other government-controlledfavorsas
prizesand with governmentofficialsas competentand usuallyhonest ref366

POLICIES

AND

PRAGMAS

erees In other economies,induding some of the HPAES,similar interventionist 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 mirade? 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 foreign ideasand technology;and let relativeprices reflecteconomicscarcities. 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 econoamic circumstances.
East Asia'sown responses to changing domestic and international circumstances put these lessons in perspective. T7h HPAEsare themselves involved 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, Thailand's promotion of breign investment,and Malaysia'sprivatization programs 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 economics 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 complicates rather than simplifies the caskof development. The challenge
for policynakers becomes twofold: to uisepolicies compatible with axistillg 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-subsrirution
policieson a
largerscale and were thereforedoomed to failure;the
morc recentinitiativesare motivatedby a desireto consoildate more outward-orientedstrategies.Subregional
arrangementsareflourishing.In Latn America,for example, theseincludethe North AmericanFreeTradeArea;
Mercosur(Argentina,Brazil,Paraguay,and Uruguay),
whichaimsat a common marketforgoodsand services
by
1994; the Andean Pact (Bolivia, Colombia, and

368

Venezuela,with Ecuadorand Peru slated to join); the

CentralAmericanCommonMarker(CostaRica,El Salvador,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.

Bibliographic Note

THIS

REPORT HAS DIRAWN ON A WIDE RANGE OF MA-

terials,includingspeificallycommissionedbackgrund
studies, outputs of related World Bank-sponsored research, 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 background 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 Duncan, 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,Universiy of Shizuoka;KazuoKoike,HoseiUniversity,LawrenceJ. Lau,Stanford
University;DavidLindauer,VassarCollege;KazushiOhkawa,International
DevelopmentCenterofJapan; KenichiOhno, UniversityofTsukuba;MoisesNaim, CarnegieEndowmentfbr IntemationalPeace;PeterPetri, BrandeisUniversity;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 capital stock series developedby the Banks International Economics Department, 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 incentives in East and South Asia, based on the methods developedin Sciff
and Valdes(1992). The discussionof export performanceis drawn primarily from Bhattacharya and Page (1992). Analysis of the demographic 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 Division of the World Bank'sPolicyResearchDepartment. This researcheffort which is led by 1.J. Sirwh is summarized in Chen, Jeffrrson, and
370

BIBLiO'

Singh (1992). The Asian Miracle ResearchProject has also funded research by professor Shigeru Ishikawa and his associateson the rapid
growth of China'scoastalprovinces.Preliminaryoutputs of that research
will be availablein the fall of 1993.
Chapter
2
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 WorldDevelopmentReport 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 discussions with VimodThomas, Robert Wade, and partcipants in the Development 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); ChrisDar. Sanvwdla,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 macroeconomic perfomance are drawn from a researchproject, How Do National 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 prepared 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 contributions. 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 Policies in East Asia. The section on tax policy draws on Shah (forchcoming) and on preliminary results ofa research projecr, Tax Policy and Tax
Administration in East Asia. Discussion of the relative price of investment 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 marker draws primariJy on Uy (1993), SEiglitz(1993a, 1993b), and on contributions 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 international 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
Behrman,JereR., and RyanSchneider."AnInternational
Perspective
on SchoolingInvesrmentin the LastQuarter Century in Some Fast-Growing Eastern and
SoutheasternCountries."
Birdsall,Nancy,and RichardH. Sabot."VirtuousCirdes:
Human CapitalGrowthand Equityin EastAsia."
Campos,Ed. "The InstimtionalFoundationsof HighAsianEconSpeedGrowchin the High-Performing
omies: Part I, Insulation Mechanismsand Public
Sectr-PrivateSectorRelations."
. "The Institutional Foundations of High-Speed
Growthin dteHigh-PerforningAsian
EconomiesPart
II, The Bureauacy asan Instruent of Development."
Chia,SiowYuC.'The Roleof Institutionsin Singaporc's

Samuudavanija,
Chaianan."HighSpeedGrowthand High
Pertfrmancein a TechnocraticPolity:The Thai Case."
Shah, Shekhar,and SubodhC. Marhur. 1992."Sources
of LaborMarkecr
Flexibilityin EastAsia:A Reviewof
SelectedIssues."
fromtheAsianMiracle."
Stiglitz,JosephE. "SorneLessonis
Uy, Marilou."The Roleof CapitalMarketsand the Role
of Government"
AsianEconomies
Studiesofthe High-Peforming

shortlya WartldBankCoaznThesepapers
willbepublished
seriesdteThe Lessonsof EasrAsia.
tryStudfiesu rsh;lje
Bhatracharya,
Amar,andMan Pangesrn.'Indonesia:DevelopmentTransformationsince1965and theRoleof
PublicPolicy."
EconomicSuccess."
Chau, Leung Chuen. "Hong KongjA Unique Case of
Development."
Corden, W. Max. "SevenAsian Miracle Economics:
Overviewof Macroeconomic
Policies.Christensen,Scott,DavidDollar,AmmarSiamwvlla,
and
PakornVichyanond."Thailand.The Insirtuionaland
Dekle, Robert."RaisingSavingRates:Lessonsfrom the
PoliticalUnderpinningsof Growth."
JapaneseExperiencc."
'Taiwan, China:
Carl,and OusaSananik-one.
Fields,GaryS. "ChangingLaborMarketConditionsand Dahlmnan,
Policiesand InstitutionsforRapidGrowth."
EconomicDevelopmenrin Hong Kong,Korea,SingaKim, Kiwan,and Danny Leipziger."Korea:A Case of
poreandTaiwan.'
Development."
Government-Led
Freeman,Richard. "Does Suppressionof Labor Contribute to EconomicSuccess?Labor Relationsand Leipziger,Danny,and VinodThomas."CountryLessons
Markersin EastAsia.'

Relationship
Kim, Chang-Shik "Government-Business
in Korea:SelectiveCaseStudies."
"Trade Policyand Industrial Transformationin
Korea."

Lawrence,RobertZ. "The GlobalEnvironmentfor the

EastAsianModel."
Murdock,K., and JosephE. Sdiglim"The Effectof FinancialRepressionin an EconomywithPosiriveReal

from EastAsia."

Petri,Peter."GovernmentPolicyandEastAsia'sSuccess."
Salleh, Ismnael,Kim Leng Yeah, and Saha Dhevan
Growth,Equity,and StrucMeyanathan."Malaysia:
rural Transformation."

Soon,Teck-Won,and C. SuanTan. "Singapore:Public

Policyand EconomicDevelopment?
Thomas,Vinod,and YanWang. GovernmentPoLicies
and ProductivityGrowth:Is EastAsiaan Exception?"

Rates."

Pack, Howard. "Induscrialand Trade Policiesin the

High-Performing
AsianEconomies."
Pack,Howard,andJohn M. Page,Jr. "Accumulaton,Erports and Growth in the High-PerformingAsian
Economies.'
Change
Page,JohnM.,Jr.,andPeterA Pctd. Productivity
andStrategicGrowthPolicyin theAsianMiracle."
Ranis,Gustav."LaborMarkets,Human CapitalandDevelopmentPerformancein EastAsia-"

Studiesof PublicPolicyinJapan
at a SeminaronPublicPoliy
Thesepaperswerepresented
Tolyo,Jzanry 1,993.
duringJapanf?RapidGrowth.
Horiuchi,Akiyoshi."GovernmentControluJ Financial
Mechanismin PostwarJapan:A TentativeReport."
Itoh, Motoshige. "Industrial Policy and Corporate
Japan'sPostwar
Growthin the AutomobileIndusaryr
Experience."

373

IflCNTE
Koike, Kazuo. "Human ResourceDevelopment Systems
in Japan.'
Kuroda, Masahiro,Takanobu Nakajima,and Kanji Yoshioka. "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 BusinessRelauionship
in Japan:A ComparativeInstitutionalAnalysis.'
Shimomura, Yasutami. "Industrial Policy in the Rapid
Growth PeriodcEvidencefrom Early Writings."
Teranishi, Jurn. "Emergenceand Establishmentof the Financial Svsremin Postwarjapan: Government Interven;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 Economic 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 University Press.
Ando, Albert, and Franco Modigliani. 1963. 'The Life
Cycle Hypothesis of Savings:AggregateImplications
Economic
Review53(1).
andTesrts."American
Arrow, KennethJ. 1969."ClassificatoryNotes on the Production and Transmission of TechnologicalKnowledge."AmericanEconomicRetiew 59(2):29-35.
374

Asian WallStreetjolrnaL 1993. 'MatsushimaBuilds Export Basein Malaysia."May 18.

Aw, Bee-Yan,and Amy R Hwang. 1993. "Productiviry
and the ExportMarker:A Firm LevelAnalysis."PennsylvaniaState University,StareCollege,Pa.
Aw, Bec-Yan,and Hong W. Tan. 1993."Training, Technological Capaciry, and Firm Level Productivity."
WorkingPaper. World Bank, PrivateSecrorDevelopment Department, Washington, D.C.
Balassa, Bela 1982. Development Strteies in SemiIndustrializing Economies. BalLiliore, Md.: Johns
Hopkins UniversityPress.
1991. EconomicPolciesin thePacifxAreaDeveloping
Countrie.tNewYork. New York UniversityPress.
Balass, Bela,and Associates.1971. The Structureof Protectionin DevelopingCountries.Baluimore,Md.: Johns
HopkinsUniversityPress.
Balassa,Bela,and Marcus Noland. 1988.Japan and the
WorldFconomy. Washington, D.C.: Institute for InternationalEconomics.
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 presented at a conference on How Do National Policies
Affec Long-Run Growth?World Bank, Washingron,
D.C. February.
Behrman,Jere P. 1987. 'Scholing in DevelopingCounrries: Which Countries Are the Under- and OverAchievers 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

BIBLIO-GR

Bank, Policy Research Department, Washington,

. 1993b."Virtuous Circles: Human Capital Growth
D.C.
and Equity in East Asia." Backgroundpaper for The
Bell, Clive, Peter Hazell, and Roger Slade. 1982. Projecr
EastAsianMiracle.World Bank, Policy ResearchDcEvaluation in RegionalPerspective.Baltimore, Md.:
partment, Washington, D.C.
Johns Hopkins UniversityPress.
Black, Stephen L., and Lance Taylor. 1974. "Practical
Berry, R. Albert, and William Cline. 1979. Agrarian
General Equilibrium Estimation of Resource Pulls
StntutreandProdiuctiviy in DevelopingCounzriesBalundcr Trade Liberalization."joUrnalof International
timore, Md.: Johns Hopkins UniversityPress.
Economics4(1): 37-58.
Berry,R Albert,and BrianLevy.1993."Indonesia'sSmall
Blomstrom, Magnus. 1989. Foreigninvestments and
and Mvledium
Industrial Exportersand Their Support
SpilloversLondon: Roudedge
Systems."Report preparedfor the World Bank, Policy
Blumenthal,1. 1970. "Savingsin PostwarJapan." East
ResearchDepartmcnr,Washington,D.C.
Asian Research Center, Harvard University, CamBerry, R. Albert, and Richard H. Sabor. 1981. "Lubor
bridge,Mass.
Market Perfbrmancein DevelopingCountries:A SiurBoadwvay,
Robin, Dale Chua, and Frank Flatters.Forthvey." In Paul Streetenand RJolly, eds., RecentIsuesin
coming. "Investment Incentives and the Corporate
WorldDevelopmentOxford: PergamonPress.
Tax Svstcm in Malaysia."In Anwar Shah, ed., Fiscal
Bhagwari.JagdishN. 1978.Anatomyand Consequencas
of
Incentivesfrrinvestmentin DevelopingCountries.New
Exchange Control Regimes. Cambri"-e, Mass.:
York: OxfordUniversityPress.
BallingerPress, for the National Bureauof Economic
Boneili,R. 1992. "Growth and Productivirvin Brazilian
Rescarch.
Industries:Impacts of Trade Orintation." Journal of
85-111.
Development&Economics39(1):
Bhagwati,JagdishN., andT. N. Srinivasan.1979.'Trade
Policyand Development.' In RudigerDombusch and
Borrus, Michael,Laura D'Andrea Tyson, and John ZvsJacob A. Frankel, eds., InternationalEconomicPoliey.
man. 1986. "Cr- ir Policyand International ComnpeBaltimore,Md.: Johns Hopkins UniversiryPress.
tition." In PawII rugman, ed., Srategic TradePolicy
Bblacachatya,Arnar, and John lvi. Page,jr. 1992. "Adand the New International Economics.Cambridge,
justnnent,investment; and Growth in High-PerformMass.:MITPress.
ingAsian Economies."Paperpresentedaraconference
Boskin,MichaelJ., and LaurenceJ. Lau. 1992. "Inrernaon the Nev DevelopmentEconomicsin Global Contional and Intertemporal Comparison of Productive
text, American Universityin Cairo, May 28-30.
Efficiency:An Application of the Meta-Producton
Bharacharya,Amar, and Mari Pangestu.1993. Indonesist
Function Approach to the Group of Five Countries."
43(4): 298-312.
Economi StudiesQucarterly
DevelopmentTransfirmationsince1965 and theRoaeof
Public Policy.World Bank Country Study, WashingBradford,Colin L, Jr. 1986. "East Asian Models:Myths
ton, D.C.
and Lessons."In John P. Lewisand ValerianaKallab,
Biggs,Tyler, and Brian Levy. 1988. "StrategicInrerveneds., DevelopmentStrates ReconsideredU.S. Third
don and the PoliticalEconomyof Industrial Policyin
World Policy Perspectives 5. Washington, D.C.:
Developing Countries." In Dwight Perkins aind
OverseasDevelopmentCouncil.
Michael Romer, eds., RefirmingEconomicSystemsin
. 990. Policy Interventionsand Markets: DevelopDevelopingCountries.Cambridge,Mass.: Harvard JnmentStrategyTypologiesand PolicyOptions."In Gary
sticute for International Development.
Gereffi and Donald L. Wyman, eds., Manufacturing
Birdsall, Nancy, Dav.d Ross, and Richard H. Saborz
Miracles:Pathsofindurialization in Latin Americaand
1993. "Underinvestment in Education: How Much
EastAsia.Princeton,N.J.: Princeton UniversityPress.
Growth Has Pakistan Forgone?"Paper presented at
Brimble,PeterJohn. 1993. "IndustrialDevelopmentand
the Ninth Annual General Meetng of the Pakistan
ProductivityChange in Thailand." Ph.D. diss.,Johns
C..ociety
of Development Economist, Islamnabad,
PakHopkins University,Baltimore,Md.
ismt, January.
Bruton, Henry J., G. Abeysekera,N. Sanderatne,and Z
Birdsall,Nancy, and Richard H. Sabor, eds. 1993a. OpA. Yusof. 1993. ThePoliticalEconomyof PoverJ Eqaportunity Forgvnc Education, Growth,and lnequaiq
uiy, and Growth:Sri Lanka and MalaysiaNew York
in Brazil World Bank,Washington, D.C
Oxford UniversityPress.
375

OG4R~KflflC-NOTE

Calomiris, Charles W., and Charles P. Himmelberg.

frInvestmentin DevelopingCgAnntries.
NewYork Ox1993. "Directed Credit Programs for Agricultureand
ford UnivcrsityPress.
Industry: Arguments fiom Theory and Fact." Papar
Chia, SiowYue. 1992. 'The Roleof Institutions in Singpresented at the Annual Bank Conference on Develapore's EconomicSuccess."
Backgroundpaper for The
opment Economics,World Bank, Washington,D.C.,
EastAsian Miracle.World Bank, PolicyResearchDeMay3-4.
May3-4.
~~~~~~~~~parmacnt,
Washington, I).C.
Canning, David, and. Marianne
Fay.
1993.
'The
Effect
of
PautnCnr,
i
onenD.C.
= .
. _
,, _
~~~~~Chiu,
Paul C.WH. 1992.
"Money
and FimnancalMarkets:
Infratucture Networks on Economic Growth." CoThe Domestic Perspective."In TaiwanfromDeveloplumbia University,New York,January.
Caprio, Gerard, Jr., IzakAnyas, and James Hanson, s
ing to Mature Economy.Colorado:Westview.
Forthcoming. Financial Reform. Theoryand F.erl-.
Cho, Yoon Je, and Thomas Hellmann. 1993. 'GovcrnFnot.Camrnidge:Canbridge URrivrrsityPress.
menr Intervention in Credit Market: An Alternative
Cic, Caridge:, Camidg e, Uni
Press. SInterpreton
of Japancse and Korean Experiences
Carroll,
Crspe
David and
LawrenceH. Sp.ca
from the Ncw Institutional Economics Perspective."
-imers.193. "savings and Growth: ArcgicerprecaPaper prepared for a World Bank researchproject on
dion." Paper presented a1tthe Carnegie-Rochester
Th Efetvns*fCei-Plce
nEs
sa
Public Policy Conference, Bradley Poliqy Research
World Bank, FinancialSrecitPoriDevelmEt Dcpa.rCen_
r April
Api
23--24.
Worlci Washington,
Bankl, FinancialSectorDvcvdopmenc
DepartCenter,
23-2.
mentc,
D.C.
tChan,Steve,and Cal Clark. 1992. FA&xiii Fior
m-siWahigon DC
(Than,
Stuev an CaizCark 1992.
Fesgt
Cho, Yoon Je, and Deena Khatkare. 1939. LessonsofFiandFotun,i
Thwa
pment Navigatingheo
nancaalLiberalizationin Asia: A ComparativeStudy.
tweenScyllaand Chanybdis.London: Roudledge.
World Bank DiscussionPaper 50. WashingEon,D.C.
Chant, John, and Man Pangestu. Forthcoming. An AsCho, YoonJe,andJoon-Kyang Im. 1993.'Cedit Polisessmentof FinancialReform in IndonesLiz
- YnJ
Polis t- 1983-90."
'
T
T
I
A v^ ttFa~~ces
and IndJoonKyg
Industralization
ofKim.1
Korea."CWedt
World Bank,
In Ge_rd Caprio,Jr., IzakAti,
andjames Hanson,
Finanal SectorDevelopmentDepartnent, Washingeds., Financia Rqornm: Theoryand Eaperince. Camt,
Fcbra.
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 Investment Decisions."In Moses Abramovit ecL, The
AllocationofEconomicResource.Sianford,Calif.:Sranford 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 Refonn in PacificBasin Countries.Lexington,Mass.:Lexington Books.
Chia, Ngee Choon, and John Whalley. Forthcoming.
"Paterns in InvesrtmentTax Incentivesamong Developing Countries.' InAmwarShah, ed., FiscalIncntive
-.

376

Chrisrensen,Scott R., David Dollar, Ammar Siamwala,

Q
and Pakor Vichyanond.1993.sInstitutional
andPolitical Bases of Grwth-hiindztng Policies in Thailand

Lessonsof East Asia World Bank Counnty Study.

Washington, D.C.
Carke, Gerge. 1992. "More Evidenceon Income Disrribution and Growth." Policy Research Working
Paper 1064. World Bank, Policy Research Department, 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 Europeand theSoviet Unionin the WorldEconomy.Washington, D.C.: Insitute for International Economics.
Corbo, Vittorio, and Klaus Schmidt-Hebbel. 1991.
"Public Policyand Savingsin DevelopingCountries."
Jownl ofDevelopmenitEconomics36(1):
89-115.

BIBLAjt

Corden, W. Max. 1971. The Themryof Protection.London: 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 Developmenr: 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. "Productiviry 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
-.

-.

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 Department, 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 Culturd 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."Moderate 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 Forthcoming. "The Macroeconomics of Public Sector
Deficits A Synthesis."In Wlliam Easterly,CarlosRodrigucz,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 In377

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 Productivity on Economic Growth.' Background paper for
the Worlk DevelopmentReport 1991. World Bank,
Office of the Vice President, Development Economics, 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 UNCTADexecuted project to assist tizc Intergovernmental
Group of Twnet-Four.
Far EasternEconomicRevitew.Various issues.
Farrell,M. J. 1957. "The Measurementof ProductiveEfficicncy."Journal of tht RoyalStadsticalSociety,Series
A 120 (part 3): 253-8 1.
Fecher, M., and S. Perelman. 1992. "Industrial Efficiency in OECD Countries." In R. Caves, ed., IndustialEEfficiernyinSilNaionss 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 ImprovedDiributon in theRepblic ofChina,1952-79.
Boulder, Colo.: Westview.
Fields, Gary S. 1992. "Changing Labor Market Condidions 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, PolicyResearchDcpartmenE,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 Inastrializing
Countries.World Bank Policy and Research Seriec
Paper7. Washington, D.C.
Fry, MaxwellJ. 1988. Money. Interest,and Banking in
EconomicDeveiopment.Baltimore, Md.: Johns Hopkins UniversityPress.
Fukasaku, Kiichiro. 1992. Economic Regionalization
and Inta-Industry Trade: Pacific-AsianPerspecaives."
TechnicalPaper 53, OECD DevelopmentCentre, Padis.
Gelb, AManH. 1989. "FinancialPolicies,Growth and Efficiency." PRE Working Paper 202. World Bank,
Washington, ''.C.
Gelb, Alan H.,Ju..n Knight, and RichardH. Sabo: 1991.
'Public Sector Emp!oyment, Rent Seekingand Eco101(408):1186-99.
nomic Growth." Economic Journal
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., MassachusettsInstiruteof Technology(Mrr),Cambridge, Mass.
1983b. "The Interest Elasticityof Savingsin Developing Countries: The Existing Evidence." WorldDevelopment11(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, Kotaro Suzumura, eds., IndustrPolicy oflapan. New
York-AcademicPress, Inc.
Granger, Clive W. J. 1969. 'Investigating Causal Reladions by Econometric Models and Cross-Spectral
Methods."Economerica37: 428-38.
Greenwald,B., and Joseph E Stiglirz. 1986. "Externalities in Economieswith Imperfect Information and In-.

BI BLOO,G.X

complete Markers." Quarter4'journal of Econonic

101(2): 229-64.
_.
1938. 'Pareto Inefficiencyof Market Economies:
Scarch and EfficiencyWage Models." AmericanEconomicReview78(2): 351-55.
Gregory, Peter. and Deepak LaL.1977. "Demand, Supply, and Structure:A Cross-SectionalAnalysisof Manuficuring Wages.' World Bank, Employment and
Rural Dcvelopment DivisionWorking Paper, Washington, D.C.
Grossman, Gcne, and Elhanan Helpman. 1991. Innovaonvand 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 EconomicDevelvPment9(2):25-34.
H-adad, Mona, and Ann E. Harrison. 1991. "Are There
Dynamic Externalities fiom Direct Foreign Investment? Evidence From Morocco." Indus-ry Series
Paper No. 48. World Bank, Industry and EnergyDeparument,Washington, D.C.
Forthcoming. "Are There PositiveSpilloversfrom
Direct ForeignInvestment?Evidencefrom Panel Dara
Economics.
for Morocco."JournalofDerelopment
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 DevelopingCounins: Eliminating the Maid-Fiber Arrangementin the
1990s Washington, D.C World Bank
Hanson,james, and Roberto Rocha. 1.986.'High Interest 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. Siregar. 1992. 'Thc Effort of FinancialLiberalizationon

Firms' Capital Structure and Investment Decisions:

Evidencefrom a Panel of Indoncsia Establishments,
1981-88." Paper presentedat a conferenceon the Impacaof 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: Structuresand So:,rcesof Finance." FinancialTimes BusinessInformation, London.
Hill, Anne M., and ElizabethKing,cds. 1991. "Womcn's
Education in Developing Countries: Barriers. Benefits, and Policy." PHREE Background Paper Series.
World Bank,Washington [D.C.
Hon, Wong Seng. 1992. "Exploiting Information TechnoAogy.ACase Study of Singapore." World Development20(12): 1817-28.
Hong, Won-Tal. 1989. "Korean Economy at Crossroads." 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 Influence 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" Research School of Pacific Studies, AustralianNational
University.
Hulten, CharlesJr. 1989. "The Embodiment Hypothesis
Revisited."Universityof Marvland,CollegePark, Md.
Huseinga, Mark, and ChrisEosKostopolous.1992. "Options 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 Technological 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 PublicCorporation." World Bank, Economic DevelopmentInscituteWorking Paper,Washington, D.C.
IMF (Inernational Monctary Fund). 1993. World Economic 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 Systemns? 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, Cambridge,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 OkunoFugiwar, and KataroSuzumura 1988.'Industrial Policy 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, Washington, 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 ProductivityCorrparisans 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 Globalization and Developing Countries, The Hague.
(Reprinted as Ccnter for Economic Policy Research
DiscussionPaper 687, London, July 1992.)
in LauinAnmrKatz,Jorge,ed. 1987. Techno Generation
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-BusinessRelationship in Korea: Selectve Case Studies.' Background paper for TheEasrAsianMiracLcWorld Bank.
PolicyResearchDepartment, Washington, D.C.
. 1993b. "Trde Policy and Industrial Tranfrmadion 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 Industrialized 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 Organizadions.? Worid Bank, Policy Research Departrment,
'Washington,D.C.
Kim, S. 1987. In Service Trainingas an Insmnrentfbr the
DevelopmentofHwman Resourcesin Korea.Paris: OECD

Konishi,Y. 1989. A Quantitative Anadis of Eduational

Policies in Postarjapan
ResearchInsticute for Economics and BusinessAdministration, Kobe Universicy,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 SconomirReview. 68(2): 270-74.
-.
1984. "ComparativeAdvantage and Development
PolicyTwenty YcarsLater." In MosheSyrquin, Lance
Taylor, and Larry E. Westphal, eds., Economic StructureandPerformance.New York:AcademicPress.
_. 1993. "East Asia: Lessons for Growth Theory.'
Paper prcsented at the Fourth Annual East Asian
Seminar on Economics, National Bureau of Economic Research,San Francisco,Calif.,June 17-19.
Krueger, Anne O., and Jagdish Bhagwati. 1973. 'Exchange Control, Liberalization Attempts, and Economic Development." American Economic Review
63(2): 419-27.
Krgrnan, Paul R 1986. Introduction: New Thinking
about Trade Policy." In Paul R. Krugman, ed., Strate-

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 Manufacturgc Trad Polcy and the New Intenational Economics.
ing.?Journd ofDevelopment Economics 9: 265-78.
Cambridge, Mass.: MIT Press.
King, Elizabeth, Katherine Anderson, and Yan Wang.
Kumar, Nagesh 1985. Cost of Technology Imports:
1993. "Feedback Effects on the Labor Market on
The Indian Experience.? Economic and Political
SchoolingChoice in Malaysia? Paperpresented at the
Weekly 20: Ml 03-114.
April 1992 Meetingsof the Population Associationof
Kunio, Y. 1988. The Riseof Ersaz Capitalism in South
America,Cincinnati, Ohio.
EsAsia. New York.Oxford UniversityPress.
King, Robert,and RossLevine.1993. "Financz,EntrpreKuo, S. 1983. The Taiwan Economy in Transition. Boulneurship and GrowdLhPaper presentedat a conference
der, Colo.: Westvzew.
on How Do National Policies Affect Long-Run
Kuo, Wan-Yong. 1976. 'Income Distribution by Size in
Growh? World Bank, Washington,D.C., February.
Taiwan Area. Changes and Causes." Indmut of Free
Knight, John, and Richard H. Sabor. 1991. Education;,
Cbina 45.
Produciviy, and Inequality. New York.cOxford UniKuznets, Simon. 1955. "Economic Growth and Income
versityPress.
Inequality. American EconomicReview 45(1): 1-28.
Komiya, Ryutaro. 1990. The Japanese Econoyn: Trade,
_.
1959 Six Lectures on Economic Growth. Glencoe,
Induatry, and Government. Tokyo: University of
Ill.: Free Press.
Tokyo Press
. 1966. Modern Economic Growth. New Haven,
Komiya, Ryuraro. Masahiro Okuno, Kotaro Suzumura,
Conn.: YaleUniversityPress.
eds. 1988. IndusialPolicy ofjapan. New York AcadLal, Sanjaya,and G. Wignaraja. 1992. 'Foreign Involveemic Press.
ment by European Firms and Garment Exports by
381

>NTE

HIID Conference on the Malaysian Economy, Kuala

Developing Countries." DevelopmentStudies WorkLumpur. Institute of Strategic and International
ing Paper 54, Queen Elizabeth House, OxfordlCenStudies.June 1-3.
tro Studi Luca d'Agliano,Torino, Imlia.
Lim, Linda Y. C., and Pang Eng Fong. 1991. Foregn DiLanyi, Anthony, and Rusdu Saracoglu. 1983. "Interest
ret:InvestmentandIndutrializtioninMa(aria.SingRate Policies in Developing Countries." Occasional
apor, 7Taiwnand ThaikladnOECD, Paris.
Paper 22. International Monetary Fund, Washingrim,Youngil. 1991. 'Upgrading industrial Base and
ton, D.C.
Trade: The Role of Government Policy in Korea."
Lawrence, Robert Z 1991. "Scenarios for the World
United Nations Industrial Development OrganizaTrading System and Thcir Implications for Develoption, New York.
ing Countries." Tcchnical Paper 47, OECD DevelopLindaucr. David L., and Barbara Nunbcrg, eds. 1993.
meentCenter, Paris. (Paper prepared for the Rcsearch
"Rehabilitating Govcrnment: Pay and Employment
Program on Globalizationand Regionalization.)
Refrmn in Sub-Saharan Africa."World Bank, Policy
1993. "The Global Environmenr for the EastAsian
Department, Washington, D.C.
Research
Model.' John F. Kennedy School of Government,
LiEte, Ian M. D. 1979. UTheExperienceand Causesof
Harvard Univcrsiy, Cambridgc, Mass.
Rapid Labor-IntensiveDevelopment in Korea, TaiLee, Chinaboon. 1985. "FinancingTechnical Education
in LDCs Economic Implicatons from a Survey of
wan,Hong Kong, and Singapore,and the Possibilities
Training Modes in the Republicof Kora? Education
of Emulation." Working Papers, Asian Employment
Bangkolk
Programme,ILO-ARTEP,
and TrainiingSeries 6. World Bank, Education and
1982. Economic Development New York: Basic
_
Training Deparmnent, Washington, D.C.
Books.
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,
Forthcoming. Boom,Cri4s,
and Sarath Rajapathirana.
Depening in
New YorakiSt. Martin'sPras
BbMatroeconondc
Th
EEpeneneofDevelopNewYo&
Press.Adyjustmen
Sr. Martin's
AduenThMarcomiEprece/DvlDanny.
87. Ko-reicManaging
Leipzrger,
mg Counties. New York.Oxford UniversityPress.
tJycIndustrial
SnivZ
Lzpzlgcr,
Danny. 19
1987.
Korca Mnnugmg the

Transition.Washington, D.C.: World Bank.

Leipziger,Danny, David Dollar, Anthony F. Shorrocks,
and Su-YongSong. 1992. The DistributionofIncome

-and

Bandh

WeaLrhin Korea-Washington, D.C.: World

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 UniversiryPress.
Liddle, William. 1991. "The RelativeAutonomy of the
Third World Politician: Soeharto and Indonesian
EconomnicDevelopmenr in Comparative Perspecive." Interational StudiesQuarterly35: 403-27.
Lim, Chee Peng 1992. "Managing Structural Change
and Indusrrialization-Heayy Industrialization: The
Malaysian Experience" Paper presented at the isis382

Lltle, Ian M. D., Tibor Scitovsky,and Maurice FG.

Scott. 1970. Induniy and Trade in Some Developing
Countnies.London: Oxford UniversityPress.
T

Alan P. L. 1987. Phoenixand the LamneLion ModBakentioznin 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 Economics Department, Washingron,D.C.
Lockheed, Marlaine, and Adriaan Verspoor. 1991. Improving 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 ProducrMty Growth in MalaysianManufaucring IndusLiu,

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 Bcnomic Development Washington, D.C.: Brookings

Insttution.
Mellor,John, ed. 1993. "Agriculntreon the Road to Industrialization." International Food Policy Research
Institute, Washington, D.C.
MiddkUton,J., A Ziderman, and K van Adams. Forthcoming. 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-Invescment 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. Boulder, 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 TechnologyDevelopment 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 Markers."AmericanEconomicReview72(2): 278-84.

Nalebuff,B., and JosephE StigLirz.1983. "Prizesand Incentives 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." PRWorking
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,Methodology, and Results."World Bank, Internatonal Economics 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 Producriviy 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 ResearchDepartment, Washington, D.C-.
1991. "Trade Policy,Market Orientation, and ProductivityChange 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 savkg behaviorof the people]."In KenkyuHakoko Gaiyo
[-Sumwy of ResearchFindings].Kokuci Nin'i Seimei
Hoken no Shorai Tenbo ni k..isuru Chosa Kenkyukai [Reserch project on the Future Prospectsof Goveinment-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 Suzumum, edmi
P
ndustria PolicyofJapan.New York.AcadlemncPress.
Okimoto, Daniel. 1989. BetweenmiT! and the Market
JapaneseIndstrial Policyfir High Tecbnolg. Stanford, 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

. 1993c. "Technology Gaps between Industrial and

DevelopingCountries:Are There Dividends for Latecomers?" In Proceedingsof the World BatnkAnnual
Bank Conferenceon DevelopinentEconomics,1992.
Washington, D.C.: World Bank.
Pack, Howard, and John M. Page,Jr. 1993. 'Accumulation, Exports, and Growrh in the High-Perfirming
Asian Economies." Background paper for The East
Asian Mirade. World Bank, PolicyResearchDepartment, Washington, D.C.
Pack, Howard, and Lary E. Wcstphal. 1986."Industrial
Stratcgyand TechnologicalChange: Theory vs. Reality."fournalofDevelopmentEconomics22(l):87-128.
Packer, Frank. 1992. 'The Role of Long-Tarm Credit
Banks within the Japanese Main Bank System." Columbia University,New York.
Page, John M., Jr. 1990. "The Pursuit of Industrial
Growth: Policy Initiatives and Economic Consequences."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. "Productivity Change and StrategicGrowth Poliqc in the Asian
Miracle" Backgroundpaper for The EastAsianMinac& World Bank, PolicyResearchDepartment, Washington, D.C.
Panagariya,Arvind. 1993. Unravelinj the Mysteries of
China's Foreign Trade Regime." WorldEconomy
16(1): 51-.
Pempel, T. J. 1984. "The JapaneseBureaucracyand Economic Development Can the Model bc Copied?'
Economic Development Institute, World Bank,
Wsigo,DC
Washington, D.C
Perry, Guillemo, and Ana Maria Herrera. 1993. "Finanzas Publicas,Esmabilizacion
y Reforma Estructural en
America Latina." Working Paper Series 165. InterAmericanDevelopmentBank,Washington, D.C.

Policy Industializaton and Development. Oxford:

Clarendon Press.
_.
1993a. "Exports and Externalities:The Source of
Taiwan's Growth in Taiwan." Universityof Pennsylvania, Philadelphia,Pa.
_. 1993b. Indtstrial and Trade Polcies in the HighPerformingAsian Economies."Backgroundpaper for
The ast Asia Miracl. World Bank, PolicyResearch
Department, Washington,D.C

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

-.

_. 1992. 'Leamningand Producdvily Change in D:evelOping Countries." In Gerald K Helliner, ed., Trade

384

BIBLO'4

Psacharopoulos, George. Jee-Peng Tan, and Emmanuel

Jimenez. 1986. Financing Educaton in Developing
CoanmierAn Exploration
ofPoliqyOption. Washington, D.C.: World Bank.
Ranis, Gusrav. 1973. "Industrial Sector Labor Absorpuon." 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 Economy. Boulder,Cola.:WstviCw.
_. 1993. "labor Markets, Human Capital, and Dcvdopmenr Performancein EastAsia."Backgroundpaper
for The East Asian Miracle.World Bank, Policy Research Department,Washington, D.C.
Ranis, Gustav, and John Fei. 1961. "A Theory of Economic Development." American Economic Reviewt
51(4): 533-65.
Ranis, Gustav,and Chi Schive.1985. "Direct Foreign Investrment in Taiwan's Development." In Walter
Galenson, ed., ForeignTrad and Investmen Economic
Developmentin &ieNewly Industriizng Asian Couentiim Madison, Wis.:Universityof-WisconsinPress.
Ranis, Gustav, and Frances Stewart. 1987. "Rural Linkages in the Philippinesand Taiwan.' In Frances Stewam, 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 America: Lessons of Experience." Occasional Paper 6.
World Bank, Latin Ameica Technical Department,
Washington, D
.
Remolona, E., and M. Lamberre. 1986. 'Financial Ref[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 Intcrnational Markets: Lessonsfrom the East Asian Experience? Industry SeriesWorking Paper 11. World Bank,
Industry and EnergyDepartmne Washington, D.C.
1989b. "Trade Finance in DevelopingEconomies."
PolicyResearchSeries5. WorldBank,Washington,D.C.

Rhee,YungWhee, and Lary F. Westphal. 1977. "A MicroeconomicInvestigation of Choice of Technique."

JowofDevelpmenrtEonomia4: 205-38.
Riedel,Jamnes.1988. "Demand for LDc Exporcsof Manufictures: Estimatesfrom Hong Kong.' Economiclournal98(389): 138-48.
Rob, Rafael.1990. "The Evolutionof Markcts niderDemand Uncertainty." University of Pennsylvania,
Philadelphia,Pa.
Romer,Paul M. 1993. "Two Srategies for EconomicDevclopmenr Using Ideas and ProducingIdeas." In Proceedings
of the Word BankAnnual Bank Conference
on
DevehopmmtEnoms 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, Congressof the Unitcd States.
Sakong,1. 1993. Koreain de WorldEonomy. Washington, D.C.: Insritute for International Economics.
Salleh,IsmaeL1991. The Privatisaion of PublicEnterprises: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
and Srcr
rat
Manatan
1993. Growth,
Meyanathan.
1993.
Growab,
Eqity, andf SDxrd
Trsmj$rmazionin Malaysi The Roleof the Pubic cctor. Lessonsof East Asi World Bank County Study,
Washington, D.C.
Samudavania Chaian. 1992. "High Speed Growth
and High PerD.Ciance in a Technoc.ric Poliy: The
ai
.n
P
acegin a TocreastAic o irT
Thai Cae." BackgroundpaperoforTec stAsarnMiracs. World Bank, Policy Research Department,
Washington, D.C
San, G- 1990. "Enrerprise Training in Taiwan: Results
fromdieVocationalTrainingNeedsSurvey? Econamic ofEducationReview9(4): 411-8.
Saxonhouse,Gary. 1983. "What Is Al This about Industrial Targeting in Japan?" World Economvy6(3):
253-73.
Scdiff, Maurice. 1993. "Agriculture Pricing Policy and
Growth in Asia.' World Bank, Policy ResearchDepartment, Washington, D.C
385

Scott, Maurice FG. 1979. "Trade." In W. Galenson, ed.,

EconomicGrowth and Stral
Canga in Taiwan
Ithaca, N.Y.: Cornell UniversityPress.

Singapore-Variousyears. YearbookofStatsticsandanEconomic Surveyof Singapore.Singpore Department of

Statistics.
. 1983. Economicand Social Statistics 1960-1982.
Singapore:Department of Staiics.
Siqgh.Ait. 1992. "'CIose' vs. 'SErategic'Integrationwith
the World Economy and the 'Marker-FriendlyApproach to Developtrnent'vs. an 'Industrial Policy':A
Critique of the WorldDevelopmentReport1991and an
AltemativePolicyPerspective."Facultyof Economics,
Universityof CanibridgecCambridge,England.
Singh, Ajit, and Javed Hamid. 1992. 'Corporate Financial Structures in Developing Countries." Technical
Paper l InternationalFinance Corporation,Washing-

Scully,MichaelT., and GeorgeJ. Viksninsa1987. Finazcing EastAsias Succesr ComparatveFinancal Develop

menctinEghtAsian Counris. London: Maanilan.
Shah, Anwar, ed. Forthcoming. FiscalIncentivesfor InNe York Oxford
vetcmenttin Devloig Coun
Uesnivrersity
PressUniverilty Prem
Shah, Shekhar, and Subodh C. Madiur. 1992. Source
of Labor bMarket Flexdbilityin East Asia:A Reviewof
SelectedIssues."Backgroundpaper for The EastAsin
Mirrd, World Bank, Policy Research Department,

ton, D.C.
Solow, Robe M 1956. "A Conterbution to he Theory
QuareeryJornalofconomics
ofEconomic6Growth."
701)65-9.T
-1957. "Tehnca Chang and die Agreae Producton
Function.' Reizew of Economicsand Sttstics
~~~~~~~~39(3):
3 12-20.
Soon, Tek-Wongand Suan Tan. 1992. Public Policy
andtEcnomic Development in Singapore.' Manuscript prepared for a Wodd Bank Conference on the
Role of Goverment and East Asian Success,Wash-

Schiff, Maurice, and Alberto Vald&. 1992. Tle Political

Eonomy of Agriulural PricingPolicy Volume 4: A
Synthesisof the Eamomicsin DevelopingCountries.Balrimore, Md: Johns Hopkins UniversityPr.
Schultz, T. Paul. 1988. "Education Invesrmentand Returns." In HoUlisB. Chenesy and T. N. Srinivasan,
eds., HandhookofDevelopmentEconomis voL 1. Arnsterdam: North Holland.
Scirovsky, Tibor. 1954. "Two Conceprs of Extemal
Economics." Journal of Politcal Economy 62(2):

143-51.

Washington,D.C.

ingon, a.C.

Spence, M. 1981. 'Te Learning Curve and CompetiShapiro, Helen, and LanceTaylor. 1990. "The State and
on BeffJourd ofEconomics 12(l): 49-70.
Industrial Strategy." World Development 18(6):
861-78.
Lvn. 1981. Employment Policy in Developing
Shea, J.D.Fothoin.DeeopenadstuSquire,
Shea, J. D. Forrhcoming. 'Developmcnt and Strruaural
CCounS: A Surveyof lssuesand Evidence.New YorkOxfordUniversityPres
Changeof the FinancialSystem."In Patrick Hugh and
Srinivasan, T. N. 1992. "Datbas for Development
Yung Chul Park, cds., Financia DevelopmentJapan
Analysis:An Overview.' Paper presentedat a ConferKorea, and Taiwan. New York: Oxford University
ence on Data Base of Dcvelopment Analysis,Yale
Press.
University,New Haven, Conn., May.
Sheng, Andrew. 1989. "Bank Restructuringin Malaysia,
Stem, Joseph,Ji-Hong Kim, Dwighr Perkins, and Jung1985-88." PRnE
Worlkng Paper 54. World Bank, FiHo Yoo. 1992. "Industrialization and dhe Stare The
nancial Sector Development Department, WashingKorean Heavy and Chemical Industry Drive."
ton, D.C.
Research reporr Harvard Institute for International
_. 1990. "The Art of Bank Resuucturing2Isues and
Techniques."World Bank, FinancialSector DevelopDevelopmenr,Cambridge,Mass.,and KoreaDevelopment Department, Washington, D.C.
mentInsitute, Seoul.
Stevenson H., and J. Stigler. 1992. The Learning Gap.
Shouda,Yasutoyo.1982. 'EffectiveRatesofProteciion in
Japan." Nikd7onieza Kmkyu 11 (March).
New York.Summir.
Shoven,John. 1988. GovernmentPoicy TowardsIndustr
Stewart, Francis, and Ejaz Ghani 1991- 'How Significant AreExternaities for Development? WorldDevelm the United Stares and Japan. Cambridge, U.K:
opmentl9(6): 569-94.
Cambridge UniversityPress.
386

Stigicz Joseph E. 1993a. "The Role of the State in FinancialMarkets."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 CauriousCase for a^
uious Cse for
a Secuwriics
SecuritiesTmnsaceian
T ransactionTax."
Ta." In
In RegvRet&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 Expanded 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.'Experiencesof FmancialDistrss in Thailand." PRE Working
Paper283. World Bank,Officeofthe VicePresidentfor
DevelopmentEconomics,Washington,D.C
Sussangkarn, Chalongphob. 1988. "Production Structures, LaborMarketsand Human CapitalInvestments:
Issues of Balancefor Thailand." Thailand Developincur ResearchInstitute,Ofijprinr#7, Bangkok
Suzuki,Yoshio.1987. ThejapaneseFinancialSystem.Oxford: 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, Accounting, 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 Economic 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
Masahiro
and
Koraro
Suzumur, ed.Komiya,
Indrutnl
PvoliyOkuno,
of/Japan.
New
Yorkc
Acadmim pes.
mc rss.
Tanzi, Vito, and ParthsarathiShome. 1992. "The Roleof
Taxation in the Development of East Asian Economies." In TakaroshiIto and Anne 0. Krueger,eds.,
hser
Poliicagl
Economyof Tax RPfrm Chicago: Universiry of Chicgo Press.
Teranishi, Juro. 1986. "Economic Growth and Regulation of Financial Markcts:Japanese Experienceduring 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 Exception?" Backgroundpaper for TheEastAsian Miracle. World Bank, Policy Research Department,
Washington, D.C.
Thurow, Lester. 1992. Head to Heatd The ComingEconomic 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., FiscalIncentivesfor 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 Revolurion Occur in England?"In C. Rowley, R. Tollison, and G. Tullock, eds., The PoliticalEconont of
Rent-Seeking.Boston: KluwerAcademicPublishers.
387

.1988b. "Rents and Rent-Seeking."In C. Rowley,R.

Tollison, and Gordon Tullock, eds., The Political
Economyof Rent-Seeking.Boston: Kiuwer Academic
Publishers.
Turnham, David. 1993.EmploymentandDevelopmenetin
New ReviewofEvidence.Paris: Development Center,

The WallStreetJournaL1990. "Taiwan,Long Notd for

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 Taiwan, 1950-87." Ph.D diss.Universityof Californiaat
OECD.
Los Angeles.
Ueda, Kazuo. 1992. Insitutiona and Regrulatory
Frame- Warr, Peter G., and Bhanupong Nadhiprabha. 1993.
worksfortheMain Bank System.Tokyo: Universityof
"MacfocconomicPolicies,Crisis,and Growth in die
Tokyo.
Long Run: Thailand." Backgroundpaperprepared for
UNESCO.
Variousyears. StatisticalYearbook.Paris.
Litde, Ian M. D., Richard N. Coopcr, W. Max Corden, and Sarath Rajapathirana,cds., Boom,Crisis,AdExpience of Developing
justnent- The Macroeconomic
United Nations. Variousyears,a. DemographicYearbook.
Countries.New York:Oxford UniversityPress.
New York.
Wei, Shang-Jin. 1993. 'Open Door Policyand China's
_.
Various years, b. Eneg Statistics Yearbook.New
Rapid Growth: Evidencefrom City-LevelData." PreYork.
_.
1976. WorldEnergySupplies195o-74. New York.
pared for the Fourth Annual East Asian Seminar on
_S1992.WorldEconomicSutnvy: Current Trendsand
Economics, National Bureau of EconomicResearch,
Policiesin the WorldEconomy.New Yorkl
Cambridge,Mass.
uNop (United Nations Development Programme).1991.
Westphal,LarryE. 1978. "The Republicof Korea'sExperience with Exporr Led Development." WorldDevelHuman DevelopmentReporr.New York: Oxford UniversityPress.
opmnent6:
347-82.
Van Adams, A. 1989. "A Study of the Economic EnviWestphal, LarryE., YungWhee Rhee,and Garry Pursell.
ronnent's Influenceon Human Capital Formation in
1988. KoreanIndustrial Competenre:Wher It Came
Korea and Egypt." George Washington University,
FmronWorld Bank Staff Working Paper 469. WashWashingon, D.C
ington, D.C.
Viner, Aron. 1988. InsideJapaneseFinaial Me.
Whalley,John. 1990. "Non-DiscriminatoryDiscriminaHomewood, Ilk:
DowaJoneserFin.
don: Special
and Diffierential
Treatrnent
under
the
Homewood,
Dow
Jons--Invin.CArr
IL:
for Developing
Countries."
Economic
Journal
100(403):1318-28.
Vitts, Dimitri. 1991. 'Measuring CommercialBank Efficicncy:Use and Misuse of Bank Operating Ratios."
WiboonchutikulaPaitoon.1987."otal FactorProductivity Growth of Manufacmring Industries in ThaiPolicy ResearchWorking Paper 806. World Bank,
land." In Productivity Changes and International
Country EconomicsDepartment, Washington, D.C.
-_
.1992. 'Poliy Issues in Financial Regulation."
Competitiveness
of ThaiIndustries.Bangkok Thailand
DevelopmentResearchInstitute.
World Bank, Country Economics Department,
Winkler, Donald R. 1990. Higher Edruation in Latin
Washington, D.C. April.
Amerca: Issuesof Efficiencyand Equity. World Bank
Vogel,Ezra F., and DavidLindauer. 1991. "Towarda Social Compact for KoreanLabor." In David Lindauer,
DiscussionPaper 77. Washington, D.C.
Wolf, Charles. 1988. Markets or Governments:Choosing
Jong-Gie Kim, Joung-Woo Lee, Hy-Sop Lim, JacbetweenImpect Altenzetives.Cambridge, Mass:Mrr
YoungSon, and Ezra F. Vogel,eds., Korec The Strains
of EconomicGrowt. Cambridge, Mass.: Harvard InPress.
World Bank. 1983. "SpecialEconomic Reporc Policies
stiture for InternationalDevelopment.
Wade, Robert. 1989. 'What Can EconomiesLearn from
fior the Financial Secaor." Europe, Middle East, and
EastAsianSuccess?"Annab oft heAmericanAcademyof
North AfricaRegion,Washingron,D.C
. 1987a. Korec Managingthe Imndrial Transition2
PoliticalScience505: 68-79.
. 1990. Goening the Market EconomicTho?y and
volt Washington, D.C.
the Rolevfthe Governmentin EastAsian Industriai_. 1987b. WorldDevelopmentReport1987 New York.:
tiomzPrinceton, NJ.: PrincetonUniversityPress.
Oxford UniversityPress.
_

388

_.

1989a. 'Indonesia: Strategyfor Growth and StrucTrade Policyand the New International Econonmia.
Cambridge, Mass.:MIT Press.
tural Change." Counny Department V, Washington,
_
1985. The Politial Economy of fapan. Stanford,
D.C
Calif.: Stanford UniversityPress.
1989b. Malaysia:MatchingRisk and Rewardsina
Yamawaki,Hideki. 1988. 'The Steel Industry." In RyuMixedEconomy.Washington, D.C.
taro Komiya, Masahiro Okuno, and KomaroSuzu_. 1989c. WorldDevelopmentReport1989. New York:
mura, eds., Indusrial Poliy of Japan. New York:
Oxford UniversityPress.
l990a. Indonesic SterategyJfra SustainedReduction
1.
AcademicPress.
Yarnazawa,Ippie. 1988. "The Textile Industry."In Ryuin Povety. Washington, D.C.
taro Komiya, Masahiro Okuno, and Kotaro SuZU. 1990b. WorldDevelopment
Report1990. NewvYork:
Oxford Universiy Press.
murm, eds., Industrial Polic of Japan. New York:
1.l99a. Colombia:IndustrialCompettionand PJeforAcademicPress.
Yan, Lin See. 1991. 'Savings Investment Gap, Financing
-mance.
Washington, D.C.
-.
199lb. WorldDevelopmentReportThe Chalengeof
Needs and Capital Market DevelopmenL"Bank NeDevelopment.
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.
Yonezawa,Yoshie. 1988. YThe Shipbuilding Industry."
Washington, D.C
-.
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 ofEcoOxford UniversityPress.
nomics,Kelo University,Tokyo, Japan.
1993a. GlobalEconomicProspectsandtheDeveloping Young,Alwyn. 1992. "A Tale of Two Cities: Factor Accumulationand Technical Change in Hong Kong and
Counties.Washington, D.C
. 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
ReporGeneva.
in Malaysia" In Gerard Caprio Jr., Izak Ariyas,and
TheoryandExJames Hlanson,eds., FinancialRefrnrn
Yaamamura,
Kozo. 1986. "CaveatEnmptorThe Industrial
PoliCy of Japan." In Paul Krugman, ed., Strategic
perience.Cambridge:Cambridge UniversityPress.

The completebacklistof publicationsfrom the World Bank is shown in the

annual hIder
ofPublications,
which containsan alphabcical tide list and
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]

,..

C't

~~

~ ~~~~~~~~~~ClHINA6
r

6r-

j,

>~~~~~~~~~~~~~~~L

%L:0

befI/

_-.

j2kmmx~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

SfDN
I

Ais,~

-;, ;; D

D-

..

1)

~
~~~~~~~~~~-

1
*.