1987 dornbusch park - korean growth policy

7/27/2019 1987 Dornbusch Park - Korean Growth Policy

1/66

RUDIGER DORNBUSCHMassachusetts Institute of TechnologyYUNG CHUL PARKKorea University

K o r e a n r o w t h P o l i c yKOREA CONTINUESto be in the limelightas one of a handfulof developingcountries that have adjusted successfullyto both the oil shocks of the1970sand the debt shock of the early 1980s.Generallycreditedwiththesuccess is Korea's superioreconomic policy. Infact, Korea now servesas a model for the export-oriented trategyof developmentthat multi-lateral nstitutionsareurgingon countriesof Africaand LatinAmerica.This paper reviews the role of policy in Korea's success thus far andinvestigatesthe currentchallengesfacing governmentpolicymakers, nparticularheemergenceof a massive currentaccountsurplus.

We havebenefited rom hecommentsof ourdiscussantsandmembers f theBrookingsPanel.We would ike to acknowledge speciallytheresearchsupportandadvicewe havereceived from Won-Am Park, Sung-Hee Jwa, and Choong-Soo Kim of the KoreaDevelopment nstitute.ThispaperwaspreparedwhileYungChulParkwaswith he KoreaDevelopment nstitute.1. The Koreanexperiencehas led to a sizablerecent iterature f whichthefollowingareamong hemore mportant:BijanB. Aghevli andJorgeMarquez-Ruarte,A Case ofSuccessful Adjustment:Korea's Experience During 1980-84," OccasionalPaper 39(InternationalMonetary Fund, August 1985); S. Arndt, "Policy AdjustmentsUnderBalance of PaymentsEquilibriumFor the Republic of Korea" (AmericanEnterpriseInstitute,March1986);BelaBalassaandJohnWilliamson,Adjusting o Success:Balanceof Payments Policy in the East Asian NICs, Policy Analyses in International Economics17(Washington,D.C.: Institute orInternational conomics,June1987);RobertBaldwin,"U.S. andForeignCompetition n the DevelopingCountriesof the AsianPacificRim,"WorkingPaper2208 NationalBureauof EconomicResearch,April1987);ThorkilCasse,The Non-Conventional Approach to Stability: The Case of South Korea: An Analysis ofMacroeconomicPolicy (Copenhagen:Center for DevelopmentResearch, 1985);SusanCollins and A. W. Park, "Korean Macroeconomicsand Debt" (NationalBureauofEconomicResearchand HarvardUniversity, 1987);VittorioCorboandSangWooNam,"Korea'sMacroeconomicProspects and Major Policy Issues for the Next Decade,"

389


2/66

390 Brookings Papers on Economic Activity, 2:1987Two broad questions about Korea's experience are particularly

interesting.-What explainsthe Koreanadjustment uccess andgrowth perfor-mance?Have particular oliciesespecially contributed o this success?-Does Koreatoday have a structural xternalsurplus?And shouldpolicy respond?At a time when theworld'sdevelopingcountries,especially those inLatinAmerica,areeconomicallystagnantanddebt-laden,Koreaenjoyshighgrowth,relatively ow inflation,and a relativelyequal distributionof income.Itis theonly majordebtor hathas overcomethe debtproblemand has done so with a vengeance:debt is being paidoff, and the tradesurplus s so largethat it invites trade rictions.Since 1968, the Korean nonfactor current account has shown anupward rend, interrupted nly by the oil shocks of 1973and1979.Overthepastfiveyearsthe externalbalancehasbeen steadily improvingandin 1986-87 there was a sizable surplus.That surplusreflects a strongtradeperformance.Korea smakingtselffeltintheareaofmanufactures,competingwithother industrial ountries n the U.S. market.Korea's success raises questions not only about how it might betransported o otherdevelopingcountriesbutabout how Koreaaffectsand is affected by the industrializedcountries, especially the UnitedStates. What,for example, are the effects of dollar-yenexchangeratemovements withouta correspondingmovementin the dollar-wonrate?ReportDRD276 WorldBank,1987);WorldBank,"TheRecentMacroeconomic volutionof theRepublicof Korea:AnOverview,"ReportDRD208 WorldBank,February1987);CharlesR. Frank,KwangSukKim, and LarryE. Westphal,Foreign TradeRegimesandEconomic Development: South Korea (Columbia University Press, 1975);Wontack Hong,Trade, Distortions and Employment Growth in Korea (Seoul: Korea DevelopmentInstitute,1979);Anne 0. Krueger,"The Importance fEconomicPolicy nDevelopment:Contrasts Between Korea and Turkey," Working Paper 2195 (National Bureau ofEconomicResearch,March1987);EdwardS. Masonandothers, TheEconomicandSocialModernization of the Republic of Korea (Harvard University Press, 1980); Yung ChulPark, "ForeignDebt, Balance of Payments,and GrowthProspects:The Case of theRepublicof Korea,1965-1988,"WorldDevelopment,vol. 14(August1986),pp. 1019-58;TiborScitovsky, "EconomicDevelopment n TaiwanandSouth Korea:1965-1981," nLawrence J. Lau, ed., Models ofDevelopment:A Comparative Study ofEconomic Growthin South Korea and Taiwan San Francisco:ICS Press and WorldBank, 1986),pp. 135-95; World Bank, Korea: Development in a Global Context (Washington, D.C.: WorldBank, 1986);World Bank, Korea: Managing the Industrial Transition (Washington, D.C.:WorldBank, 1987).


3/66

Rudiger Dornbusch and Yung Chul Park 391If a won appreciationbenefitsprimarilyJapanandTaiwan,which as aresult become morecompetitivein the U.S. market, should the UnitedStates seek more specificallytargetedpolicy concessions fromKorea,such as import liberalization n areas of particular nterest to U.S.exporters?How much of the present Korean surplusis due to U.S.overspendingandwillvanishwithbudgetcorrection,andhow much,ifany, should be eliminated by policy action? How much pressure foradjustmentsn their bilateralbalancewith the UnitedStates can Koreaand other newly industrializedcountries (NICs) expect? Will Koreafollowpoliciesandperformance fJapan,steadilygaining hare nworldmarketswhile maintaininghehome marketsubstantially losed?This paper cannot answer all these questions. But it will set outanswers to the two main questions about Korea's experience raisedabove and, in doing so, lay the groundwork or answeringthe widerrangeof questions one can ask aboutsuccessful NICs concerning heirrole in the worldeconomyinthe comingyears.Thepaper allsbroadly nto threeparts.Thefirstpartreviews Koreangrowthhistoryand the structureof theeconomy. Thesecond partoffersexplanations or Korea's superiorperformance.A centralpoint is thatKoreanwages are exceptionallylow by international tandards,giventhe skill level of thelaborforce, andhenceprovide continuing cope fortrade success. Finally, we argue that Korea may well tend toward astructural urplus.But we also arguethatdramaticgovernmentactionto eliminatethe surplus would be premature.Uncertaintiesabout theworldeconomy, about domesticlabormarketdevelopments, andaboutthe forthcomingU.S. budgetadjustmentspoint to the possibility of asignificantdecline in the Korean current account surplus.Given thatpossibilityand the obvious difficultyof reversingreal appreciationorexpansiononce it has occurred, we conclude that policy initiativesshouldbe limited o selective import iberalization.But we do see roomfora major rade nitiative n the directionof free tradewith the UnitedStates.

A Reviewof Long-Term PerformanceDuring he past thirty-fiveyears Koreangross nationalproducthasincreasedmorethansevenfold. Althoughthe countryremainspoor by


4/66

392 Brookings Papers on Economic Activity, 2:1987Table 1. Comparative Levels of Real per Capita Income, Various Years, 1955-85aIndex, United States = 100Country 1955 1960 1970 1977 1985Korea 12 12 17 23 31Brazil 15 18 18 24 23Portugal 19 23 32 40 37Spain 31 33 49 53 48Israel 32 41 52 53 48Japan 23 33 64 67 77

Source: Robert Summersand Alan Heston, "Improved nternationalComparisons f Real Product and ItsComposition:1950-1980,"Reviewof Income atid Wealth,vol. 30 (June 1984), pp. 207-62. Data for 1985wereprovidedby the authors.a. Realgrossdomestic ncomeper capita n eachcountry adjustedorchanges n the termsof trade)relative oU.S. grossdomestic ncomepercapita n eachyear.

comparisonwith the industrializedworld,it hasplaceditselfamongthetop developing countries, just behind Singaporeand Taiwan and notmuchbehindPortugal.Table 1 shows a comparisonof real per capitagross domestic product usingpurchasing-power-adjusted easures ofincome. The purchasing-powerdjustments essential for internationalcomparisonsbecause systematicdifferences nrelativeprices otherwiselead to anunderestimate f the real income of poor countries.TheU.S.level of real income per capitain each year serves as a benchmark orcomparisons.2Koreanrelative growthstartedonly in the 1960s. Until thenrealpercapitaincomegrowthparalleled hat in the United States, butwas notstriking.But even with extremely high growthrates over an extendedperiodthe level of real incometodayinKorea is less than a third hat ofthe United States and does not yet matchthat of Spainor Portugal.Therelative evel of per capita ncomeinJapanand Korea has stayednearlyconstant since 1960.Korea today has the same 40 percentof Japan'sstandardof livingthat it had in 1960.But it has alreadyovertakenandmoved faraheadof Brazil,LatinAmerica'sstrongestgrowthperformer.Table 2 summarizes he growthand transformation f Korea sincetheearly 1960s.Emerging learly romthese dataarefive characteristicsof the economy:a sustained,exceptionallyhigh growthrate of output;

a structuralransformationf theeconomy, intermsof bothoutput2. These purchasing-power-adjustedealincome measures akeinto accountthe factthat n poorcountries he realpricesof servicestend to be low andthat,accordingly,GNPin dollars s not an appropriatemeasureof the actualstandard f living.


5/66

Rudiger Dornbusch and Yung Chul Park 393Table 2. Macroeconomic and Structural Patterns, Korea, Various Periods, 1963-86Percent

Measure 1963-73 1974-80 1981-86Annual inflation ratea 16.6 21.9 6.0Annual real GNP growth rate 9.7 7.2 8.7Manufacturing product 19.7 14.1 10.0Share in GNPAgricultureb 35.2 20.9 15.6Manufacturingc 15.7 28.0 33.1Other 49.1 51.1 5i.3Share in employment

Agricultureb 54.5 41.2 28.6Manufacturingc 12.6 21.5 23.5Other 32.8 37.2 47.9Exports as share of GNP 14.5 30.6 38.6Taxes as share of GNP 11.8 16.8 19.2Investment as share of GNP 22.0 30.4 30.2Foreign saving as share of GNP 8.3 7.2 4.3

Source: Bank of Korea, Econiomic Statistics Yearbook, various issues.a. GDP deflator.b. Includes orestryandfishing.c. Includesmining.

and employment,with a substantial decline in agriculture,a rise inmanufacturing,nd a growing mportanceof trade;-a significantncreaseinpublicsectorresources;-a sustainedhigh rate of investment;and-large, but declining,external inancing.A striking haracteristic f the Koreaneconomyis thegrowingshareof exports in GNP and the changingcompositionof its trade. Table 3shows thatin little morethantwenty years Korea moved frombeingacommodity exporterto being a net importerof commoditiesand anexporterof manufactures.The reorientation f the economy toward rade s equallyapparent nimport penetrationand export ratios. In manufacturing, he ratio ofexportsto total production ncreasedduring1970-83from11percent to21 percent,while the importcontent of the m-anufacturingector rosefrom 17.3 percent to 22.2 percent. The ratio of imports to domesticproduction n the manufacturingector has declinedover the past fifteenyears fromalmost20percentto only 15percent.3

3. Theimport atio s definedas the ratioof importedntermediate oodsto domesticproduction.The datacomefromBankof Korea,Input-OuitputTables,various ssues.


6/66

394 Brookings Papers on Economic Activity, 2:1987Table 3. Composition of Exports and Imports, Korea, 1962, 1970, 1985Percent of total

1962 1970 1985Indiustry Exports Imports Exports Imports Exports Imports

Commodities and processed foodsa 75.6 33.7 21.5 37.3 5.1 17.5Mineral fuelsb 5.0 7.3 1.0 6.9 3.1 23.6Chemicalsc 1.8 22.4 1.4 8.3 3.1 9.0Manufactured goodsd 17.6 36.6 76.1 47.6 88.7 49.9Machinery and transport equipment 2.6 16.5 7.4 29.7 37.6 34.2Source: Bank of Korea, Economic Planning Board (EPB), MaijorStatistics of Korean Economy,, various issues.Numbers may not add to totals because of rounding.a. Standard industrial trade classification (SITC) 0-2 and 4.b. SITC 3.c. SITC 5.d. SITC 6-9.

Finally, scarcelyless remarkable han Korea's transformations thatit took place underconditionsof relatively modest inflation.In the pasttwenty years inflationaveraged9percenta year andreacheda maximumof30percent.The nflationwas comparable o thatof the UnitedKingdomor Italy, very remotefrom the LatinAmericanexperienceof inflationrates of 100percent,as in Mexico, or the 1,000percentplus of Brazil orArgentina.Moreover,for the past few years inflationhas been less than3 percent.

ExplainingSuccessful Growth and TransformationEconomistsandpolicymakers eekingto applythelessons of Korea'ssuccess to poorly performingcountriesin Latin America should note

that the recipe is definitelynot simply "hands off, give free reign tomarket orces." Governmentnterventionhasbeenintense,andrestric-tions on trade and capital flows are the rule. Thus if any generaldescription s appropriate,t is thatthegovernmenthas forthemost partrunatight ship,sailingvery close to the wind.Inthe appendixwe reviewthemainphasesof Koreaneconomic historyandthe directionsof policysince the 1950s. In this section we identify policies that have spurredhigh growth.BROAD EXPLANATIONS FOR KOREA S GROWTHIn identifying the differences or similarities between Korea anddevelopingcountries n LatinAmerica,asummary omeseasily. Korea's


7/66

Rudiger Dornbusch and Yung Chul Park 395Table 4. Macroeconomic Performance, Korea and LDCs, Selected Periods, 1967-85Percent per year

Middle LatinMeasure Africa Asia Eu,opea East America KoreaInflation1967-76 8.5 9.4 9.0 8.7 24.5 7.21977-85 17.2 7.9 25.1 14.1 77.6 11.1Real GDP growth1967-76 5.0 5.2 6.0 9.3 5.9 10.31977-85 1.9 6.5 3.1 0.7 2.6 6.4

Sources: International Monetary Fund, World Economic Ouitlook, various issues, and EPB, Major Statistics ofKorean Econo,ny, various issues. Averages of country growth rates are weighted by the average U.S. dollar valueof GDPs over the preceding three years.a. Developing countries in Europe, that is, Southern Europe including Portugal, Spain, Greece, and Yugoslavia.

labor orce is better trainedand worksharder.Itspeople save more andborrow wisely. Policies are perhaps as activist but not grossly mis-directed.Budgetdeficitsaremoderate,andthe realexchangeraterarelygets overvalued. The differences add up to a performancestrikinglybetter thanthat of LatinAmerica, though not thatof otherSouth EastAsian countries,as table 4 shows.Korea has not been without macroeconomicdifficulties.In 1980-81,in the aftermathof the second oil shock, inflation ncreased sharply,outputdeclined,andthe externalbalancewas in disarray.Korea is alsoa majorLDC debtor. But, unlike the less successful countries, Koreanever allowed these problems to get far out of hand, or for long.Adjustment nvariablycame rapidly,before economic agents becameaccustomedand adjusted o instabilityand inflation.Economistshave tried o come up with a generalized ecipefor growththat might explain why some countries prosper in spite of adversitywhileothers do not. In an authoritative eview of whatis known aboutgrowth successes, Dervis and Petri compare a group of seven high-growth countries(Taiwan, Korea, Brazil,Thailand,Portugal,Greece,andYugoslavia)with a group of thirteen ess successful cases, rangingfromTurkeydown to theIvory Coast. Forthe period 1965-85the moresuccessful group shows annualper capita growth of 5.0 percent, asagainstonly 2.5 percent for the less successfulgroup. Dervis and Petriconclude that no single explanationsets winners apartfrom the rest.Theynote:Countries hatgrew rapidly hroughout hepasttwo decadeshave had to excelinseveraldimensions.Earlyon, highratesof investmentandfavorabledomestic


8/66

396 Brookings Papers on Economic Activity, 2:1987Figure 1. Baumol's Catching-Up Hypothesis

Per capita realGDP growth rate, 1960-81(percent)

8Korea ?7 .)

6-05 0 0

o 0 04OO4- 0 000 c00 o 03 000 0000 0 02 -o o00 0 01 gco 0?0 00G 0

0.6 1 1.4 1.8 2.2 2.6 31960 per capita income(thousands of 1970 "international" dollars)aSource: Authors' calculations based on William J. Baumol, "Productivity Growth, Convergence, and Welfare:What the Long-Run Data Show," American Economic Reviewt',vol. 76 (December 1986), fig. 3. The data are fromRobert Summers and Alan Heston, "Improved International Comparisons of Real Product and Its Composition:1950-1980," Review of Income and Wealth, vol. 30 (June 1984), pp. 207-62.a. Adjusted for changes in the terms of trade as in Summers and Heston.

preconditionswere the most significant orrelateswith success. Between 1978and 1979 ast growthcalledfor highinvestmentandfrugal iscal policies. After1979, debt and especially the financingof debt throughhighexports becameparamount.4

Baumol has offered the hypothesis that rapid growth is a reflection ofcatching up.' He argues, drawing on the hundred-year evidence of

4. Kemal Dervis andPeter A. Petri, "TheMacroeconomicsof SuccessfulDevelop-ment:Whatare the Lessons?" in StanleyFischer,ed., NBERMacroeconomicsAnnual,1987 MITPress, 1987),pp.211-54.5. See WilliamJ. Baumol, "ProductivityGrowth, ConvergenceandWelfare:Whatthe Long-RunData Show," AmericanEcononmic Review, vol. 76 (December 1986),pp. 1072-85.


9/66

Rudiger Dornbusch and Yung Chul Park 397Table5. Sourcesof EconomicGrowth n Korea,1963-82Percentper year

Measure 1963-72 1972-82RealGDP 8.2 8.0Totalfactor input 4.2 5.6Labor 3.1 3.5Capital 1.1 2.1Outputper unitof input 4.0 2.4

Sources: Korea Development Institute, Quarterly Economic Review (Seoul: KDI, 1986), p. 33; and K. S. Kimand J. K. Park, Souirces of Economic Growth in Korea: 1963-1982 (Seoul: KDI, 1985), pp. 61-62.

industrialcountries,that there is a tendency towardconvergence: thecountrieswith thelowest levels of productivityhave thehighestrates ofgrowth.An essential ngredient ortheirhighgrowth s theability o takeadvantageof existing knowledgeandtechnology. Since thatprocess issubjectto diminishingreturns, convergenceultimatelysets in and thecatching-up ountries slow to the commonrate of long-termgrowth ofthe most advancedcountries.Figure1 shows per capita incomein 1960andincomegrowth during1960-81 for a group of middle-incomecountriesranging romLesothoto Greece, Portugal,and Spain. AlthoughKorea supportsthe Baumolhypothesis, the entire sampledoes not.Absenta single broad explanation or growth,we turnto look moreclosely at the individualbuilding blocks of Korea's success. As apreliminary,we briefly review Denison-style growth accounting forKorea. Table 5 reports estimates of the sources of growth in Koreabetween 1963 and 1982. More than half the growth rate of output isexplainedby increasesin factorinputs, labor and capital. Slightlylessthanhalf is due to growthof total factor productivity,which dependsmainlyon scaleeconomies and advancesinknowledge.Inourview, thecornerstoneof Koreangrowth s a highlytrainedandproductiveabor orcewhose wagesarelow by international tandards.Korea's laborforce is a precondition or the highrates of investmentandcapacityexpansion hat makethe export expansionpossible.

THE LABOR FORCE AND WAGESThebroadest,vaguest explanation or Korea's success drawsatten-tion to the people, the extent of their work effort, their education, andthe distribution f income among hem.


10/66

398 Brookings Papers on Economic Activity, 2:1987Table 6. Hours of Work in Manufacturing,1960-85Hours per week

UnitedP'eriod Korea Greece Mexico Japan States Germany1960-69 55.5a 43.9 45.9 45.4 40.6 44.01970-79 51.8 42.9 45.6 41.1 40.1 42.11980-85 53.8 39.1 46.1 41.2 40.0 40.9

Source: International Labor Office, Yearbook of Labor Statistics, various issues.a. Average for 1963-69.

Table6 shows weekly work hours in Korea and a numberof othercountries. The simple fact is that Koreans work longer. The Koreanwork week, which has actually ncreased since the 1970s, is 35 percentlonger than that in industrialized ountriesand 17 percent longer thanthat in Mexico. The extent of work effort is described by Hong in thefollowing erms: "Even nowadays,atypicalKoreanwhitecollarworkerleaves home before 7 o'clock in the morningand leaves his office after 8o'clock inthe eveningevery day. . .. He usuallyworks ateon Saturdayafternoons and, if something goes wrong, has to go to work onSunday. " 6Duringthe several decades of Japanese occupation before WorldWarII, Koreancitizens suffered from a poor educationalsystem. Inresponse, they placedan enormousvalue on education n thereconstruc-tion period following independenceand the Korean War.The commit-mentto educationhas, ifanything, ncreasedovertheyears. The averageeducation level of employedmales was 7.2 years in 1960,9.3 years in1970,and 10.3years in 1980.7Table7 showsaninternational omparisonof enrollment evels in secondaryschools and highereducation.In 1960Korea alreadyexceeded by a substantialmargin he averageforupper-middle-incomecountries. By 1983the countrywas well on the way toeducationalstandardsof industrial ountries.

6. See Wontack Hong, "Export-OrientedGrowth of Korea: A Possible Path toAdvancedEconomy," SeminarPaper382 (Institute or International conomics,Stock-holm University, 1984).7. These datarefer o maleemployees n thenonagriculturalector. Thecorrespondingnumbers or the entire abor orceare4.2, 6.4, and8.0. The datacome fromK. S. KimandJ. K. Park, Sources of Economic Growth in Korea: 1963-1982 (Seoul: Korea DevelopmentInstitute, 1985),p. 18.


11/66

Rudiger Dornbusch and Yung Chul Park 399Table 7. Educational Enrollment Levels, 1960 and 1983

Secondary school Higher education(percent of (percent ofage group) age group)Countuy 1960 1983 1960 1983

Middle-income ountriesLower 10 40 3 12Upper 20 55 4 14Industrial ountries 64 85 16 37Korea 27 89 5 24Source: World Bank, World Develop,nenit Report, 1983 and 1986 (World Bank, 1983 and 1986).

Until recently, laborunrest andunionactivity have not been majorissues in Korea. Culturalcharacteristicsmay contribute to peacefullaborrelations, andpolitics, certainly,has left littleroomfor organizedlaborandeven less forunionmilitancy.Butgovernmentpolicy has alsohelpedbringabout a relativeequality in income distribution hat mayhavehelpedavoid laborproblems.Improvements n Koreanincome distributionsince 1965, shown intable 8, are the result of strong growth in employment, and hencedecliningunemployment.Exactly how income distributionhas influ-enced growth, other than by promotingsocial stability, is open todiscussion. Butitwouldcertainlyshapethedomesticmarket irms ace,it mayinfluencesavingbehavior, must influencepolitics, andmay haveimportantmplications ortheease withwhichthegovernmentcanshifteconomicpolicies.Rapid economic growthis often accompaniedby a deterioration nTable8. IncomeDistribution,Korea,VariousYears,1965-85Percent

Item 1965 1970 1980 1985RuralhouseholdsIncomeshare of bottom40 percent 22.6 21.2 17.5 19.7Incomeshare of top 20 percent 38.0 38.6 42.2 38.7Percentbelow poverty linea 10.0 3.4 11.2 7.5UrbanhouseholdsIncomeshareof bottom40 percent 14.1 18.9 15.3 16.7Incomeshare of top 20 percent 47.0 43.0 46.9 45.6Percentbelow povertylinea 17.9 7.0 15.1 7.8

Source: Korea Development Institute.a. Poverty line defined as one-third of average household income.


12/66

400 Brookings Papers on Economic Activity, 2:1987Table 9. Income Distribution ComparisonPercent

HongPopulation Koreaa Brazilb Mexicoc Thailandd Konge Hungary' SpaingIncome shareof bottom40 percent 16.9 7.0 9.9 15.2 16.2 20.5 19.4Income shareof top20 percent 45.3 66.6 57.7 49.8 47.0 35.8 40.0

Source: World Bank, World Developmenit Report, 1986.a. 1976. b. 1972. c. 1977. d. 1975-76. e. 1980. f. 1982. g. 1980-81.

incomedistribution ue to shortagesof skilledandeducated abor.Koreaexperienced these shortages to some extent in the 1970s, but theexpansionof educationsincethe 1950smayhavehelpedcontainadverseconsequences for growth.Income distribution n Korea is similarto that in other East AsianNICs and ndevelopedcountries.Itis extraordinarily ifferent romthatin Latin America,as table 9 shows.A finaldimension hat s muchharder o quantify s entrepreneurship.Korea has rapidlybuilt up large, Japanese-styleproductionand tradingconglomerates.But Korean growth and export success also has de-pended on massive gambles, especiallyin the late 1970s.The outward-looking strategy would not have succeeded without Korea's dynamicand highlytrainedentrepreneurs,yet anotherlegacy of Korea's com-mitment o education.8The punchline in this discussionof the labor force appears n table

10. As that comparisonof hourly compensation n major ndustrializedcountries andin Korea, Taiwan,and Singaporeshows, Korea's dollarwage is the lowest of all. The data in that table must be somewhatqualified,however, because wagesdiffersignificantlyromone industryto another. In some export-sector industries, wages in 1986 weresignificantlyhigher hanthose showninthe table andmuchcloser to thecorrespondingwages in Taiwan. For example, in 1986 n the iron andsteel sector the Koreanwage was $2.17, as against $2.29in Taiwan; nmotorvehiclesandequipmentmanufactureshe Koreanwagewas$2.12,comparedwith$2.21forTaiwan.Thus nthetradedgoods sector, wages8. See Noel F. McGinn and others, Education and Development in Korea (HarvardUniversityPress, 1980).


13/66

Rudiger Dornbusch and Yang Chul Park 401Table 10. Hourly Compensation in Manufacturing, 1986-87U.S. dollarsper hour Country 1986 1987a

United States 13.21 13.21Germany 13.35 15.92Japan 9.47 11.03United Kingdom 7.50 8.32Korea 1.39 1.52Taiwan 1.66 2.02Singapore 2.23 2.29Hong Kong 1.88 1.88Source: U.S. Department of Labor, Bureau of Labor Statistics, and authors' calculations.a. The 1987 data represent 1986 levels of hourly compensation evaluated at June 1987 exchange rates.

among Asian competitorsaremuch closer thanindicatedby the manu-facturingaverage, which, for Korea,even more than for the moreopenAsian economies, includes a share of lower-skilled abor in the homegoods sector.9Still, the wage level is strikingly ow by internationalstandards.THE DEVELOPMENT STRATEGYGovernmentpolicy support or Korea's growth goes beyond educa-tion andwagerestraint. nparticular, overnmentpolicieshavesustainedrelative financialstability by never allowing massive, money-financeddeficits and by following a real exchange rate policy that sustainedprofitabilityof the traded goods sector rather than, Argentinian-orMexican-style, nvitingcapital light.Beyondthat,thegovernmentused

subsidies and preferentialcredit allocation to channel capital to thetradedgoods sector. Thestrategywas clearly activist,but it avoidedthepitfalls of protection so apparentin some Latin Americancountrieswhereimportsubstitutionoften came at the price of exports and henceled sooner orlater to balance of paymentscrises.Korean growth in the past thirty years resembles that of postwarGermany, specially n the 1950sandearly1960s,or ofJapan.Anample9. There are no up-to-date ndexesfor hourly compensation n Latin America.Thelatestavailabledata are for 1985.At that time hourlycompensationwas $1.73 n Brazil,$2.66 nMexico,and$2.07 nVenezuela,as against$1.44 n Korea.ForPortugal ndSpainthecorresponding 985numbersare $1.53 and $4.79. The data are from the U.S. Bureauof LaborStatistics.


14/66

402 Brookings Papers on Economic Activity, 2:1987supply of labor, first rom agriculture ndsubsequently hroughmigrantworkers, put steady pressure on wages, and policy, too, discouragedextravagantwage settlements. As a result, wages that were low relativeto productivitytranslated nto profitability,high investment, and sus-tained growth.10Low relative wages and high investment assured pro-ductivitygrowthand hence steady, highgrowthof realwages, as shownfor Germany and Japan in table 11.11Despite the broadsimilarities n the wage-productivity elationshipof GermanyandJapan,policies in the two countrieswere not quitethesame. Both successfullypursuedwage restraint,a "realistic"exchangerate policy, and, as we will see below, a currentaccount surplus. ButGermanyrelied more on import iberalization,both unilaterallyand inthe context of the Common Market,while Japan never opened up inmanufacturing. apan'sgrowth lasted through the 1960s and into theearly 1970s.InGermany,realwage demandsand a strongemphasisonconsumptionn the mid-1960smarked he endof thehigh-growth hase.Today Korea stands roughly in the same position relative to theUnited States in which Japanstood in 1960. The standardof living isone-third hatof theUnitedStates,andthe level of hourlycompensationis about one-tenth. Like Japan n the 1950sand 1960s, South East Asia,and Korea inparticular,practices wage restraint,high saving,andhighinvestment.Thestrategydelivershighgrowthrates frealwagesbecauseproductivitygrowth nvariablyrunsslightlyahead of wage increasesatthe competitivemargin.We considernext how exchangeratepolicy and subsidyand creditpolicies helpedreinforce hisgrowth strategy.

Outward-Oriented rowth. Japanhas developedwith a nearlycon-stant ratio of exports to GNP. In Korea, by contrast, the export-GNPratio increasedfrom less than6 percentin the early 1960sto around40percentin the 1980s. Korea's German-style"outward-oriented" trat-egy is the characteristicof its developmentmost commonly singledoutas thekeyto success. Rather hanpushing nefficient mport ubstitutionfor its small domestic market,Korea has opted for outward-oriented10. HerbertGiersch,"Arbeit,LohnundProduktivitat,"Weltwirtschaftlichesrchiv,vol. 119, no. 1 (1983), pp. 1-18, has discussed extensively this paradigm or the case ofGermany.11. By comparison he averageannualgrowthrate of unit labor costs in the UnitedStates during1950-65was 1.9 percent.


15/66

Rudiger Dornbusch and Yaing Chul Park 403Table 11. Growth in Manufacturing in Germany and Japan, 1950-65Percent per year

Dollar unit RealCountiy Productivity Emnployment labor cost wagesa InvestmentbGermany 7.0 3.3 2.8 7.2 22.8Japan 7.2 5.7 1.4 5.7 26.8

Source: U.S. Department of Labor, Bureau of Labor Statistics.a. Hourly compensation deflated by the CPI.b. Gross fixed capital formation as a fraction of GNP; Japan, 1952-64, Germany, 1950-64.

growth.Except in 1979-81,governmentpolicy has avoided anoverval-ued exchange rate.By standing nthe way of a strongunionmovement,thegovernmentsanctionedthe marketpressureon wages generatedbytherural-to-urbanabormigration.Wages husroseslowly despitestrongproductivitygrowth in manufacturing.Growth in employment andcontinuingprofitabilityof the export sector rather than much fastergrowthin manufacturing eal wages were the result. Industrialpolicyandprotectioncombined o yield an incentivestructure hatfavored anexport-orientedndustrialization.Korea's policies clearly do not representa laissez-faireapproach:intervention in the form of trade restrictions, subsidies, and creditallocation is pervasive.12 Intervention has also been used in LatinAmerica,but withmixed results. In the 1930sLatinAmericadevelopedimportsubstitutionas a response to the GreatDepression. As AngusMaddisondocuments, the import-substitution trategy was initiallysuccessful: Latin America grew, whereas industrial countries stag-nated.13ButfollowingWorldWarIIfurther mportsubstitution an nto

12. See Anne 0. Krueger, The Developmental Role of the Foreign Sector and Aid(HarvardUniversityPress, 1979);Krueger,"Export-ledIndustrialGrowthReconsid-ered," in WontackHong and Lawrence B. Krause, eds., Tradeand Growthof theAdvancedDeveloping Countries in the Pacific Basin (Seoul: Korea Development Institute,1981), pp. 3-27; Krueger, Trade and Employment in Developing Countries: Synthesis andConclusionsUniversityofChicagoPress, 1983);Krueger,"TheImportance fEconomicPolicyin Development";DeepakLal, "IdeologyandIndustrializationn IndiaandEastAsia,"ReportDRD218 WorldBank,January 986);DeepakLalandSarathRajapatirama,"ForeignTradeRegimesandEconomicGrowth nDevelopingCountries,"ReportDRD217 (WorldBank, September1986);and WontackHong, "Export-OrientedGrowth nKorea," ordiscussionsof theoutward-orientedmodelas applied nKorea.13. See Angus Maddison, Two Crises: Latin America and Asia: 1929-1938 and 1973-1983 Paris:OrganizationorEconomicCooperation ndDevelopment,1985).


16/66

404 Brookings Papers on Economic Activity, 2:1987the limitationsofoverly smalldomesticmarkets.Moreover, ncreasinglypervasive importprotection soon became an impedimentto exports.Slowgrowth,overvaluedexchangerates,and nefficientndustrieswerethe legacy of thatpolicy everywhere except Brazil.Perhapsbecauseofits largerdomesticmarket,Brazilsuccessfullyused protectionto buildup a highlyefficient industrial tructure.Brazilalso, unlike otherLatinAmericancountries, avoided the externalbottlenecks that arise fromimplicit taxes on exports. Exchange rate and tax policies stronglysupported xportsandthusavoidedrecurrentpaymentscrises withtheiradverse macroeconomiceffects on confidence, nflation,andrecession.The Koreanstrategy s muchthe same, withpervasiveprotection ofan infant-industry indgoinghand in hand withfavorabletreatmentofthe export sector through ax incentives andcredit.The economythusmaintainsa constantactivecontactwith theworldeconomyon both theimportandexportside. Korea goes further n its outwardorientationbyallowing firms to take advantage of intermediategoods imports toenhance theirexport competitiveness.Such an outward-oriented trategy is supportedby the efficiencyadvantageof freertrade over a restrictivetraderegime.Export promo-tion is moreclosely relatedto free trade than is importsubstitution.Inaddition, he effects of enlarged rade on savingandinvestment,and ontechnologyand firms'behavior,and thepossibilityof structural hangecoming romopening ndustries oworldmarketsalladdto theargumentfor export promotion.14However, while the high positive correlationbetween exportperformanceandeconomic growth is an accepted factindevelopmenteconomics,recent studiesfailto confirm hattheformercauses the latter. 5

14. See W. M. Corden,"The Effectsof Tradeon the Rateof Growth," nJagdishN.Bhagwati and others, eds., Trade, the Balance of Payments and Growth: Papers inInternational Economics in Honor of Charles P. Kindleberger (Amsterdam: North-Holland,1971),pp. 117-43;Krueger,"Export-led ndustrialGrowthReconsidered"; ndHong,"Export-Oriented rowthof Korea."15. Bela Balassa, "Exportsand Economic Growth:FurtherEvidence,"JournalofDevelopmentEconomics,vol. 5 (June 1978),pp. 181-89;Corden,"The Effects of Tradeon the Rateof Growth";A. F. Darrat,"AreExportsanEngineof Growth?AnotherLookat theEvidence," AppliedEconomics,vol. 19(February1987),pp. 277-83;Hong, Trade,Distortions and EmploymnentGrowth in Korea; Wontack Hong, "Import Restriction andImportLiberalization n Export-OrientedDevelopingEconomy" (KoreaDevelopmentInstitute,1986);Hong,"Export-Oriented rowthandTradePatternsof Korea," inColin


17/66

Rudiger Dornbusch and Yung Chul Park 405The chief differencebetweentheKoreanstrategyofoutward-oriented

growth and the failed import-substitution olicies in LatinAmerica inthe 1950s is that in LatinAmericaprotection was too comprehensive,andtoo littleattentionwas paidto thepossibilityof exportingmanufac-turesto complement mportsubstitution.Thepolicybecame inefficientinpartbecausehighvalueaddedsectors (atworldprices)weresacrificed,but perhapseven more because the export sector atrophied.The Latineconomieswere thereforemore crisisprone,and theirmacroeconomicperformanceworsened.In Korea, an industrialization trategy that began as pure importsubstitution,with aidfinancinghe tradegap, expanded o includeexportpromotion.By around1960,Korea hadvirtuallyexhaustedthepossibil-ity of rapidgrowth hrough mportsubstitutionof nondurable onsumergoods and intermediate nputs. Additional mportsubstitutionof ma-chinery, consumerdurables,and theirintermediatenputswas rejectedbecause the domestic marketwas too smalland the capitalrequirementsof such ventures too large, especially given the chronic shortage offoreignexchanges.The rationale orprotecting nfant ndustries s thatindustriesof highgrowth potential but subject to externality stemming from marketimperfection, conomies ofscale,orcapitalmarketmperfectiondeserveencouragement,preferablyby productionsubsidiesbut, second best,by protection.Thatargumentprovidestheunderlying ogicas much forimport ubstitution sforexportpromotion.Koreahasoperatedon bothfronts. It has used tariffsand licensing to createa shelteredmarket orthe developmentof infants. And as these industries have developed,

Koreahas turned them toward the world marketby subsidies, credit,and exchange rate policy. The credit system has channeledfinancialresourcesat subsidizedrates to preferredactivities. The tax system hasprovided an exemptionfrom importduties for export content (oftenI. Bradford and William H. Branson, eds., Trade and Structural Change in Pacific Asia(Universityof ChicagoPress, 1986),pp. 273-305;Woo S. JungandPeyton J. Marshall,"Exports, Growthand Causality n Developing Countries," Journal of DevelopmentEconomics,vol. 18(May-June1985),pp. 1-12;RostamM.Kavoussi,"ExportExpansionandEconomicGrowth:FurtherEmpiricalEvidence," Journalof DevelopmentEconom-ics,vol. 14 January-February984),pp. 241-50;MichaelMichaely,"ExportsandGrowth:An Empirical Investigation," Journal of Development Economics, vol. 4 (March 1977),pp. 49-53.


18/66

406 Brookings Papers on Economic Activity, 2:1987amounting o muchmorethanadrawback),avorable axrates on profitsand incomes, anddirect cash subsidies. The implicit subsidy, from allsources, per dollarexportis shownin table 12. The table separatesnetexport subsidiesandgross;the latterincludeexemptionsfromindirecttaxes and tariffs.Thecombinedeffect of the taxsystem andthe creditsystem is difficultto estimate,inpartbecausevariousactivitieseven withina given sectoraretreateddifferently.In a systemwithpreferential reditallocation,atleast as important s the rateof interest s wherethe creditgoes. Becausethe export sector receives a majorshareof availableofficial credit, thesubsidydata in table 12 aresurelyunderestimated.As an economy becomes more industrializedand more advanced,conflicts between importsubstitutionand an active export sector in-crease. Exportersoften requireaccess to lower-cost or higher-qualityinputs of intermediategoods than the home marketcan yet deliver. Inthatsituation,protectionmustbeflexible. Koreamovedinthatdirectionin the early 1980swhen it opened its markets o certain mportscrucialto the export sector, whilecontinuing o protect infant ndustries.

ExchangeRate Policy. A policy of export-ledgrowthwouldinmostcircumstances include a wage in dollars that, in combination withtechnology, capacity, and productivity,would make a country highlycompetitive.At firstsightit mightappearthatwage andexchangeratepolicies in Korea did not infact combineto producethisresult.Figure2shows Koreanunitlabor costs in dollars relative to an averageof U.S.andJapanesedollarunit laborcosts. Theindexnearlydoubledbetween1973 and 1979,and the real depreciationof the won in the early 1980srolled back only a small part of the increase. The sharp increase inKorean relative unit labor costs during 1973-79 might suggest thatcompetitivenessmusthavesuffered,butthe tradeperformancendicatesno such thing.Theexplanations thatKoreanhourlycompensationnmanufacturingremains even today extraordinarilyow by comparison with that inindustrialized ountries.Ina considerable angeof activities, productiv-ity differentialsmay now be negligible.The existing wage differentialthus representsan open invitation or industrialexpansionand exportexpansion.As new industriesopen up, implementingoreign echnologyanddrawingontheworldcapitalmarket orfinancial esources ofinancecapital expansion, a rise in the average relative unit labor cost is


19/66

Rudiger Dornbusch and Yung Chul Park 407Table 12. Export Subsidies, Korea, 1961-80Percentagesubsidy per dollarforeign exchangeSubsidy 1961-69 1970-79 1980

Gross 25.3 22.2 21.3Net 12.8 3.3 3.3Sources: Larry E. Westphal and Kwang-Suk Kim, "Korea," in Bela Balassa, ed., Development Strategies inSemi-Induistrial Economies (Washington, D.C.: World Bank, 1982), pp. 212-79; C. H. Nam, "Trade, IndustrialPolicies and the Structure of Protection in Korea," in Wontack Hong and Lawrence B. Krause, eds., Trade andGrowth of the Advanced Developing Countries in the Pacific Basin (Seoul: Korea Development Institute, 1981), pp.168-89; and Kwang-Suk Kim, "The Timing and Sequencing of a Trade Liberalization Policy-The Case of Korea"(Seoul: Korea Development Institute, 1986).

warranted.Policy haspushedindustries ntorangesof increasingvalueaddedrather hanpulling aborout of suchindustriesas, say, textiles orrubber ootwear.Thusif labor s beingreallocated owardproductionofcars and electronics, away from low value added activities, exportcompetitivenesssurvives even withrisingunitlaborcosts.This pointcan be developedin terms of a Ricardianmodelof exportof technology, as shown in figure3.16The modeldetermines or a two-countryworldthe equilibrium elativewageand the geographicpatternof specialization.Let w/w*be the wage of the poor country relativetothe rich one and let A(z) = a*(z)la(z) represent the relativeunit laborrequirementof commodity z in the rich countryrelative to that in thepoor country. Along the vertical axis we measure the relative wagew/w* and the relative unit labor requirements, a*(z)la(z). Along thehorizontalaxis we align the range of goods, z, with the poor countryrelativelymore efficient in the productionof commoditiesnearer theorigin.Geographic pecialization s determinedby relative unitlaborcosts.The homecountrywillproduceallthosegoodsforwhich unit aborcostsare ess thanthe unit aborcost of the samegoodproducedabroad.Thusfor a particular ood z, productionwill be at home if wa(z) < w*a*(z).16. R. Dornbusch,S. Fischer,andP.A. Samuelson,"ComparativeAdvantage,Trade,and Payments n a RicardianModelwith a Continuum f Goods," AmericanEconomicReview, vol. 67 (December1977),pp. 823-39; Paul Krugman,"A Modelof Innovation,TechnologyTransfer nd the WorldDistribution f Income,"Journalof PoliticalEcon-

omy, vol. 87 (April1979),pp. 253-66;Krugman,"TechnologyGaps, TechnologyTrans-fers, andthe ChangingCharacter f U.S. Trade" MIT, 1982);Krugman,"A TechnologyGap Model of InternationalTrade" (MIT, 1982); and Susan M. Collins, "TechnicalProgress n a Three-CountryRicardianModel with a Continuum f Goods," JournalofInternational Economics, vol. 19 (August 1985), pp. 170-79.


20/66

408 Br-ookings Papers on Economic Activity, 2:1987Figure 2. Relative Unit Labor Costs, Korea, 1964-86aIndex, 1980 = 100110

100 -

90

8070-

60 -

50 1965 1970 1975 1980 1985Source: U.S. Department of Labor, Bureau of Labor Statistics.a. Korean manufactturingunit labor costs in dollars relative to an average of U.S. and Japanese unit labor costs.

For a given relativewage, w/w*,we thus obtain the competitivemarginbetween productionat home and abroad. But the relative wage isendogenousand s determinedby demandandrelative evels of spending.Demandconditionsare shown by the schedule OB, alongwhich demandfordomesticallyproducedgoodsis equalto thefull-employment upply.An increase nthe rangeof goods producedbythehome country movingalong the horizontalaxis to theright)createsanexcess demand orlaborand hence leads to an increasein the equilibrium elativewage. PointErepresents the general equilibriumwhere goods markets clear andproductionoccurs in the lowest-cost location. The home country pro-duces goods in the range Ozo,and the foreign country, the range ofproductsto the rightof zo.We now use this framework o ask what happensto relativewagesand to tradepatternswhen superior oreign technologyis introduced nthe poor country. At point H in figure 4, w/w* = 1; at the initialequilibrium t point E, wlw*< 1, andthe foreign countryhas superiortechnologyforgoodsthat thepoor country s alreadyproducing.As this


21/66

Rudiger Dornbusch and Yung Chul Park 409Figure 3. The Ricardian Model

wlw*, A(z)

B

(w/w*)o ___ _ _ _

IA(z)

G Zo

technology is imported,the poor country'srelative unit labor require-ment declines and the A(z) schedule rotates upward.The improvedtechnologyreduces labor costs in the export industry and allows thepoor country o expandtherangeof goodsit cancompetitivelyproduce.Butthe attempt o expandtheexport sectorinevitablycreatesanexcessdemand or laborand hence leads to a rise in the relativewage. A newtradeequilibrium esultsatpointE' with anincrease ntherelativewagefromxo o x' andanexpansion romzo oz' intherangeofgoodsproducedin the poor country.Now considerwhat the importof superiortechnology does to unitlabor costs. For marginal ndustriesthe relative unit laborcost of thepoor countrydeclines.This is true,forexample,nearpointE'. But it isnot necessarilytruefor the average. For activities to the left of point Hthere s nochange ntechnologyand a rise in the average relativewage.


22/66

410 Brookings Papers on Economic Activity, 2.]1987Hence it is quite possible in this model that a country shows a rise in itsrelative unit labor cost and yet shows a gain in competitiveness andhence an expansion n the rangeof goodsproduced. 7Indeed, the larger he gainin relativeunitlaborcosts, the largertheloss to the rich country rom ts export of technology.Thereason is thattraditional xport industries hat have notbenefited romthe transferoftechnology now pay higherwages. As Korea, for example, moves intocars and computersas a result of productivitygains, other tradablesectors experienceincreasedcosts andhencechargehigherprices. Inamulticountry ontext Korea is a middle-income ountry,whose gaininproductivityhas spillovereffects on poorercountries who now gainincompetitiveness n the middle-income ountry'straditional xportsec-tors. 18Thereare severalotherways in which a NIC canbecome moreexportcompetitive and yet show a rise in the relative unit labor cost. Eachcomplementsthe technology-transfer xplanation.One obvious possi-bility is reduced-costaccess to imported ntermediategoods. Here thetax incentives and credit subsidiesalreadydiscussedclearlyplaya role.Another possibility is to become a supplierof intermediategoods inmore advancedcountries. The finalpossibility is to move in Japan'stracks, pickingup industries hatinJapanhave becomeoverly costly.Theexplanationsdrawn rom Ricardian radetheoryareparticularlysuitablefor Korea because they highlightproductivitygrowth, whichhas played such a central role in Korea's growing export competitive-ness. But to what extent are these resultsthe outcome of market orcesand to what extent do they dependon policies? Since the early 1970sKorea has experienced both a risingshareof exports in GDP and anincreasing share in world manufacturesexports. From 1973 to 1985Korea's manufactures xports increasedin volume termsat an annualaveragerateof 14.6percent,while its share n themanufactures xportsof developingcountries rose from 11percent to 18.5percent.19Wageand exchange rate policy did not stand in the way of these market

17. When here s a nontradedmanufacturingectorwithoutproductivity rowth, hepresumptionhatthe averagerelativeunit aborcost rises is further trengthened.18. See Collins, "TechnicalProgress,"for a model thatdevelopsthese effects in athree-countryetting.19. See UnitedNations,GeneralAgreement nTariffsandTrade, nternationalTrade1985-86(GATT,1986),p. 16.


23/66

Rudiger Dornbusch and Yung Chul Park 411Figure 4. Importing Superior Technologywlw*, A(z)

B

\ / ~~~~~~~~A'(z)

/ \ ~~~~~~A(z)

0L I z

1-%~~~~~~~~z

developments n the manner hathappens requently n LatinAmerica:by discouragingunion activity and strikes, the governmentpromotedemploymentgrowth and investment in the export sector ratherthanastill fastergrowthin real wages. And by avoidingovervaluation,unitlaborcosts in dollarswere kept fromrisingfaster than a broadrange ofnewand arising ndustriescould afford.

SAVING, INVESTMENT, AND FINANCIAL POLICYFor Koreato takeadvantageof its exportopportunities, t needed anexpandedmanufacturing ase. For that, it neededhighratesof invest-ment.Figure5 shows the growthof investment as a ratioof GNP since1970.Table 13, which shows the financingof investmentby domesticand foreign saving, makes clear that Korea amply used external re-


24/66

412 Brookings Papers on Economic Activity, 2:1987Figure 5. Investment to GNP Ratio, Korea, 1970-86aRatio0.34

0.32 -

0.30 -

0.28

0.26 -

0.24

0.22

0.20 1975 1980 1985Source: Bank of Korea, Economic Statistics Yearbook, various issues.a. Gross fixed capital formation in the national income accounts.

sources. Until 1986the currentaccount was continuously n deficit. Insome years the deficitreachedmore than 10percent of GNP. But unlikeMexico or Argentina,Koreausedexternal inance or investmentratherthanfor consumptionor capitalflight.High rates of investmentare not sufficient or success. If investmentis misallocated so that it has a low social rate of return, then a highinvestmentratecan ultimately by way of debt service problems)becomea difficulty ather hana source of growth.There seems tobewidespreadagreement hat the heavy and chemical industry nvestmentcampaignof the 1970s nvolved a misallocationof resources.20We know today thatworld excess capacity in these industries made investment in themdubious.But there is no hard evidence that Korea's investments werein fact poor. The udgment hat the investmentdrivewas a poor idea wasmade in 1980-82, when excess capacity was large and the second oil

20. See World Bank, Korea: Development in a Global Context; and World Bank,Korea: Managing the Industrial Transition.


25/66

Rudiger Dornbusch and Yung Chul Park 413Table 13. Korean Saving and Investment, 1960-86Percent of GNP

Gross National savingdomestic ForeignPeriod investment Total Business Personal savinga1960-69 18.2 8.9 b b 9.11970-79 27.6 20.5 8.5c 10.5c 6.81979 36.0 26.5 8.8 11.2 8.91980 32.1 20.8 8.8 6.6 11.51981 30.3 20.5 8.2 6.7 9.81982 28.6 20.9 8.0 6.8 7.01983 29.9 25.3 10.5 7.6 4.71984 31.9 27.9 11.0 9.9 4.01985 31.1 28.6 11.1 10.6 3.11986 30.2 32.8 n.a. n.a. - 2.8

Source: EPB, Major Statistics of Korean Economy, various issues.n.a. Not available.a. Equals gross domestic investment minus total national saving. The identity may not be exact because thestatistical discrepancy is omitted.b. Average private saving for 1960-69 was 7.4 percent of GNP.c. Average for 1975-79.

shockhurt he chemical ndustry nparticular.Todayit is apparent hatmany of these industrieshave gainedin export share.The automobileindustry s a case in point. In any event, the investmentportfolio wassufficientlywellchosenthat realwages increasedandexports expandedenoughto payinterestandeven principalon theexternalresources thathelpedfinance he investment.Improvedtechnology has come with high investment levels. Tech-nologicaldevelopmentnaturallybeginswiththeimportation fadvancedforeign echnology,andproceedsthrough he developmentof domesticvariantsof this importedtechnology, and eventually to technologicalself-reliance. As successive five-year economic development plansunfolded, Korea came to recognize that technology was an essentialingredient n enabling ndustry to producefor the world market.Thegovernment husmadea majoreffort to digest, adopt,andadapt oreigntechnology.21One of the pillars of Korea's science and technology

21. See LarryE. Westphaland KwangSuk Kim, "Korean IndustrialCompetence:Where It Came From," in Bela Balassa, ed., Development Strategies in Semi-Industrial-ized Economies (Johns Hopkins University Press, 1982), pp. 212-79; and H. S. Choi,"Science and TechnologyPolicies for IndustrialDevelopment," n Industrialization ndDevelopmentStrategies KoreaDevelopment nstitute,1986).


26/66

414 Brookings Papers on Economic Activity, 2:1987developmentpolicy was refining heability o identifyappropriateoreigntechnology,properlyselect, and thenadapt t.The Budget and Financial Stability. The chief contrast betweenKorea and Latin Americalies no doubtin the budgetand in financialmarkets. In Korea budget and financial market policies helped keepinflation n check and avoided capital flight duringthe early 1980s, aperiod of highinflationandlargebudgetdeficitsworldwide.Despite theoil andcommoditysupplyshocksof thelate 1970sandthe realdeprecia-tion of 1979,inflationneverquite reached 30percent-certainly not anextremelevel, againby Latin American standards.The unifiedbudgetdeficit,althoughswingingwidely, never reached5 percent of GNP andnever stayed very highfor more thantwo years in a row. Figure6 showsthe behavior of the deficit from 1970to 1986, and figure 7 shows thesteady postwarexpansion ntaxation-thats one source of Korea's fiscalstability.Because of that stability, Korea has sufferedno majorbuildupofdomestic debt. Nor has the budget deficit at any point become largeenoughto necessitate rapidmoney creation.Thislatterpointis broughtoutby a simplemodel of money-financed udgetdeficits.22 upposethatthe budgetdeficit is a fractiong of real output and that it is entirelyfinancedby base moneycreation.Supposefurther hat the base moneyvelocity is a linear unctionof therateof inflation.Under these assump-tionswe canderiveasimplerelationbetween thedeficitratio,thegrowthrateof money,and the rateof inflation:23(1) F = g (a + ow),

where Vj is the growthrate of money, cx s a constant in the velocityequation,and0 represents he responsivenessof velocity to the rateofinflation,wT.Next we use the steady-staterelationshipbetweeninflation,moneygrowth,and thegrowthrateof real income:(2) s = > - ay,

22. This model follows Robert A. Mundell, Monetary Theory: Inflation, Interest andGrowth n the WorldEconomy(PacificPalisades,California:GoodyearPublishingCom-pany, 1971).23. Deficitfinance mpliesthatMIP= g Y,whereM is the nominalmoney expansionand Ythe evel ofoutput.Thiscanbe rewritten s pL(M/P)= g Yorp.= g Y/(M/P).Assumingmonetary equilibrium nd using the velocity equation Y/(M/P) = a + OI, we obtainequation1in the text.


27/66

Rudieer Dornbusch and Yung Chul Park 415Figure 6. Unified Budget Deficit, Korea, 1970-86Percent of GNP5.04.54.0-3.5-3.0-2.5-2.0-1.51.0

.?b 111975 1980 1985Source: Bank of Korea, Economic Planning Board, Major Statistics of Korean Economy, various issues.

where a is the income elasticityof money demandand y is the growthrate of real income. Combining hese two relations yields an equationfor the rate of inflation:(3) IT = (ag - cy)/(1 - Og).

Three points emerge from this simple equation. First, the budgetdeficit influencesthe inflationrate in a highly nonlinearfashion. Theaddedinflation rom an extra 1percentdeficit is greaterthe higherthedeficit.Second, thehigher he growthrateof real output,and the higherthe incomeelasticityof money demand,the lower the rate of inflation.Third,the interceptof the velocity equation,a, influencesthe inflationimpactof a given deficit.Other hingsequal,the availability f substitutesfor domesticmoney, such as dollar deposits, a possibilityof externalassetholdings,or financial iberalizationhatreducesbankdepositsandhence reserves,tends to raisevelocity and hence the inflation mpactofa givendeficitratio.Eachof thesepointsis relevantto a comparisonbetweenthe Koreanand Latin American economies. In Korea the deficit never reached


28/66

416 Brookings Papers on Economic Activity, 2:1987Figure 7. Taxes as a Percentage of GNP, Korea, 1953-86Percent of GNP2018-16 -

14

12108-641 1 1 I-1955 1960 1965 1970 1975 1980 1985Source: Bank of Korea, National Accounts, various issues.

exceptionally high levels and certainly not for long. Growth has beenhigh and financial iberalization,until recently, moderate.Dollarizationnever occurred. As a result the inflation rate never reached LatinAmericanranges, and because it did not, pressurefor financial iberali-zation (which nturn ncreases nflationunless there s deficitcorrection)was never strong.The revenue from base money creation, or seignorage,is well ex-plained n Korea by two determinants:nflationandgrowth.To explorethis relationwe rana regressionof the ratio of base money creation tonominalGDP on inflationand growth.Theresults, using ordinary eastsquareswithannualdatafor theperiod 1970-86,were as follows:

Seignorage= -1.38 + 0.23 Growth + 0.065 Inflation,GDP (-1.75) (3.87) (2.16)R2 = 0.45; Durbin-Watson 2.0,

with t-statisticsin parentheses.Over the sample periodon averagethegovernmentderived 1.35percentof GDPin revenue from base moneycreation. During he 1970s he revenue reached2.2 percentof GDPand


29/66

Rudiger Dornbusch and Yung ChuilPark 417thus financed the major part of the budget deficit. If seignorage isconsidered ust anothertax, the government inancedthe majorpart ofoutlays in one formor anotherby taxationratherthan by borrowing.The rising share of outright axes in GDPin the 1970sreduced budgetdeficitsandmade t possibleto shift to lower rates of inflation.Togetherwith the shift to lowerinflation he governmentalso startedfinancingdeficitsincreasinglyby domestic debt ratherthan by moneycreation.Thedomestic publicdebtratio,whichrangedbetween3 and 4percent of GDPin the 1970s, has startedrising moderately,although tis still far below 10 percent of GDP. With levels of real interest rates inexcess of 10percenta year, a largerdebtburdenwould easily becomeaproblemby itself, as has been the case in Brazil or in Mexico. In Koreathe small size of the debt makesthis issue unimportant.Just as Korea's domestic debt was small, so was its external publicsector debt(not, though,publiclyguaranteeddebt).24As a consequencetheworld nterestrateshocksof theearly 1980sandthe realdepreciationdid not have a direct impacton the budget via the debt channel. Bycontrast,LatinAmerica,where much fnotmost of thedebtwasdirectlyin the public sector, suffered a major budget deterioration, sharplyincreaseddeficitfinance,and inflation.Financial Repression. Financial repression helped finance budgetdeficits in a relatively noninflationaryway, as already mentioned, bykeeping money substitutes out of reach. But it supported financialstabilityand growth n other importantways.25With Korea's domesticfinancialmarket underdevelopedand market informationnot readilyavailable, the Korean government stepped in, exerting far-reachinginfluence hroughownershipof financial ntermediaries nd control ofaccess to foreigncapital.By controllingcapitaloutflowsand thus forestallingcapital flight nmomentsof economic and political uncertainty,such as 1980-81, the

24. By the IMF definition (line 89a.h of International Financial Statistics), the externalpublicdebt n 1985hadreached9.7 percentof GDP. TotalKoreanexternaldebt amountsto more han40 percentof GDP.25. See especially David C. Cole and Yung ChulPark, FinancialDevelopment nKorea, 1945-1978 HarvardUniversityPress, 1983);Park,"FinancialRepression,Lib-eralization, ndDevelopmentn DevelopingCountries" KoreaUniversity,1985); ndY.J. Cho and David Cole, "The Role of the Financial Sector in Korea's StructuralAdjustment"KoreaDevelopment nstitute,1986).


30/66

418 Brookings Papers on Economic Activity, 2:1987governmenthas avoidedthe extrareal depreciation equired o generatethe foreignexchange that financescapital light.Avoidingreal deprecia-tion is tantamount o avoidingmajor nflationary hocks,alesson learnedpainfullyin Mexico and Argentina,where capital flight forced majorexchange depreciation.The government'sfinancialrepressionhas also mobilizedresourcesfor investment n targetedareas.By paying depositors ow real interestratesandby controlling apitaloutflows,the government mplicitly axeddepositors,then channeledthe proceedsto favored sectors for invest-ment.AlthoughKorea s oftencitedasanexampleof successfulfinancialliberalizationwithhighandpositivereal interestratesmobilizing inan-cial resourcesfor investmentandgrowth,table 14 shows that realrateshave not been high, except in 1965-69, in the immediateaftermathoffinancial eform26 Moreover,even lapses into smallnegativereal nterestrates did not interferewith a steady increasein the ratio of M3 to GDP,as showninfigure8.Therelationship etweensavingand nterestratesremainsunresolvedin Korea,just as everywhere else. Some authorshave argued hat highreal interest ratescausedthe saving spurtandthatthe financialreformin 1965spurred he expansionin intermediationn the latter halfof the1960s.27Others, however, show that Korea's saving respondslittle tointerestrates.28Overall, heKoreanexperience suggeststhat there is noneed for high positive real interest rates to mobilizesaving through he

26. See A. LanyiandR. Saracoglu,"InterestRatePolicies n DevelopingCountries,"Occasional Paper 22 (InternationalMonetary Fund, October 1983);Vicente Galbis,"Financial ntermediation nd EconomicGrowthnLess-DevelopedCountries:A Theo-retical Approach," Journal of Development Studies, vol. 13 (January 1977), pp. 58-72;Ronald McKinnon, "FinancialRepressionand the LiberalisationProblemwithin LessDevelopedCountries," nSven Grassman ndErikLundberg, ds., TheWorldEconomicOrder:Past and Prospects (St. Martin'sPress, 1982);RonaldMcKinnonandDonaldJ.Mathieson,"How To Managea RepressedEconomy," PrincetonEssays inInternationalFinance, 145 (Princeton University, 1981); and Mathieson, "FinancialReform andStabilizationPolicy na DevelopingEconomy,"Journal ofDevelopment Economics, vol.7 (September1980),pp. 359-95.27. See John Gurley,Hugh Patrick,andEdwardShaw, "TheFinancialStructurenKorea"(StanfordUniversity, 1967);and RonaldI. McKinnon,ed., MoneyandFinancein Economic Growth and Development: Essays in Honor of Edward Shaw (New York:MarcelDekker, 1976).28. See, for example, Alberto Giovannini,"The InterestElasticityof Savings inDevelopingCountries:TheExistingEvidence,"WorldDevelopment,vol. 11(July 1983),


31/66

Rudiger Dornbusch and Yung Chul Park 419Table 14. Real Interest Rates, Korea, 1960-86Percent

Export sectorPeriod Curb marketa Depositsa loansb1960-64 31.1 - 6.7 n.a.1965-69 44.4 26.9 n.a.1970-74 28.2 -0.2 - 16.31975-79 24.0 -4.5 - 12.51980 16.3 -2.4 - 10.31981 14.0 3.8 -0.41982 23.4 4.2 4.71983 22.4 1.3 6.11984 22.5 5.3 6.21985 21.5 5.8 5.91986 20.8 7.7 7.7

Source: EPB, Major Statistics ofKorean Econo,tny, various issues; IMF, GovernmienitStatistics Yearbook, variousissues;andWorldBank,Korea, vol. 22 (Washington, .C.: WorldBank, 1987).n.a. Not available.a. Nominal nterestrate ess consumerprice nflation.b. Nominal nterestrate ess inflation f the GNPdeflator.

financial ystem; as long as largenegativereal interestratesare avoided,the real nterestrate is relatively nsignificant.FAVORABLE WORLD ECONOMIC ENVIRONMENTThe final elementin Korea's success was the worldeconomic envi-ronment.Korea's exposure to world economic influences differs littlefrom that of Taiwan, Singapore,or Hong Kong, all of which have the

same trade structure, importingoil and commodities and exportingmanufactures.But it differs significantly rom that of Latin Americancountries, uchas BrazilorArgentina, hatarenetcommodityexporters.That differencebecame crucially mportant n the late 1970sand early1980s,because real oil prices moved up andcommodity prices moveddown. As aresult,Brazilhada much arger erms-of-trade eteriorationthan did Korea. Opportunities n the Middle East for constructionprojects,of which Korea took significant dvantage,wereanadditionalpp. 601-07; S. vanWijnbergen,"Macro-economicEffects of Changes n BankInterestRates:SimulationResults orSouthKorea,"JournalofDevelopmnentconomics,vol. 18(August 1985), pp. 541-54.


32/66

420 Brookings Papers on Economic Activity, 2:1987Figure 8. Ratio of M3 to GDP, Korea, 1960-86Percent90

80 -

70 -

60-

50-

40-

30-

20 -

10 _

0 I I*1965 1970 1975 1980 1985Source: Economic Planning Board, Major Statistics of Korean Economy, various issues.

offset to higherreal oil pricesand an extrasource of foreignexchangerevenue.The late 1970supturnof inflation n Korea,as inLatinAmerica,wasin partdue to oil and to the necessary exchangedepreciation.But unlikeLatinAmerica, Koreamade a rapid iscaland externaladjustment venbeforethe 1982debtshockplayeditself out. Korea'sability to restrainwages may be as importanthere as the dampeningof externalshocks


33/66

Rudiger Dornbusch and Yung Chul Park 421through ts particular radestructure.Wagerestraintand the large shareof trade nGNP imply hat realexchangeratescouldbemovedatalowerinflation ost and with a moresignificantmacroeconomic mpact.29Anotherdifferencebetween Latin America and Korea may be thelatter'sresponse to interestrateshocks. Since muchof Korea's externaldebt is private,the debt shock affectedprimarily irmsratherthanthegovernmentbudget.As aconsequence,the riskof aninflationary udgetdeficitdid not arise, and the downwardpressureon real wages at thefirm evel was much stronger.Marketaccess, especiallyto the UnitedStates,has beenanimportantadvantage for Korean export-led growth. But other Asian NICs andLatin America have the same opportunity.One might think thatprox-imity to the largeJapanesemarketwould have been anadvantage.But,just as has the United States, Korea hasfoundthat marketsubstantiallyclosed.We now turnto the policy questions. Whatis a structuralsurplus,does Koreahave one, shouldit be corrected,andwhat is the best wayto do so?

Does Korea Have a Structural External Surplus?In 1986, Korea had its first currentaccount surplus, amounting osome 3 percent of GDP. Beforethat,as figure9 shows, the only time inthe pastquarter enturythatKorea had come even close to balancewasin 1976-77. Otherwise, consistently large external deficits were the

rule.30Has that patternnow beenreversed,andis a history of surplusesin the making?Certainly t appearsthat since the late 1970s there hasbeen a steadymove towardsurplus,withthe surplusactuallymaterial-izing in 1986.If thesurpluswereto provetransitory, herewouldbenojustificationfor policy changesto trimit. But if it provespersistent,theremaybe a29. See Frederick aspersen,"AdjustmentExperienceand GrowthProspectsof theSemi-Industrializedconomies,"WorkingPaper477 (WorldBank, 1981); ndJaspersen,"AdjustingoExternalShocks:TheNewlyIndustrialized evelopingEconomies n1974-76 and1979-81" WorldBank,1981), oranaccountingrameworkhatevaluatesexternalshocksandpolicy responses.30. We assume he 1987currentaccountsurplus o be 4 percentof GDP.


34/66

422 Brookings Papers on Economic Activity, 2:1987Figure 9. Korean External Balance, 1954-86aPercent of GDP

6.0

4.0

2.0 Nonfactor current accountb/

0.0 Current accountA ~I}

-2.0 l

-4.0 I

-6.0

- 8.0

-12.CL - I IJ1955 1960 1965 1970 1975 1980 1985Source: International Monetary Fund, International Financial Statistics, various issues.a. Computed s percentof GDPand as a three-year enteredmovingaverage.The 1987currentaccountsurplusis assumed o be 4 percentof GDP,and the 1987nonfactor urrentaccountsurplus s assumed o be 7 percentofGDP.b. The current ccount ess net factorpaymentsabroad.

policy issue. At least it is worthaskingwhatthe costs and benefits of along-termsurplus would be. And, to the extent that the surplusis theresultof deliberatepolicies,it isappropriateoaskwhether hesepoliciesare ustifiedon a cost-benefitanalysisof the surplus.Hencethe need fora closer look atwhat a structural urplus s andwhetherKoreahas one.STRUCTURAL SURPLUSESThere is no accepted definitionof a structural urplus n the externalbalance. But apragmatic newilldo. A structural urplus s one thatcan


35/66

Rudiger Dornbusch and Yung Chul Park 423Table 15. Long-Term Foreign Saving Patterns, Germany, Japan, United States, Korea,1950-86Percent of GNP

Country 1950_59a 1960-69 1970-79 1980-86bGermany 2.7 2.1 2.6 2.5Japan 0.1 0.2 0.8 1.9United States 0.2 0.2 -0.5 -1.8Korea - 8.7 - 10.1 -5.9 - 1.6

Source: Net exports n the national ncomeaccounts romIMF, nternational Finantcial Statistics, variousssues.a. Japan,1952-59;Korea, 1953-59.b. Japan,1980-85.

be expected to persistovera few years fworldmarketconditionsremainbroadlyunchangedand thehome economy does not experienceunusualshocks, such as earthquakes, evolution,orthe like.Table 15 shows the long-term current account patternsof Japan,Germany, and the United States. The Germanpattern meets mostobviously he definition fastructuralurplus Japan hows an ncreasingtendency owardsurplus,andthe United States,ashifttowardpersistentdeficits.The long-runbehaviorof the current account depends on nationalsaving and investment. Investment opportuinitiesn the world capitalmarket,public finance,and demography ogether determinewhetheracountryis a net lender or a net borrower.Although ong-termcurrentaccount patternscan be temporarily bscuredby a boomoranexternalshock, such interruptions rerelatively nsignificant.Therole of demog-raphy in explainingsaving and the current account remains almostunexplored,except for an interestingcontributionby George von Fur-stenberg.31 n emergingdemographic ife-cycle interpretation f Japanargues that the shifting age distribution mplies that at the turnof thecenturythe Japanesepopulationwill have a higheraverage propensityto spend.A nation's integrationwith the world capital market determineswhether a given domestic savingis captured or domestic investment,with a resulting endency towardcurrentaccount balance,or whether tis available for investment abroad,with a resultingtendency towardforeign endingandsurpluses.Integrationwiththeworldcapitalmarket

31. See George M. von Furstenberg, "Domestic Determinants of Net U.S. ForeignInvestment," International Monetary Fund Staff Papers, vol. 27 (December 1980), pp.637-78.


36/66

424 Brookings Papers on Economic Activity, 2:1987will tend to create structuraldeficitswhen investmentopportunitiesareample at home (say, because labor s abundant)andcan be financed nthe world capitalmarket.Absence of Ricardian quivalencealso has a bearingon the externalbalance. Long-termswings in the government budgetwill affect con-sumption spendingand investment and hence influence the nationalsaving-investmentbalance.A swingtowardbudgetdeficits, U.S. style,leads to external deficits, while a surplus policy leads to externalsurpluses.Thesamephenomenonmay applyat the level of firmsby wayof undistributed arnings hat are used forforeigndirect nvestment.Incountriessuchas Germany,where firmsretainearningsand use themtoinvest abroad, stockholdersmay not spend fully the capital gains and,as a result, the current account will show a surplus. Thus a foreigninvestmentmotive, in the absence of Ricardianequivalence, may wellengender persistent current account surpluses. The tendency will bestronger hemorefirmsrelyonundistributedarningsandthe moretheyrely on direct ownershipof external investments as a means to exploittheir monopolisticmarketpositions. In a sense, then, these currentaccount surplusesreflect mperfectionsof goods and assets markets.Thesaving-investmentnterpretation f the externalbalancecanalsobe appliedto the questionof how openness to tradeaffects the tradebalanceandhence the currentaccount. WouldJapan,with moreopenmarkets, have a smaller current account surplus? Not necessarily.Germanyis a relatively open economy whose current account hasconsistently shown a surplus. An opening by Japanwould certainlyincrease mportsand thusfreeresources,whichcouldbeused eitherforproductionof extraexportsorfor increaseddomesticabsorption.Howtheadjustmentwould occurdependsamongotherthingsonfiscalpolicyandpublicsector spendingreactions.If a tax on savingwere to financean expansion in public sector infrastructurenvestment, the surplusmightvanish. But if therewere no fiscalresponse, the long-runcurrentaccountmightnotchange much,and increased mportswouldbe offsetby higherexports.

THE KOREAN CASEDoes Korea have a structural urplus?Table 16 shows the Koreancurrentaccount, the current account excluding net factor payments


37/66

Rudiger Dornbusch and Yung Chul Park 425Table 16. The Korean External Balance, 1960-86Percentof GDPMeasure 1960-69 1970-79 1980-86 1986Currentaccount -9.1 -6.8 - 5.3 2.7Nonfactorcurrentaccount - 10.1 - 5.9 - 1.6 5.8Net factorpaymentsabroad - 1.0 0.9 3.7 3.1

Source:IMF,International Financial Statistics, various ssues.

abroad,andnetfactorpayments, each expressedas a fractionof GDP.Like figure9 above, table 16 shows a pattern of decliningdeficits anidincreasingsurpluses.The figure,especially, bringsout clearlyhow theoil shocks of 1974and 1979-80 temporarilyset back this patternofimprovement.But despite these shocks, the forces driving he currentaccounttowardsurpluswere strongenough to restorenear-balance n1977 and againin 1984-85. Three worldwideeconomic developmentshelpedpush the currentaccount ntosurplus n 1986.First, interestratesdeclined rom heirpeaksof 1982-83,andhence debtservicefell sharply.Second, real oil prices and real commodity prices declined after 1980.Third, the decline of the dollar, and Korea's decision to stay with thedollar, helped improve Korea's internationalcompetitiveness at theexpense of Japan. These three developments were largelyresponsiblefor Korea's 12percentgrowthand$4.6 billion currentaccount surplusin 1986.Ro estimates hatabsentthesethreebenefits, hecurrentaccountsurpluswould have beenonlyaround$0.3billion.32Theemergenceof the surpluscan also bedescribed ntermsof savingand nvestment.Thebudget mprovement ince 1982 ncreasesnationalsavingandhence leads to an improvementn the externalbalance thathasnot beenoffset by anincrease in investment.Since the highgrowthis temporary,beingdue to transitoryexternaladvantages,it will haveonly a minoreffect on consumption,which is linkedat least in parttopermanentncome. Withconsumptionresponding luggishly o currentreal ncome,personalsaving ncreases.There is a short-termstructural endency towardcurrentaccountimprovementnthat nvestment,whichwasconcentratedduring helate1970soncapitalgoods witha significantmportcontent,hassinceshiftedtowardprojectssuchasconstructionand nfrastructurehathave amuch

32. See S. T. Ro, "FavorableExternalConditions nd the KoreanEconomy n 1986"(KoreaDevelopment nstitute,1987).


38/66

426 Brookings Papers on Economic Activity, 2:1987lower importcontent. Butonly in a context of unusedresources wouldan expenditure-switching owardhome goods lead to a long-term n-crease in outputandnationalsaving.Withfull employment,the excessdemandfor resourcesin the home goods sector will bringabouta realappreciationthat leads to a reductionof exports and an increase inimports. To the extent that the currentaccountsurplusreflects at leastin part he shift ninvestment,the realappreciationprocess may alreadybe underway, and one significant omponentof the surplusmay thus bemerely transitory.But one reasontothink hatKoreamaybeheading owardastructuralsurplus is that Korean firms may take the view that potential tradeconflicts make it relatively unprofitable o locate all new capacity inKorea and mayinvest directlyin targetmarketssuch as Europeor theUnited States. As a result,domesticinvestmentwould declinewithoutan offsetting fall in saving, and the current account would show atendency towardsurplus.TheNationalPensionSystem thatwillgo intoeffect in 1988may be regardedas a forced savingdevice. Butits impacton nationalsaving s estimated o be small.33

We conclude that Korea is likely to have a structuralbalancein theexternalaccount,perhapsa smallsurplus.The size is difficult o judge,and some of the presentsurplus s likely to be transitory.Afterall, thenational saving rate in 1986reachedthe highestlevel ever, andeven ifthetrend nsaving s upward,partof thatsaving s surelytransitory.Thenextquestion s whether hestructuralurpluscallsfor correctivepolicy.Inconsidering hatissue we emphasizethat one must look at theoverallcurrentaccount, not at bilateraltrade balances, the aggregatetradebalance,or even the currentaccountexcluding actorpayments.

Policy Responsesto the StructuralSurplusThere are two opposingschools of thoughtaboutexternal balancesof developingdebtorcountries.One, thatexport-ledgrowthand reduc-tionsindebt burdensare desirablestrategies,hasbeenargued orcefully

by Corboand Nam.34The other, that East Asian NICs' currentaccount33. See J. S. Min, "The Master National Plan and its Socioeconomic Effects,"ResearchMonograph 605 KoreaDevelopment nstitute, 1986).34. See CorboandNam, "Korea'sMacroeconomicProspects."


39/66

Rudiger Dornbusch and Yung Chul Park 427surplusesare a provocationto a liberalworldtradingsystem and mustbe reduced, was espoused as the lowest common denominatorat thelatest economicsummitandhas becomea mainpoint nU.S. commercialdiplomacy.The question whether surpluscountries should adjust is difficult.Policymakerswho look back on recent balance of paymentsproblemsrightlyview an improvement n the noninterestcurrentaccount as anachievement hat should notbe readilygiven up. Theyexpress concernthat the surplusesmerely representcushions againstexternal shocksand that it would be frivolous to sacrificethe protectionthey afford.Moreover,they pointout thatit is mucheasier to give away surplusesby appreciation r wage increasesthanto generate hemby restraintanddepreciation.Makingsurpluses,when this becomes essential becauseof terms-of-trade hocksor creditrationingnworldmarkets, nvariablyinvolves inflation and recession such as Korea experienced in 1980.Hencethetendencyto hangon to surpluses.But there are two other sides to the argument.First, on stricteconomics, consumption s the ultimateobjective:policies that favorgrowth at the expense of currentconsumptioncannot get high marksforever.There s littleargumentor open-endedsurpluses nthe style ofJapan or Germany.The only exception is the case where net foreignlendingreflectsthe transitorydemographiceffects of life-cycle savingwhen theage structureof thepopulation s changing.Second, firmsandpoliticians n the exportmarketswhere the tradesuccesses are scoredoppose the invasion and ask at least for full reciprocity, meaninginparticularmport iberalization.

BelaBalassaandJohn WilliamsoncontendthatKorea should elimi-nateits currentaccount surplus:We haveargued hatacontinuing urplus s undesirable tKorea'spresentstageof development. t is unnecessary n termsof providingKoreawith anadequatesafety margin gainst oreignshocks.... A continuing urpluswould create anunnecessary hoice betweenlimiting nvestment,andtherebycurtailinguturegrowth,and continuing o holdconsumption-and thereforerealwages-at anunnecessaryow level.35The sameview has beenexpressedrecently n WorldFinancialMarkets:Chronic,excessive surplusesserve no productivenationalpurpose. On thecontrary, hey needlesslypostponeimprovementof domesticlivingstandards

35. See Balassa and Williamson, Adjusting to Success, pp. 70-71.


40/66

428 Brookings Papers on Economic Activity, 2:1987and, by aggravatingnternationalrade rictions, eopardizecontinuedaccess toforeign markets. Specifically,[Taiwanand Korea]would be well advised-intheir own self-interest-to adoptfiscalpolicies thatexpanddomesticdemand;allow significantcurrencyappreciation o boost purchasingpower and actualliving standards;move rapidly oward mportregimeswith minimal estrictionsanduniformly ow tariffsof 10% rless; liftexchangecontrols hatblockprivate-sector lending and investmentoverseas; and fireetheirbanking systems andcapitalmarkets rom crippling egulation.36

These policy suggestions are a by-product of a search for a solutionto the U.S. trade problem, given the unwillingness of Congress and thePresident to agree on U.S. fiscal policy and the unwillingness of Europeand Japan to expand. Only a few years ago Korea was urged in thedirection of policy adjustments thatwould enable it to service the externaldebt and to be dropped from the list of problem debtors. One way tointerpret the new attitude is that Korea overshot the target.Two questions need to be asked at this stage. One is whether anadjustment in policies should be pursued to trim the external surplus;the other, what particularpolicy should best be used.

THE DOUBTFUL CASE FOR UNILATERAL SURPLUS REDUCTIONThe strongest argument in favor of surplus reduction is political. The

United States is experiencing extraordinary trade deficits. Since 1980the U.S. manufacturing trade balance with developing countries hasshifted by more than $50 billion. Partof these deficits has as a counterpartthe surpluses of East Asian NICs, including Korea, as table 17 shows.The surpluses may be the outcome of adjustments to the debt crisis orof superior tradeperformance. But one way or another they area politicalproblem. To avoid costly U.S. trade restrictions Korea should takemeasures to cut down the bilateral surplus.A more sophisticated version of this argument would emphasize thatKorea, by staying with the dollar rather than the yen, undermines U.S.exchange rate adjustment. As Japan becomes less competitive in thecourse of yen appreciation, Korea picks up the business without muchof an improvement in the U.S. external balance. Again, for Korea to

36. See MorganGuarantyTrustCompany,WorldFinancialMarkets January1987),p. 11.


41/66

Rudiger Dornbusch and Yung Chul Park 429Table 17. U.S. Bilateral Trade Balance and Exchange Rates, 1981-86Billions of dollars except as notedItem 1981 1982 1983 1984 1985 1986U.S. balancewith East AsianNICsa -7.0 -8.2 -12.6 -21.4 -25.0 -30.8U.S. balancewith Japan - 18.1 - 19.0 -21.7 - 36.8 - 49.7 - 58.6U.S. balancewith Korea -0.4 -0.5 - 1.7 -4.0 -4.8 -7.1Exchange rates (index, 1981= 10O)bWon-dollar 100 107 114 118 128 130Yen-dollar 100 113 108 108 108 76Taiwandollar-dollar 100 106 109 107 107 103

Sources: U.S. Department of Commerce, Bureau of Economic Analysis, Survey of Current Blisitness, variousissues; and Morgan Guaranty Trust Co.a. Hong Kong, Singapore, Taiwan, and Korea.b. The exchange rates for August 1987 are as follows: Korea, 119; Japan, 64; Taiwan, 82.

avoidtraderestrictions, hereneeds to be animprovementn the bilateraltradebalance.These politicalargumentshave obvious force. Butthey makethe casefor bilateraladjustmentsratherthan broad-basedpolicy moves. If theU.S. interestfocuses on the bilateraltradebalance, thenit is not clearthatbroad,overallbalanceofpaymentsadjustmentpoliciesarethe mosteffective way to forestall U.S. policy action. On the contrary, policiestargetedspecifically to increasing mportsfrom the United States andreducing exports to the United States would be more effective. Forexample,special ncentivesmightbe used to shift mports romJapanesesuppliers o U.S. sources, especiallyin industrieswhere U.S. manufac-turerswouldbe most appreciativeandhence politicallymost supportive.A more importantargumentagainst initiatingmajor policy adjust-ments at this stagedraws attentionto U.S. budgetbalancing.Overthenext few years the United States will undoubtedly sharply reducedomestic demand and the external deficit. It is therefore essential toconsider Korea's adjustment n the context of the realignmentof theworldeconomy attendantupon U.S. tradeandbudgetadjustment.The direct expenditure effects of U.S. budget correction may bythemselves eliminate a good part of Korea's surplus. If Korea, byexpansionarypolicies or real appreciation,eliminates its surplus overthe nexttwo years,beforeU.S. budgetcuttingactuallygets underway,the U.S. cuts would drive Korea's current account into deficit andrequirea correctivereal depreciation.If fiscal expansionhad been usedto eliminate he surplus, fiscal contractionwould be necessary just as


42/66

430 Brookings Papers on Economic Activity, 2:1987export industrieswere losing their markets.Clearlythe Koreanadjust-mentquestioncannotbe solved independentlyof the U.S. expenditurecuts. The importanceof the U.S. marketfor Koreanexports (andthefact thatJapaneseandEuropeanmarketsaresubstantially losed)makesKorean adjustments dependent on U.S. policies and timing. As acomponentsupplier o other countrieswho in turnexportto the UnitedStates, Korea also shares in their losses of exports in the course of aU.S. budgetcut. Moreover,reducedactivityin thirdcountrieswill alsoadverselyaffect Korea's exportsthere.It is difficult o quantify he impactof U.S. budget cuttingon Korea.Total Koreanexports in 1986were $34.7 billion, of which $9.5 billionwent directly to the United States. If indirect exports to the UnitedStatesamount o another$5.5 billion, a 7 percentreduction nU.S. totalimports, across the board,would reduceKoreanexports by $1 billion,not counting the reduced levels of demandin thirdcountries. Clearlythis expenditure shock would not eliminate the present surplus, butdisturbances nvolving interest rates and oil could combine with theexpenditureshock to eliminatemost of it. The possible combinationofadollarcollapse andattendantncreases ninterestrates, realcommodityprices, and oil prices implies a major shock to the Korean externalbalance thatwould certainlybe unwise to reinforceby a prematureanddifficult-to-reverse issipationof the externalposition.

How to Adjust?If, the above argumentsnotwithstanding,Korea seeks some overalladjustment n the balance of payments, how might it best make theadjustment?Twoapproachesarepossible:expenditurencreasesversusexpenditureswitching, on one hand,andpolicies towardcapitalflowsandcapitalmarkets,on the other hand.CAPITAL MOBILITY

Although,as table 18 shows, nonbank Korean residents at presenthave almost no external depositholdings, liberalizingprivateportfoliocapital outflows is the least desirableway to reduce the surplus.Theexperience of Latin America has demonstratedthat capital mobility


43/66

Rudiger Dornbusch and Yung Chul Park 431Table 18. Cross-Border Bank Deposits of Nonbanks by Residence of Depositor,Various Years, 1981-86Billions of U.S. dollars except as noted

Dollars percapitaCountty 1981 1983 1985 1986 (1986)Argentina 6.4 7.9 8.5 8.5 274Brazil 3.5 8.1 9.8 11.7 86Mexico 9.4 12.7 16.1 15.8 199Venezuela 15.6 10.9 14.0 12.8 720Korea 0.3 0.3 0.7 0.7 17Italy 12.0 10.6 10.9 11.3 189France 12.7 11.8 11.0 13.5 244Japan 1.9 2.0 4.4 7.3 60

Source: IMF, International Financial Statistics, various issues. End-of-year figures for each year.

readily underminesexchange stability and emphasizes finance at theexpense of productive activity. Of course, these developments do notoccur independentlyof poor budget performanceand adverse shocks.But the case for the social benefits of hot money mobility remainsdoubtful.37Moreover, iberalizing apitaloutflowswould not see

1987 dornbusch park - korean growth policy

Documents