7 The Latin American Debt Crisis (1982–1989)
7.1 What Happened
7.1.1 Mexico’s Phone Call
On August 12, 1982, Mexico’s Finance Minister Jesús Silva Herzog telephoned the United States Treasury and the International Monetary Fund with a message that stopped financial markets across two continents: Mexico could no longer service its foreign debt. The announcement was not entirely unexpected among those watching the deteriorating arithmetic of Latin American borrowing, but the suddenness and the scale created immediate shock. If Mexico — Latin America’s second-largest economy — was defaulting, so might Brazil, Argentina, Venezuela, Chile, and a dozen other countries that had borrowed heavily throughout the 1970s.
The nine largest American banks held combined loans to Latin America equal to two hundred and eighty percent of their capital. A generalized Latin American default would have rendered every major American bank technically insolvent. The question of whether to call it a sovereign debt crisis or a global banking crisis depended entirely on one’s vantage point.
7.1.2 The Petrodollar Recycling Machine
The origins of the crisis lay in the petrodollar recycling boom of the 1970s. The OPEC oil price increases of 1973 and 1979 transferred enormous wealth from oil-importing economies to oil-exporting states. Saudi Arabia, Kuwait, the UAE, and other Gulf states suddenly held vast dollar surpluses that vastly exceeded their domestic investment capacity. These surpluses were deposited in Western commercial banks — primarily American banks operating through their international divisions in New York and London.
The banks, awash in deposits and under pressure to deploy capital, recycled these petrodollars into loans to Latin American sovereign borrowers. The logic seemed sound: sovereign governments do not go bankrupt, or so the banking wisdom of the era held. The loans were made at variable interest rates — tied to LIBOR, the London interbank offered rate — which meant that the borrower’s debt service costs would float with market interest rates rather than being fixed at origination.
Latin American governments borrowed for ostensibly productive purposes: industrialization programs, infrastructure development, agricultural modernization. Mexico borrowed to develop its newly discovered oil fields in the Gulf of Mexico. Brazil borrowed to build its steel industry and hydroelectric dams. Argentina borrowed to finance infrastructure. But borrowing also papered over fiscal deficits, subsidized politically connected industries, and financed imports that domestic industry could not match in price or quality. The distinction between productive and wasteful borrowing was easier to make in retrospect than in real time.
7.1.3 The Volcker Shock Arrives in Latin America
As long as interest rates remained low and commodity prices remained high, the arithmetic of Latin American debt was manageable. When both changed simultaneously, the mathematics of solvency reversed overnight.
Paul Volcker’s anti-inflation campaign — the Federal Reserve’s decision in October 1979 to prioritize the destruction of American inflation over the costs imposed on borrowers — raised US interest rates to levels not seen since the early twentieth century. The federal funds rate reached twenty percent in mid-1981. LIBOR, to which Latin American sovereign loans were pegged, rose correspondingly. A country that had borrowed at six percent found its interest payments tripling.
Simultaneously, the global recession that the Volcker shock contributed to cut commodity prices — the primary source of export earnings for Latin American debtors. Oil prices, which had driven Mexico’s borrowing confidence, began falling in 1981. Agricultural commodity prices fell. The terms of trade — the ratio of export prices to import prices — turned sharply against Latin American exporters. Debt service was consuming export revenues that were simultaneously shrinking. By 1982, forty countries had sought to reschedule their debts.
7.1.4 The Scale of Exposure
By August 1982, when Mexico made its call to Washington, the exposure of the international banking system to Latin American sovereign debt had reached levels that would have been difficult to justify to regulators or shareholders had they been transparent. The nine largest American banks — Citibank, Bank of America, Chase Manhattan, Manufacturers Hanover, Morgan Guaranty, Chemical Bank, Continental Illinois, Bankers Trust, and First National Bank of Chicago — had collectively lent two hundred and eighty percent of their capital to Latin American sovereigns. Citibank alone had lent sixty-three percent of its capital to Brazil, sixty-one percent to Mexico, and significant sums to Argentina and Venezuela.
These were not the margins of error. These were the central positions of the American banking system, held with full regulatory awareness. The exposure had accumulated gradually, each individual loan decision defensible in isolation, the aggregate becoming dangerous only when the underlying conditions changed. By 1982, the conditions had changed completely.
Brazil followed Mexico in declaring debt service difficulties within weeks. Argentina, already in political turmoil following the Falklands War, followed. By 1983, the entire structure of Latin American sovereign finance had collapsed. The question was how to resolve it — and the first several years of the answer would prove to be wrong.
7.2 The Failures
7.2.1 Three Interlocking Failures
The Latin American debt crisis exposed three interlocking structural failures that would take a decade of experimentation and pain to fully diagnose and address. Understanding all three was necessary before a genuine solution could be constructed; the crisis’s first seven years were spent addressing only the first while the second and third went untreated.
The first failure was in loan origination. Variable-rate dollar loans to sovereign borrowers whose revenues were commodity-dependent was a structural mismatch that no competent risk manager should have approved. A Latin American government’s ability to service dollar-denominated debt depended on its capacity to earn dollars — through exports priced in world commodity markets. Those commodity prices were determined by global supply and demand, not by the borrower’s policies. The interest rate on the loans was determined by US monetary policy, also outside the borrower’s control. The borrower bore the full risk of adverse movements in both variables simultaneously, with no hedge and no adjustment mechanism beyond default.
Regulators had the information to identify this mismatch. They did not do so, partly because the loans were to sovereign governments, which were assumed not to default, and partly because the loans were profitable and growing rapidly, creating institutional incentives to accommodate them rather than restrict them. The assumption that sovereign governments could not go bankrupt — taken as an axiom in international banking practice through the 1970s — was simply wrong. Mexico demonstrated that it was wrong in August 1982.
7.2.2 Austerity in a Collapsed Economy
The second structural failure was in the initial policy response. When Mexico and other debtors approached the IMF for assistance in 1982–1983, the Fund’s standard prescription was fiscal austerity: reduce government spending, raise taxes, restrict credit, correct the current account deficit. The logic was conventional: a country that has borrowed too much must tighten its belt to restore solvency.
The economics of this prescription in crisis conditions were perverse. In a severely depressed economy, fiscal austerity does not simply reduce the government deficit — it reduces GDP through the fiscal multiplier, which reduces tax revenues, which reduces the denominator and numerator of the debt-to-GDP ratio in ways that do not improve solvency. The IMF’s models assumed that each dollar of spending cuts would reduce GDP by roughly fifty cents — a multiplier of 0.5. The actual multiplier in these conditions was closer to 1.5, meaning each dollar of cuts reduced GDP by a dollar and fifty cents.
Bolivia’s experience illustrated the extremes. By 1985, Bolivian annual inflation had reached fifty thousand percent — hyperinflation triggered partly by the collapse of commodity export revenues and the government’s fiscal desperation. Peru imposed price controls and refused IMF conditionality, producing a different set of disasters. Brazil endured multiple unsuccessful stabilization plans through the decade. The 1980s became genuinely lost for economic development across the region.
7.2.3 The Debt Overhang Trap
The third structural failure was the most subtle and the last to be recognized: the debt overhang itself had become an obstacle to recovery. Countries carrying debt burdens that exceeded any plausible capacity to repay had no rational incentive to implement structural reforms that would increase economic output, because all additional output would flow to foreign creditors rather than to domestic investment or consumption.
The economic theory, developed through the 1980s by economists including Jeffrey Sachs, Paul Krugman, and Rudiger Dornbusch, demonstrated that there existed a “debt Laffer curve”: beyond a certain debt burden, reducing the debt actually increased the expected repayment to creditors, because it removed the overhang trap and allowed the debtor to grow. Banks holding Latin American loans at fifty or sixty cents on the dollar in the secondary market — where the bonds traded because institutional investors had largely given up on full repayment — had already implicitly acknowledged this arithmetic. The challenge was to translate the secondary market’s price discovery into an official debt reduction that would end the crisis.
7.3 The Response
7.3.1 Extend and Pretend
The initial response to Mexico’s August 1982 announcement was what would later be characterized as “extend and pretend”: banks would roll over existing loans, the IMF would provide bridge financing, debtor countries would implement austerity programs, and the crisis — diagnosed as a liquidity problem rather than a solvency crisis — would resolve as growth returned and debt service ratios improved.
This diagnosis was incorrect but politically convenient. Acknowledging that the debt was unsustainable would require banks to write down the value of their loan portfolios — recognizing losses that would have rendered many major American banks technically insolvent. Regulators had a strong incentive to avoid forcing that recognition. The IMF and the US Treasury had a strong incentive to avoid the political and financial implications of declaring the major American banks insolvent. The debtors had a strong incentive to keep receiving new loans, even at the cost of continued austerity. Everyone had an incentive to pretend that the problem was temporary.
The IMF’s initial program for Mexico, negotiated in November 1982, required Mexico to reduce its fiscal deficit from seventeen percent of GDP to five point five percent in a single year — a contractionary adjustment of staggering magnitude in an economy already in deep recession. In exchange, the Fund provided three point nine billion dollars in loans and arranged a debt rescheduling with the commercial banks. Similar programs were negotiated with Brazil, Argentina, and other debtors.
The programs stabilized the immediate crisis. They did not produce recovery. Mexico’s GDP contracted in 1982 and 1983. Brazil’s manufacturing output fell sharply. Unemployment and poverty rose across the region. The austerity was being delivered; the growth that was supposed to follow was not materializing.
7.3.2 The Baker Plan
By 1985, the failure of extend-and-pretend was sufficiently obvious that a new approach was required. Treasury Secretary James Baker unveiled the Baker Plan at the IMF and World Bank annual meeting in Seoul in October 1985. The plan acknowledged, at least implicitly, that growth rather than just austerity was needed to resolve the crisis.
Baker proposed that the fifteen major debtor countries would receive fifteen billion dollars in new lending from commercial banks over three years, in exchange for implementing structural economic reforms. The reforms — trade liberalization, privatization of state enterprises, deregulation of domestic markets — reflected the emerging “Washington Consensus” on development economics that would define the 1980s and 1990s.
The Baker Plan failed to generate the commercial bank lending it proposed. Banks, which were already recognizing their Latin American loans as non-performing in their internal accounting, had no appetite for additional exposure to countries that had demonstrated inability to service existing debt. The plan’s failure demonstrated the fundamental problem that extend-and-pretend had created: by maintaining the fiction that the old loans would be repaid in full, no one was willing to extend new credit on normal terms. The market had figured out that the loans were impaired even if the official architecture had not.
7.3.3 The Brady Plan
The genuine turning point came in March 1989, when Treasury Secretary Nicholas Brady unveiled the Brady Plan — an approach that acknowledged, for the first time at the official level, that the debt burden was too large to be repaid in full and that some degree of debt reduction was necessary.
The Brady Plan’s mechanism was elegant. Existing commercial bank loans to sovereign debtors would be exchanged for new securities — “Brady bonds” — that were partially backed by US Treasury zero-coupon bonds purchased specifically for this purpose. The zero-coupon bonds, which required no interest payments until maturity, guaranteed repayment of principal at the bond’s maturity date; the interest payments in the interim were guaranteed by a separate reserve fund. In exchange for these collateral protections, the face value of the debt was reduced — typically by thirty to thirty-five percent — and interest rates were set at levels that reflected the debtor’s actual capacity to service.
The Brady bonds achieved what extend-and-pretend could not: they transformed illiquid bank loans — assets that banks were unable to sell or mark to market accurately — into tradeable securities that could be valued, sold, and diversified. Banks could book their losses, remove the assets from their balance sheets, and stop having their capital constrained by the non-performing exposure. Investors who believed Latin American economies would recover could buy the bonds in secondary markets at prices reflecting distressed levels. The mechanism allowed the banking system to end its crisis exposure while creating a market-based mechanism for pricing and holding the residual debt.
Mexico’s Brady deal, concluded in July 1990, was the template. Mexico exchanged approximately forty-eight billion dollars in bank debt for new bonds at a face value reduction of approximately thirty-five percent, with the principal guaranteed by US Treasury zero-coupon bonds. Venezuela, Costa Rica, Uruguay, Argentina, and Brazil followed with their own deals through 1990–1994. The crisis formally ended not through repayment of the original debt but through restructuring that acknowledged its partial impairment.
7.4 The Legacy
7.4.1 The Human Ledger
Latin America’s “lost decade” is documented in statistics that, while dry in presentation, encode enormous human suffering. Per capita income across the region fell approximately eight percent between 1980 and 1990. In some countries the decline was far more severe: Bolivia’s per capita income fell by more than thirty percent. Poverty rates rose from roughly forty percent to nearly fifty percent of the Latin American population over the decade. These aggregate figures mask the variation in experience: Mexico’s per capita income declined approximately ten percent in real terms; Argentina’s fell even further.
Health indicators deteriorated alongside economic ones. Child mortality rates, which had been falling steadily through the 1970s, stagnated or reversed in the early 1980s across much of the region. School enrollment fell as families pulled children out of education for work or could not afford fees and materials. Nutritional status among children under five deteriorated measurably in the countries most severely affected by austerity. The costs of the adjustment were borne asymmetrically by those least able to bear them.
7.4.2 The Brady Plan’s Success
The Brady Plan’s implementation through 1990–1994 enabled the recovery that the preceding seven years of extend-and-pretend had failed to produce. Once debt was restructured to sustainable levels — face value reductions of roughly thirty to thirty-five percent, combined with interest rate adjustments — capital flows resumed, domestic investment recovered, and growth returned.
Mexico grew at an average of three percent annually through the early 1990s. Brazil’s stabilization, culminating in the Plano Real of 1994, finally defeated the hyperinflation that had plagued it through the crisis decade and launched a period of growth that continued through the commodity boom of the 2000s. Argentina recovered through the late 1980s and early 1990s, though its subsequent dollarization experiment would produce a further crisis in 2001–2002.
The mechanism of recovery — debt relief rather than structural adjustment alone — became central to subsequent IMF doctrine, though the institution was slow and resistant to fully incorporate the lesson. The World Bank’s subsequent work on debt relief for heavily indebted poor countries, culminating in the HIPC Initiative of 1996, applied the Brady Plan logic to a broader category of sovereign debtors.
7.4.3 The Washington Consensus and Its Aftermath
The Latin American debt crisis generated a specific policy framework that dominated development economics through the 1990s: the Washington Consensus, a term coined by economist John Williamson in 1989 to describe the package of policies that Washington-based institutions — the IMF, the World Bank, and the US Treasury — recommended for developing economies in crisis.
The Washington Consensus prescribed fiscal discipline, tax reform, trade liberalization, privatization of state enterprises, deregulation, and protection of property rights. These prescriptions reflected the diagnosis that Latin American governments had borrowed excessively, spent inefficiently, and protected industries from competition in ways that reduced productivity. Some elements of the consensus were well-founded: fiscal discipline was genuinely necessary; some privatizations improved efficiency. Others proved more ambiguous: financial liberalization without adequate regulatory frameworks contributed to the Mexican peso crisis of 1994 and the Asian financial crisis of 1997–1998.
The post-Washington Consensus debate of the 1990s and 2000s was, in substantial part, a reckoning with what the consensus had gotten wrong. The Latin American debt crisis was its origin point.
7.4.4 The Emerging Market Bond Market
The Brady Plan’s financial innovation had an unintended consequence that proved more durable than the crisis resolution itself: it created the emerging market bond market. By transforming illiquid bank loans into tradeable securities, Brady bonds established the infrastructure — pricing conventions, settlement mechanisms, legal documentation — for sovereign bond markets in developing economies.
By the late 1990s, dozens of developing countries were accessing international capital markets directly through bond issuance rather than through bank lending. This democratization of sovereign finance was, on balance, beneficial: it provided access to capital at competitive rates, created market-based discipline on sovereign borrowers through real-time bond pricing, and gave developing economies a more diversified funding base than exclusive dependence on bank lending.
It also created moral hazard. The Brady Plan’s implicit message — that sovereign debts, when they became clearly unsustainable, would eventually be restructured rather than repaid in full — reduced the incentive for prudent borrowing. This tension between the demonstrated willingness to provide debt relief in extremis and the incentive effects of that willingness has never been resolved in international finance. It remains the unresolved core of the IMF’s debt crisis mandate.