Asian Financial Crisis: How Women Sustained the Hidden Recovery

Chapter 1 — Before the Crisis: Growth Built on Mobile Capital

During the years before 1997, several East and Southeast Asian economies were celebrated for rapid industrialization, export growth and rising living standards. That record was real. Poverty had fallen sharply in parts of the region, education had expanded and millions of households had benefited from employment linked to manufacturing and urban growth. The crisis did not prove that all of this progress was an illusion.

It did reveal, however, that some of the financing behind the expansion was more fragile than it appeared. Banks and corporations borrowed heavily from abroad, often at short maturities and in U.S. dollars or other foreign currencies. Domestic lending then financed property, industrial expansion and corporate investment. As long as exchange rates remained stable and foreign lenders continued refinancing those obligations, the structure appeared manageable.

Why short-term foreign debt mattered

A company earning revenue in baht, rupiah or won could face debt denominated in dollars. If the local currency depreciated, the amount of local revenue needed to repay the same dollar debt rose sharply. Short maturities created another vulnerability: borrowers depended on foreign banks and investors continually renewing their loans.

This was not simply a story of governments spending beyond their means. Much of the vulnerability involved private banks, finance companies and corporations. Financial liberalization had increased access to international funding faster than regulation, risk management and disclosure had developed. Managed exchange rates also encouraged the belief that currency risk was limited.

The result was a mismatch at the center of the system: long-term domestic investments were supported by foreign funding that could leave quickly. Radelet and Sachs estimated that the swing from private capital inflows to outflows across five crisis economies exceeded $100 billion between 1996 and 1997. A funding model that had supported rapid expansion became a channel for rapid contraction (Radelet & Sachs, 1998).

Growth did not create equal protection

Strong growth had improved well-being, but formal safety nets remained uneven. Many workers depended on employers, extended families or community ties rather than comprehensive unemployment protection. Women were concentrated in particular export industries, services, family work and informal occupations. Their exposure therefore depended on which sectors contracted, whether they held formal contracts, and how household responsibilities were divided.

This distinction matters. The crisis did not meet a uniform population with equal savings, bargaining power or access to public support. It struck societies that had achieved major gains while retaining differences by gender, class, region, industry and employment status.

Chapter 2 — The Sudden Stop: From Currency Pressure to Banking Crisis

Thailand became the first major break in the regional system. After sustained pressure on its currency and financial sector, the Thai government abandoned the baht’s peg on July 2, 1997. The depreciation made foreign-currency obligations more expensive and deepened concern about banks, finance companies and corporate borrowers.

Investors then reassessed risks across the region. Funding was withdrawn not only from institutions that were already weak but also from economies now perceived as similar. Exchange rates fell in Indonesia, Malaysia, the Philippines and South Korea, although the scale and timing differed. What began as pressure on one currency became regional contagion.

The feedback loop

The mechanism was destructive because its parts reinforced one another:

1. Foreign lenders refused to renew short-term loans or withdrew capital.
2. Demand for foreign currency increased and local currencies depreciated.
3. Foreign-currency debts became much larger in local-currency terms.
4. Corporate defaults weakened banks that had financed those companies.
5. Banks restricted lending, even to businesses that might otherwise have survived.
6. Production, investment and employment contracted, further weakening borrowers.

This sequence explains why calling the episode only a “currency crisis” is insufficient. It was simultaneously a capital-flow, corporate-debt, banking and employment crisis. The International Labour Organization described exchange-rate instability, capital outflows and the liquidity crunch as pathways from financial disruption to factory closures and worker displacement (ILO, 2009).

Why confidence was not a harmless abstraction

Financial accounts can make confidence sound psychological and remote. In this crisis, confidence determined whether banks and businesses could refinance obligations due within weeks or months. When lenders treated an entire market as unsafe, selling and withdrawal could worsen the very conditions they feared.

The reversal did not distribute losses evenly. Owners lost equity, banks accumulated bad loans and governments assumed costly restructuring responsibilities. Workers lost jobs or hours. Families faced higher prices for imported food, fuel and medicine. The crisis moved downward through contracts and institutions until it reached people who had never borrowed abroad or traded a currency.

Chapter 3 — One Regional Crisis, Different National Experiences

The phrase “Asian financial crisis” can hide important national differences. The five economies most commonly examined—Thailand, Indonesia, South Korea, Malaysia and the Philippines—had different financial systems, political conditions, industrial structures and policy responses. Women’s experiences also varied within each country.

Thailand and Indonesia

Thailand was the initial center of the currency break. Property and finance-company distress fed a severe contraction, and falling urban employment affected rural areas through migration and reduced remittances. Workers returned from cities, while households adjusted spending and work arrangements.

Indonesia experienced the most profound social and political disruption. Currency depreciation, banking distress, inflation and corporate failure overlapped with drought and political instability. The price of food and medicine rose sharply. Formal nonagricultural employment contracted, while agriculture absorbed displaced workers at lower productivity. ILO data later showed that between May 1997 and May 1998, paid nonagricultural employment fell by 2.1 million. Men experienced especially large losses in manufacturing and construction; women’s losses were concentrated in manufacturing and in wholesale, retail, restaurants and hotels (ILO, 2009).

South Korea

South Korea’s crisis centered on highly indebted corporations, short-term foreign obligations and a weakened banking system. The economy contracted sharply and unemployment rose from historically low levels. The country also expanded employment insurance and active labor-market support as the crisis exposed the limitations of existing protection.

Recovery in output and exports was relatively fast, but labor restructuring left a longer legacy. Non-regular, temporary and part-time employment became more prominent. Women seeking work faced particular barriers to reemployment, even though male-dominated industries also suffered major layoffs. This is one reason a single claim that “women lost the most jobs” would be misleading: total job losses and the difficulty of returning to secure work are different measures (Tutnjevic, 2002; ILO, 2012).

Malaysia and the Philippines

Malaysia rejected part of the policy path followed elsewhere and imposed capital controls in September 1998. Those controls distinguished its response from the IMF-supported programs in Thailand, Indonesia and South Korea. The Philippines entered the crisis with its own banking and external vulnerabilities but experienced a less severe output collapse than Indonesia, Thailand or South Korea. Overseas employment and remittances also shaped how Filipino households experienced regional instability.

These contrasts prevent two analytical errors. First, the region should not be treated as one economy. Second, women should not be treated as one social group. A salaried worker in Seoul, a displaced factory worker in Bangkok, an agricultural worker in Java and a migrant worker supporting a family in the Philippines faced different risks and options.

Chapter 4 — Stabilizing Markets Without Rebuilding Every Household

Governments and international institutions responded to an emergency that threatened payment systems, foreign reserves, banks and major corporations. Policy packages included financing, higher interest rates, bank closures or recapitalization, corporate restructuring, changes in financial supervision and fiscal measures. The IMF organized large support programs for Indonesia, South Korea and Thailand.

Some initial fiscal targets were restrictive because officials feared that public deficits would deepen external pressure. As recessions became more severe than expected, targets were relaxed and greater attention was given to social spending and safety nets. It is therefore more accurate to describe an evolving and contested policy response than to suggest that every country followed one fixed austerity program.

Bank repair was necessary—but not sufficient

Closing insolvent institutions and recognizing bad loans were important to restoring a functioning financial system. Yet bank repair could not immediately restore a closed factory, replace lost household income or reverse interrupted schooling. Credit also tended to remain difficult for smaller businesses after larger institutions regained access.

High interest rates could help defend currencies and limit capital flight, but they also raised borrowing costs inside already contracting economies. Corporate restructuring could remove insolvent firms, but it also produced layoffs. Policies could therefore be necessary for stabilization while still imposing unequal short-term costs.

Different meanings of recovery

A currency can strengthen before wages recover. A bank can meet capital requirements while former employees remain in informal work. GDP can return to growth even though a family has depleted savings, sold assets or withdrawn a child from school. These outcomes operate on different timelines.

This is the central distinction of the crisis. Financial stabilization reduces the immediate danger of bank runs, currency collapse and frozen credit. Social reconstruction restores secure work, income, education, health and the capacity to plan. The first can create conditions for the second, but it does not automatically complete it.

For a wider comparison of how different financial emergencies required different institutional responses, see History of Global Financial Crises: 400-Year Timeline. The Asian episode remains distinct because volatile cross-border funding and foreign-currency exposure were central to its transmission.

Chapter 5 — How the Financial Shock Reached Families

Households did not experience the crisis as an abstract decline in investor confidence. They experienced it through lost jobs, fewer hours, lower real wages, higher prices and reduced access to services. Even families that kept an employed member could be affected when the purchasing power of wages fell or relatives returned from cities needing support.

Income, prices and essential spending

Currency depreciation raised the price of imported goods and inputs. In Indonesia, food and medicine prices rose sharply during the most severe phase of the crisis. The World Bank recorded households postponing medicines, reducing meals and struggling with school costs. These responses were not merely lower consumption in a statistical sense. They affected nutrition, health and education.

Families prioritized immediate survival. They reduced nonessential purchases, changed foods, shared housing, delayed care, used savings and sold assets. Some sent additional household members into paid work. Others moved between cities and rural communities. The specific combination depended on employment, assets, location and access to relatives or public programs (World Bank, 1998).

Formal unemployment understated the disruption

In economies without comprehensive unemployment insurance, losing a formal job does not always lead to recorded unemployment. A worker may accept fewer hours, lower wages, agricultural work or street vending because remaining without income is impossible. The result can be a large deterioration in work quality without an equivalent rise in the official unemployment rate.

Indonesia illustrates this pattern. Agriculture absorbed many displaced workers, including women, but output per worker fell. Movement into agriculture or informal activity prevented some households from losing all income, while also spreading limited earnings among more workers.

Family networks redistributed—not eliminated—losses

Extended families shared food, housing, money and care. Urban workers returned to rural homes. Relatives financed school expenses or helped watch children. These networks reduced immediate hardship for some people, but they could not create new resources at a national scale. They redistributed pressure within relationships.

When every member of a network faced the same recession, assistance could weaken. Households with few assets, limited social connections or marginalized status had less capacity to draw on informal support. Community resilience therefore depended on resources and power, not goodwill alone.

Chapter 6 — Women’s Employment: Layoffs, Informality and Added Work

The effect on women’s employment cannot be summarized by saying that women either lost jobs first or entered work to save their families. Both patterns occurred, and neither was universal.

Job losses depended heavily on industrial segregation. Where male-dominated construction or heavy industry contracted most, men could represent a larger share of layoffs. Where garments, electronics, retail, hospitality or other female-intensive sectors contracted, women were highly exposed. Employers’ assumptions about men as primary earners could also make women appear more disposable, while women seeking reemployment encountered narrower opportunities.

The added-worker response

When a principal earner lost wages or hours, another family member might seek work. Economists call this an added-worker effect. In Indonesia, labor-force participation increased as households tried to replace lost income. Women entered or expanded agricultural work, petty trade, food production, domestic service and other activities compatible with immediate need and limited capital.

Entry into work did not necessarily mean empowerment or improved earnings. A woman might work more hours while the household’s total real income continued to fall. Informal activities offered flexibility and fast entry but usually lacked contracts, benefits, predictable pay and legal protection.

Informality was not a universal safety valve

Informal work could absorb some displaced labor, yet informal sellers and producers also depended on customers whose incomes were falling. Currency depreciation raised the cost of imported materials. More people competing in the same low-capital activities could reduce earnings. Informality helped families survive, but it did not stand outside the recession.

In South Korea, formal restructuring contributed to the longer-term expansion of non-regular employment. In Indonesia, agriculture and informal activity absorbed workers at lower productivity. In Thailand, returning migrants did not always find adequate rural employment. These were different labor-market adjustments, not one regional model.

The gender lesson is therefore precise: preexisting divisions of labor shaped exposure to layoffs, access to replacement work, pay and responsibility for care. The crisis interacted with those structures rather than producing one identical outcome for all women.

Chapter 7 — Unpaid Care and Community Networks as Hidden Infrastructure

When income falls, households often replace purchased goods and services with time. Meals are prepared more cheaply at home. Illness is managed longer without professional care. Family members provide childcare, eldercare and transportation. Budgets require more comparison, improvisation and negotiation. This work is economically important even when no wage is paid.

Contemporary World Bank assessments reported that households under stress increased mothers’ workloads and sometimes withdrew children from school. Women and girls could also reduce their own food intake or absorb additional care. Evidence was incomplete and varied across communities, so these observations should not be presented as a uniform regional statistic. They nevertheless identify a pathway that conventional measures often miss (World Bank, 1998).

From household care to community support

Women also participated in neighborhood groups, religious organizations, cooperatives, savings arrangements and mutual aid. These networks helped circulate information, food, small loans, childcare and work opportunities. They mattered most where formal services were limited or slow to respond.

Community support should not be romanticized. Organizing assistance requires time, transportation, trust and administrative labor. When women perform this work without recognition, community resilience can rest on their overextension. Networks may also exclude people who lack social connections, migrant status, family support or accepted membership in the community.

Why unpaid work changes the meaning of recovery

Suppose a public clinic reduces services and a daughter provides more care to an older parent. Recorded public expenditure falls, while the required care does not disappear. It changes location—from a paid institution to unpaid household labor. If she reduces her work hours as a result, the family absorbs both the care responsibility and the loss of earnings.

This transfer is central to understanding women’s resilience during crises. The household appears to adapt, but the adaptation may consume women’s time, health and future income. Resilience that protects the system while exhausting the person providing it is incomplete.

A related HerMoneyPath article examines how this transfer operated in a different place and period: Unpaid Labor in Hard Times: Why Women Took on More at Home During the 2008 Recession. The comparison is useful, but the mechanisms should not be collapsed: the 1997–1998 Asian crisis began with cross-border capital reversal and currency-banking distress.

Chapter 8 — Why Social Recovery Lagged Behind GDP

Several crisis economies returned to positive growth relatively quickly. Exports became more competitive after depreciation, international demand remained available, financial restructuring progressed and capital eventually returned. South Korea’s macroeconomic rebound was especially rapid. Yet the recovery recorded in national accounts did not reverse every household loss.

Losses create a longer timeline

A period of unemployment can reduce lifetime earnings even after a worker finds another job. Time in informal or non-regular work may interrupt benefits, training and promotion. Savings used for food or rent are not automatically replenished when GDP rises. Assets sold during an emergency cannot always be repurchased. Interrupted education and delayed health care may have effects that appear years later.

Women could face several of these consequences at once: a job loss, added informal work, increased care and greater difficulty returning to a secure occupation. Women with fewer assets or less education had less room to wait for formal opportunities. Migrant women and women in home-based work could be undercounted in both employment and recovery statistics.

Institutional changes were uneven

The crisis did produce reforms. Banking supervision strengthened, corporate governance changed and some social-protection systems expanded. South Korea broadened employment-insurance coverage and active labor-market programs (ILO, 2012). Thailand later established unemployment insurance. Regional governments accumulated larger foreign-exchange reserves, and Asian financial cooperation eventually included the Chiang Mai Initiative.

These changes improved some forms of institutional resilience. They did not erase labor-market segmentation or guarantee that care would be distributed more equally. A stronger reserve position can help a country manage capital-flow pressure; it does not by itself provide childcare, replace lost wages or protect an informal worker.

Recovery must be measured at more than one level

A complete evaluation asks at least four questions:

• Have currencies, banks and credit markets stabilized?
• Have production and employment recovered?
• Have households restored income, assets, education and health?
• Has the unpaid work created by the crisis been reduced or merely normalized?

The answers may arrive in different years. That gap is not a statistical inconvenience. It is part of the crisis itself.

Chapter 9 — Institutional Resilience and Invisible Resilience

The Asian financial crisis changed how governments understood volatile capital flows, short-term foreign debt and financial supervision. It also encouraged larger reserve buffers and regional financial cooperation. These are lessons about institutional resilience: how states and financial systems prepare for, contain and recover from market disruption.

A second form of resilience received less attention. Families protected children, cared for relatives, shared housing, found replacement income and maintained communities. Women performed a substantial share of that work, although the form and scale differed by country and social position.

Resilience is not the same as absorbing unlimited loss

Calling women resilient can recognize skill, persistence and collective action. It can also hide an institutional failure. If stability depends on women continually adding unpaid hours, accepting insecure work and sacrificing their own consumption, the system has transferred risk rather than resolved it.

True social resilience reduces the damage people must absorb. It includes timely income support, accessible health and education, labor protection, childcare, unemployment insurance and policies designed with informal and migrant workers in view. It also requires data that distinguish outcomes by gender, industry, employment status, income and location.

The historical lesson

The crisis should not be reduced to a universal rule that all financial collapses follow the same sequence. For the broader recurring mechanisms of leverage, confidence and contraction, see Why Financial Crises Keep Coming Back: Lessons for Women. The distinctive lesson of 1997–1998 is narrower and more concrete: when economies depend on short-term international funding, a sudden reversal can turn foreign-currency debt into domestic banking and employment crises with extraordinary speed.

The social lesson follows from that mechanism. Banks and currencies may stabilize before households recover. The distance between those two moments is filled with work—some paid, much informal or unpaid. Recognizing that work changes the story from one of a successful technical rescue to a more complete account of who sustained recovery and what it cost them.