Responding to the East Asian Crisis - World...

IN THE SECOND HALF OF 1997 SEVERAL EAST ASIAN CRISIS COUNTRIES

experienced a massive reversal of the large foreign private capitalinflows they had enjoyed through much of the 1990s. The net swingfrom inflows to outflows between 1996 and 1997 amounted to more

than $100 billion for the five crisis countries—Indonesia, the Republic ofKorea, Malaysia, the Philippines, and Thailand—or 11 percent of theirgross domestic product (GDP) before the crisis. The reversal precipitatedsteep devaluations of currencies, large increases in interest rates, andsevere declines in stock and other asset prices, initiating the deep financialand economic crises that have gripped these countries since. Contagioneffects from the crisis spread throughout the developing world. Butinstead of dying away quickly, as they did after the 1994 Mexico pesocrisis, they were the precursor of currency and financial crises in theRussian Federation in August 1998, followed by a more general with-drawal of private capital from emerging markets.

55

.

2Responding to the

East Asian Crisis

G L O B A L E C O N O M I C P R O S P E C T S

While financial crises in emerging mar-kets have many common elements—capitaloutflows, falling currencies, failing banks,and loss of confidence—their specificcauses and, in particular, the factors precip-itating loss of confidence, often differ andmerit different policy responses. This chap-ter tries to advance our understanding ofcrisis response by focusing on the featuresof the East Asian countries that—with ben-efit of hindsight—made them so vulnerableto the outbreak of financial crisis, as well asthe reasons for the emergence of these vul-nerabilities. Different views on the causesof the crisis lead to different policy pre-

scriptions for dealing with it. The chapterthen analyzes the evolution and propaga-tion of the crisis since its outbreak, dealingespecially with reasons for its severity andduration, and the role of the policyresponses implemented by governments tostem, manage, and resolve it. The discus-sion of policy responses reviews macroeco-nomic policies and financial and corporatesector restructuring and reforms, includingefforts to work out the foreign debt diffi-culties of the private sector. The last sectionconsiders the social impact of the crisis andpolicy measures to mitigate these effects.

This chapter’s key messages:• The crises in several East Asian coun-

tries highlighted the extent to whichtheir integration in global financial mar-

kets had outpaced the building ofdomestic institutions necessary to super-vise and regulate the financial sector.The interaction of these institutionalweaknesses with international capitalmarket imperfections, and the use ofinconsistent macroeconomic policies tomanage surging capital inflows, gener-ated crucial vulnerabilities that laid thegroundwork for the subsequent finan-cial crises—and ensured that their con-sequences would be severe.

• The critical immediate vulnerabilitycame from an excessive buildup ofshort-term foreign currency debt on thebalance sheets of private agents. Thisdebt made countries vulnerable to sud-den swings in international capitalmarket sentiment. Macroeconomicpolicies to manage large-scale privatecapital inflows tended to create upwardpressure on local interest rates. Largelocal interest rate premiums overfalling international rates encouragedunhedged short-term foreign currencyborrowing. In the euphoric business cli-mate of the time, many market partici-pants ignored the risk of exchange ratedepreciation, lulled as well by the sta-bility of exchange rates in the region,resulting from policies of pegging cur-rencies. Surging capital inflows andweaknesses in financial regulation andsupervision in the wake of financial lib-eralization in the 1980s and early1990s also contributed to booms indomestic lending. These credit boomsaugmented already high levels of cor-porate leveraging, fostered speculative,low-quality investments, and weigheddown banks’ portfolios with doubtful

56

Financial crises in emerging marketshave many common elements—capitaloutflows, falling currencies, failingbanks, and loss of confidence—but their specific causes often differ.

R E S P O N D I N G T O T H E E A S T A S I A N C R I S I S

loans collateralized on assets whosevalue had been inflated by asset pricebubbles. Banks and corporationsbecame highly vulnerable to shocksaffecting their cash flow and net worth.

• An ailing financial sector and largeincreases in central bank credit to fail-ing banks helped trigger the run on theThai baht and the collapse of the peg.The crisis quickly spread to other coun-tries in the region, in part because ofcommon vulnerabilities (such as highshort-term debt and financial sectorweaknesses) and spillovers throughinternational trade linkages, but alsothrough the contagion effects of a sud-den change in capital market senti-ment. Real activity began a sharpdecline as private investment suffered amassive shock because of increaseduncertainty, withdrawal of externalfinancing, and the impact of interestrate increases and currency devalua-tions on corporate and bank cash flowand balance sheets. Personal consump-tion, especially of durables, also fellsharply. A strong export response,which had helped recovery in Mexicoafter the December 1994 peso crisis,failed to materialize because of theweakness in overseas market demanddue to the regionwide downturn, sharpdeclines in export prices, and the creditproblems facing firms.

• Given the large declines in privateinvestment and consumption, the initialfiscal policy responses to the crisisturned out, contrary to design, to becontractionary and would have beenstrongly contractionary if fully imple-mented. As the severity of recessions

became apparent, fiscal policies wererelaxed in favor of a more stimulativestance. The use of monetary policy alsothrew up extremely difficult tradeoffsbetween macroeconomic and financialsector stabilization objectives. Ortho-dox monetary policies in defense ofcurrencies are appropriate in many cir-cumstances, but not necessarily in all.Some initial policy responses in the cri-sis countries stressed raising interestrates to stabilize exchange rates (thuslimiting the damage to the balancesheets and cash flow of firms withheavy foreign currency debts). Buthigher interest rates were themselveslikely to damage economywide balancesheets and cash flows, and so weakenthe real economy. Also, by increasingthe risk of default on instrumentsissued by financially weak banks andcorporations, higher rates could under-mine rather than strengthen currencies.In the event, none of the initial policyresponses had much immediate effectin stemming pressures on currencies—much of the decline occurred afterthese measures were taken.

• The primary role of fiscal and mone-tary policy now is to alleviate the col-lapse in aggregate demand, expand thesocial safety net, and recapitalize thefinancial system in a noninflationarymanner. Financial support from theinternational community is vital. Theinitial reform packages in some coun-tries were notable for their focus onstructural reforms. Some of these (suchas those dealing with the financial sec-tor or corporate governance issues)addressed significant causes of the cri-

57


sis; others, however, while importantfor medium-term progress, have raisedsome questions about priorities inshort-term crisis management.

• By mid-1998 large parts of the financialand corporate sectors in the mostaffected East Asian countries were insol-vent or suffering severe financial dis-tress. In several countries the cost ofrecapitalizing banking systems isexpected to rise to 20–30 percent ofGDP or more. Cross-country experiencesuggests that restructuring on this scalewill require government interventionwithin a comprehensive plan for thefinancial sector, including the injectionof substantial public funds. To reduceincentives for excessive risk taking(moral hazard), restructuring shouldallocate a substantial share of losses tobank shareholders, managers, and oth-ers who benefited the most from pastrisk taking. But these longer term goalswill need to be balanced against theimmediate priority of not exacerbatingthe credit difficulties facing viable firms.

• The success of bank restructuring andof restructuring debts of local corpora-tions is intimately linked. Orderlyworkouts—less formal ways to achievethe same economic objectives as bank-ruptcy proceedings, by bringing credi-tors and debtors together for voluntarynegotiation—will be important forboth domestic and foreign debt.Although this is a difficult and pro-tracted process, experience suggeststhat strong government leadership canplay a critical role. OECD govern-ments, in particular, can facilitatetimely workouts between debtors and

external private creditors, for example,by not holding out the possibility ofmore favorable bailouts for creditors inthe future. Expanded flows of foreigndirect and equity investment can alsocontribute to successful financial andcorporate restructuring.

• The crisis has had an enormous socialcost, placing a huge burden on thepoor and, in some countries, exacer-bating social conflict. Many of theseconsequences are likely to be pro-tracted. Social policy concerns thusneed to be an integral part of policyresponses to the crisis. While not asubstitute for sound, pro-growth,macroeconomic policies, social safetynets can play a major role in mitigatingthe social effects of crises. In recentdecades East Asian countries havereduced poverty and improved livingstandards at a pace unrivaled in his-tory. Nevertheless, cross-countryresearch suggests that protracted criseslead to more poverty, greater incomeinequality, and deteriorating healthindicators such as infant malnutri-tion—trends that can have enduringeffects on people’s health and theirability to participate in the economy.An important lesson from this crisis istherefore the importance of establish-ing appropriate ex-ante social safetynets in all countries, prior to a crisis.

• Unemployment in Indonesia, theRepublic of Korea, and Thailand isexpected to more than triple between1996 and 1998. Real wages are likelyto fall dramatically in Indonesia. Con-servative estimates put the number ofpeople falling into poverty in 1998 at

58


25 million in Indonesia and Thailandalone, and the number could be muchlarger if income inequality rises. Prior-ity actions to protect the poor includeensuring food supplies through directtransfers and subsidies, generatingincome through cash transfers andpublic works, preserving the humancapital of the poor through basic healthcare and education services, andincreasing training and job search assis-tance for the unemployed.

Buildup of vulnerabilitiesleading to crisis

What the crisis was notUnlike the debt crisis of the 1980s, the cri-sis in East Asia was not driven by severemacroeconomic imbalances or instability,particularly those originating in large,money-financed public sector fiscal deficits.Fiscal positions in the East Asian crisiscountries were either balanced or in surplusthroughout the 1990s1 (table 2-1). Public

sector debt was also generally low andfalling as a share of gross domestic product(GDP) in the 1990s. Despite large privatecapital inflows and rapid output growth,there was little evidence of economic over-heating in the form of inflation, whichremained moderate, ranging from around 4percent in Malaysia to 9 percent or so inIndonesia and the Philippines. In severalcountries inflation was actually falling inthe year to mid-1997.

The main expressions of excessdemand were large and widening currentaccount deficits in Thailand, Malaysia, andKorea, and somewhat smaller ones inIndonesia and the Philippines. By and large,these were not accompanied by a deteriora-tion in countries’ ability to service their for-eign debts from export revenues. A com-parison of debt-service ratios (amortizationand interest on long-term and short-termexternal debt expressed as a percentage ofexport revenues) at the start of the debt cri-sis of the 1980s and on the eve of the latestcrisis, shows that for most East Asian coun-tries debt-service ratios were relatively low

59

The crisis countries were in fiscal surplus for most of the 1990sTable 2-1 Central government fiscal balances, 1977–96(percentage of GDP)

1977–81 1982–86 1987–91 1992–96 1996

Indonesia –2.4 –1.4 –1.0 0.9 1.2Korea, Rep. of –2.1 –1.3 0.0 0.2 0.5Malaysia –7.9 –9.7 –3.9 0.7 0.7Philippines –1.8 –3.1 –2.6 –0.1 0.3Thailand –3.7 –4.6 2.1 2.4 2.3Argentina –3.9 –4.7 –1.1 –0.7 –1.8Brazil –1.6 –7.3 –10.4 –6.6a ..Chile 2.3 –2.0 1.3 2.2 2.3Colombia 0.2 –2.1 –1.2 –2.6 –4.3Mexico –3.7 –9.6 –6.4 0.2b ..

a. 1992–94.b. 1992–95... implies data is not available.Source: International Monetary Fund.


in 1982 and had fallen even further by1996 (figure 2-1). Only in Indonesia did theratio rise, although even there it remainedbelow the 1982 levels in many Latin Amer-ican countries.

The buildup of short-termforeign debtThe chief external source of vulnerabilityarose not from major macroeconomicimbalances but from a rapid buildup fromthe late 1980s onward of risky forms ofleverage on the balance sheets of financialinstitutions and nonfinancial corpora-tions, in particular short-term foreign cur-rency debt in excess of foreign currencyresources available on short notice. Mis-matches between the currency and matu-rity of liabilities and assets made firmsvulnerable to sudden swings in interna-tional investors’ confidence and to the

possibility of being unable to borrow frominternational capital markets to roll overshort-term debt or meet other current debtservice obligations.2

The ratio of short-term debt to foreignreserves is a rough measure of a country’sability to meet its current obligations fromits own liquid resources.3 These increasedsharply in 1994–97 in most of the crisiscountries (figure 2-2). In the three worstaffected countries—Indonesia, Korea, andThailand—short-term debt to reservesratios had risen to well over 100 percent bymid-1997. Malaysia and the Philippines,with short-term exposures less than 100percent, avoided the need for emergencyfinancial support packages from multilat-eral institutions.4 High short-term debtratios in these countries were also associ-ated with a broader measure of vulnerabil-ity, the ratio of M2 money to reserves (fig-ure 2-3). This ratio indicates the potentialfor a run on the foreign exchange reservesof a country with a fixed exchange rateregime by its own residents when confi-dence in the local currency fails. (The figurealso shows that the Russian currency crisisin the second half of 1998 was foreshad-owed by very high levels of short-term for-eign debt.)

The buildup of foreign liabilities by pri-vate agents took different forms. In Thai-land banks and finance companies were aprincipal conduit of external loans. Theirnet foreign liabilities rose from 6 percent ofdomestic deposit liabilities (M2) in 1990 toone-third by 1996 (table 2-2). Financialinstitutions’ net foreign liabilities also rosesignificantly in Korea. In Indonesia, how-ever, direct foreign borrowing by nonfinan-cial corporations was more prominent.

60

Most East Asian countrieslooked in fine shape toservice debt in 1996…Figure 2-1 Debt service ratios,1982 and 1996

Source: World Bank 1998a.

■

■

■■

■■

■■

■

■

■■

0

30

60

90

0 30 60 90

Indonesia

China

India

BrazilArgentina

MexicoChile

Turkey

PhilippinesThailand

Malaysia

Hungary

1996 (percent)

1982 (percent)


Several factors furthered the emergenceof these vulnerabilities in the 1990s. Macro-economic policies adopted to manage thelarge-scale private capital inflows to theregion tended to emphasize monetary policyas a way of sterilizing inflows. This createdupward pressure on local interest rates at atime when international rates were falling inresponse to modest growth, low inflation,and accommodative monetary policies inindustrial countries (figure 2-4). The largerate differentials created incentives forunhedged foreign currency borrowing, espe-cially at short maturities, which carried thelowest rates. In the euphoric financial cli-mate of the time, market participantsignored or discounted the associated risk ofexchange rate depreciation. These exces-sively risky financial strategies were alsofostered by the exchange rate stability inmany countries that had pegged currenciesto the dollar or to baskets of currencies witha high dollar weight.

High interest-rate differentials and lowvariability in exchange rates may have been

61

…but short-term debt was high…Figure 2-2 Ratio of short-term debt to foreign reserves, 1994 and 1997

Rep

. of K

orea

Indo

nesi

a

Thai

land

Phi

lippi

nes

Mal

aysi

a

Arg

entin

a

Mex

ico

Bra

zil

Col

ombi

a

Per

u

Eas

t Asi

a an

d P

acifi

c (5

)

Latin

Am

eric

a an

d

the

Car

ibbe

an (5

)

Percent

East Asia average Latin America averageJune 1994: 100 June 1994: 93June 1997: 134 June 1997: 87

0

50

100

150

200

250

June 1994 June 1997

Source: Bank for International Settlements, International Monetary Fund.

…and vulnerabilityincreased…Figure 2-3 Vulnerability indicators,June 1997

■

■

■

■

■

■

■

■

■

■

■

■

0

100

200

300

400

500

600

700

0 50 100 150 200 250

Ratio of short-term debt to foreign reserves (percent)

Ratio of M2 to foreign reserves (percent)

Colombia

BrazilMalaysia

Chile

Peru

RussianFederation

Rep. of Korea

Indonesia

ThailandPhilippines

Mexico

Argentina

Source: Bank for International Settlements, Interna-tional Monetary Fund.


especially important to the short-term debtaccumulation in Indonesia and Thailand(table 2-3). Obstfeldt (1998) notes that simi-lar conditions fostered short-term foreigncurrency over-borrowing by Mexican finan-cial institutions in the early 1990s. When thepeso crisis struck at the end of 1994, a pri-vate sector financial crisis emerged alongsidethe problems arising from the government’sown dollar-linked foreign borrowing. Diaz-Alejandro (1985) describes a similar processbefore the Chilean financial crisis of theearly 1980s. The buildup of short-term debt

in East Asia occurred during several wavesof international capital market enthusiasmfor emerging market debt in the 1990s. Themost recent wave, from late 1995 well into1997, was marked by an unprecedented fallin spreads on such debt, which, several stud-ies argue, was unjustified by observable eco-nomic trends and represented a significantunderpricing of risk.5

Finally, the liberalization of domesticfinancial systems and external capitalaccounts that took place in the late 1980sand in the 1990s occurred without an ade-

62

…while international rates fellFigure 2-4 G-3 interest rates and yield curve, 1988–98

Sep

t. 19

88

May

198

9

Jan.

199

0

Sep

t. 19

90

May

199

1

Jan.

199

2

Sep

t. 19

92

May

199

3

Jan.

199

4

Sep

t. 19

94

May

199

5

Jan.

199

6

Sep

t. 19

96

May

199

7

Jan.

199

8

-4

-2

0

2

4

6

8

10Percent

Government bond yield

Six-month LIBOR

Yield curve (LIBOR-bond)

Note: Averages of United States, Japan, and Germany.Source: World Bank.

Banks were the biggest borrowers in Korea and ThailandTable 2-2 Foreign exposure of banks and finance companies(percent)

Ratio of foreign liabilities to M2 Ratio of foreign liabilities to assets

1990 1994 1996 1990 1992–96 1996

Indonesia 1.2 7.0 3.2 108 193 143Korea, Rep. of 4.4 8.3 14.1 140 149 174Thailand 6.1 25.1 32.8 265 519 775Argentina 33.7 10.1 9.5 313 197 158Brazil 20.6 10.0 17.3 207 177 282Mexico 55.3 66.8 44.7 901 750 498

Source: IMF International Financial Statistics.


quate strengthening of prudential regula-tion and supervision, facilitating excessiverisk taking by financial institutions on boththe liability and asset sides of their balancesheets. Several countries created offshorefinancial markets with tax and regulatoryadvantages aimed at fostering the develop-ment of regional financial centers. Thesebecame channels for so-called out-in trans-actions—that is, external bank funding forlocal firms. In Thailand foreign bank loansthrough the Bangkok International BankingFacility soared from $8 billion in 1993, thefirst year of its operation, to $50 billion in1996, $30 billion of it out-in transactionsand $20 billion out-out transactions(Kawai 1997). Financial liberalization isalso likely to have contributed to thebuildup of short-term debt relative to otherexternal financing. Some East Asian coun-

tries had welcomed or accepted long-termforeign capital in the form of foreign directinvestment (FDI) or long-term debt forsome time. Liberalization therefore tendedto focus on removing barriers to short-termflows. Others, like Korea, maintained con-trols on long-term flows like FDI, while lib-eralizing short-term ones.

Private sector debt, corporatevulnerability, and financialfragilityThe buildup of short-term foreign debt wasonly one element, though the most impor-tant, in a wider increase in corporate sectorvulnerability and financial sector fragilityin many East Asian countries. Strong eco-nomic growth, buoyant domestic savings,booming private capital inflows, and lowerreserve requirements resulting from finan-cial liberalization fostered surges in domes-tic lending in the 1990s, especially in Thai-land, Malaysia, and the Philippines (table2-4). In Indonesia the boom in credit to theprivate sector occurred with financial liber-alization in the second half of the 1980s,and bad debt problems had already sur-faced in the early 1990s (Caprio andKlingebiel 1996a).6 Marked credit boomsalso preceded the outbreak of financial cri-sis in Chile in 1982 and, to a lesser extent,in Mexico in 1994. Although financialdepth rises systematically with per capitaincome on a cross-country basis, in severalEast Asian countries the amount of privatecredit relative to GDP had risen by 1996 tolevels well above those suggested byincomes alone (figure 2-5). By contrast,credit levels in major Latin American coun-tries by this time were close to or belowthose suggested by their income levels.

63

Incentives for unhedged foreignborrowing rose in the 1990sTable 2-3 Macroeconomic conditions related tounhedged foreign currency borrowing in EastAsia, January 1991–June 1997

Average Interest annual Exchange

rate appreciationb rate Country spreada (+) vs. US$ volatilityc

Indonesia 11.5 –3.8 0.7Korea, Rep. of 4.1 –3.2 3.4Malaysia 1.6 1.2 2.6Philippines 6.5 0.9 3.8Thailand 4.0 –0.3 1.2

Memorandum itemsGermany 1.2 –2.0 6.4Japan –2.2 2.5 10.7

a. Local deposit rate less LIBOR (US$) for East Asiancountries. Local LIBOR less LIBOR (US$) for Japan and Ger-many. Interest rate spread in percentage points.b. Relative to the U.S. dollar (in percent); a minus signindicates depreciation.c. Standard deviation of percentage deviation of exchange rate(US$) from regression on a time trend.Source: International Monetary Fund and World Bank dataand estimates.


High levels of bank credit werereflected in high leverage, or debt to equityratios, in East Asian corporate sectors.Highly leveraged firms are especially vul-nerable to sharp fluctuations in cash flowand net worth as a result of external shocksor macroeconomic volatility, a fact thatgoes a long way to explaining the severeimpact of the crisis on real output andgrowth. Over the past decade, the debt ofnonfinancial corporations was two to three

times higher than equity in the most seri-ously affected countries (Korea, Indonesia,Thailand), and that leverage was generallyrising in 1995–96, the runup to the crisis(figure 2-6). By 1996 the median value offoreign debt alone ranged from 70 percentof firms’ equity in Indonesia and 80 percentin Thailand to 150 percent in Korea. Evenmore striking, short-term foreign debtranged from about 40 percent of equity inIndonesia to nearly 100 percent in Korea.7

64

There were marked lending booms in several crisis countriesin the 1990sTable 2-4 Credit to private sector, selected years, 1975–96(percentage of GDP)

Country 1975 1982 1990 1994 1995 1996

Indonesia 20 14 47 52 54 56Korea, Rep. of 42 55 65 69 69 75Malaysia 33 47 71 115 130 142Philippines 32 46 22 36 45 54Thailand 28 46 83 128 139 100Argentina 16 34 16 18 18 19Brazil 55 44 38 51 35 31Chile 9 84 47 51 53 55Mexico 27 16 21 47 36 22Venezuela 34 55 25 13 12 10

Source: World Bank 1998c.

Private credit could not be justified by East Asian income levelsFigure 2-5 Credit to private sector and per capita GDP, 1996

Note: Logarithmic scales.Source: World Development Indicators 1998, World Bank.

6

7

8

9

10

11

0 1 2 3 4 5 6

Credit to private sector (percent GDP)

BrazilMalaysia

■

■

■ ■

■

■

■

■■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■■

■■■

■

■

■■

■

■

■

■

■

■

■

■

■

■

■

■

■■

■

■

■

■

■

■

■

■

■

■

■

■■

■

■

■

■■

■

■ ■

■

■

■■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■

■ ■

■

Chile

Kenya

Thailand

Philippines

MexicoArgentina

Ethiopia

Japan

South Africa

ChinaIndonesia

Rep. ofKorea

Per capita GDP (PPP)


While the excessive buildup of leverageon corporations’ balance sheets was themain source of their increased vulnerability,some cyclical deterioration in returns toinvestment likely contributed as well. Incre-

mental capital-output ratios (the amount ofinvestment needed to generate an extra unitof output) rose in the 1990s, implying somedecline in returns to capital (figure 2-7).Such declines are consistent with the sus-

65

Corporate debt was excessive in the most seriously affectedcrisis countriesFigure 2-6 Debt to common equity ratios of nonfinancial corporations, 1988–96

Source: Claessens, Djankov, and Lang 1998.

Rep

. of K

orea

Japa

n

Hon

g K

ong

(Chi

na)

Thai

land

Indo

nesi

a

Ger

man

y

Phi

lippi

nes

Uni

ted

Sta

tes

Taiw

an (C

hina

)

Sin

gapo

re

Mal

aysi

a

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

1988–94 1995–96

Investment efficiency declined in the 1990s…Figure 2-7 Incremental capital output ratios, 1973–96

1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1996

2

3

4

5

6

7

8

9

10

Five-year moving averages

Indonesia Rep. of KoreaMalaysia Thailand

Source: World Bank data and staff estimates.


tained, exceptionally high rates of invest-ment in East Asia in the 1990s, as well aswith some deterioration in investment qual-ity during a credit and investment boom.8

Credit booms occur in times of generalprosperity and rising asset prices, condi-tions that increase the information prob-lems facing banks by temporarily boostingborrower collateral, making most firmsappear profitable and blurring differencesbetween good and bad long-run risks. Thelikelihood of credit flows into poor invest-ment projects increases (Gavin and Haus-mann 1996). The low levels of expertise inscreening, selecting, and monitoring loanscommonly found in recently liberalizedfinancial systems tend to be further strainedby the rapid increase in loan activity.Increased competition in the banking sectorin the wake of financial liberalization alsotends to reduce the franchise value of banks,which also encourages more risk taking.

Incremental capital output ratios in therecent period had not significantly exceededthe upper end of their range over the past 25years, however. Similarly, while accountingrates of return on assets of nonfinancial cor-porations in some East Asian countries fellsharply in 1995–96, average returns in mostEast Asian countries in 1988–94 had run ina relatively high 5 to 8 percent range (figure2-8). Claessens and others (1998) commentthat “These ROAs [returns on assets] can becompared to ROAs in mature marketeconomies of about 1–3 percent, providingsupport to the notion that the corporate sec-tor contributed significantly to the EastAsian Miracle during most of this period.”Indeed, it is only in Korea and Japan,among East Asian economies, that returnson assets were persistently low over theperiod. These patterns make it more diffi-cult to argue that the decline in investmentproductivity in the mid-1990s was in all

66

…while returns on assets slumped in 1995–96Figure 2-8 Return on assets for nonfinancial corporations, 1988–96

Indo

nesi

a

Thai

land

Mal

aysi

a

Phi

lippi

nes

Sin

gapo

re

Ger

man

y

Japa

n

0

1

2

3

4

5

6

7

8

9

Percent

Rep

. of K

orea

Hon

g K

ong

(Chi

na)

Uni

ted

Sta

tes

Taiw

an (C

hina

)

1988–94 1995–96

Source: Claessens, Djankov, and Lang 1998.


cases a long-run trend related to long-stand-ing structural distortions or rigidities associ-ated with a single East Asian model ofdevelopment, rather than a characteristicelement in the buildup of vulnerabilities in acredit and investment boom.9

The combination of high leveragingand deteriorating returns among nonfinan-cial firms in the mid-1990s translated into adecline in the quality and an increase in theriskiness of bank loans to the private sector.In East Asia, bank fragility rose as lendingto higher risk firms and sectors increased,especially loans for real-estate and stockmarket speculation. Property is estimatedto have accounted for 25–30 percent oftotal bank loans in Indonesia and 30–40percent in Thai land and Malays ia(J.P. Morgan 1998a). But weaknesses inbanks’ loan books were concealed so longas there was strong growth, and asset pricescontinued to rise. Indeed, nonperforming(bad) loans relative to total loans were gen-erally estimated to be at modest levels andin several cases to be falling in the runup tothe crisis, even though underlying condi-tions were worsening (IMF 1997).10 Whenmacroeconomic conditions became lessfavorable, however, bad debt problemsquickly came to the fore and played animportant part in precipitating the crisis.

The links between financial sector liber-alization, credit booms, and banking crisesin East Asia is corroborated in a growingbody of cross-country research. A recentstudy of 53 countries in 1980–95 finds astrongly significant association between theprobability of a banking crisis and earlierfinancial liberalization (proxied by theremoval of interest rate controls) (Demirgüç-Kunt and Detragiache 1998). The study also

finds an association between the probabilityof banking and other characteristics of thefinancial system, such as rapid credit growthand vulnerability to currency crises, (as rep-resented by high levels of M2 to foreignreserves), and macroeconomic factors thatincrease corporate distress among borrow-ers, such as slower economic growth, fallingterms of trade, and high real interest rates.The study does find, however, that the likeli-

hood of a crisis following financial liberal-ization is lower where there is a strongerinstitutional environment for the properoperation of financial markets, as measuredby indexes for the rule of law, corruption,the quality of contract enforcement, and thequality of the bureaucracy.11

Outbreak and spread ofthe crisis

Two main events triggered the crisis inThailand in 1997. The first was the

bursting of the bubble in asset prices. Realestate values had turned down as early as1992, while stock market prices, especiallythose for property and financial companyshares, began to decline in 1994 (figure 2-9).Asset price declines dragged down the valueof borrowers’ net worth and collateral, pre-cipitating a deterioration in the quality ofbanks’ and finance companies’ loan portfo-lios and balance sheets. Construction activ-ity began to fall sharply in 1996.

67

In East Asia, bank fragility rose aslending to higher-risk firms increased,especially loans for real-estate and stock market speculation.


The second trigger was the abruptslowdown in export growth in Thailandand many other East Asian countries in1996. The slowdown appears to have beencaused mainly by cyclical factors: fallingworld import demand, a sharp cyclicaldownturn in world semiconductor demand,and appreciation of many pegged Asiancurrencies as the U.S. dollar rose againstthe yen and other major currencies.12 Aswing toward fiscal contraction in Japan in1997 contributed to slowing demandgrowth in the region, as well as to acceler-ating weakness in the yen. This harsh exter-nal shock reduced sales revenue and corpo-rate cashflow growth in the export-orientedmanufacturing sector, leading to furtherdeterioration in bank asset portfolios. Pri-vate investment and consumer demandgrowth slowed, the latter reflected, forexample, in large declines in Thai automo-bile and department store sales in 1996.

A more flexible currency regime (afloating rate or a floating rate inside broadtarget bands) could have allowed the realexchange rate to adjust to weaker externalconditions and, more important, couldhave made local borrowers more aware ofthe true foreign currency risk. Banks andcorporations, assuming that the currencypeg was there to stay, piled on more short-term foreign debt as a (supposed) low-costand low-risk financing option to tide over a(presumably) temporary economic down-turn. Thailand’s short-term borrowingsfrom Bank for International Settlementbanks rose by $15 billion between the endof 1994 and the end of 1996.

Significantly, Thailand had been oneof the countries more seriously affected bycontagion effects in the wake of the Mexi-can crisis at the end of 1994. Speculativepressure on the currency recurred severaltimes thereafter, and intensified in 1996

68

Thai property and financial company shares tumbleFigure 2-9 Thai stock market indexes, 1993–98

July

199

3

Apr

il 19

94

Dec

. 199

4

Aug

. 199

5

Apr

il 1

996

Jan.

199

7

Sep

t. 19

97

May

199

8

0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

Index (local currency; Dec. 31, 1993 = 1)

Financial

Property

Total market

Source: Datastream.


and 1997 as the difficulties of Thai finan-cial intermediaries and corporationsbecame more obvious. The authoritiesdefended the exchange rate peg duringthese episodes by raising interest rates andusing foreign exchange reserves. Thesedefenses lacked credibility, however, sinceinvestors understood that the authoritieswould ultimately be unwilling to inflictthe kind of damage on the financial andcorporate sector that would be caused bythe sustained high interest rates necessaryto defend the peg indefinitely. Highlight-ing this implicit contradiction, the centralbank began massive liquidity infusions tosupport ailing commercial banks andfinance companies in December 1996. Inthe next s ix months, these creditsincreased four times in real terms, provid-ing the backdrop for the intensifying pres-sure on the exchange rate that culminatedin the floating of the currency in July 1997(figure 2-10). Central bank credit to finan-

cial institutions rose from 2 percent ofGDP at the end of 1996 to 15 percent atthe end of 1997. Central bank credit to thefinancial system also rose sharply inIndonesia and Korea immediately before,or at the time of, significant declines intheir currencies.

It is common for central banks to pro-vide short-term loans as an immediate mea-sure to support a troubled banking system.The problem is to distinguish between lend-ing that may be justified as temporary liq-uidity support to solvent banks, and lend-ing that is an unsustainable attempt to propup fundamentally insolvent banks. The cen-tral bank’s ability to maintain an exchangerate commitment erodes as it monetizes itssupport for the banking system or issuesdebt to finance it, raising expectations of afuture monetization. Even without a centralbank bailout, chronic banking difficultiesmay reduce the central bank’s ability to raiseinterest rates or take other measures to

69

Central banks prop up ailing banks and finance companiesFigure 2-10 Real central bank credit to banks and other financial institutions,January 1996–June 1998

Jan.

199

6

Mar

. 199

6

May

199

6

July

199

6

Sep

t. 19

96

Nov

. 199

6

Jan.

199

7

Mar

. 199

7

May

199

7

July

199

7

Sep

t. 19

97

Nov

. 199

7

Jan.

199

8

Mar

. 199

8

June

199

8

0

1

2

3

4

5

6

7

8

Index (January 1997 = 1)

Thailand

Indonesia

Malaysia

Philippines

Rep. ofKorea

Source: International Monetary Fund, IFS, deflated by CPI.


maintain an exchange rate peg. Consistentwith these arguments, there is considerableevidence that banking crises are significantprecursors of subsequent balance of pay-ments or currency crises (Kaminsky andReinhart 1996; Calvo and Mendoza 1996;Sachs, Tornell, and Velasco 1996). In a sam-ple of 20 countries in 1970–95, for example,56 percent of banking crises were followedby a currency crisis within three years,although only 12 percent of currency criseswere followed by banking crises in the sameinterval (Kaminsky and Reinhart 1996).

Spread of the crisisPressure on the currencies of neighboringSoutheast Asian nations built quickly afterthe fall of the Thai bhat, leading to thefloating of the Malaysian ringgit andPhilippine peso in mid-July 1997 and theIndonesian rupiah in mid-August. All fourcurrencies declined by 25–30 percentagainst the U.S. dollar between June and

the end of November. After that the rupiahentered a second phase of deep decline,while the other three currencies, movingclosely together, began to stabilize (fig-ure 2-11). The rupiah fell by a further two-thirds between the end of November 1997and the end of January 1998, accompaniedby a two-thirds increase in central bankloans to commercial banks aimed at prop-ping up a failing financial system, and atwo-thirds increase in loans by commercialbank to their own faltering private sectorcustomers. Mounting political instabilityadded to the pressure on the exchange rate.In Korea, growing bankruptcies of majorconglomerates in 1997 fueled concernsabout the health of the corporate sector.The won was forced to devalue at the endof October, as international banks failed toroll over large volumes of maturing short-term debts. The won fell 50 percent in twomonths before partially stabilizing. Withthe attack on the won, capital outflows and

70

Most crises currencies have moved together closelyFigure 2-11 Exchange rate indexes, March 1997–September 1998

Mar

. 199

7

May

199

7

June

199

7

Aug

. 199

7

Sep

t. 19

97

Oct

. 199

7

Dec

. 199

7

Jan.

199

8

Mar

. 199

8

Apr

. 199

8

May

199

8

July

199

8

Aug

. 199

8

0.2

0

0.4

0.6

0.8

1.0

1.2

Index (March 31, 1997 = 1)

Indonesia

Rep. ofKorea

Thailand

Malaysia

Philippines

Sep

t. 19

98

Note: U.S. dollar per local currency exchange rates.Source: International Monetary Fund.


speculative pressure on currencies spreadwithin the region—to Hong Kong (China)and Taiwan (China)—and then, outside theregion—to Argentina, Brazil, Mexico, andRussia. Although these pressures werelargely fended off, they returned withredoubled energy later in 1998, leading toRussia’s devaluation of the ruble and decla-ration of a debt moratorium in August; thisthen contributed to a severe, more general-ized downgrading of emerging marketfinancial instruments and intense pressureon currencies in Latin America and otheremerging markets.

The sequence of events in East Asiaconfirms several patterns seen in earliermajor currency crises of the 1990s, such asthe European Monetary System crisis of1992 and the “tequila crisis” of 1994–95.First, currency crises do not occur randomlyin time but are clustered, appearing to passcontagiously from one country to another.Second, a currency crisis in one country sig-nificantly increases the probability of a crisisin other countries, even after controlling fordomestic macroeconomic fundamentals.Third, crises also tend to be clustered geo-graphically, for example in Europe in 1992,

in Latin America in 1994–95 and, at leastinitially, in East Asia in 1997.

Thus in East Asia most currencies fellby large and similar amounts (except forIndonesia’s, which fell furthest) despitesubstantial differences in conditions (table2-5). Current account deficits varied con-siderably (highest in Thailand, lowest inIndonesia), as did overall external debtburdens. Ratios of short-term debt to for-eign reserves, a critical indicator of vulner-ability to speculative attack, were high inThailand, Indonesia, and Korea, but lowerin Malaysia and the Philippines. Exportgrowth had slowed dramatically through-out much of the region in 1996, but theslowdown was most severe in Thailand,where it persisted through 1997. However,countries like the Philippines maintainedhigh double-digit export growth through1996 and 1997, while others, such asIndonesia and Korea, saw a recovery inexport growth in 1997. In the domesticeconomy, the extent of financial sectorweakness as reflected in measures of creditbooms, exposure to the real estate sector,and nonperforming loans also variedsignificantly.

71

Despite different conditions, most currencies fell by similar amountsTable 2-5 East Asia: selected macroeconomic and financial sector conditions, 1996–98(percent)

Ratio of Ratio of Ratio of Change in Change incurrent nonperforming real-estate Change in real exchange real exchange

account deficit loans to bank exposure to bank reserve money rate rate to GDP 1996 assets 1997 assets 1997 Jan. 1997–Jan. 1998 June 1995–June 1997 June 1997–Jan. 1998

Indonesia –3.8 11.0 25:30 362.2 14.0 –68.0Korea, Rep. of –4.7 16.0 15:25 4.4 2.5 –41.6Malaysia –4.9 7.5 30:40 25.8 9.3 –34.2Philippines –4.8 5.5 15:20 12.1 20.0 –30.2Thailand –7.9 15.0 30:40 13.4 16.1 –44.2

Source: World Bank data; International Monetary Fund; J.P. Morgan 1998a.


While a full analysis of contagioneffects in the East Asian crisis still remainsto be done, the wide range of conditions inthe region is consistent with earlier resultsthat the transmission of currency crises isonly partially explained by domestic eco-nomic fundamentals. The East Asian crisesare also consistent with earlier evidencethat currency crises tend to be transmittedbetween countries with strong trade links.13

Among the five crisis countries, exports tothe East Asia region (excluding Japan) aver-age 31 percent of total exports and rangefrom 24 percent in the Philippines to 40percent in Malaysia. East Asian countries

also tend to be more represented in eachothers’ export markets than in the worldmarket, suggesting a greater degree of com-petition with each other than with coun-tries from other regions (Bhattacharya andothers 1998). These channels provide a“fundamentals” rationale for how a crisisin one country can change macroeconomicconditions in others, and make transmis-sion of currency crises more likely. Reces-sion in one country can reduce demand forother countries’ exports, and devaluation inone can increase competitive pressures onothers’ exports in world markets.

Still, the quantitative impact of thesefundamental trade channels in the transmis-sion of the crisis from Thailand to the otherEast Asian countries is likely to have been

limited. For example, Thailand accounts foronly 2 percent of the exports of Korea andIndonesia, 4 percent of Malaysia’s, and5 percent of Philippine exports. Similarly,Thailand’s share as a competitor in theexport markets of the other four countriesamounts to only 1–2 percent. Thus Thai-land’s devaluation creates little pressure forcompetitive devaluations by those countries.The statistical significance of trade links incontagion may rather reflect the fact thatcountries near each other geographicallytend to trade more,14 and it is this physicalproximity that coordinates and focusesinternational investors’ fears about countryvulnerabilities, such as high short-term debtto foreign reserve ratios. These ratios hadbeen high for several years without provok-ing much concern, but, once Thailand wasattacked, a creditor panic or run on inade-quate foreign reserves in Korea and Indone-sia became much more likely. In addition,the importance of fundamental trade chan-nels rises with each additional neighboringcountry affected by contagion, so that oncea number of neighbors have suffered crises,trade channels become more important inamplifying pressure on currencies in theregion.

The issue of contagion has importantimplications for policy. First, if an attack onone country has adverse effects on othercountries unrelated to any fundamentalweaknesses, there is a stronger argumentfor coordinated multilateral action to stemthe spread of the contagion. Second, thesame circumstances mean that there can bea stronger rationale at the national level forgreater caution on full liberalization of thecapital account, or for the use of carefullydesigned controls on short-term capital

72

Recession in one country can reducedemand for other countries’ exports, and devaluation in one can increasecompetitive pressure on others’ exportsin world markets.


inflows to developing countries. Third, theregional dimension of contagion and therole it plays in intensifying currency andother economic pressures in each countrymay provide a stronger rationale for coor-dinated regional responses to the crisis.

Impact on the real economyExperience in many countries shows thatfinancial crises can produce deep and pro-longed downturns in the real economy. Evena relatively mild financial crisis can initiateor exacerbate a downturn, as seen in thecontribution of the U.S. savings and loanepisode to the U.S. recession of 1990–91.One implication is that policymakers need totake account of the high degree of uncer-tainty during financial crises. In particular,they need to take account of the downsiderisk of a shortfall in demand, pushing aweak economy into deep recession through avicious circle of bankruptcy, financial mar-ket collapse, and further declines in demand.

Most initial estimates of the impacts ofthe financial crisis in East Asia on the real

economies of the region (not least those inlast year’s Global Economic Prospects)vastly underestimated their depth and dura-tion. Consensus estimates of GDP growthfor 1998 were progressively cut over theyear since the outbreak of the crisis (figure2-12). This widespread over-optimism wasat variance with a range of indicatorsshowing a weakening in domestic demandeven before the outbreak of the crisis, aswell as with accepted macroeconomic rea-soning and empirical evidence on theimpact of financial crises on the real econ-omy. Such optimism may have been basedon the 1994–95 crisis in Mexico, whereoutput fell sharply for two quarters butbegan growing thereafter, led by a powerfulexpansion in exports (figure 2-13). But dif-ferences in the Mexican case may havemade it a misleading analogy, and basingthe initial policy responses in East Asia onan overly optimistic scenario may havemade it more difficult to protect against theworst outcomes—and may even have exac-erbated the downturn.

73

Last year’s forecasts on GDP growth were over-optimisticFigure 2-12 Evolution of consensus forecasts for 1998 GDP growth in East Asian crisiscountries, June 1997–August 1998

June

199

7

July

199

7

Aug

. 199

7

Sep

t. 19

97

Oct

. 199

7

Nov

. 199

7

Dec

. 199

7

Jan.

199

8

Feb.

199

8

Mar

. 199

8

Apr

. 199

8

May

199

8

June

199

8

July

199

8

Aug

. 199

8

-16

-12

-8

-4

0

4

8

Percent

Indonesia

Malaysia

PhilippinesThailand

Rep. of Korea

Source: Consensus Forecasts, Inc.


The collapse in the near-term growthoutlook in East Asia has been driven bylarge declines in private aggregate demand,private investment in particular, even with-out taking into account possible effects ofthe monetary and fiscal policies taken as ini-tial responses to the crisis. Contributing tothe contraction of private demand were thedecline in external financing to the region,and the impact of falling asset prices, deteri-orating balance sheets, rising interest rates,and high uncertainty on credit supply, andon private investment and consumption.Moreover, various domestic and interna-tional factors resulted in a weaker exportresponse to currency devaluation in some ofthe crisis countries than initially expected.

Private investment. Indicators showed aweakening in private investment in severalcountries even before the Thai crisis. InThailand, construction activity and theBank of Thailand’s composite monthlyindex of private investment activity weremoving downward from early 1996, whilein Korea large corporate bankruptcies were

on the rise and gross fixed capital formationwas weakening through 1997 (figure 2-14).Rates of decline in investment acceleratedsharply with the currency crises, as externalfinancing was suddenly closed off, reducingthe availability of savings for investmentand of foreign exchange for imports of capi-tal equipment. In 1996 current accountdeficits (an approximate measure of capitalinflow) represented 10–20 percent of grossdomestic investment in the crisis countries,being channeled to firms either directlythrough corporate foreign borrowing orindirectly through the domestic banking sys-tem. Other things being equal, the suddenshutting down of capital inflows would bereflected in East Asian firms facing muchtighter quantitative constraints on creditand a higher cost of capital.15

The macroeconomic adjustment forcedby the loss of external financing is vividlyillustrated by the size and rapidity of themove from external current account deficitsto surplus in the crisis countries. There wasa net swing in Korea’s quarterly current

74

Mexico bounced back quickly from the tequila crisis…Figure 2-13 Mexican real GDP and its components, 1994–97

Index (Q1 1994 = 1)

GDP

Fixed investment

Government consumption

Private consumption

Exports

Imports

Q1 1994 Q1 1995 Q1 1996 Q1 1997

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Source: Datastream.


account balance of about $15 billionbetween the second quarter of 1997 andthe second quarter of 1998, or 12 percentof its precrisis (1996) average quarterlyGDP. The net swing for Thailand amountedto 13 percent of precrisis GDP. Both weresubstantially larger than the 7 percentswing in Mexico between the last quartersof 1994 and 1995 (figure 2-15).16 Fixedinvestment in Korea fell almost 60 percentin the first quarter of 1998, three times thefall in Mexico in the first quarter of 1995.The current account adjustment was moresevere in East Asia in part because lessemergency funding was made available.Indonesia, Korea, and Thailand togetherreceived about $118 billion, about the sameamount relative to GDP as the $50 billionpackage Mexico received. Some $44 billionof the Korean and Indonesian packages wascontingent second line of defense fundsfrom individual governments, with littlechance of early disbursement. Indeed, onlyabout $35 billion was disbursed by the end

of the first quarter of 1998, mostly by themultilateral institutions (IMF 1998;Radelet and Sachs 1998b).

Other factors worsening the financialconditions facing East Asia firms includethe bursting of asset price bubbles, leadingto enormous declines in real estate andstock prices, and the huge increase in thelocal currency cost of servicing foreign debtbrought on by devaluation. Together thesefactors dramatically reduced the value offirms’ net worth, rendering many of theminsolvent; reduced the value of security onbank loans to firms; and greatly increasedbanks’ stock of bad loans. Private sectorestimates of nonperforming loans of thedomestic banking system range from 25–30percent of total loans for Korea and Thai-land, to 30–35 percent for Indonesia (J.P.Morgan 1998b). These conditions increasebanks’ difficulties in finding out about,selecting, and monitoring borrowers, lead-ing to less credit availability at given inter-est rates—a “credit crunch.”17

75

…but things were very different in East AsiaFigure 2-14 Private investment indicators in East Asia, 1990–98

0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Index (Q1 1991 = 1)

Thailand: construction area

Thailand: BOT private investment index

Rep. of Korea: gross fixed investment

Q1 1990 Q1 1992 Q1 1994 Q1 1996 Q1 1998Q1 1991 Q1 1993 Q1 1995 Q1 1997

Source: Datastream.


When firms are already highly lever-aged, as in East Asia, these effects are magni-fied, since a fall in asset prices reduces networth by a greater proportion. Emergencysales of assets by credit-constrained firmscan cause a vicious circle of further assetprice and net worth declines that affect moreand more firms, including those that had fol-lowed a relatively prudent, lower leveragepolicy. Some such process of amplification offinancial shocks (such as the bursting of anasset price bubble or a devaluation) appearsnecessary to explain the astonishing assetprice declines in East Asian countries, wherestock prices, especially for property andfinancial companies, have fallen to less than20 percent of their level 4–5 years ago innominal local currency terms.18

Adverse selection problems in financialmarkets are also likely to have been wors-ened by higher interest rates in the immedi-ate aftermath of the currency crises. Higherrates affect firms’ net worth by putting fur-ther downward pressure on asset prices and,where firms have high volumes of short-term

or floating-rate debt, by reducing firms’ freecash flow. More uncertainty, as a result ofhigh volatility in exchange rates, asset prices,and macroeconomic variables, is anotherkey factor that tightens credit conditions byincreasing the informational difficulties ofsorting out good risks from bad (Mishkin1996). Higher uncertainty also tends tolower investment (where the investment isirreversible) by increasing the value of wait-ing for more information about the invest-ment environment.19 According to Rama(1993), virtually all studies of investment indeveloping countries that look at uncertainty(measured as the volatility of various macro-economic indicators) find it to have a signifi-cant adverse impact.

Direct evidence suggests the existenceof a credit crunch to varying degrees in thefive most severely affected East Asian coun-tries.20 Data on credit aggregates are ofonly limited value, however, since they arethe outcome of both demand and supplyfactors. Lower bank lending may reflect afall in demand for credit, as firms scale

76

Thailand’s and Korea’s current account adjustments weremuch greater than Mexico’sFigure 2-15 Mexico, Republic of Korea, and Thailand: current account balances, 1994–98

Q1 1994 Q1 1995 Q1 1996 Q1 1997 Q1 1998-10

-5

0

5

10

15

Billions of U.S. dollars

Mexico

Thailand

Rep. of Korea

Mexico crisis Thai crisis

Source: Datastream.


back investment plans, as much as a greaterunwillingness by banks to lend. In the firsthalf of 1998, volumes of real bank credit tothe nonfinancial private sector in Indone-sia, Korea, and Thailand were flatteningout, though in a highly erratic manner, andin Indonesia only after the credit explosionat the start of the year noted earlier (figure2-16). Studies in the United States and else-where suggest that at first lending continuesto grow, as firms attempt to smoothdeclines in internal cash flow by borrowingmore, earlier in the cycle. Credit begins tofall only six to nine months after a mone-tary tightening.21 In East Asia much newlending is also likely to represent a rolloverof bad and doubtful debts, capitalization ofinterest arrears (which themselves will risebecause of higher interest rates), and exten-sion of new loans to bad debtors, to avoiddeclaring them bankrupt and acknowledg-ing large losses. Overall credit aggregatescould then conceal a significant tighteningof credit to new borrowers.

Some evidence for difficult credit condi-tions comes from rising risk premiums.Spreads between rates charged to high-riskand low-risk borrowers widened—forinstance, between yields on corporate bondsand government housing bonds in Korea(figure 2-17). Other evidence comes fromthe sharper impact of the recession on capi-tal spending by small firms, which generallyhave weaker balance sheets, have less accessto corporate bond markets, and rely moreon bank financing. Survey data from Koreashows small- and medium-size firms expect-ing to raise a significantly smaller share oftheir total funding from financial institu-tions, and a decline of around two-thirds intheir fixed investment spending in 1998.

How much is investment in East Asialikely to fall? Using estimated coefficientsfrom an empirical study of private invest-ment in East Asia, a rough calculation canbe made of the potential impact of the crisison private investment in Thailand in 1998(Larrain and Vergara 1993). This indicates a

77

Real bank credit to corporations flattened out in the first halfof 1998Figure 2-16 Real credit extended to nonfinancial private sector, December 1996–June 1998

Dec

. 199

6

Feb.

199

7

Apr

. 199

7

June

199

7

Aug

. 199

7

Oct

. 199

7

Dec

. 199

7

Feb.

199

8

Apr

. 199

8

June

199

8

1.0

1.1

1.2

1.3

1.4

1.5

1.6Index (Dec. 1996 = 1)

Thailand

Indonesia

Rep. of Korea

Source: International Monetary Fund, World Bank.


hefty 9.3 percentage point fall in the ratio ofprivate investment to GDP (table 2-6).22

Over half this fall is associated with height-ened uncertainty (proxied here by the coeffi-cient of variation of the real exchange rate),which worsens information problems incredit markets and increases the optionvalue of delaying investment. Another 3percentage points of the decline is associ-

ated with weakening economic activity.Activity affects investment in two ways—byits impact on expectations of future prof-itability and by its impact on the currentprofitability, cash flow, and net worth ofcredit-constrained firms (Romer, D. 1996).In common with most studies on investmentin developing countries, credit to the privatesector has a significant positive association

78

Spreads between high-risk and low-risk debt surge in theRepublic of Korea in 1997Figure 2-17 Korean corporate and government housing bond yields, April 1993–August 1998

0

5

10

15

20

25

Percent

Corporate

HousingSpread

Apr

. 199

3

Dec

. 199

3

Aug

. 199

4

Apr

. 199

5

Dec

. 199

5

Aug

. 199

6

Apr

. 199

7

Dec

. 199

7

Aug

. 199

8

Note: Korean corporate bond yield is three-year maturity. Government housing bond yield is five-year maturity.Source: Datastream.

More uncertainty could account for much of the fall in privateinvestment in Thailand in 1998Table 2-6 Determinants of private investment in East Asia as applied to Thailand, 1997–98

Coefficient T-statistic 1997 1998 Change

Dependent variable: real private investmenta 23.50 13.8 -9.3

Explanatory variablesUncertainty (coefficient of variation of real exchange rate)b –1.402 –2.94 0.04 0.2 –5.2Activity (previous year change in per-capita real GDP) 0.019 4.76 5.00 –2.0 –3.1Real public investmenta 0.178 2.78 10.00 10.0 0.0Debt overhang (external debt)a –0.153 –3.39 59.60 72.5 –0.7Credit to private sectora 0.177 2.48 114.00 110.0 –0.1Real interest rate –0.015 –2.20 10.00 12.0 –0.1

Note: Adjusted R-squared: 0.919; S.E. of regression: 0.082; number of observations: 54.a. Percentage of GDP, shown in levels in table but calculated in logs in estimation.b. Coefficient of variation over three years.Source: Larrain and Vergara 1993, World Bank data and estimates; Thailand: Memorandum of Economic Policies, May 28, 1998.Econometric estimates use panel data for four East Asian countries in 1975–88.

with investment. Since actual credit volumesare not in themselves an appropriate mea-sure of the severity of credit constraints,however, they are left out of the calculation.Real interest rates had risen significantly in1997 and in this model had the bulk of theireffect on investment in that year, with only asmall further effect in 1998. A direct esti-mate of the impact of worsening credit con-ditions cannot be made without appropriatemeasures. However, the role of credit inmagnifying the impact of shocks is implicitin the importance of uncertainty and declin-ing activity as the major forces leading to acontraction in investment (table 2-6).

Personal consumption. In Thailand,and to a lesser degree in Korea, personalconsumption indicators were also weaken-ing before the outbreak of the currency cri-sis (table 2-7), with rates of decline quickly

accelerating after it. Retail sales volumesfell 10–15 percent below levels a year ear-lier by the first quarter of 1998. Again, thesize of the declines was not inconsistentwith accepted macroeconomic models andcross-country evidence. Greater imperfec-tions in developing country credit marketsmean that consumers are less able to bor-row to maintain consumption during tem-porary downturns. Shocks to currentincome thus have strong impacts on currentconsumption, as do fluctuations in wealth,a factor of obvious importance in East Asiagiven the huge declines in stock market andreal estate values.23

A study of household savings in10 developing countries (including Korea,the Philippines, and Thailand) estimatesthat a 10 percentage point fall in the ratioof financial wealth to disposable incomeproxied by the ratio of M2 money to dis-posable income is associated with a 2 per-centage point fall in the ratio of consump-tion to income (Schmidt-Hebbel and others1992). It also finds that a reduction in theavailability of foreign savings (a reductionin the current account deficit) has a signifi-cant negative impact on household con-sumption, presumably as a result of tightercredit. Finally, the massive increase inuncertainty during the financial crisis isalso likely to have a negative impact onconsumer durable purchases, as peopleincrease precautionary savings to guardagainst bad times, and as they delay pur-chases to gather more information. C.Romer (1990–1993), for example, arguesthat the uncertainty generated by the U.S.stock market crash of 1929 and the subse-quent financial volatility was a major forcein the sharp fall in U.S. personal consump-


79

Personal consumption wasweakening in Thailand andKorea before the crisisTable 2-7 East Asia: personal consumptionindicators(percentage change from year ago)

Rep. ofThailand Korea

Retail Sales

Newcar Bangkok— Entire Retail

sales 38 storesa kingdom sales

Q1 1996 11.3 0.9 10.9Q2 1996 3.1 –2.9 10.1Q3 1996 3.6 –14.4 11.2Q4 1996 –1.8 –16.8 9.4Q1 1997 –8.8 –14.7 6.4Q2 1997 –19.9 –24.0 6.5Q3 1997 –48.7 –8.0 5.6Q4 1997 –73.5 –7.3 0.3Q1 1998 –70.7 –12.3 –9.9Q2 1998 –10.3b

a. Discontinued.b. April to May.Source: World Bank.


tion in 1930 and the onset of the GreatDepression. For the 1980s and 1990s, Has-sler (1996) finds a significant negative asso-ciation between increased financial volatil-ity in Sweden and the United States, and themost important element in consumerdurable purchases—automobiles.

Exports. One expected outcome of alarge devaluation (combined with a contrac-tion in domestic demand) is a strong exportresponse, which would strengthen the abil-ity of firms to service foreign debts. Such atextbook response indeed occurred in the

aftermath of the Mexican peso crisis, whenexport growth in U.S. dollar terms almostdoubled from 17 percent in 1994 to 31 per-cent in 1995 (table 2-8). The acceleration inMexican export earnings growth came froma doubling in export volume growth, whileexport prices in U.S. dollars remainedroughly flat. The export boom was helpedby continued economic expansion in Mex-ico’s major market, the United States (whichtook 85 percent of Mexico’s exports), andthe recent conclusion of the North Ameri-can Free Trade Agreement (NAFTA), which

80

The U.S. economy witnessed three major businesscycle contractions between 1920 and 1940: therecession of 1920–21, the Great Depression of1929–1933, and the recession of 1937–38. In

response to severe loan losses in the early 1930s and highcosts of raising new capital, U.S. banks faced pressures to

reduce depositors’ perceptions of risk. They cut dividendsbut found it difficult to raise new capital. The primarymeans to reduce depositor risk and prevent withdrawalswas contraction in the supply of loans. Banks substitutedriskless assets for loans, a process that took place overseveral years. This is consistent with the view that the

Credit crunch in the Great Depression; New York Fed member banksBalance sheets of New York City banks, 1922–40

Deposits Loans Cash and government securities

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

Millions of U.S. dollars

1922 1929 1933 19361925 1931 1934 1940

Source: Calomiris and Wilson 1998.

Box 2-1 U.S. banks and the credit crunch duringthe Great Depression


spurred a boom in export-oriented foreigndirect investment in Mexico.

Factors suggesting a similar favorableoutcome in East Asia included the large sizeof the currency devaluations, flexible factormarkets (allowing for speedy redeployment ofresources into export industries), and thestrong export marketing channels and experi-ence built up over decades of successfulexporting. Set against these positive condi-tions, however, was the fact that East Asiancountries’ principal regional export mar-kets—other East Asian countries and Japan—

81

contraction in bank credit during the Depression waslargely a result of a capital crunch that forced banks tolimit loan portfolio risk. An alternative explanation is thatbanks became more risk-averse.

Relative to other recessions, the Great Depressionsaw extreme declines in loan ratios and capital ratios. Itwas also unusual in another respect: a reduction indeposits by more than 30 percent between 1930–32. Thehistory of American interwar business cycles is reflected inthe balance sheets of New York city banks (see box figureleft)—rapid loan growth, reductions in the ratio of liquidassets, and large capital injections during the boom of1922–29—whereas recessions are associated with declinesin lending activity, increases in riskless assets, (cash plusgovernment securities), and falls in bank capital.

Clearly, the primary means banks employed for con-trolling their asset risk was the decline in ratio of risky assetsto riskless assets. This variable declined steadily throughoutthe 1930s. The ratio of the book value of loans to cash,reserves, and government securities rose first from 2.06 in1922 to 3.33 in 1929, and then fell to 1.89 in 1931 andcontinued falling, eventually reaching 0.25 in 1940 (see fig-ure right). The alternative would have been for banks toraise new capital. But banks did not replace the capital theylost during the Depression. Virtually no stock was offeredafter 1930. This is consistent with the view that in the after-math of the Depression, as the potential for loan lossesloomed large, the cost of new stock issues was prohibitive,

and thus banks sought to satisfy the depositor risk con-straint by continuing reductions in portfolio risk. Reductionin bank lending was a response to the need to avoid deposi-tor “discipline” and the adverse selection costs of raisingequity. The banks’ capital crunch contracted the supply oflending and thus worsened the severity of the Depression.

Source: Calomiris and Wilson 1998

A significant decline in lendingNew York City banks, loans to cash plusgovernment securities ratio, 1922–40

1922

1925

1929

1931

1933

1934

1936

1940

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Ratio

Source: Calomiris 1998.

Dollar export revenues in EastAsia are flat or only modestlyhigherTable 2-8 Growth merchandise export revenues(U.S. dollars), 1994–97(percent change)

Rep. ofYear Mexico Korea Indonesia Thailand

1994 17.4 16.3 8.7 22.81995 31.2 31.0 13.7 25.61996 20.9 4.6 8.6 –0.61997 15.1 5.1 8.9 3.3July 1997–June 1998 12.6 6.9 2.6 0.4

Note: Average of 12 months over previous 12 months.Source: World Bank.


were slowing down and entering recession atabout the same time. About 50 percent of theexports of the most seriously affected crisiscountries go to Japan and the rest of Asia.

Added to this weakness were fallingforeign currency export prices. The reces-sion in the region has reduced worlddemand and prices for primary commodi-ties, such as oil, rubber, and timber,exported by the crisis-affected countries.Foreign currency prices of manufacturedexports were also falling sharply as region-wide devaluations and weakness in the yenshifted large new supplies from the regioninto highly competitive (and in part reces-sion-affected) world markets. In Korea,almost entirely an exporter of manufactures,export volumes in June 1998 were estimatedto be 15 percent higher than a year earlier,while export prices in U.S. dollar terms weredown 17 percent, resulting in a year-on-yeardecline in dollar revenues. The picture wasthe same among the crisis countries gener-ally, with export volumes increases of 15–25percent in the year since the Thai crisis offsetby price declines. The result was flat or onlymodestly higher dollar export revenues.

Thus, while growth in export volumes pro-vides important near-term support for out-put and employment in an otherwise bleakdemand picture, weak foreign currency rev-enues hamper firms’ abilities to resolve their

external debt problems, making a return tofinancial health, revived investment, andsustainable growth more difficult.

Immediate policy responses

In contrast with the crises in Mexico andArgentina in 1994–95, one of the great

surprises in East Asia was how little imme-diate effect the initial policy responsesappeared to have had in reducing pressureon currencies or stabilizing investor confi-dence. To the contrary, much or even mostof the depreciation in currencies occurredafter these measures were taken (figure 2-18). This was the case whether the initialpackage entailed new agreements with themultilateral institutions (Indonesia, Korea,and Thailand) or not (Malaysia and thePhilippines).24

Financial and real economic conditionsdeteriorated much more than expected,requiring several quick changes in initialpackages. Some stabilization of currenciesoccurred in the first part of 1998 in mostcountries, other than Indonesia. Nominalinterest rates have come down in recentmonths—some to below precrisis levels. InKorea negotiations between the govern-ment and foreign commercial banks (sup-ported by the U.S. government) led to astandstill on short-term foreign debt servic-ing and then to an agreement to roll overand restructure short-term debt. This movetoward an orderly workout may havereduced immediate balance of paymentspressures on the currency and eased credi-tor panic. Some have argued that currencystabilization in Thailand may have beenencouraged by government guarantees onbank liabilities, including those to foreign

82

In contrast with crises in Mexico andArgentina in 1994–95, initial policyresponses had surprisingly little effectin reducing pressure on currencies orstabilizing investor confidence inEast Asia.


83

Much of the depreciation in currencies came after initial policy responsesFigure 2-18a Effects of policy responses:Thailand

Exchange rate index (left scale)

Interest rate (right scale)

1 32

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆

◆◆

◆◆

◆◆◆◆◆◆

◆◆◆

◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆

◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆

◆◆◆◆◆◆

◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

Mar

. 199

7

June

199

7

Aug

. 199

7

Nov

. 199

7

Jan.

199

8

Apr

. 199

8

June

199

8

Sep

t. 19

98

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

●●●●●●●●●●●●●●●●

●●●●●●●●

●●●●●●●●

●

●●

●●●●●●

●●●●●

●●●●●●●●●

●●●●●●

●●●●

●●

●

●●

●●●●●

●●●●●●●

●●●●

●●●●●●●●●●●●●●●

●●●●

●●●●●●●●

●

●●●●●

●●●●●●●●

●●●

●●●●●●●●

●●●●●●●●●●●●●

●●●●

●●●●●●●●●●●●●

●●●●●●●

●●●●●●●●●●

●●●●●●

●●●●●●●●●

●●●●●●

●●●●●●●●

●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●

●●●●●●●●●●●●●

●●●●●●●●●

●●●●●●

●●●●●●

●●●●●●●●●

●●●●●●

●●●●●●●●●●●●●

●●●●●●●

●●●●●●●●●●●●●●●●●●●

●●●●●●●●

5

10

15

20

25

30

Dollar/local currency unit; March 31, 1997 = 1 Percent

*

1 – Initial package 2 – First revision 3 – Second revision * Bank liabilities guaranteeSource: International Monetary Fund, World Bank.

Figure 2-18b Effects of policy responses: Indonesia

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆

◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆

◆◆

◆◆◆◆◆

◆◆◆◆

◆◆◆◆

◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

0

0.2

0.4

0.6

0.8

1.0

1.2

●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●

●

●●●●●●●●●●●●●●●●●●●

●●●●●

●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●

●

●

●●

●●

●●●

●

●●

●

●●●●●●●●●●

●●●

●

●●●●●

●

●

●●●●

●●●●●●●●●●●●●●●●●

●●●●●

●

●●

●●●●●●●●●

●

●

●●●●●

●

●●●●●●●●

●

●●●●●●●●●●

●●●●●●●●●

●●●●●

●●●●●●

●

●●●●

●●●●●●●●●●

●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●

●●●●

●

●●

●

●●●●●

●●●

●●

●

●●●●●●●●●●

●●●●●●

●

●●●

●

●

●

●

●●

●

●

●

●●●●●

●●●●●●●●●●●●

●

●●●●●●

●

●

●●●●

●

●●

●●●●●

●

●●●●●●●

●●

●

●

●●●●

●

●●●

●●

●

●●●●

●

●●●●●●●●●

0

20

40

60

80

100

120

Mar

. 199

7

June

199

7

Aug

. 199

7

Nov

. 199

7

Jan.

199

8

Apr

. 199

8

June

199

8

Sep

t. 19

98



1

*

3

2


1 – Initial package 2 – First revision 3 – Second revision * Debt restructuring agreementSource: International Monetary Fund, World Bank.


84

Figure 2-18c Effects of policy responses: Republic of Korea

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●

●●

●●●●●●●●●●●●●

●●

●

●

●●●●●●●●●●

●●●●●●●●●●●●●●●●●●

●●●●●●●●●

●●

●

●●●●

●●●●●●●●●●●●●●●●●●●●●●

●●●●

●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●

●●●●●●●●

●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

10

12

14

16

18

20

22

24

26

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆

◆

◆◆◆◆◆

◆◆◆◆◆◆

◆

◆

◆

◆◆

◆

◆◆◆◆

◆

◆◆◆

◆◆

◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆

◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1



1

3

2

Dollar/local currency unit; March 31, 1997 = 1

Mar

. 199

7

June

199

7

Aug

. 199

7

Nov

. 199

7

Jan.

199

8

Apr

. 199

8

June

199

8

Sep

t. 19

98

Percent

*

1 – Initial package 2 – First revision 3 – Second revision * Debt restructuring agreementSource: International Monetary Fund, World Bank.

Figure 2-18d Effects of policy responses: Malaysia

●●●●●●●●●●●●●●●

●●●●●●●●●

●●●●●●●●●

●

●

●●●●●

●

●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●

●

●

●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●

0

5

10

15

20

25

30

35

40

45

50

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆

◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆

◆◆◆

◆◆◆◆◆

◆◆

◆◆◆◆◆

◆◆◆◆◆◆◆

◆◆◆◆◆◆

◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

Mar

. 199

7

June

199

7

Aug

. 199

7

Nov

. 199

7

Jan.

199

8

Apr

. 199

8

June

199

8

Sep

t. 19

98



1

3

2


*

1 – Initial package 2 – First revision 3 – Second revision * Foreign exchange controlsSource: International Monetary Fund, World Bank.


85

creditors (Radelet and Sachs 1998b). InIndonesia, however, a debt workout initia-tive in June 1998 did little to reverse whatby then amounted to an 80 percent nomi-nal devaluation.

Initial programs agreed with multilat-eral institutions called for some $118 billionin loan commitments from multilateral andbilateral sources to help stabilize exchangerates and provide resources for repaymentof debts to international creditors, althoughEast Asia appears to have had less immedi-ate access to disbursements than Mexico in1995. Countries agreed to fiscal tighteningto preserve budget balance or surplus, andto adopt more restrictive monetary policies(higher nominal interest rates, quantitativetargets for domestic credit aggregates) tosupport exchange rates and curb inflation.Other important elements of initial pro-grams included commitments to restructurethe financial sector and adopt structural

reforms to further liberalize foreign tradeand investment and deregulate domesticmarkets, including elimination of monopo-lies and other restraints on trade, privatiza-tion, and removal of price controls and sub-sidies. Initial responses in countries withoutagreements with the multilateral institu-tions, such as Malaysia, agreed with theemphasis on fiscal tightening, but differedon other provisions, such as the stance ofmonetary policy.

Fiscal policyFiscal balances in the affected countries hadall been in surplus or virtual balancethrough the 1990s, so that excess demandgenerated by fiscal imbalances was notamong the main sources of the crisis.Indeed, given the plunge in private invest-ment and consumption, weak rather thanexcess aggregate demand became the prin-cipal macroeconomic characteristic of the

Figure 2-18e Effects of policy responses: Philippines

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●

●

●

●

●●●●●●●●●●●●●●●●●●●●

●

●

●

●

●

●

●●

●

●●●●●

●

●

●●●●●●●

●

●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●

●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●

0

20

40

60

80

100

120

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆

◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆

◆◆◆◆◆◆

◆◆◆◆◆◆

◆◆◆◆

◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆◆

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

Mar

. 199

7

June

199

7

Aug

. 199

7

Nov

. 199

7

Jan.

199

8

Apr

. 199

8

June

199

8

Sep

t. 19

98


Interest rate (right scale)1

2


1 – Initial package 2 – First revisionSource: International Monetary Fund, World Bank.


86

crisis economies in late 1997 and in 1998.From the onset of the crisis through early1998, fiscal policies (contrary to theirdesign and with benefit of hindsight) werecontractionary. If the initially announcedfiscal targets had been implemented, theywould, indeed, have been strongly contrac-tionary. As the severity of recessionsbecame apparent, however, the aim ofachieving fiscal balance or surplus wasquickly relaxed and more stimulative mea-sures adopted (table 2-9).

While adopting fiscal policies that aremore supportive of economic activity inthe near-term as a first priority, policymak-ers also face the daunting task of makingprudent medium-term provisions for theexpected high fiscal costs of financial sec-tor restructuring. Estimating the fiscalcosts of restructuring has just begun andwill likely not be completed for someyears. Preliminary private sector estimatesof bank recapitalization needs range from20 percent of GDP for Indonesia and

Malaysia, to 30 percent for Korea andThailand (table 2-10).25

Fortunately, the most seriously affectedcountries are relatively well placed to takeon these costs because of their earlier fiscalprudence and resultant low levels of publicdebt (figure 2-19). These relatively lowstarting points make it more feasible tofinance restructuring through increased gov-ernment borrowing, either foreign (forexample, from multilateral institutions) ordomestic, since the resulting higher govern-ment debt to GDP ratios could be stabilizedby primary fiscal balances not too differentfrom those before the crisis (table 2-10).26 InThailand, for example, government borrow-ing to finance the entire financial restructur-ing would raise the ratio of governmentdebt to GDP from 4 percent in 1996 to 34percent. The primary fiscal surplus neededto stabilize this debt depends on how far thereal rate of interest in Thailand exceeds thegrowth rate of the economy. But for a rea-sonable range the implied primary surpluses

Initial fiscal targets were contractionary, but were relaxed asrecession worsenedTable 2-9 Central government fiscal balances, 1987–98(percentage of GDP)

1998

Initial First Second Third1987–91 1992–96 1996 1997 package revision revision revision

Indonesia—Fiscal balance –1.0 0.9 1.2 –0.2 1.0 –1.0 –3.0 –8.5Primary Balance 1.6 2.7 2.4

Korea, Rep. of—Fiscal balance 0.0 0.2 0.5 0.3 0.0–0.3 0.0–0.3 –0.8 –1.8Primary balance 0.7 0.8 1.1 0.8

Malaysia—Fiscal balance –3.9 0.7 1.1 2.6 2.5 0.5 –3.5Philippines—Fiscal balance –2.6 –0.1 0.3 –0.9 0.0 –1.0 –3.0Thailand—Fiscal balance 2.1 2.4 2.4 –0.9 1.0 1.0 –1.6 –2.4

Primary balance 4.3 3.0 2.5 –0.6

Note: The primary balance includes interest on government debt. Data above for Indonesia refer to fiscal year starting that year(1997 refers to FY1997/98 April to March) and that for Thailand ending in that year (1997 refers to FY1996/97 October toSeptember).Source: IMF International Financial Statistics; World Bank data and staff estimates: Rep. of Korea 1997a, 1997b, 1998a, 1998b;Indonesia 1998a, 1998b; Thailand 1998a, 1998b. 1998 excluding projected interest costs of financial restructuring.


are all substantially less than the 3–4 per-cent of GDP that Thailand averagedbetween 1987 and 1996.27 Restoring theeconomy to fast growth and full employ-ment is, therefore, an important conditionto help medium-term fiscal stabilization.

Monetary policyThe use of monetary policy to manage theEast Asian crises threw up extremely diffi-cult tradeoffs between macroeconomic andfinancial sector stabilization objectives. In

several cases initial adjustment programsadopted by the East Asian crisis countriesstressed tightening monetary policy to sta-bilize exchange rates and curb inflationarypressures, the benefit of a more stableexchange rate being to limit the damage tothe balance sheets and cash flow of banksand firms with heavy foreign currencydebts. Orthodox monetary policies indefense of currencies are still appropriate inmany circumstances. But higher interestrates could themselves be expected to have

87

Bank recapitalization costs will be high, but fiscal surpluses neededto finance them are not largeTable 2-10 Fiscal costs of financial restructuring(percentage of 1996 GDP)

Primary fiscal surplus needed to stabilize newCost of New debt-to-GDP ratio assuming a gap between

Government financial government real interest and growth rates of:debt restructuring debt 1% 2% 3% 4%

Indonesia 24 19 43 0.43 0.86 1.29 1.72Korea, Rep. of 9 30 39 0.39 0.78 1.17 1.56Malaysia 36 20 56 0.56 1.12 1.68 2.24Thailand 4 30 34 0.34 0.68 1.02 1.36

Note: End-1996 government debt numbers are used because 1997 data are not yet fully available.Source: IMF International Financial Statistics; J.P. Morgan 1998b; Daniel 1997.

Governments are well-placed to meet the costs of financialrestructuringFigure 2-19 Central government debt, 1975–96

1975 1978 1981 1984 1987 1990 1993 19960

20

40

60

80

100

120

Percentage of GDP

Rep. of Korea

Malaysia

Indonesia

Thailand

Source: IMF International Financial Statistics.


damaging effects on already weak bank andfirm balance sheets and real economicactivity, as discussed above. In evaluatingthis difficult tradeoff, policymakers wereconcerned about how reliably, and underwhat conditions, higher interest rateswould in fact serve to stabilize exchangerates during a currency crisis, a subject onwhich there had been surprisingly little for-mal research.

Higher interest rates should strengthenexchange rates by making it more attractiveto hold financial instruments denominatedin the particular currency.28 Under somecircumstances, however, this mechanismmay not represent the full (or general equi-librium) impact of interest rates onexchange rates. If tighter credit and higherinterest rates worsen the financial conditionof already weak banks and corporations,an increase in the probability of default onfinancial instruments issued by them couldweaken the exchange rate by increasing therisk premium attached to the currency.Monetary tightening could also weaken theexchange rate by reducing expectations offuture output, demand for money, andinterest rates.29

The relative importance of thesediverging influences in the East Asian crisisis something of an open question, especiallygiven the difficulty of establishing counter-factual scenarios of what would have hap-

pened without a given policy action. Casualinspection of the evolution of interest ratesand exchange rates in the East Asian coun-tries does not suggest any simple connec-tion between the two (figures 2-18a to 2-18e). This perception is confirmed at aslightly more systematic level by considera-tion of correlations over a rolling 30-dayinterval between exchange rates and inter-est rates in Korea, Malaysia, the Philip-pines, and Thailand (figures 2-20a and 2-20b: a positive correlation indicates anassociation between higher interest ratesand exchange rate appreciation).30 Clearly,negative correlations are at least as com-mon as positive ones, although the signifi-cance of this fact is not clear, since it is con-sistent with both interest rate increasesweakening exchange rates through thechannels noted above, or with fallingexchange rates inducing defensive increasesin interest rates.

The cross-country movement of exchangerates over time also provides interesting obser-vations. Leaving aside the special case ofIndonesia, where the exchange rate fell morethan 80 percent against the dollar in the yearto mid-1998, currencies in the other four crisiscountries moved closely together, each depreci-ating 35–40 percent by mid-1998. This simi-larity in exchange rate paths is notable, sinceseveral of these countries pursued quite differ-ent interest rate policies. Thailand and Koreatightened monetary policies, Malaysia did notraise interest rates apart from a brief effort inJuly 1997, and the Philippines reduced interestrates to near precrisis levels after increases inthe second half of 1997.

Kraay’s (1998) study of interest ratepolicies in some 186 speculative attacks oncurrencies in 75 middle- and high-income

88

The use of monetary policy tomanage the East Asian crises threw upextremely difficult tradeoffs betweenmacroeconomic and financial sectorstabilization objectives.


89

There appears to be no simple connection between theevolution of interest rates and exchange ratesFigure 2-20a Thirty-day rolling correlations between interest rates and exchange rates,March 1997–September 1998: Malaysia and Thailand

Mar

. 199

7

May

199

7

July

199

7

Sep

t. 19

97

Nov

. 19

97

Jan.

199

8

Mar

. 19

98

May

199

8

July

199

8

Sep

t. 19

98

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1.0

Thailand

Malaysia

Correlation coefficient

Source: Kraay 1998.

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1.0

Rep. of Korea

Philippines

Mar

. 199

7

May

199

7

July

199

7

Sep

t. 19

97

Nov

. 199

7

Jan.

199

8

Mar

. 199

8

May

199

8

July

199

8

Sep

t. 19

98Correlation coefficient

Figure 2-20b Thirty-day rolling correlations between interest rates and exchange rates,March 1997–September 1998: Philippines and Republic of Korea


countries in 1960–97 found that interestrate increases are not always necessary toward off a speculative attack (table 2-11).Rates were not increased in 50 of the 117cases (43 percent) in which attacks failed,and increases failed to foil speculativeattacks in 35 of the 102 instances wherethey were used (33 percent of the cases).The study was unable to reject the hypothe-sis of no significant relationship betweeninterest rate policy, and the success or fail-ure of speculative currency attacks. Itcould, however, be that, where rateincreases were not necessary, the specula-tive attacks were less serious. Similarly,where rate increases were not sufficient toprevent a successful attack, it could be thatthe speculative attacks were more severe.These considerations were addressed bycontrolling for various fundamental condi-tions that earlier studies have found to begood predictors of speculative currencyattacks. It was still not possible to reject thehypothesis of no significant association.

Goldfajn and Gupta (1998) find moreevidence for a positive link between interestrates and exchange rates in the generalcase, but this link is reversed when a coun-try is facing a banking crisis. For a large set

of undervaluations of real exchange rates inthe aftermath of currency crises, they findtight monetary policy significantly increasesthe probability of reversing the undervalua-tion through an appreciation of the nomi-nal exchange rate, rather than throughhigher inflation. But the opposite is truewhen the country is facing a banking crisis.In that event, a monetary tightening isfound to significantly reduce the probabil-ity of reversing undervaluation through anappreciation of the nominal exchange rate(table 2-12). They conclude that “the rela-tionship between high interest rates andstable currencies hinges crucially on thestate of the banking system.”

Overall, the still early state of theresearch into the behavior of interest ratesand exchange rates during crises may notallow firm conclusions. There is, however,more evidence about the adverse impact ofhigh interest rates on real economic activity,confirming the importance of undertakingmonetary policy in a flexible and nuancedway that gives due consideration to the pol-

90

One study found interest ratehikes are not always successful infending off speculative attacks…Table 2-11 Discount rate policy in 186 specula-tive attacks on currencies, 1960–97(number of instances)

Attack AttackPolicy succeeds fails Total

Discount rate raised 35 67 102Discount rate not raised 34 50 84

Total 69 117 186

Source: Kraay 1998.

…another study finds a morepositive link between interestrates and exchange rates, butthis is reversed in countries inbanking crisesTable 2-12 Conditional probability of reversing anundervaluation of the real exchange rate throughappreciation of the nominal exchange rate

Tight Nontight money money

All countries 0.37 0.26Countries with banking crises 0.29 0.45

Note: Estimates are for cases in which the real exchange rateis undervalued by at least 15 percent, and at least 50 percentof the undervaluation is removed by a rise of the nominalexchange rate.Source: Goldfajn and Gupta 1998.


icy dilemmas that arise, such as in EastAsia, where the financial system is fragile,corporations are highly leveraged, andshortfalls in aggregate demand are large.

Structural reformsAdjustment programs in East Asia are dis-tinguished by the importance they place onstructural reforms. Many of these focusedappropriately on financial and corporaterestructuring, and strengthening of financialregulation, supervision, and corporate gov-ernance areas. Especially for Indonesia,adjustment efforts also included a variety ofother structural reforms (table 2-13). Thesefocus on liberalizing domestic markets andforeign trade and, in fiscal affairs, removingpublic subsidies. The valid general rationale

for these structural measures is that theywill reduce impediments to long-run growthby increasing efficiency, improving resourceallocation, and enhancing competition.

From the perspective of successfulimplementation, however, it may also beuseful to evaluate what the most promisingtimes and conditions for initiating suchreforms might be. On the one hand, “ifthere is one single theme that runs throughthe length of the political economy literatureit is the idea that crisis is the instigator ofreform.” (Rodrik 1996. See also Williamson1994). Thus the study of trade reform in theperiod up to the 1980s debt crisis by Papa-georgiou, Choksi, and Michaely (1990)finds that most “strong” trade reforms tookplace in the context of a general perception

91

In Indonesia, structural reforms focused on domestic and foreigntrade liberalization and removing subsidiesTable 2-13 Structural reform programs in East Asia

Reform Indonesia Rep. of Korea Thailand

Domestic trade Eliminate monopolies in food, plywood, and clove distribution.

Foreign trade Reduce tariff on a broad range of items, including agricultural products, chemicals, and metal products. Phase out all nontariff barriers. Eliminate or reduce export taxes and quantitative restrictions on palm oil, leather, cork, minerals, logs, timber, and other exports.

Fiscal issues Eliminate subsidies and raise prices on domestic fuel, electricity, rice, soybeans and soymeal, sugar, wheat flour, corn, and fishmeal. Discontinue tax, trade, and credit privileges for national car and aircraft projects.

Privatization Privatization of state enterprises overthe medium term.

Note: Financial and corporate restructuring and reform programs are not included.Source: Indonesia: 1998b. Rep. of Korea: 1998b, 1998c. Thailand: 1998b.

Privatize and restructureenergy, telecommunica-tions, water, andrailways.

Phase out Import DiversificationProgram. Align import certificationprocedures with internationalpractice. Permit foreigners to engagein securities dealing, insurance,leasing, and other property-relatedbusinesses.

Review and rationalize subsidyprograms.

Privatize five state-owned enterprisesimmediately, another six by 2002.


92

of complete economic collapse, a change ofpolitical regime, or both, and the close asso-ciation between crisis and reform was, ifanything, even more evident in the 1980s.Various arguments have been proposed toexplain this empirical observation: forexample, that crises create a sense ofurgency and public solidarity and sostrengthen the hands of reformist govern-ments relative to obstructive special interestgroups, or, even more simply, that lack ofresources obliges governments to accept theadvice of external leaders.

Researchers have noted a number ofqualifications to the significance of theempirical association between crisis andreform, however. First, there is great varia-tion in the necessary intensity and durationof crises needed to bring about reform.Haggard and Webb (1993) note that at var-ious points in the 1980s, the Thai, Colom-bian, and Indonesian governments under-took reforms as preemptive responses towarning signs of impending difficulties,while, at the other end of the response con-tinuum, several African countries enduredfull-blown economic disasters year afteryear without taking action. They state,however, that we do not really understandwhy governments behave in these differentways. They also note that “a crisis in noway guarantees that any remedial actionstaken will be sustained or institutional-ized,” as evidenced by the main instanceswhere reforms are reversed once a crisis is

over, an event especially likely where gov-ernments are not themselves firmly commit-ted to, or “have ownership” of, reforms.31

Rodrik (1996) also notes that “theemphasis on crisis has in itself little predic-tive content as to what form the responsewill take,” and that, in particular, somestructural reforms undertaken duringmacroeconomic crises have little logical con-nection with the task of macroeconomic sta-bilization and indeed can even complicatethat effort. Classical examples are wherereductions in tariffs result in loss of fiscalrevenue and higher fiscal deficits, or wherethe “compensating” devaluation required bytrade liberalization conflicts with a nominalexchange rate anchor adopted to stabilizeinflationary expectations. Here, as in otherareas of response to the crisis, difficult trade-offs become apparent.

Crises may open windows of opportu-nity for structural reforms and should betaken advantage of in many circum-stances. However, more research is alsoneeded on circumstances in which struc-tural reforms may complicate immediatecrisis management tasks during financialcrises. For example, dealing with struc-tural impediments that are genuine butrelated in only a limited way to the causesof the immediate financial and macroeco-nomic crisis may undermine confidence,and aggravate uncertainty and informa-tional problems, and thereby make crisismanagement appear even more difficultthan it already is. The fact that well-con-sidered structural reforms are often diffi-cult to design and monitor, and take yearsto implement and assess, may also add tothe difficulties of economic agents in eval-uating the immediate stabilization pro-

Adjustment programs in East Asia aredistinguished by the importance theyplace on structural reforms.


gram. In some circumstances, could theinherently higher social contentiousness ofsome structural reform policies make itmore difficult to achieve social consensuson immediate adjustment tasks during afinancial crisis?32

Financial and corporaterestructuring and reform

By the middle of 1998, large parts of theprivate corporate and financial sectors

in the five crisis countries were insolvent orsuffering severe financial distress. Eco-nomic recovery policymakers need toundertake what experience has shown to bethe unusually long, complex, and arduoustask of nurturing these sectors back tohealth, as well as strengthening institutionsof prudential supervision, regulation, andgovernance that would reduce the likeli-hood of such crises in the future.

Given the systemic nature of the crisis,financial restructuring will require stronggovernment leadership within a clearstrategic framework, including, inevitably,the injection of substantial public funds.Returning viable corporations to healthwill mean restructuring their often enor-mous domestic and foreign debts, byrescheduling, writing down, or convertingdebt to equity. The involvement of foreigninvestors, who can provide new equity andrisk capital, will be important for bothfinancial and corporate restructuring.There is much that OECD governments cando to speed the resolution of debt over-hang, especially with external private credi-tors. The need for resolution of domesticdebt problems is equally compelling ifeconomies are to move ahead. Restructur-

ing on the scale needed in the East Asiancrisis countries is relatively unexplored ter-ritory, however, and new approaches maywell be needed.

Dimensions of financial sectorrestructuringThe dimensions of financial restructuringare staggering. In Indonesia, Korea,Malaysia, and Thailand, nonperformingloans are thought to be so extensive thatwriting them off against bank capital willresult in a negative net worth in the bank-ing system. Recapitalizing banking systemsto achieve the 8 percent capital adequacyratios recommended by the BIS will cost anestimated 20 to 30 percent of GDP (fig-ure 2-21).

Systemic banking crises—often definedas a situation of negative net worth in thebanking system (Caprio and Klingebiel1996b)—exact large real economic costs.Under such conditions, banks are especiallysusceptible to sudden losses of confidenceand runs by depositors. Contagion effectscan result in shutting down significant por-tions of the payments system and the criti-cal information-collecting and -processingfunction performed by banks, with severeconsequences for economic activity. Ulti-mately, even more harm can result wheninsolvent and poorly regulated banks

93

Given the systemic nature of the crisis,financial restructuring will requirestrong government leadership within aclear strategic framework, including,inevitably, the injection of substantialpublic funds.


remain open, protected against bank runsby explicit or implicit government guaran-tees of deposits. This kind of “silent, pro-tracted financial distress,”33 common indeveloping and transition economies,results in mounting misallocation ofresources as banks continue to lend newmoney to insolvent debtors, to avoid hav-ing to write off bad debts, or gamble on new high-risk investments to recoup ear-lier losses.

A prompt, comprehensive program offinancial restructuring aims to avoid or

minimize these costs by shutting down ormerging insolvent banks and restructuringand recapitalizing those judged to be viable.Restructuring can involve market-basedsolutions or government intervention—orboth. Market-based solutions include liqui-dating insolvent banks without more com-pensation for depositors than that providedunder existing deposit insurance schemes(for example, the liquidation of BCCI in theUnited Kingdom in 1991) or selling (ormerging) sick banks to other banks withoutgovernment financial assistance. Thesemethods are evidently better suited to situa-tions where bank insolvency or distress islimited or localized.

In systemic financial crises, however,where the risk of significant economic disrup-tion and loss of confidence is high, govern-ment intervention, involving the use of publicresources for bank restructuring, has beencommon. In a study of eight bank restructur-ing exercises in the 1980s and the early 1990s(Argentina, Chile, Colombia, Ghana,Malaysia, the Federal Republic of Yugoslavia,Spain, and the United States), although mar-ket-based solutions were tried in most cases,there was government intervention in all eight(Sheng 1996). Mechanisms used include pub-lic recapitalization and sale to new owners,government-assisted merger with a viablebank, temporary nationalization, or even,failing all else, bailouts.

Financial restructuring programsneed, however, to create strong incentivesagainst future moral hazard, typically bywriting off bad debt against the capital ofexisting shareholders, replacing bankmanagement, and otherwise ensuring thatthose who benefited from earlier riskybehavior bear a significant part of the cost

94

Recapitalizing banks inIndonesia, Korea,Malaysia, and Thailandcould be 20–30 percentof GDPFigure 2-21 Estimates of non-performingloans, capital asset ratios, and costs ofrecapitalizing banking systems

Hon

g K

ong

(Chi

na)

Indo

nesi

a

Japa

n

Mal

aysi

a

Phi

lippi

nes

Sin

gapo

re

Rep

. of K

orea

Thai

land

-20

-10

0

10

20

30

40

Cost of bank restructuring (percent of GDP)Indonesia: 19 Rep. of Korea: 30Japan: 20 Thailand: 30Malaysia: 20

Percent

Non-performing loans (percent) 1998Capital asset ratio (percent) after NPL write-off

Note: NPLs are mid-points of ranges cited in source;recapitalization assumes 8 percent capital asset ratios.Source: J.P. Morgan 1998b.


of restructuring. In addition, probleminstitutions must not be allowed to con-tinue to expand credit to high-risk ordelinquent borrowers.34 Of course, closingdown insolvent financial institutions orallocating losses to bank owners, share-holders, managers, or employees is politi-cally difficult. The credibility of restruc-turing programs will depend heavily onthe willingness of governments to dealwith existing majority bank shareholdersfrom influential business groups.

These priorities need to be balancedagainst other, more immediate, ones. Thefirst is to maintain public confidence in thefinancial system when closing down insol-vent banks. This requires a comprehensiveand credible plan for financial restructuringthat demonstrates to the public that theremaining banks will be solvent, well-capi-talized, and will have adequate access tolender of last resort liquidity. The failure todo this when closing 16 Indonesian com-mercial banks in November 1997 under theinitial program with multilateral institu-tions sparked bank runs and large transfersfrom local private banks to state- and for-eign-owned banks, as well as capital flightoffshore, and contributed to a secondround of currency collapse.

To buttress confidence among deposi-tors and creditors, Indonesia, Korea, andThailand announced public guarantees ofdeposits as well as of other domestic andforeign liabilities. Such socialization ofbank liabilities is not without its costs,however, such as increasing the fiscal costof restructuring by forgoing some contri-bution from depositors.35 Although it isoften feared that imposing losses on depos-itors will lead to bank runs, that does not

appear to be the case, at least when theaction is taken within a comprehensiveplan. This is the conclusion of Baer andKlingebiel (1994) from five episodes wheredepositors were asked to share in the costof resolving banking crises—in the UnitedStates (1933), Japan (1946), Malaysia(1985–88), Argentina (1980–82), andEstonia (1992).

Restructuring also needs to be done inways that ease the stringent credit crunchfor viable corporate borrowers, for exam-ple, by facilitating credit to exporting firms(that confirm their viability by earning for-eign exchange in world markets). In thiscontext, the implementation of capital ade-quacy standards needs to be handled flexi-bly and not on a rigid timetable, so that itcomes about through injections of new cap-ital into banks rather than through banksfurther cutting back on loans, which wors-ens the credit squeeze.

Financial sector restructuringand reform effortsMost of the crisis-affected countries haverevised legislation to strengthen prudentialregulation and supervision of the financialsector. They have tightened up loan classifi-cation and provisioning requirements (gen-erally aiming to achieve at least the 8 per-cent BIS capital adequacy standards by theyear 2000) and improved disclosure,accounting, and auditing standards to

95

Restructuring also needs to be done inways that ease the stringent creditcrunch for viable corporate borrowers.


international levels. They have set limits onlending to shareholders and other con-nected parties and strengthened rules tolimit maturity and currency mismatches onexternal borrowing.

Countries have also created institu-tions to carry through financial restructur-ing—such as the Indonesia Bank Restruc-turing Agency, the Korean FinancialSupervisory Commission, and the ThaiFinancial Sector Restructuring Author-ity—although in cases such as Indonesiatheir operations are seriously hampered bya shortage of trained personnel. A strong,independent, public agency with the politi-cal and legal clout to implement difficultdecisions is essential for successful finan-cial restructuring (for example, the Reso-lution Trust Corporation in the UnitedStates to handle the savings and loan cri-sis). Inadequate powers and dispersal offunctions can be disastrous; pooling offinancial talent may be essential given thethinness of such human resources. Theseagencies are evaluating banks, their port-folios, systems, and management to sortinstitutions into sound, well-managedones that could form the core of a newrevitalized banking system; the nonviablethat need to be shut down; and the weakthat could be restructured and recapital-ized under the direction of the supervisoryauthority to regain viability.

In Thailand, 56 finance companieswere shut down and their assets were auc-tioned. By the end of August 1998, sixcommercial banks had been temporarilynationalized (with the aim of later privati-zation) through writing off bad debtsagainst shareholder capital, replacement ofmanagement, and recapitalization through

conversion of short-term credit from thecentral bank into equity. Initially, the gov-ernment adopted a market-based approachto recapitalization of the rest of the bank-ing system, encouraging voluntary infu-sions of private capital. The need for publicsupport was recognized in the second halfof 1998, however, when some 300 billionbaht in public funds were made available tosupport tier 1 and tier 2 bank capital,largely on condition that bad debts be writ-ten off or provisioned on an acceleratedtimetable (implying acceptance of painfulwritedowns of shareholder equity).

In Korea, the government had byAugust 1998 provided for up to 75 trillionwon of public support for financial restruc-turing, about 16 percent of GDP. Some 10merchant banks were shut down, and twocommercial banks temporarily nationalizedfor later privatization. Five of 12 commer-cial banks that did not meet the 8 percentcapital adequacy rule at the end of 1997were to be taken over by other banks aftermuch of their bad debts were acquired bythe newly formed Korea Asset ManagementCorporation. The remaining seven under-capitalized banks were to present voluntaryrehabilitation plans with sufficient infu-sions of new private capital, failing whichthey will be subject to temporary national-ization, mandatory merger, or closure. Inthe rest of the banking system, workouts ofloans to the corporate sector are to be han-dled under a voluntary framework.

In Indonesia, 16 banks were closed inNovember 1997. In the absence of a com-prehensive plan, this action provoked runson other banks. Many other weak bankswere placed under control of the bankrestructuring agency. In August 1998,

96


plans were announced to nationalize,merge, or close 11 banks, involving consid-erable writedowns of shareholders’ capital,and efforts to recover intergroup or con-nected loans made by banks to sharehold-ers from these shareholders’ other assets.

Mustering the political will to raise ade-quate public funds to recapitalize banks in anoninflationary way is one of three basicprinciples in managing financial crises(Rojas-Suárez and Weisbrod 1996).36 Thiswill not be easy where countries are cor-rectly expanding fiscal deficits to buoyaggregate demand and where the interestcosts of financial restructuring are likely tobe significant. One solution is to attractexternal capital to recapitalize banks anddistressed corporations. Thailand, which hasattracted about $3 billion in foreign invest-ment and sold four banks to foreigninvestors, has made the most progress in thisdirection. Korea and Indonesia have alsoattracted some external investment on thecorporate side. Foreign investment inflowswill also make stronger managerial and tech-nical capabilities and skills available to thefinancial sector. Success in attracting foreigncapital will depend in part on the credibilityof official programs, as well as on the legalease, transparency, and speed of the foreigninvestment process. One reason for LatinAmerica’s better economic performance inrecent years (and Mexico’s quick recoveryafter its crisis) has been success in privatizingand in attracting new FDI.

Corporate restructuring,bankruptcy law, and debtworkoutsThe success of financial restructuring inEast Asia will be closely bound to the suc-

cess of restructuring in corporate sectors,which hold most loans made by domesticbanking systems. The indebtedness of localcorporations to local banks is one corner ofa triangle of debt relationships that willneed to be simultaneously addressed. Theother two are the debts of local corpora-tions and local banks to external creditors,mostly foreign banks.

The relative weight of these debt rela-tionships varies considerably. In Indonesia,direct borrowing by the corporate sectormakes up the bulk of external borrowing bythe private sector (figure 2-22). At exchangerates of the 10,000–15,000 rupiah to thedollar prevailing in the first half of 1998,the cost of debt servicing was so high that

97

In Indonesia, mostexternal borrowing wasby the corporate sectorFigure 2-22 Structure of external debtby class of borrower, year-end 1997

Indonesia Rep. of Korea Thailand

0

10

20

30

40

50

60

70

80

90

100

Percent of total debt

Public sector

Private financial sectorPrivate nonfinancial sector

Total debt: $102 billionTotal debt: $137 billion Total debt: $154 billion

Source: J.P. Morgan 1998b.


98

virtually the entire Indonesian corporatesector was probably insolvent, and theimprovement in exchange rates since thenpermits more leeway. In Korea, where cor-porate regulations made such borrowingdifficult, a much larger proportion of exter-nal debt was taken on by local banks forlending to local corporations. In Thailand,both banks and corporations were bigexternal borrowers. Given the insolvency oflarge parts of the local financial and corpo-rate sectors, the resolution of debt over-hangs on each corner of the triangle is likelyto require restructuring of the original termsof borrowings, either through court-orderedcorporate rehabilitations or reorganizations,or through orderly debt workouts betweendebtors and creditors.

One obstacle in working out debt diffi-culties is the weakness of bankruptcy lawand administration in East Asian countries.A well-developed bankruptcy law recog-nizes that, left to themselves, debtors andcreditors may take a long time to reach avoluntary reorganization and that delays inreaching agreement have real economiccosts for society.37 One is the failure toexploit profitable new investment opportu-nities because of conflicts of interestsamong the parties (for example, amongshareholders of the debtor firm, existinglenders, and potential new lenders).Another is to encourage investments to beoverly risky, when they are made. Yet

another is the possibility of “grab races”between creditors, which force a borrowerinto liquidation even if it is in the interest ofall lenders to maintain the borrower’senterprise as a going concern.

Bankruptcy law attempts to overcomesome of these problems through a collec-tive proceeding in which the court pro-vides binding arbitration between theclaims of the parties, balancing two broadaims. The first is to maximize the incomeand growth prospects of firms that appearable to earn at least their economic cost ofcapital (through debt rescheduling, writingdown unserviceable debt, or introducingnew debt with priority over older claims),while liquidating unviable firms. The sec-ond is to create incentives for strict adher-ence to debt contracts in future, for exam-ple, by penalizing debtors for resorting tobankruptcy.

Since most claims in the East Asian cri-sis are between private parties, a well-devel-oped bankruptcy system could provide onechannel for sorting out problems in a waythat would be in the long-run interest ofboth debtors and creditors. Resort to bank-ruptcy is unlikely by itself to provide acomplete framework for dealing with EastAsian debt overhangs, however. First, in asystemic economic and financial crisismany even well-managed firms becomeinsolvent as a result of factors such as steepdevaluations, interest rate hikes, and dropsin aggregate demand. Without a coordi-nated approach to the debt problem, acase-by-case treatment could lead to the liq-uidation of many firms that would beviable under more normal macroeconomicconditions. The loss of the information andknowledge capital represented by these

One obstacle in working out debtdifficulties is the weakness ofbankruptcy law and administration inEast Asian countries.


firms could then have a harmful long-runeffect on growth.

Second, bankruptcy law systems inmost East Asian countries are antiquatedand lack adequately trained personnel.Legal processes are so costly, prolonged,unpredictable and, in some cases, open topolitical influence and corruption, thatresort to the courts is usually not attempted.In Indonesia, for example, virtually all com-panies have stopped servicing debt, but fewhave applied for bankruptcy court protec-tion (see box 2-2). Until this year, bank-ruptcy law was based on a one-page Dutchedict of 1906, which emphasized liquidationbut did not allow for court-supervised cor-porate rescues. More significantly, corrup-tion and personal influence were perceivedto pervade the administration of bankruptcylaw. In Korea, filing bankruptcy papers isstraightforward, but progress thereafter isslow. Eight Korean chaebol (conglomerates)filed for bankruptcy court protection in1997. Since then the companies have con-tinued to operate while in default on debt,but there has been little progress in present-ing business plans for corporate restructur-ing to the court. In Thailand, too, lengthyproceedings and other weaknesses have cre-ated a credit culture in which companieshave been able to default on debt withimpunity.38

In the aftermath of the crisis, manycountries have moved ahead to revamp andmodernize their bankruptcy and fore-closure law, and to strengthen corporategovernance. In Thailand and Indonesia,amendments to bankruptcy laws strengthenthe capacity of the courts to approve reor-ganization (rather than just liquidation),reduce the discretionary power of the

courts, and increase transparency, certainty,and efficiency in court proceedings. Indone-sia, recognizing the low level of profes-sional skills, is also making provisions forimproved training, licensing, and selectionof judges, receivers, and administrators.

These efforts, vital as they are for long-run prosperity, are unlikely to provide aquick fix for the problems at hand. Passinglegislation is one thing: developing the nec-essary human skills, knowledge, expertise,credibility, and respect for the rule of law inthe day-to-day administration of bank-ruptcy law is a task that will take years, ifnot decades, to complete. Even a workingbankruptcy system would not deal with theproblems of insolvency in a systemic eco-nomic crisis. Thus, what will becomeimportant are orderly debt workouts—thatis, less formal ways in which governmentsor other arbitrators attempt to achieve thesame economic objectives as formal bank-ruptcy court proceedings by bringingtogether debtors and creditors to negotiateresolution of debt problems.

99

Box 2-2 Filing a bankruptcypetition in Indonesia

Foreign creditors gained some experience of the Indone-sian bankruptcy system after 1991, when the Bentoel cig-arette company declared insolvency, owing $300 millionin equal parts to foreign and domestic banks. After the

failure of voluntary negotiations with the company, the foreignbanks filed a bankruptcy petition in 1992. The court first ruledthat creditors could not enforce a personal guarantee on debtgiven by a Bentoel shareholder until the company entered liquid-ity proceedings. It then ruled there were no grounds for puttingthe company into liquidation. The case continues.

Source: J.P. Morgan 1998b.


Unlike bankruptcy courts, whose deci-sions are binding, orderly workouts mustrely on voluntary agreements betweendebtors and creditors, although in practicemoral suasion, political pressure, or finan-cial incentives play an important part. Butvoluntary negotiation can take a long time,especially in the case of default on debt. Anumber of earlier debt crises illustrate theproblems (Aggarwal 1998). The debt crisisof the 1980s was not resolved until the U.S.government threw its weight behind mar-ket-based debt restructuring with the Bradyplan. Often, allocating losses between andamong debtors and creditors requiresstrategic direction from some such centralplayer.

By mid-1998 the East Asian crisis hadled to two main initiatives on workouts forexternal debt. In January, Korea worked outa deal with creditor banks to reschedule $24billion of maturing short-term debt owed byKorean banks, converting it to government-guaranteed loans with maturities of one tothree years. Several features are noteworthy:first, the high proportion of Korean debtowed by local banks provided a strongmotivation for averting disruption to thedomestic payments system. Second, the rela-tively small number of bank debtors andcreditors helped coordination. Third, thedeal, while not envisaged in the initialadjustment program with multilateral insti-tutions, had behind it the strong support ofthe United States government. Fourth, thedeal seems to have eased immediate pres-sures on the won in foreign exchange mar-kets. The currency stabilized and graduallystrengthened in the following months.

Indonesia’s June 1998 agreement withits creditor banks attempts to create a

framework to deal with the much morecomplicated question of the $65 billion orso of foreign debt owed by thousands of itscorporations. This initiative provided anincentive for voluntary debt restructuringthrough a government guarantee of foreignexchange for debt service at a fixed rate, tobe determined by the actual course ofexchange rates over a 12-month periodending in 1999. It was modeled on the planto restructure Mexican private corporateexternal debt in the wake of the 1982 debtcrisis. Running from 1983 to 1992, theMexican plan helped restructure $12 bil-lion of debt. However, the plan came at thepotentially high cost of a substantial publicsubsidy to the private sector in the eventthat the exchange rate depreciated againstthe dollar by more than that implied byMexican-United States interest rate differ-entials. It remains to be seen whether theIndonesian scheme (described in WorldBank 1998b) can succeed without a sub-sidy. Initial results were not encouraging,with very few transactions. The rupiahexchange rate also initially remainedextremely weak, averaging 14,000–15,000to the dollar in the month after the deal andfalling below 15,000 in July; however, byOctober 1998, the rupiah had strengthenedsharply to about 7,000–8000, thanks to thefalling value of the dollar. Overall, theIndonesian experience so far is consistentwith earlier experiences that orderly inter-national debt workouts are more likely tosucceed when they have strong supportfrom the governments of creditor countries.

Some East Asian governments are alsopromoting voluntary workouts for corpo-rate debt owed to domestic banks. They areremoving tax disincentives to debt restruc-

100


turing, removing legal barriers to debt-equity swaps, and encouraging creditorcommittees and workout units at commer-cial banks. Korea has gone furthest towarda formal framework, bringing some 200financial institutions to sign a corporaterestructuring agreement. Voluntary debtworkouts with corporations will followLondon Rules for extrajudicial resolution ofclaims, and disagreements will be submittedto formal arbitration. Potential incentiveproblems of purely voluntary approach areaddressed to some extent by guidelines tocurtail the scope for emergency or rescueloans to troubled corporate debtors and tocurb cross-guarantees within industrialgroups, which allow weak affiliates to con-tinue borrowing via stronger affiliates.

Given the high corporate leveraging inEast Asian countries, a strong focus ondebt-to-equity conversions may be the keyto resolution of the debt problem—espe-cially if supplemented by policies to liberal-ize conditions for foreign equity ownershipand to foster the development of domesticcapital markets, through mutual funds, pri-

vately managed pension funds, and so on.The purchase from the banks of newly cre-ated equity positions in corporations bystrongly managed foreign and domesticmutual funds and pension funds wouldhelp to clean up bank balance sheets. Itwould also strengthen corporate gover-nance by breaking down the concentrationof corporate control by a few insider familygroups and provide much stronger indepen-dent monitoring of management. Limits onforeign investment have been almost com-pletely removed in Indonesia and signifi-cantly loosened in Korea and Thailand.

Social impact of the crisis

In recent decades East Asia has reducedpoverty and improved living standards

and social conditions at a pace unrivaled inhistory (table 2-14).39 Per capita incomegrowth averaged a remarkable 5.5 percenta year over the past 30 years. In the mid-1970s, six out of every 10 people in theregion lived on less than $1 a day. By themid-1990s, only two of 10 did.

101

Living standards improved dramatically throughout East Asia in the20 years to 1995Table 2-14 Living standards in East Asia, selected years, 1970–96

Number of people Net primaryin poverty Headcount indexa Life expectancy Infant mortality rate school enrollment(million) (percent) (at birth) (per 1,000 live births) (percent)

Country 1975 1995 1975 1995 1970 1996 1970 1996 1970 1995

China 568.9a 269.3 60a 22 62 70 69 33 76 99Indonesia 87.2 21.9 64 11 48 65 118 49 76 97Korea, Rep. of 61 72 46 9 >99 100Malaysia 2.1 <0.2 17 <1 62 72 45 11 84 91Philippines 15.4 17.6 36 26 57 66 71 37 >99 100Thailand 3.4 <0.5 8 <1 58 69 73 34 79 88

Note: All estimates of poverty are based on $1 per person per day poverty line at 1985 PPP prices.a. Data are for 1978 and apply only to rural China.Source: World Bank 1998c.


Growth has led to larger and fasterreductions in poverty in East Asia than inother regions.40 One reason is that theregion started off 30–40 years ago with arelatively equal distribution of income andwealth, and with a strong policy focus onwidely spread human capital gains, ruraldevelopment, and labor-intensive growththat has protected this distribution. Integra-tion with the world economy, coupled withflexible labor markets, translated into highrates of job creation and rising real wages.Between 1986 and 1993, employment andmanufacturing productivity growthallowed real wages to rise 3–6 percent ayear. Countries also invested in humandevelopment, focusing on areas that benefitthe poor most directly: basic curative andpreventive health services, primary educa-tion, and others. One study found thathigher school enrollment rates in East Asiaaccounted for 38 per cent of the predicteddifference in economic growth betweenEast Asia and Latin America in 1960–90.

The social achievements of the EastAsian miracle are genuine and undeniable.Nevertheless, many of the social conse-quences of the present economic crisis arelikely to be protracted. Most countries lackformal mechanisms to protect people fromjob losses and their consequences, while pri-vate savings and informal safety nets may beinsufficient to deal with economy-wide

shocks. Governments do not have systemsin place to adequately track the impact ofshocks or policy interventions on householdwelfare or the institutional capacity to dealwith mass layoffs (for example, throughretraining schemes or massively scaled-uppublic works programs).

Social impact of crises andstabilization policiesWhat are the social costs of crises? Whichgroups will be hurt more than others? Theimpact of crises on household welfare iscomplex and often difficult to isolate fromthe impact of policies to manage crises.Important dimensions are a loss in house-hold income due to layoffs, unemployment,and reduced hours of work; a decline inpurchasing power due to price increasesand a fall in real wages; and reduced accessto social services because of lower personalincomes and, in some cases, public spend-ing. Output losses, unemployment, andhigher inflation are associated with anincrease in poverty. There is also a worsen-ing in the condition of those who arealready poor, including lower living stan-dards and greater malnutrition (see box 2-3). Income inequality often rises as well.41

How much poverty increases duringcrises depends on initial levels of povertyand inequality, the mix of stabilization poli-cies, the structure of the economy, the flexi-bility of output and factor markets, andother factors. Wage and price rigidities inthe modern sector of the economy may shiftthe burden of adjustment to the informalsector, and thus onto the poor. Differentpolicies also affect the poor differently.Devaluation may have a positive effect onthe poor who are employed in sectors pro-

102

In recent decades East Asia hasreduced poverty and improved livingstandards and social conditions at apace unrivaled in history.


ducing tradable goods and services. Layingoff public sector employees may create anew group of poor. Eliminating consumersubsidies and reducing social expenditureswould affect the well-being of all the poor.42

Both poverty and inequality increasedin Sub-Saharan Africa during most episodesof economic crises and attempted adjust-ment (Demery and Squire 1996).43 Thesame thing happened in Eastern Europeduring the transition of the 1990s (Raval-lion and Chen 1996; Milanovic 1995).Poverty increased in 55 of 58 recessions inLatin America during the 1980s anddecreased or remained unchanged in 25 of32 recoveries (Morley 1994). Recessionswere also associated with higher incomeinequality in Latin America (figure 2-23).High and variable inflation was particu-larly damaging to the poor, who have lim-ited access to mechanisms for protectingconsumption. Workers were forced to

accept large wage cuts, unemployment, andlow paid jobs in the informal sectorbecause of the absence of social safety nets.

The impact on unemploymentand poverty in East AsiaUnemployment in the crisis countries hasrisen sharply as firms have reduced outputor shut down operations. Unemploymentcould rise to about 13 million in Indonesiaby the end of 1998, 3.5 million in Thailand,and 1.6 million in Korea, for a total of 18million, compared with 5.3 million in1996.44 Construction and financial sectoremployment are being especially hard hit.Inevitably, given the uncertainties andpotential structural shifts associated withfinancial crises, the band for plausible esti-mates of the impact on unemployment (aswell as other social indicators) is likely tobe a wide one. Official Indonesian esti-mates suggest unemployment of 20 million

103

Economic crises hurt poor and rich. The poor,however, are much less able to respond to a non-diversifiable risk like a recession. If domestic capi-tal markets were perfect, all economic agents

could borrow to smooth consumption and maintain wel-fare during a crisis. But capital markets are imperfect:credit or insurance is not available to the poor, and thereis little they can do to smooth out consumption and wel-fare. Thus, crises and recessions may result in irreversibledamage to the poor: malnutrition or death from starva-tion (in extreme cases) and lower schooling levels(Thomas and others 1996).1

In Côte d’Ivoire, higher food prices in the stabiliza-tion program of the 1980s may have increased malnutri-tion (Thomas and others 1996; Grootaert 1994). Suddenfluctuations in income or food availability can be fatal toalready malnourished children. The most common types

of malnutrition (iodine, vitamin A, and iron deficiency)may lead to lower IQ, retarded physical growth, mentaldisabilities, reduced learning capacity, and lower resis-tance to infections. These conditions are associated withincreased repetition and dropout rates in school. Theeffects of malnutrition on child mortality are severe:results from 53 developing countries indicate that 56 per-cent of child deaths were attributable to the collateraleffects of malnutrition and 83 percent of these were dueto mild to moderate malnutrition.

1 Alderman and Paxson (1994); Bourguignon, Lambert and Suwa(1996); Cornia et al. (1987). See also Jolly and van der Hœvern(1991).

Source: World Bank (1993b); Pelletier and others (1995).

Box 2-3 The impact of crises on the poor maybe irreversible


104

in 1998, for example. In Thailand andIndonesia, where agriculture employs 40percent and 55 percent of the labor force,respectively, and where underemploymentis also a significant problem, the impact ofthe crisis was greatly exacerbated bydrought (box 2-4).45 In Thailand, at the endof 1997, 70 percent of the unemployedwere in agriculture, most in the poor north-east area. Migrant workers from neighbor-ing countries will also be severely affected:

about a million will go back to Bangladesh,Cambodia, and Myanmar in 1998. Realwages are expected to fall dramatically inIndonesia, with the sharp depreciation ofthe currency since June 1997 and the col-lapse of industrial activity, exceeding evenreal wage declines in Latin America duringthe crises of the 1980s (table 2-15).

How this affects poverty in the crisiscountries will depend on how much outputdeclines and how income distribution

Recessions meant higher income inequality in Latin AmericaFigure 2-23 Latin America: income inequality before and after recessions

0

10

20

30

40

50

60

70

Before After

Gini coefficient

Argentina1986

Brazil1983

Chile1982

Costa Rica1981

Mexico1982

Mexico1986

Note: Income inequality is measured by the Gini coefficient two to three years before and after the relevant recessions:Argentina 1982; Brazil 1982–83; Chile 1982; Costa Rica 1981; Mexico 1982 and 1986.Source: World Bank Development Prospects Group.

El Niño—an abnormal warming of sea-surface tem-peratures in the central and eastern Pacific Oceanoff the coast of South America—affected atmos-pheric conditions worldwide and contributed in

1997 to one of the worst droughts reported in 50 years inIndonesia, Papua New Guinea, some areas in the Philip-pines and Thailand, and the central and northeastern partsof China. Forest fires, smoke, and haze aggravated envi-ronmental conditions. Irregular rains and floods were seen

in Bangladesh, Myanmar, and Sri Lanka. Indonesia wasparticularly affected by El Niño. In addition to low rainfallduring the 1997 cropping season, the late onset of rains inthe last quarter of 1997 delayed rice planting by one totwo months. Rice production in 1997 fell by 4 percent, butsome areas saw larger declines. Food shortages and out-right hunger have emerged in the most affected areas.Much of the poverty increase in Asia in 1998 will be deter-mined by the drought rather than the financial crisis.

Box 2-4 There will be more poverty in Asia because of El Niño


105

evolves. Let’s look at projected increases inpoverty in some East Asian countries(assuming no change in the distribution ofincome and using a $1 a day poverty linefor Indonesia and the Philippines and $2 aday for Malaysia and Thailand) (figure 2-24). Almost 17 million more people inIndonesia are expected to fall below thepoverty line in 1998. Moreover, many peo-ple in Indonesia live only slightly above thepoverty line, so measured poverty is quitesensitive to the poverty line chosen. With apoverty line of $1.25 a day, for example,the number of poor people would rise by22 million, bringing the total number ofpoor to 56.5 million. In other countries theincrease would be less pronounced, but stilllarge: 2.3 million in Thailand, 665,000 inthe Philippines, and under 500,000 inMalaysia.46

Poverty calculations are also very sensi-tive to changes in distribution. A deteriora-tion in both growth and income distribution(a 10 percent worsening in the Gini coeffi-cient) would bring the poverty incidence in

most countries back to the levels of the early1990s, practically eliminating the effects of10 years of growth (figure 2-25).

Real wages are expected to collapse in IndonesiaTable 2-15 Real wages and unemployment during crises in East Asia and Latin America

Real wages (percent change) Unemployment rate (percent)a

One year Year of One year One year Year of One year Country (year of crisis) before crisis crisis after crisis before crisis crisis after crisis

East Asia and PacificIndonesia (1997) 13.5 5.5 –40 to –60 4.9 5.9 13.8Korea, Republic of (1997) 7.3 –1.4 –0.4 2.0 2.6 7.5Thailand (1997) 2.3 2.1 –10.3 1.5 3.5 10.9

Latin America andthe CaribbeanArgentina (1982) –11.0 –10.1 26.3 4.8 5.3 4.7Chile (1982) 9.0 0.0 –11.0 25.0 26.2 21.4Costa Rica (1981) n.a. –12.0 –19.3 5.9 8.8 9.4Mexico (1995) 0.0 –13.1 –8.2 3.7 6.2 5.5

a. Urban unemployment rate only for Latin America and the Caribbean.Source: International Labour Organisation; Central Banks; World Bank staff estimates; U.N. Economic Commission for LatinAmerica and the Caribbean; Economic Survey of Latin America (various issues); World Bank 1994.

The total number of poorpeople could risethroughout East AsiaFigure 2-24 Poverty projections in EastAsia with no change in inequality

Indonesia Malaysia Philippines Thailand

0

5

10

15

20

25

Poverty rates (percent)

1997 1998 1999 2000

Note: Poverty lines ($/day): Indonesia (1.0);Malaysia (2.0); Philippines (1.0); Thailand (2.0).Source: Ravallion and Chen 1996.


In Indonesia (using the local currencydefinition of the poverty line), poverty inthe next couple of years would almost dou-ble in urban areas and increase by half inrural areas (figure 2-26). The sharpestincreases are expected to be among workersemployed in trade and manufacturing.

Poverty is just one dimension of wel-fare. Will life expectancy or education andhealth indicators deteriorate as well as aresult of the crisis? Certain to come is areduction in access to social services,because of losses in household income,higher costs of imported drugs, anddecreasing nongovernmental organization(NGO) activities. Anecdotal evidence sug-gests that this is happening everywhere in

East Asia. But whether social indicatorswill worsen is not straightforward. First,indicators such as education or the infantmortality rate are determined by past

106

Poverty will increase fasteramong Indonesian workersin trade and manufacturing,but rural poverty willincrease mostFigure 2-26a Projected poverty trends inIndonesia

2

6

10

14

18

22

Percent

Rural

Urban

All Indonesia

Jan. 1,1996

Mar. 31,1997

Mar. 31,1998

Mar. 31,1999

Mar. 31,2000

2

4

6

8

10

12

14

16

18

20

22

Percent

Agriculture

Trade

Manufacturing

Services

Jan. 1,1996

Mar. 31,1997

Mar. 31,1998

Mar. 31,1999

Mar. 31,2000

Figure 2-26b Projected poverty trendsin Indonesia, by sector

Lower growth anddistribution wouldeliminate the effects of10 years of growthFigure 2-25 Poverty projections for theyear 2000 with no change in inequalityand with a 10 percent increase ininequality

Indonesia Malaysia Philippines Thailand

0

5

10

15

20

25

30

PercentNo change in inequality 10% increase in Gini index

Note: Poverty lines ($/day): Indonesia (1.0); Malaysia(2.0); Philippines (1.0); Thailand (2.0).Source: Ravallion and Chen 1996.

Source: World Bank staff estimates.


investments in physical and human capital.Second, increased public spending in socialsectors and special programs for the mostvulnerable can counteract the negativeimpact of a drop in personal income. InLatin America, infant mortality, lifeexpectancy, and primary enrollment contin-ued to improve in the 1980s, even aspoverty and inequality increased. It is con-ceivable that they would have improvedeven more without the economic crisis. Butin Chile, where the crisis was especiallysevere, reduction in infant mortality accel-erated because of public actions to safe-guard the poorest. By contrast, in 1986–87,education spending in Indonesia was notmaintained: enrollment rates fell dramati-cally for the poor and it took almost adecade to return to previous levels (Lustig1995; Pritchett 1998). Preliminary reportsfrom Thailand and Indonesia indicate thata growing number of children are notreturning for the new school year in 1998as parents lose jobs and cannot affordschool and transportation fees. The healthstatus of women and children is alsoreported to be deteriorating, as medicineand preventive care become more expen-sive. Reported increases in child labor,prostitution, and domestic violence in allthe countries affected by the crisis may havelong-lasting effects on the social fabric.

Responding to social impact ofthe crisisThe policies and instruments selected forcrisis management can have a considerableimpact on the welfare of the poor. The con-sideration of policies to mitigate adversesocial impacts in the immediate aftermathof macroeconomic crises needs to be an up-

front consideration in crisis management.Counterfactual experiments in a study ofstabilization in seven countries in the 1980s(Ecuador, Indonesia, Malaysia, Chile, Côted’Ivoire, Ghana, and Morocco) suggest thatsome stabilization measures have highersocial costs than others. For example, amoderate reduction in the pay of publicemployees has less damaging effects onpoverty and inequality than laying off pub-lic employees or increasing indirect taxa-tion. Public expenditure can be reducedwhile still maintaining public spending foressential public services (Bourguignon andMorrison 1992; Bourguignon, de Melo, andMorrison 1991). Moreover, while socialsafety nets are not a substitute for good

macroeconomic policies, they mitigate thesocial costs of crises, especially in the shortrun. The advantages, however, need to bebalanced against potential costs—for exam-ple, diversion of resources from investment,thereby compromising future growth,reducing work incentives, or displacing pri-vate transfers. There is thus an importantdistinction between short-term crisis man-agement tasks and those in the longer term.During a crisis, some immediate prioritiesarise, which are described below. In thelonger-term, these should be replaced byother instruments, such as fostering fastergrowth, human capital improvements of thepoor, and building more flexible labor mar-kets. Going further, an important lesson

107

The policies and instruments selectedfor crisis management can have aconsiderable impact on the welfare of the poor.


from this crisis is the need for ex-ante socialsafety nets, that ensure appropriate, tempo-rary, responses during crises to protect thepoor. The appropriate instruments anddesign of such social safety nets will varywith the economic, institutional, and socialstructures and fiscal capacity of countries atdifferent levels of development—a subjecton which work is beginning and wheremore effort will be needed in the future.

Priority actions to protect the poor dur-ing crises47 include:• Generating income for the poor

through direct cash transfers and pub-lic works for the unemployed. Cashtransfers are effective when the poorare easily identified or when a self-selection method is employed, forexample, by linking transfers to childnutrition programs or some workrequirement. If well designed, publicworks are one of the most efficientways of reaching the poor during eco-nomic downturns. Experience (Chile inthe early 1980s, for example) showsthat public works can achieve the mul-tiple objectives of providing incomesupport, creating local infrastructure,and keeping people in their villages—and the programs can be phased out asthe economy strengthens. Their effec-tiveness and cost, however, depend onhow they are designed, monitored, andevaluated. In particular, the wage ratemust be low enough not to attract thenonpoor, and the share of wages intotal costs must be high—internationalexperience suggests between 50 percentand 66 percent (World Bank 1997).

Project design and implementationdifficulties can make a huge difference

in the cost of public works programsand their impact on poverty. Ravallion(1998) estimates that the cost of a $1gain to the poor of a typical workfareprogram is $2.50 in both middle-income and low-income countries. InIndonesia, transferring a dollar to thepoorest 15 percent of the population isestimated to cost between $2.08 and$3.81. This range reflects differences insuch factors as the administrative leak-age of funds to officials and mistarget-ing of wage opportunities to the non-poor. Lack of coordination andmonitoring and corruption could pushcosts even higher.

In the present crisis, public workprograms are expected to create about50,000 temporary jobs in the publicsector in Korea. In Thailand, they areexpected to create about one millionperson/months of employment andover 900,000 person/ months of train-ing in 1998. Another 700,000 person/months of employment will be createdover the following three years. InIndonesia a large share of the budgethas been devoted to employment cre-ation, through funds provided to localauthorit ies and communities orthrough line agencies (including min-istries). Their impact, however, willdepend on design and implementation,which are highly uncertain: thus it isestimated that employment generationcould reduce the unemployment rate byanywhere between 1 percent and 5 per-cent (Pritchett 1998).

• Ensuring food supplies through directtransfers and price subsidization ofessential commodities. Direct food dis-

108


tribution is the last resort for reachingareas where people are starving becauseof drought or the collapse of local mar-kets. Where markets function, a moreefficient transitional way to sustain foodconsumption by the poor is to subsidizeessential commodities. In Indonesia,where the price of rice has practicallydoubled since June 1997, the govern-ment has brought in huge quantities ofimported rice to stabilize the price. Thisis an effective way of sustaining the poorin the short run, since the poorest 20percent of the population spend 25 per-cent of their total budget on rice, com-pared to 5 percent for the richest 10 per-cent. Price subsidies are a tax onfarmers, however, and in the medium-term may severely distort price incen-tives in agriculture. Rice subsidies are anexpensive way of transferring resourcesto the poor. In Indonesia, the cost oftransferring $1 to the poorest 15 percenthas been estimated at $8.20. The samenutritional adequacy could be main-tained—and targeting the poor could beimproved—by subsidizing lower-qualityrice. This could reduce the cost of trans-ferring $1 to the poor to about $3.60(World Bank 1998b).48

• Preserving the human capital of thepoor. An important consideration ismaintaining basic health care servicesfor a population whose income hasfallen and whose health status has suf-fered because of worsening nutrition,homelessness, and other factors. Thehealth and nutrition of pregnantwomen and girls is particularly at risk.Possible remedies include waiving usercharges for the poor and extending

health care to workers dismissed fromtheir jobs. Studies show that in mostcountries in East Asia public spendingin primary education is pro-poor(though this is less true for secondaryeducation).49 Maintaining or increasing

real public spending to keep children inschool should therefore be a preferredpolicy choice. In Indonesia the govern-ment launched a “Stay in School Cam-paign” culminating in the “NationalSchool Enrollment Week” beforeschool began in July. Other interven-tions include block grants targeted tothe poorest 40 percent of primary andjunior secondary schools, to compen-sate for the increased costs due to thecrisis, and a targeted scholarship pro-gram for the poorest primary andjunior secondary-school children. Realspending in education and health willbe increased (relative to 1997) in bothThailand and Korea. Changes in legis-lation allow laid-off workers to con-tinue their health insurance coveragefor twelve months, instead of six.

• Provide training opportunities, jobsearch, and other assistance to theunemployed, who may not benefit frompublic work programs. In Korea, thesocial crisis is primarily one of unem-ployment, not poverty. The 1998 bud-get includes measures to strengthenpublic employment services and job cre-

109

Ensuring basic food supplies andmaintaining basic health care are vitalwhere people’s incomes have fallensharply.


110

ation. A new program will fund startuploans for some of the unemployed. Newtraining and redeployment policies willrelax restrictions on private employ-ment and personnel leasing services.Unemployment insurance coverage hasbeen extended, and minimum unem-ployment benefits raised. 50 In Thailand,the Labor Protection Act of January1998 raised severance payments andmandated the creation of providentfunds for the purpose.

Notes1. Thailand’s fiscal position moved to a small 0.9

percent of GDP deficit in 1997, as the economy slowed.2. Chang and Velasco (1998) and Radelet and

Sachs (1998a and b) discuss international illiquidity asa source of financial crises.

3. Data on short-term foreign debt are notori-ously imperfect. Figures 2-2 and 2-3 use Bank forInternational Settlements (BIS) data on short-termloans by banks resident in its member countries (BIS1998) and various earlier issues). By definition, thesedata provide only incomplete estimates of total quanti-ties of short-term debt.

4. The Philippines’ extended arrangement withthe International Monetary Fund was expiring at thetime the Thailand crisis started; this was extended andaugmented as a precautionary measure, and a newtwo-year arrangement was subsequently approved andagreed to in March 1998.

5. The role of macroeconomic policies in manag-ing capital inflows and in the buildup of vulnerabilitiesin East Asia is discussed further in chapter 3 of thisreport, as well as in World Bank (1998a) and in Bhat-tacharya and others (1998). Recent empirical studiesfind a significant part of the reduction in spreads onemerging market debt unexplained by economic fun-damentals include Eichengreen and Mody (1998) andCline and Barnes (1997).

6. Earlier banking crises were also experienced inMalaysia in 1985–88 and in Thailand in 1983–87.

7. Discussion of financial structure and perfor-mance variables in this and subsequent paragraphs

draw on Claessens, Djankov, and Lang (1998), whodraw on Worldscope and Extel databases of 5550 EastAsian firms in nine countries.

8. The ratio of gross domestic investment to GDPin Indonesia, Korea, Malaysia, and Thailand in1991–96 averaged 5–10 percentage points higher thanin the preceding 15 years.

9. Studies of total factor productivity (TFP), ameasure of aggregate economic productivity, also donot provide clear-cut evidence of a secular deteriorationin economic performance in the 1990s. Sarel (1997)estimates that in most cases TFP growth in the ASEANcountries in 1991–96 actually increased relative to thewhole period 1978–96. Collins and Bosworth (1996)also find average East Asian annual TFP growth(excluding China) to have increased from 0.5 percentin 1973–84 to 1.6 percent in 1984–94, with most indi-vidual countries also experiencing increases.

10. For example in Indonesia, which had sufferedserious banking difficulties in the early 1990s, in thewake of a credit boom in the second half of the 1980s,nonperforming loans were estimated to have fallenfrom 12 percent in 1994 to 8.8 percent in 1996.

11. Other evidence on financial liberalization,credit booms and banking crises is presented in Gavinand Hausmann (1996) and Kaminsky (1998).

12. Dasgupta and Imai (1998).13. Eichengreen, Rose, and Wyplosz (1996a and

b) find transmission more closely associated with tradelinks than with macroeconomic fundamentals. Onestudy that does include the 1997 East Asian crisis in itsdata set is Glick and Rose (1998). It finds that in the1997 crisis, as well as in earlier episodes, the probabil-ity of crisis transmission is significantly increased bythe extent of trade links between transmitter andtransmittee, while indicators of macroeconomic fun-damentals are generally insignificant.

14. As in standard gravity models of trade.15. The empirical literature on the determinants

of investment in developing countries (surveyed inRama 1993) confirms that the cost of capital has agenerally significant inverse relation with investment.In addition, those studies that took into account theavailability of foreign exchange always found it tohave a strong, positive association with investment.

16. Expressed as a percent of average quarterlyGDP in 1994.


17. In particular, borrowers, having less to lose interms of net worth, have an incentive to proposeriskier projects for funding (adverse selection) and totake greater risks after obtaining it (moral hazard).Stiglitz and Weiss (1981), Bernanke and Gertler(1995), and Mishkin (1996) discuss the role of asym-metric information in financial markets.

18. As described by Kiyotaki and Moore (1997)and Edison and others (1998).

19. As discussed for example by Dixit andPindyck (1994).

20. Ding, Domaç, and Ferri (1998) and Dwor-Fre-caut (1998) provide a detailed review of the evidence.

21. Bernanke and Gertler (1995), Bernanke andBlinder (1993).

22. Compared to a 7 percentage point declineestimated in the Thai Government Memorandum onEconomic Policies, 5/28/98.

23. Studies of the determinants of consumptionin developing countries, including those measuring theincidence of liquidity constraints, are reviewed inAgenor and Montiel (1996) and Schmidt-Hebbel andothers (1992).

24. See also footnote 4.25. Given a smaller bad debts problem, the

Philippines is currently not expected to face large bankrestructuring costs.

26. The primary balance is the fiscal balancebefore deducting interest on government debt. Whiletable 2-10 draws on private estimates of costs of bankrestructuring, these are not the only ones available.For instance, Indonesia’s Supplementary Memoran-dum of Economic and Financial Policies (MEFP) ofApril 10, 1998, released to the public, contained aninitial estimate of costs of bank restructuring of 15percent of GDP, subsequently noted to have risen fur-ther in the Supplementary MEFP of June 24, 1998.

27. The primary balance required to stabilize agiven debt to GDP ratio is that ratio multiplied by thegap between the real interest rate and the growth rateof the economy. Daniel 1997 and sources cited thereinprovide additional details.

28. While there is no reason to believe that a tem-porary rise in interest rates will lead to a permanentstrengthening of the exchange rate, the argument isthat it may help stability in the short term.

29. See Goldfajn and Gupta (1998), Kraay(1998), and Stiglitz (1998) for discussion of these

points. The argument in the text is set out in the stan-dard monetary model of exchange rates:

m - p = αy - βi (money demand)i = i* + e(t+1) - e(t) + θ(t) (modified uncoveredinterest parity)p = p* + e + ϖ (purchasing power parity)

where variables (except i and i*) are measured in logs,θ(t) is the composite risk premium, and ϖ is a realexchange rate shock. The solution for the nominalexchange rate is given by:

where i*, p* and ϖ are set equal to 0. A monetarytightening (dm < 0) would lead to a depreciation of theexchange rate (de > 0) if it is also expected to lead to asufficiently large decline in output (y) or increase inthe risk premium (θ), the latter reflecting a greater riskof default, corporate bankruptcy, and the like. Themonetary model of exchange rates was introduced byFrenkel (1976) and Mussa (1976), and is expoundedin Obstfeldt and Rogoff (1996, 526–8). The model asset out here follows Ghosh (1998).

30. Following Kaminsky and Schmukler 1998.31. Indeed several World Bank studies of struc-

tural adjustment lending concluded that without suchcommitment external lending can undermine ratherthan fortify reform efforts (World Bank 1990a, 1992).

32. Rodrik (1994) argues that structural reformpolicies like trade liberalization generate a high degreeof income redistribution relative to net social benefit,and are therefore inherently more contentious thanpolicies (ending very high inflation, for example) witha lower “political cost-benefit ratio.”

33. Described by Caprio and Klingebiel (1996b)as “perhaps the most pernicious type of insolvency.”

34. Modes of bank recapitalization and thetradeoffs between objectives and constraints in finan-cial restructuring are discussed by Claessens andKlingebiel (1998), Rojas-Suárez and Weisbrod (1996)and Daniel (1997), among others.

35. Depositors may also lose their incentive tomonitor the activity of bank owners and managers,thereby increasing the moral hazard problem inbanking.

36. The other two being, first, to ensure that theparties who benefited from risk taking bear a large

111

e(t) 11

01

(m j - ay j + j) (1)

j

j

te

te

te

= +

+

ββ

β

θ

=∑

+ + +

∞


portion of the cost of restructuring and, second, toprevent problem institutions from continuing toexpand credit.

37. Such delays occur because of disagreementamong the parties about contingencies unforeseen ininitial contracts, the costs of renegotiating contracts,and the existence of asymmetric information about thestate of the business and its assets, which may encour-age the parties to adopt time-consuming bargainingstrategies. The costs of delay and the economic ratio-nale for bankruptcy law and orderly debt workoutsare discussed in Eichengreen and Portes (1995), Cor-nelli and Felli (1995), Krugman (1988), Sachs (1989),and Williamson (1988).

38. Asian bankruptcy law systems before the cri-sis are discussed in Gamble (1998); see also The Econ-omist (1998).

39. The incidence of poverty is the share of acountry’s population whose consumption level fallsbelow a poverty level—a threshold level of consump-tion based on minimum food and nonfood require-ments. Poverty lines vary among countries, makinginternational comparisons difficult. For the latter pur-pose, a ‘dollar a day’ poverty line is taken, defining aninternationally comparable minimum level of privateconsumption per person—usually $1 a day, measuredin purchasing power parity–adjusted 1985 dollars.

40. The link between growth and poverty is wellestablished in the literature (World Bank (1990–95),Ravallion (1995). Estimates of the responsiveness ofpoverty to growth in mean incomes (assuming distribu-tional neutrality), indicate that in the late 1980s a 1 per-cent increase in growth was associated with a 3.5 per-cent reduction in poverty incidence in Malaysia andThailand, 2.8 percent in Indonesia, less than 2 percent inmost of Sub-Saharan Africa, and less than 1 percent inBrazil. See Demery, Sen, and Vishwanath (1995), Ahujaet al (1997), Watkins (1998), and World Bank (1993).

41. Morley (1994). The association between reces-sion and higher inequality may not be so strong in devel-oped countries where wages are sticky and firms maystockpile labor if the recession is perceived to be transi-tory, and where there are unemployment insurance andother benefits to protect the incomes of the poor.

42. Deininger and Squire (1997) find a strongand significant relationship between initial inequalityof assets (measured by land) and the elasticity ofpoverty with respect to growth. However, they could

not confirm the hypothesis of a negative relationshipbetween initial income inequality and subsequentgrowth, as reported in Persson and Tabellini (1994)and Alesina and Rodrik (1994). Addison and Demery(1994) discuss the impact of factor market rigidities inthe formal sector. The differential impacts of policiesare discussed in Bruno, Ravallion, and Squire (1996).

43. It is important to note, however, that adjust-ment remained incomplete in many countries. WorldBank (1994) found that only 6 of 26 countries inAfrica achieved a significant improvement in policiesduring the 1980s. On the positive side, Demery andSquire (1996) found that poverty declined in fiveAfrican countries (out of the six for which householdconsumption data were available) where there wasan improvement in an index measuring policyperformance.

44. Estimates of employment and unemploymentare only indicative, in part because they are derivedfrom past elasticities of employment to output, whichmay not hold in crisis situations when there are largeand abrupt changes. For Thailand the 0.74 employmentelasticity prevailing in 1986–94 is used. For Indonesiaan elasticity of 0.29 from the period 1985–95 is used.Official unemployment forecasts for Korea are used.(Projections of output are reported in footnote 54.) Inaddition, as is well known, employment and unemploy-ment data in developing countries with large rural andinformal sectors are not very reliable.

45. In Indonesia, for example, 37 percent ofthose employed in urban areas and 50 percent in ruralareas work less than a 35 hour week.

46. Poverty estimates in this section (Chen andRavallion [1998]) are made using the a dollar a daypoverty line, in purchasing power parity, to allow com-parison across countries. All poverty lines are arbitrary,and alternative definitions will yield different estimates.These poverty forecasts use data on the distribution ofconsumption (or income) from the most recent house-hold surveys. They represent a first, quick approxima-tion based on the assumption that household consump-tion falls at the same rate as GDP per capita. Theyconsider neither differences in the impact on subgroups(some of the poor may lose while others gain), nor thedifferential effects of different patterns of relative pricechanges, access to credit, sectoral spending, and so on.The assumed GDP growth rates for 1998–2000 are:Indonesia, –15 percent, –2 percent, 2 percent; Thai-

112


land, –7.2 percent, –0.2 percent, 3 percent; Philippines,1 percent, 3 percent, 4.5 percent; Malaysia, –2.6 per-cent, –1 percent, 3 percent. The exercise was not per-formed for Korea because of a lack of recent householdsurvey information. (However, poverty incidence inKorea was already as low as 13 percent in the early1980s. At the end of the 1980s Korea also had one ofthe most equal income and land distributions.)

47. Measures to protect the poor are being finan-cially supported by the international community. Dur-ing 1998 specific interventions by the World Bank toaddress the social crisis have included, for Indonesia:two rural development projects to promote incomegeneration activities in some of the poorest areas ($37million); the Kecamatan Development Project target-ing investment priorities in the poorest subdistricts($225 million); the West Java Basic Education projectproviding basic education improvement programs($104 million); the Fifth District Health project ($54million) to improve the health status of rural and poorpopulations; the Early Child Development project($21.5 million) to protect educational needs of thepoorest children. Moreover, part of a $1 billion policyreform support loan will help Indonesia to financeimported food and drugs, and subsidize the price ofthese foods and provide intensive job creationschemes. For Thailand, the World Bank has approveda $300 million Social Investment project to establishsafety nets and fund job creation schemes, low-incomehealth insurance schemes, and training for the unem-ployed. A $300 million economic recovery and socialsector loan will help the government of Malaysiamaintain spending on social sectors. In Korea a $2 bil-lion Structural Adjustment Loan will, among otherobjectives, strengthen social safety nets.

48. In the short run the cost of subsidizing therice is simply the fiscal cost of the subsidy and itsadministration. It is assumed that there are no pro-ducer income losses.

49. Particularly in Indonesia; see World Bank(1993b).

50. Under the “Basic Employment Act” (Febru-ary 1998), the ministry of labor is authorized to imple-ment unemployment measures; under the “ManpowerLeasing Act” (February 1998) manpower leasing ser-vices are introduced for selected jobs that require pro-fessional knowledge and skills. The “Wage BondGuarantee Fund Act” (February 1998) entitles those

laid off from a bankrupt firm with more than fiveworkers to a sum equal to three months’ pay from astate-managed compensation fund from July 1; andthe “Labor Welfare Fund for Small and MediumEnterprises Act” (February 1998) extends the use ofthe Labor Welfare Fund to cover school and medicalcare expenses. The “Employment Insurance Act” (Feb-ruary 1998) reduces the minimum contribution periodfrom 12 to 6 months temporarily until June 30, 1999;increases minimum duration of benefit period to 60days and minimum amount of Job Search Allowancefrom 50 to 70 percent of minimum wage.

ReferencesAddison, Tony, and Lionel Demery. 1994. “The

Poverty Effects of Adjustment with Labor MarketImperfections.” In Susan Horton, Ravi Kanbur,and Dipak Mazumdar, eds., Labor Markets in anEra of Adjustment. EDI Development Studies.Washington, DC: World Bank.

Agenor, Pierre-Richard, and Peter J. Montiel. 1996.Development Macroeconomics. Princeton, NJ:Princeton University Press.

Aggarwal, Vinod. 1998. “Exorcising Asian Debt:Lessons from Latin American Rollovers, Workoutsand Writedowns.” Paper presented at the WorldBank seminar on the Aftermath of the East AsianCrisis. Washington, DC.

Ahuja, Vinod, Beru Bidani, Francisco Ferreira, andMichael Walton. 1997. “Everyone’s Miracle?Revisiting Poverty and Inequality in East Asia.”Washington, DC: World Bank.

Alderman, Harold, and Christina H. Paxson. 1994.“Do the Poor Insure? A Synthesis of the Literatureon Risk and Consumption in Developing Coun-tries.” In E. Bacha, ed., Economics in a ChangingWorld. Proceedings of the Tenth World Congressof the International Economic Association.4:48–78. London: Macmillan Press.

Alesina, Alberto, and Dani Rodrik. 1994. “Distribu-tive Politics and Economic Growth.” QuarterlyJournal of Economics. 109:465–90. (May).

Asian Development Bank and World Bank. 1998. Man-aging Global Financial Integration in Asia: Emerg-ing Lessons and Prospective Challenges. OverviewVolume. Proceedings of an ADB and World BankSenior Policy Seminar. Manila, Philippines. (March).

113


114

Baer, Herbert, and Daniela Klingebiel. 1994. “Sys-temic Risk when Depositors Bear Losses: Five CaseStudies.” In George G. Kaufman, ed., Banking,Financial Markets and Systemic Risk. Greenwich,CT: JAI Press.

Bank for International Settlements. 1998. The Matu-rity, Sectoral and Nationality Distribution of Inter-national Bank Lending. (January).

Bernanke, Ben, and Alan Blinder. 1993. “The FederalFunds Rate and the Channels of Monetary Trans-mission.” American Economic Review. (September).

Bernanke, Ben, and Mark Gertler. 1995. “Inside theBlack Box—The Credit Channel of Monetary Pol-icy Transmission.” Journal of Economic Perspec-tives 9(4).

Bhattacharya, Amar, Stijn Claessens, Swati Ghosh,Leonardo Hernandez, and Pedro Alba. 1998.“Volatility and Contagion in a Financially Inte-grated World: Lessons from East Asia’s RecentExperience.” Paper presented at CEPR/World BankConference on ‘Financial Crises: Contagion andMarket Volatility.’ London. (May).

Bourguignon, François, and Christian Morrisson.1992. Adjustment and Equity in DevelopingCountries. Paris: OECD Development Centre.

Bourguignon, François, Jaime de Melo, and ChristianMorrisson. 1991. “Poverty and Income Distribu-tion during Adjustment: Issues and Evidence fromthe OECD Project.” World Development 19:1485–1508.

Bourguignon François, Sylvie Lambert, and AkikoSuwa-Eisenmann. 1996. “Distribution of ExportRisk in a Developing Country.” Discussion Paper.London: Centre for Economic Policy Research.1482:1–35 (September).

Bruno, Michael, Martin Ravallion, and Lyn Squire.1996. “Equity and Growth in Developing Coun-tries.” Policy Research Working Paper 1563.Washington, DC: World Bank.

Calomiris, Charles W., and Berry Wilson. 1998. “BankCapital and Portfolio Management: The 1930s‘Capital Crunch’ and Scramble to Shed Risk.”(mimeo).

Calvo, Guillermo, and Enrique G. Mendoza. 1996.“Mexico’s Balance of Payments Crisis: A Chronicleof a Death Foretold.” Journal of InternationalEconomics 41:235–64 (November).

Caprio, Gerard, and Daniela Klingebiel. 1996a. “BankInsolvencies. Cross-country Experience.” PolicyResearch Working Paper 1620. Washington, DC:World Bank.

____. 1996b. “Bank Insolvency: Bad Luck, Bad Policy,or Bad Banking?” World Bank Economic Review.Annual Bank Conference on Development Eco-nomics. Washington, DC: World Bank.

Chang, Roberto, and Andres Velasco. 1998. “TheAsian Financial Crisis in Perspective.” (mimeo)Background Paper prepared for Global EconomicProspects and the Developing Countries. 1998.Washington, DC: World Bank.

Chen, Shaohua, and Martin Ravallion. 1998.“Poverty Rates in East Asia. (mimeo) Washington,DC: World Bank.

Claessens, Stijn, and Daniela Klingebiel. 1998. “Sys-temic Bank Restructuring: Scenarios, Options andQuestions.” (mimeo) Washington, DC: World Bank.

Claessens, Stijn, Simeon Djankov, and Larry Lang.1998. “East Asian Corporates: Growth, Financingand Risks Over the Last Decade.” (mimeo) Wash-ington, DC: World Bank.

Cline, William, and Kevin Barnes. 1997. “Spreads andRisk in Emerging Market Lending.” ResearchPaper 97/1. Washington, DC: Institute for Interna-tional Finance.

Collins, Susan M., and Barry P. Bosworth. 1996.“Economic Growth in East Asia: Accumulationversus Assimilation.” Brookings Papers on Eco-nomic Activity 2:1996.

Cornia, Giovanni Andrea, Richard Jolly, and FrancesStewart. 1987. Adjustment with a Human Face.New York: UNICEF.

Cornelli, Francesca, and Leonardo Felli. 1995. “TheTheory of Bankruptcy and Mechanism Design.” InEichengreen and Portes, eds., Crisis? What Crisis?Orderly Workouts for Sovereign Debtors. London:Centre for Economic Policy Research.

Daniel, James A. 1997. “Fiscal Aspects of BankRestructuring.” Working Paper WP/97/52. Wash-ington, DC: International Monetary Fund.

Daniel, James A., Jeffrey M. Davis, and Andrew M.Wolfe. 1997. “Fiscal Accounting of Bank Restruc-turing.” Paper PPAA/97/5. Washington, DC: Inter-national Monetary Fund.

Dasgupta, Dipak, and Kumiko Imai. 1998. “The EastAsian Crisis: Understanding the Causes of Export


115

Slowdown, and the Prospects for Export Recov-ery.” Washington, DC: Development ProspectsGroup, World Bank.

Demery, Lionel, and Lyn Squire. 1996. “Macroeco-nomic Adjustment and Poverty in Africa: an Emerg-ing Picture.” World Bank Research Observer.11:39–59 (February). Washington, DC: World Bank.

Demery, Lionel, Binayak Sen, and Tara Vishwanath.1995. “Poverty, Inequality and Growth.” ESP Dis-cussion Paper. Washington, DC: World Bank.

Deininger, Klaus, and Lyn Squire. 1997. “New Waysof Looking at Old Issues: Inequality and Growth.”(mimeo) Washington, DC: World Bank.

Demirgüç-Kunt, Asli, and Enrica Detragiache. 1997.“The Determinants of Banking Crises: Evidencefrom Industrial and Developing Countries.” PolicyResearch Working Paper 1828. Washington, DC:World Bank Development Research Group andInternational Monetary Fund Research Depart-ment, World Bank and International MonetaryFund. (September).

____. 1998. “Financial Liberalization and FinancialFragility.” Paper prepared for Annual World BankConference on Development Economics. Washing-ton, DC. April 1998.

Diaz-Alejandro, Carlos. 1985. “Goodbye FinancialRepression, Hello Financial Crash.” Journal ofDevelopment Economics. September/October.

Ding, Wei, Ilker Domaç, and Giovanni Ferri. 1998. “IsThere a Credit Crunch in East Asia?” PolicyResearch Working Paper 1959. Washington, DC:World Bank.

Dixit, Avinash K., and Robert S. Pindyck. 1994.Investment under Uncertainty. Princeton, NJ:Princeton University Press.

Dwor-Frecaut, Dominique. 1998. “The Credit Crunchin East Asia—Policy Options and Tradeoffs.”(mimeo) Washington, DC: World Bank.

ECLAC (United Nations. Economic Commission forLatin America and the Caribbean). Economic Surveyof Latin America and the Caribbean. Various issues.

The Economist. 1998. “Finance and Economics: TheLiving Dead.” (January 24).

Edison, Hali J., Pongsak Luangaram, and MarcusMiller. 1998. “Asset Bubbles, Domino Effects and‘Lifeboats’: Elements of the East Asian Crisis.”Discussion Paper 1866. London: Centre for Eco-nomic Policy Research. (April).

Eichengreen, Barry, and Ashoka Mody. 1998. “WhatExplains Changing Spreads on Emerging MarketDebt: Fundamentals or Market Sentiment?” Work-ing Paper 6408. Cambridge, MA: National Bureauof Economic Research.

Eichengreen, Barry, and Richard Portes. 1995. Crisis?What Crisis? Orderly Workouts for SovereignDebtors. London: Centre for Economic PolicyResearch.

Eichengreen, Barry, Andrew Rose, and CharlesWyplosz. 1996a. “Contagious Currency Crises.”Working Paper 5681. Cambridge, MA: NationalBureau of Economic Research.

____. 1996b. “Contagious Currency Crises: FirstTests.” Scandinavian Journal of Economics 98(4).

Frankel, Jeffrey A., and Andrew K Rose. 1996. “Cur-rency Crashes in Developing Countries: An Empir-ical Treatment.” International Financial DiscussionPaper 534. Federal Reserve.

Frenkel, Jacob. 1976. “A Monetary Approach to theExchange Rate: Doctrinal Aspects and EmpiricalEvidence.” Scandinavian Journal of Economics 78(May).

Gamble, William. 1998. “Restructuring In Asia.”Emerging Markets Strategies Company. (http://www.Xensei.com/users/gamblet).

Gavin, Michael, and Ricardo Hausmann. 1996. “TheRoots of Banking Crises: The Macroeconomic Con-text.” In Banking Crises in Latin America. Wash-ington, DC: Inter-American Development Bank.

Ghosh, Swati. 1998. “Note on Credit Crunch Analysisfor Korea.” (mimeo) Washington, DC: World Bank.

Glick, Reuven, and Andrew Rose. 1998. “Contagionand Trade: Why are Currency Crises Regional?”Paper presented at CEPR/World Bank Conferenceon “Financial Crises: Contagion and MarketVolatility.” London. May 8–9.

Goldfajn, Ilan, and Poonam Gupta. 1998. “Over-shootings and Reversals: the Role of MonetaryPolicy.” (unpublished) Washington, DC: Interna-tional Monetary Fund (August).

Grootaert, Christiaan. 1994. “Poverty and BasicNeeds Fulfillment in Africa during structuralchange: evidence from Côte d’Ivoire.” WorldDevelopment 22:1521–34. (October).

Haggard, Stephan, Jean-Dominique Lafay, and Christ-ian Morrisson. 1995. The Political Feasibility ofAdjustment in Developing Countries. Paris: OECDDevelopment Centre.


Haggard, Stephan, and Steven B. Webb. 1993. “WhatDo We Know About the Political Economy of Eco-nomic Policy Reform?” World Bank ResearchObserver. Washington, DC: World Bank. (July).

Hassler, John. 1996. Risk and Consumption. Stock-holm: Institute for International Economic Studies.

Indonesia, Government of. 1998. Memorandum ofEconomic and Financial Policies, 1/15/98. Supple-mentary Memorandum of Economic and FinancialPolicies. 4/10/98. Memorandum of Economic andFinancial Policies, 6/24/98.

International Monetary Fund. 1997. InternationalCapital Markets. Developments, Prospects, andKey Policy Issues. Washington, DC (November).

____. 1998. Annual Report. Washington, DC.J.P. Morgan. 1998a. “Asian Financial Markets, First

Quarter 1998.” Singapore: Morgan Trust Com-pany. Economic Research.

____. 1998b. “Asian Financial Markets, Second Quar-ter 1998”. April 24, 1998. Singapore: MorganTrust Company. Economic Research.

Jolly, Richard, and Rolph van der Hoeven. 1991.“Adjustment with a Human Face—Record andRelevance.” Special Section, World Development.19:1801–64.

Kaminsky, Graciela L. 1998. “Currency and BankingCrises: The Early Warnings of Distress.” Washing-ton, DC: Board of Governors of the FederalReserve. (March).

Kaminsky, Graciela L., and Carmen M. Reinhart.1996. The Twin Crises: the Causes of Banking andBalance of Payments Problems. Seminar Series1996–12: 1–26. Washington, DC: InternationalMonetary Fund. (March).

Kaminsky, Graciela L., and Sergio Schmukler. 1998.“The Relationship Between Interest Rates andExchange Rates in Six Asian Countries.” (mimeo).Washington, DC: World Bank.

Kawai, Masahiro. 1997. “East Asian Currency Turbu-lence: Implications of Financial System Fragility.”(mimeo) Washington, DC: World Bank.

Kiyotaki, Nobuhiro, and John Moore. 1997. “CreditCycles.” Journal of Political Economy 105:211–48. (April).

Korea, Government of. Memorandum on the Eco-nomic Program, 12/3/97. Letter of Intent,12/24/97. Memorandum on Economic Program,2/7/98. Updated Memorandum on the Economic

Program for the Second Quarterly Review, 5/2/98.Memorandum on the Economic Program, 7/24/98.

Kraay, Aart. 1998. “Do High Interest Rates DefendCurrencies Against Speculative Attacks?” (mimeo)Washington, DC: World Bank. (August).

Krugman, Paul. 1988. “Financing vs. Forgiving aDebt Overhang.” Working Paper 2486. Cam-bridge, MA: National Bureau of EconomicResearch (January).

Larrain, B. Felipe, and Rodrigo M. Vergara. 1993.“Inversion y ajuste macroeconomico: el caso deles te de As ia .” El Tr imes t re Economico60:293–351. (April–June).

Lustig, Nora. 1995. “Coping with Austerity: Povertyand inequality in Latin America,” in GerryRodgers and Rolph van der Hoeven, “The PovertyAgenda: Trends and Policy Options,” ILO.

Masson, Paul. 1998. “Contagion: Monsoonal Effects,Spillovers and Jumps between Multiple Equilib-ria.” Prepared for a CEPR/World Bank conferenceon Financial Crises: Contagion, and MarketVolatility. London. May.

Milanovic, Branko. 1995. “Poverty, Inequality, andSocial Policy in Transition Economies.” PolicyResearch Working Paper 1350. Washington, DC:World Bank.

Mishkin, Frederic S. 1991. “Anatomy of a FinancialCrisis.” Working Paper 3934. Cambridge, MA:National Bureau of Economic Research (December).

____. 1996. “Understanding Financial Crises: A Devel-oping Country Perspective.” Working Paper 5600.Cambridge, MA: National Bureau of EconomicResearch.

Morley, Samuel A. 1994. Poverty and Inequality inLatin America: Past Evidence, Future Prospects.Overseas Development Council, Washington, DC.

Mussa, Michael. 1976. “The Exchange Rate, the Bal-ance of Payments, and Monetary and Fiscal Policyunder a Regime of Controlled Floating. Scandina-vian Journal of Economics 78. (May).

Obstfeldt, Maurice. 1998. “The Global Capital Mar-ket: Benefactor or Menace?” Working Paper 6559.Cambridge, MA: National Bureau of EconomicResearch. (May).

Obstfeldt, Maurice, and Kenneth Rogoff. 1996. Foun-dations of International Macroeconomics. Cam-bridge, MA: MIT Press.

116


Papageorgiou, Demetris, Armeane Choksi, andMichael Michaely. 1990. Liberalizing ForeignTrade: Lessons of Experience in the DevelopingWorld. Washington, DC: World Bank.

Pelletier David L., E. A. Frongillo, D. G. Schroeder,and Jean-Pierre Habicht. 1995. “The Effects ofMalnutrition on Child Mortality in DevelopingCountries.” Bulletin of the World Health Organi-zation 73.

Persson, Torsten, and Guido Tabellini. 1994. “IsInequality Harmful to Growth?” American Eco-nomic Review 84(3). June.

Pritchett, Lant. 1998. “Addressing the Social Impactof the Crisis in Indonesia.” (mimeo) Washington,DC: World Bank.

Radelet, Steven, and Jeffrey D. Sachs. 1998a. “TheOnset of the East Asian Financial Crisis.” Pre-sented at NBER Currency Crisis Conference and tothe U.S. Agency for International Development.(January).

____. 1998b “The East Asian Financial Crisis: Diagno-sis, Remedies, Prospects.” Cambridge, MA: Har-vard Institute for International Development.(April 20).

Rama, Martin. 1993. “Empirical Investment Equa-tions for Developing Countries.” Policy ResearchWorking Paper 563. Washington, DC: WorldBank.

Ravallion, Martin. 1995. “Growth and Poverty: Evi-dence for the Developing World.” Economics Let-ters 48.

____. 1998. “Appraising Workfare Programs.” (mimeo)Washington, DC: World Bank.

Ravallion, Martin, and Shaohua Chen. 1996. “WhatCan New Survey Data Tell Us about RecentChanges in Distribution and Poverty?” (mimeo)Washington, DC: World Bank. (October).

Rodrik, Dani. 1994. “The Rush to Free Trade in theDeveloping World: Why so Late? Why Now? Willit Last?” In Haggard and Webb, eds. Voting forReform: Democracy, Political Liberalization, andEconomic Adjustment. Washington, DC: WorldBank.

____. 1996. “Understanding Economic Policy Re-form.” Journal of Economic Literature. (March).

Rojas-Suárez, Liliana, and Steven K. Weisbrod. 1996.“The Do’s and Don’ts of Banking Crisis Manage-ment.” In Ricardo Hausmann and Liliana Rojas-

Suárez, eds., Banking Crises in Latin America.Washington, DC: Inter-American DevelopmentBank.

Romer, David. 1996. Advanced Macroeconomics.New York: McGraw-Hill Companies.

Romer, Christina. 1990. “The Great Crash and theOnset of the Great Depression.” Quarterly Journalof Economics 105 (August).

____. 1993. “The Nation in Depression.” Journal ofEconomic Perspectives 7/2. Spring.

Sachs, Jeffrey D. 1989. “The Debt Overhang of Devel-oping Countries.” In Guillermo Calvo et al., eds.Debt, Stabilization and Development. Essays inMemory of Carlos Diaz-Alejandro. Oxford, UK/New York: B. Blackwell for WIDER.

Sachs, Jeffrey D., Aaron Tornell, and Andrés Velasco.1996. “Financial Crises in Emerging Markets-theLessons from 1995.” Working Paper 5576. Cam-bridge, MA: National Bureau of EconomicResearch (May).

Sarel, Michael. 1997. “Growth and Productivity inASEAN countries.” WP/97/97. Washington, DC:International Monetary Fund.

Serven, Luis, and Andrés Solimano. 1989. PrivateInvestment and Macroeconomic Adjustment: AnOverview. Washington, DC: Country EconomicsDepartment, World Bank.

Schmidt-Hebbel, Klaus, Steven B. Webb, and Gian-carlo Corsetti. 1992. “Household Savings inDeveloping Countries: First Cross-Country Evi-dence.” World Bank Economic Review 6:3.

Sheng, Andrew. 1996. Bank Restructuring. Lessonsfrom the 1980s. Washington, DC: World Bank.

Stiglitz, Joseph E. 1998. “Knowledge for Development:Economic Science, Economic Policy and EconomicAdvice.” Paper presented at the Annual WorldBank Conference on Development Economics.

Stiglitz, Joseph E., and Andrew Weiss. 1981. “CreditRationing in Markets with Imperfect Informa-tion.” American Economic Review 71(3):393–410.(June).

Thailand, Government of. Letter of Intent, 11/25/97.Memorandum on Economic Policies, 2/24/98.Memorandum on Economic Policies, 5/26/98.

Thomas, Duncan, Victor Lavy, and John Strauss.1996. “Public Policy and Anthropometric Out-comes in the Côte d’Ivoire.” Journal of Public Eco-nomics 61: 155–92.

117


118

Watkins, Kevin. 1998. Economic Growth with Equity,

Lessons from East Asia. Oxford, UK: Oxfam

Publication.

Williamson, John. 1988. “Voluntary Approaches to

Debt Relief.” Policy Analyses in International Eco-

nomics 25. Washington, DC: Institute for Interna-

tional Economics.

Williamson, John, ed. 1994. The Political Economy of

Policy Reform. Institute for International Econom-

ics. Washington, DC.

World Bank. 1990a. Adjustment Lending Policies for

Sustainable Growth. Washington, DC.

____. 1990b, 1995. World Development Report. New

York: Oxford University Press.

____. 1992. Adjustment Lending: Private and Public

Resources for Growth. Washington, DC.

____. 1993a. The East Asian Miracle: EconomicGrowth and Public Policy. New York: OxfordUniversity Press.

____. 1993b. Investing in Health. World DevelopmentReport.

____. 1993c. Indonesia: Public Expenditures, Pricesand the Poor. Washington, DC.

____.1994. Labor Markets in an Era of Adjustment.Washington, DC.

____. 1997. Safety Net Programs and Poverty Reduc-tion: Lessons from Cross-country Experience.Washington, DC.

____. 1998a. Global Development Finance 1998.Washington, DC.

____. 1998b. Indonesia in Crisis. A MacroeconomicUpdate. Washington, DC.

____. 1998c. East Asia: The Road to Recovery. Wash-ington, DC.

Responding to the East Asian Crisis - World...

Documents

Transcript of Responding to the East Asian Crisis - World...