Abstract

Following a robust recovery from the 1995 Tequila crisis, and strong growth in 1997, the economy slid into depression in the latter half of 1998. This depression lasted throughout the entire pre-crisis period and deepened, as the policy dilemma became increasingly apparent to the markets and binding to the government.

Following a robust recovery from the 1995 Tequila crisis, and strong growth in 1997, the economy slid into depression in the latter half of 1998. This depression lasted throughout the entire pre-crisis period and deepened, as the policy dilemma became increasingly apparent to the markets and binding to the government.

Initial Downturn—1998

The downturn was triggered by a variety of factors, including a cyclical correction, political uncertainty, and contagion from Russia and then Brazil (Box 7). These factors resulted in a severe contraction of private consumption and investment spending (Figure 9). Output growth, which had reached 8 percent in 1997, declined to less than 4 percent in 1998, falling at an annual rate of more than 10 percent during the fourth quarter. The downturn was to some degree a correction, given that the economy had been running above potential for some time, with GDP an estimated 3 percentage points above its full employment level in early 1998. A related cyclical factor was the unwinding of adjustments in the stock of durables.

Figure 9.
Figure 9.

Contributions to Output Growth

(Seasonally adjusted, in percent per quarter)

Sources: Argentine authorities; and IMF staff estimates.

The proximate trigger for the downturn was financial contagion from the Russian crisis in August 1998. Beginning in the fourth quarter of 1998, private debt-creating capital inflows, which had averaged more than 2 percent of GDP in 1996–97, turned negative (Figure 10).25 At the same time, spreads on Argentine sovereign bonds rose from about 500 basis points in July to 750 basis points in December. This had repercussions for domestic interest rates. Between the first half of 1998 and early 1999, prime lending rates on both U.S. dollar- and pesodenominated loans rose by 2 to 3 percentage points (to 10 percent and 12 percent, respectively), implying an increase in real interest rates by some 5 percentage points to 13 percent (11 percent on dollar loans). The currency board arrangement, however, muted the impact of the external shock on the domestic economy. Argentina’s spreads rose by less than half of the average increase in emerging market bond spreads (EMBI), and the increase in real domestic lending rates was significantly lower than in Mexico, for example, though comparable with that in Chile.26

Figure 10.
Figure 10.

Capital Market Indicators

Sources: IMF, Balance of Payment Statistics, International Financial Statistics; J.P. Morgan; and IMF staff estimates.1 On 30-day loans to prime customers.

Heightened political uncertainty also helped to puncture consumer confidence and dampen private sector spending. Historically, political crises in Argentina have had a significant impact on economic growth.27 Although President Carlos Menem’s ascendancy to power in 1989 brought some political stability, the situation became more uncertain in 1998 with his attempts to remain in power for an unprecedented third term, despite an explicit constitutional prohibition and public opinion polls indicating widespread opposition. The success of the Alianza party—an electoral coalition between the two main opposition parties—in the congressional elections of 1997 (winning 46 percent of the vote compared with the Peronists’ 36 percent) was a threat to the ruling party. Indeed, the opposition’s success was a key factor in Menem’s decision to rule himself out of the race for the Peronist leadership in July 1998, when it became clear that his re-election proposal risked splitting the party. This contributed to a further weakening of the government’s command of the country’s political and economic affairs.

The adverse consequences of political uncertainty and financial contagion were compounded by a series of external shocks, including the decline in demand for Argentina’s exports associated with Brazil’s 1998–99 crisis; the effect on competitiveness of the depreciation of the real; and a 6 percent terms-of-trade deterioration—stemming, in part, from a sharp decline in world prices for Argentina’s commodity exports. As a result, Argentina’s export earnings fell by 10 percent in the second half of 1998 (relative to the first half), while growth in volume terms was reduced to less than 2 percent from 15 percent in 1997.

Factors Contributing to the 1998 Downturn

Various factors gave rise to the sharp downturn in Argentina that began in mid-1998, including a cyclical correction following rapid growth in 1996 and 1997, a sharp downward adjustment in purchases of durables, an increase in political uncertainty, and dipping consumer confidence. Appendix I indicates that various measures of potential output yield an excess demand gap of more than 3 percent of GDP in 1998. Moreover, while automobile consumption rose rapidly in 1993 and 1994, and again in 1996 and 1997, in response to the stable inflationary environment, a sharp downward adjustment took place in 1999 and 2000, associated with the fact that purchasers of automobiles had by then upgraded their stock. The sharp decline in durables consumption had a noticeable effect on total private consumption.

One of the triggers for the growth slowdown was the increase in political uncertainty associated with President Menem’s desire to remain in power for an unprecedented third term. In recent years, a number of researchers have considered the relationship between political instability and growth, arguing that political instability increases uncertainty and, thereby, has negative effects on productive economic decisions such as investment and saving, and ultimately on growth (Alesina and others, 1996; Alesina, and Perotti 1995). To quantify such an impact, the Ghosh and Phillips (1998) growth model was re-estimated with the addition of a variable incorporating the number of government crises in each country based on Burns’ dataset. Argentina has had many political crises over the past 30 years, and including separate political crisis coefficients for Argentina and for the rest of the world yields significant differences. Indeed, the estimates reveal that a political crisis would lower per capita growth by 2 percentage points in Argentina compared with only ¾ percentage point in the rest of the world. Of course, these calibrations make no allowance for the severity of different political crises, and to the extent that the 1998 crisis was comparatively mild, its isolated impact on GDP growth would be smaller.

Another trigger for the slowdown was the effect of the Russian default on emerging market bond prices, which affected not only private investment in Argentina, through higher financing costs, but also consumption, in the context of a close relationship between consumer confidence and the sovereign debt spread (see figure).

uch03fig01

Consumer Confidence and the Sovereign Debt Spread

While the direct impact on output growth of the slowdown of exports was limited by the low share of exports in the economy, the combination of shocks heightened concerns about the resilience of the Argentine economy in general. With exports comprising only one-tenth of the economy, their weakening had little direct effect on real GDP growth. Nevertheless, to the extent that the external shocks contributed to rising real interest rates and a weakening in confidence, they played an important role in the sharp contraction in domestic demand. More specifically, the external shocks and subsequent real effective appreciation of the peso following the devaluation of the real in early 1999 (over 14 percent in real terms, on the basis of both consumer price inflation and unit labor cost) threw into sharper relief the precariousness of the public sector and external indebtedness, and the trap the economy had fallen into. As long as the nominal exchange rate remained strong in effective terms (reflecting both the peg to an appreciating U.S. dollar and the depreciation of the Brazilian real), restoring competitiveness would require price deflation. This, of course, would weaken economic activity to the extent that nominal wages and prices in the economy were downward inflexible (Box 8). Moreover, as elaborated in the following section, a real depreciation, whether through an exchange rate adjustment or price deflation, would have likely worsened the debt ratios. Such concerns were one factor behind the persistence of the increase in spreads on Argentine bonds, even as spreads in other emerging market countries in the region began narrow in mid-1999. In turn, these wider spreads and higher interest rates contributed to the continuing economic malaise.

Deepening Depression and Policy Dilemma: 1999–2000

Once the economy entered depression, it seemed unable to escape. Output shrank by 3½ percent in 1999 and by a further 1 percent in 2000, with deflationary pressures leading consumer prices to fall by some 1 to 2 percent both in 1999 and in 2000. Reflecting labor market rigidities, real wages rose by 2½ percent in 1999 despite the economic downturn and by a further 2 percent in 2000. Unemployment increased from 12½ percent in the second half of 1998 (its lowest level since the Tequila crisis) to 14½ percent in the first half of 1999, and 15 percent by 2000. At the same time, the public sector deficit rose to more than 4 percent of GDP in 1999, and 3½ percent in 2000, while the public debt ratio increased by 10 percentage points to more than 50 percent of GDP. However, in contrast to earlier years, when off-budget expenditures had played a major role, the debt dynamics were now driven mainly by an adverse differential between interest rates and growth.

The government’s initial attempts at curbing its growing deficit were at best timid, and the fiscal policy stance became clearly contractionary only in 2000. Public expenditure grew by some 4¼ percent in real terms in 1999 and by more than 5 percent at the provincial level. Helped by a rise in the revenue ratio, the fiscal impulse was mildly contractionary, but due to the operation of automatic stabilizers the primary balance deteriorated by 1¼ percentage points to a deficit of ¾ percent of GDP. Along with rising interest payments, the overall deficit doubled to more than 4 percent of GDP—a slippage of 2½ percentage points relative to the program target (Table 4). In light of the fiscal deterioration in 1999, the 2000 IMF-supported program sought to reduce the overall public sector deficit to 2¼ percent of GDP (and 1½ percent of GDP in 2001). However, despite a sizable structural tightening of 2 percentage points of GDP, in line with the program target, the deficit fell by merely ½ percentage point of GDP, and the government’s gross financing needs rose to US$26 billion, or more than 9 percent of GDP.28

Table 4.

Programmed and Actual Fiscal Balances and Impulses, 1999–2001

(In percent of GDP, unless otherwise indicated)

article image

Impulse that would have resulted if the program balance had been achieved (through expenditure adjustments) with the actual growth outcome.

Figures in percent of GDP deviate from those in IMF staff reports due to subsequent revisions in GDP data.

As a breakdown between revenue and expenditure is not available, the impulse is estimated by assuming that the revenue ratio in 2001 is the same as the projected ratio for 2000.

Deflation and Depression

Pinpointing the precise mechanisms through which the currency board may have contributed to the recession is difficult. The behavior of real interest rates or monetary aggregates, for instance, seems to provide only a peripheral explanation. Nevertheless, the fixed exchange rate regime implied that corrections to the real exchange rate could take place only through price deflation. One hypothesis, therefore, is that since a little inflation may help “grease the wheels” of the economy, deflation itself contributed to weakness in economic activity. Consistent with this, Ghosh and Phillips (1998) find that, while high inflation is associated with poor growth performance, at very low inflation rates, increases in the rate are associated with higher output growth. They estimate a “kink” at around 2½ percent per year (model 1). An alternative formulation (model 2) puts the kink at negative inflation rates. Using an annual cross-country panel dataset, these two models are estimated.

The standard growth determinants are generally of the expected signs. Intermediate exchange rate regimes are associated with higher growth, as is a higher investment ratio, greater trade openness, positive terms of trade shocks, a higher human capital stock (as measured by the average number of years of schooling), a lower average tax ratio, lower initial income (due to “convergence”), lower population growth, and larger economic size (as proxied by population). Higher inflation is associated with lower growth, but deflation is also associated with lower growth (model 2), with the deleterious effect stronger the higher the deflation rate.

Since Argentina experienced deflation of (at most) 2 percent per year, these estimates suggest that due to the deflation, real GDP growth was between ½ and 1½ percent per year lower than it otherwise would have been. However, some caution is required in applying these estimates. Since the regression commingles all the effects that tight monetary policy might have, including for instance through high real interest rates, this effect should not be added to other effects.

article image

Note: Asterisks indicate statistical significance at the 1(***), 5(**), and 10(*) percent levels.

Against this background, the government faced a harsh policy dilemma, as the buildup of debt in the earlier years had eliminated any room for stimulating the economy through fiscal expansion. From a cyclical perspective, fiscal and monetary expansion would clearly have been appropriate, but the former would have been associated with a larger fiscal financing requirement that would have driven up interest rates even further, while the latter was limited by the currency board arrangement.29 Indeed, even without such a rise in interest rates, and the optimistic assumption of a large fiscal multiplier effect on output, a wider deficit would almost surely have increased not just the overall level of public debt but also its ratio to GDP (Box 9). In the event, IMF-supported programs in 1999–2001 targeted primary surpluses in the range of 1 to 2¾ percent of GDP, though actual outcomes fell far short of these targets. The result was perhaps the worst of both worlds: fiscal policy failed to provide a decisive positive impulse to the economy, and confidence was probably undermined by the continual slippages of fiscal targets, resulting in rising public sector indebtedness and higher interest rate premiums.

A key aspect underlying the fiscal underperformance was the disappointing outturn of both real GDP growth and inflation relative to projections. Both program projections and consensus forecasts in 1999–2001 repeatedly assumed that growth would recover and that there would be some positive inflation (Figure 11). Had it been known that growth and inflation were going to be lower, the need for fiscal adjustment might have been clearer since the projected debt ratio (given the lower nominal GDP) would have been correspondingly higher. Conversely, to the extent that a prolonged recession was projected, there would have been more calls for providing a fiscal stimulus to boost activity. Thus, whether more accurate projections would have resulted in a very different policy stance is debatable. That said, the continued fiscal underperformance probably undermined confidence in the authorities’ ability to deliver on agreed targets.30 The error in projecting positive inflation may have been of even greater consequence given that, under the program, nominal wages were supposed to adjust downward to restore external competitiveness. Inasmuch as positive inflation was projected for 2000 and 2001 by both IMF staff and private forecasters, workers were likely to hold similar expectations, making them less willing to accept lower nominal wages.

Figure 11.
Figure 11.

Evolution of Forecasts for Real GDP Growth and Consumer Price Inflation: Program Projections and Consensus Forecasts

(In percent per year)

Sources: IMF, World Economic Outlook; Consensus Forecast; and IMF staff estimates.

Could Expansionary Fiscal Policy Have Stabilized the Debt Dynamics?

Some commentators have claimed that fiscal tightening, by further weakening activity, exacerbated Argentina’s debt dynamics, hastening the onset of the crisis.1 To assess this claim, it is useful to consider under what circumstances a widening of the deficit could reduce the debt ratio. Starting from the standard expression for debt dynamics:

d˙=(r-g)d+p

where d is the debt-to-GDP ratio, r is the real interest rate, g is the real growth rate of the economy, and p is the primary deficit, a larger primary deficit reduces the debt-to-GDP ratio if:

(d˙)/p<0(rpgp)d+1<0.

A key parameter is the response of interest rates to the wider deficit. Assuming that there is no response, the above condition reduces to: gp > (1 / d). That is, the response of output growth to a larger primary deficit must be greater than the reciprocal of the debt ratio; at end-1999, the latter was 1/0.47 = 2.12.

The most generous framework to the case for a fiscal expansion is a Keynesian model in which output is fully demand-determined. While such a framework does not lend itself easily to addressing output growth, by evaluating the policy around an initial growth rate of zero (a reasonable assumption for this exercise), it is possible to interpret the increase in output from the boost to aggregate demand as the change in the growth rate.

Under a Keynesian model, the output response to an increase in government spending is given by

Δyy=11c+mΔgy

where c is the marginal propensity to consume and m is the marginal propensity to import.2 Point estimates for the marginal propensities to consume and import (estimated over the period 1991–2001) are

0.657 and 0.187, respectively, yielding

11c+m=110.657+0.187=1.88

—shy of the critical value of 2.12.

In other words, even in the most favorable circumstances, a fiscal expansion would not have helped stabilize the debt ratio. Moreover, to the extent that the recession was not purely an aggregate demand shock, but also reflected slower growth of potential output, the response to the fiscal expansion would have been smaller.

The above calculation is also predicated on the larger fiscal deficit not triggering a rise in interest rates. Although effective interest rates on government debt rose steadily from 6.3 percent per year in 1997 when the public sector deficit was 2 percent of GDP to 9.0 percent per year by 2001 when the public sector deficit reached almost 7 percent of GDP, it is difficult to distinguish the impact on interest rates of the widening deficit from the effects on confidence of the general deterioration of macroeconomic conditions. In particular, it is at least theoretically possible that the announcement of a wider deficit—by presaging a stimulus to economic activity—would boost confidence so much that interest rates would fall. Conversely, the wider deficit might erode confidence further, implying higher interest rates. While the evidence is equivocal, an “event study” analysis suggests that Argentina’s spreads relative to benchmark U.S. bonds increased in response to news of wider fiscal deficits (see figure), although the pattern of Argentina’s spreads relative to the emerging market bond index (EMBI) is more supportive of the opposite conclusion (arguably, however, the latter spread is of less relevance for the evolution of Argentina’s debt dynamics).

As is readily verified, to the extent that the wider deficit elicits higher interest rates, rp > 0, the condition for an expansionary fiscal policy to stabilize the debt ratio becomes more difficult to satisfy.

uch03fig02

Argentine Bonds Relative to U.S. Bonds, Three-week Moving Average

1 See Stiglitz (2002).2 If there are income taxes, the multiplier is smaller: Δ yg = (1 − c (1 − τ) + m)-1.

Other than the fiscal stance, the key decision facing the authorities was whether to abandon the currency board and adopt an alternative monetary and exchange rate regime. Clearly, the authorities’ ability to address the crisis would have been greater in the absence of a currency board—that is, if it had taken the opportunity to exit in non-crises times earlier in the decade. Moreover, given that the currency board was ultimately abandoned, it is not clear that there was any benefit from delaying the exit. But by the time the recession was under way, changing the currency regime would, by itself, have given little relief from the crushing debt dynamics; on the contrary, it would have immediately worsened the situation given the foreign-currency exposures of public and private balance sheets.

Abandoning the currency board in 1999 or 2000 would have provided little relief on financing costs, and would have been unlikely to rescue the economy from the slump. A convenient way of characterizing the monetary policy stance as constrained by the currency board is through the use of a monetary conditions index, which is a weighted average of real exchange rate and real interest rate movements relative to a base period when the economy is in internal equilibrium at full employment (Box 10). The first panel in the Box 10 figure presents this index for Argentina and shows that it has a close relationship with the output gap. Indeed, the fit is remarkably strong and swift so that sudden movements in the monetary policy stance are rapidly mirrored in output movements. The procyclical tightening of monetary conditions during the recession is illustrated by the substantial rise in the monetary conditions indices (MCI) beginning in 1998, with an even sharper rise during 2001. The second panel in the box figure qualifies this result, comparing MCI indices for Argentina, Chile, and Mexico; here, it is noteworthy that in Mexico, with a flexible exchange rate, monetary conditions also tightened from 1998 through the end of 2000, although they did not mirror Argentina’s sharp tightening during 2001. This suggests that the tightening Argentina experienced during the recession was a reflection not solely of the currency board but also, in part, of the broader environment facing emerging markets. As such, it is unclear to what extent Argentina could have engineered substantial easing in its monetary stance in 1999–2000, even if it had abandoned the currency board. Already at this stage, an orderly exit from the currency board would have been very difficult to achieve, and would have required, at a minimum, significantly higher interest rates to curb capital outflows, suggesting little chance for monetary easing.

The stance of monetary policy during this period can also be viewed in relation to the behavior of money and credit aggregates: the growth of both broad money and credit to the economy in real terms decelerated substantially in 1998–2000, with real credit expansion becoming negative from mid-1999 through 2001 (Figure 12). Estimates suggest that this credit contraction was not a classic “credit crunch,” but reflected in large part the negative effects on credit demand of the rising cost of funds and the shrinking economy (Box 11).

Figure 12.
Figure 12.

Monetary Aggregates

Sources: Central Bank of Argentina; and IMF staff estimates.

If the currency board did not exist, a nominal depreciation of the exchange rate could have helped offset the earlier shock to exports without the need for painful deflation. But a very substantial real depreciation would have been required to have an appreciable impact on growth in the short term due to low price responsiveness of exports (explained by the predominance of primary commodities) and the low share of external trade in the economy (Box 12). On the basis of the model described in Box 12, a 60 percent depreciation (broadly in line with the actual fall in Argentina’s real effective exchange rate that occurred in 2002) would raise export volumes by a mere 5 percent in the short run, translating into a ½ percentage point contribution to real GDP growth. In reality, the growth of real exports of goods and services was close to zero in 2002 (on a national accounts basis), but has picked up since. In addition, as demonstrated by the events of 2002, a substantial depreciation would have boosted the peso value of Argentina’s large foreign-currency liabilities. Even though the banking system itself held a long position in U.S. dollars, there would have been an adverse effect on banks through the increase in nonperforming loans, as borrowers with pesodenominated income streams would have started to default on dollar-denominated loans.31 Together, these effects meant that any significant devaluation would have caused severe economic disruption.

A move to full dollarization would not have solved the dilemma either. The advantage of full dollarization over a currency board arrangement is that it implies an even stronger commitment to the exchange rate peg, since it is more difficult to reverse. This stronger commitment would generally be reflected in a lower exchange rate risk premium on interest rates. However, in Argentina, not much would have been gained from this credibility effect, as serious doubts about the currency board’s sustainability did not surface until 2001. Although the spread between onshore dollar and peso interest rates occasionally reached as much as 3 percentage points, on average it was no more than 1½ to 2 percentage points prior to 2001—or about ½ percentage point greater than the average during 1996–97. More fundamentally though, no exchange rate regime could have alleviated the pressure on the government to restore debt sustainability. To the extent that dollarization would have simply perpetuated the existing inconsistencies, it would have been very unlikely to restore the confidence of either Argentine consumers and firms or foreign investors.

Monetary Conditions Indices for Argentina, Chile, and Mexico

A number of central banks have derived a monetary conditions index (MCIs) to help represent the stance of monetary policy. The basic premise of the MCI is that it captures the effects of changes in monetary policy on real output through the real interest rate and real exchange rate channels. The concept was initially developed at the Bank of Canada with weights of 3:1 attached to the real interest rate (lending rate) and the real exchange rate (CPI based), corresponding to the relative impact of these variables on the Canadian economy.1 The relative weights approximate the ratio of exports and imports to output, measured in real terms. Since Argentina is considerably more closed to trade than Canada, a ratio of 9:1 was chosen for the relative weights for the Argentine MCI, broadly corresponding to its export ratio at 10 percent of GDP; for Chile and Mexico, relative weights of 2.5:1 were chosen, corresponding to export ratios for these two countries of about 30 percent of GDP. To smooth out movements in the real lending rate, a four-quarter moving average was used.

The MCI is presented relative to a baseline value when the monetary policy stance is neutral, defined as a point at which the economy is operating at its full potential. Since the Tequila crisis in 1995 severely disrupted these economies, it seems appropriate to consider a more recent period as the reference point. A standard Hodrick-Prescott filter revealed that each economy was operating at its full employment level in the first two quarters of 1997 and therefore this date was chosen as the reference point. This definition of a neutral policy stance is not without caveats, since a zero output gap does not necessarily imply that the real interest rate and the real exchange rate are at their long-term equilibrium levels. Moreover, to the extent that equilibrium rates change, movements in the MCI do not always correspond to shifts in the policy stance.

With the above caveats in mind, the top panel of the figure shows the close relationship between the MCI and the output gap in Argentina, with the tightening of the monetary policy stance in early 1995 associated with a sharp output decline later in the year. Similarly, the easing of monetary conditions over the period 1996–98 was associated with a booming economy, which was reversed in late 1998. Of course, it could be argued that the close relationship between the MCI and growth highlights the cost for Argentina of not being in control of monetary policy over this period.

The bottom panel shows a comparison of the monetary condition indices for Argentina, Chile, and Mexico, all presented relative to the common baseline. The chart shows that monetary conditions in Mexico were tight in late 1994, immediately prior to the devaluation, and that monetary conditions were similarly tight in Argentina in 1995 following the Tequila crisis. However, by early 1997, monetary conditions were comparable in the three countries (by assumption), and these conditions did not diverge until early 1999 when monetary conditions in Argentina tightened considerably relative to Mexico, and especially Chile. By early 2000 the MCI in Mexico had risen to Argentina’s level and it was not until early 2002 that Argentina diverged again significantly.

uch03fig03

Monetary Conditions Index

Sources: IMF, International Financial Statistics; and IMF staff estimates.

While it could be argued that Argentina should have adopted a floating exchange rate between 1996 and early 1998, it is clear that by late 1998 it was too late. Since monetary conditions in Mexico were comparable to those in Argentina by this time, it is unlikely that Argentina would have been able to adopt a looser monetary policy stance with a floating exchange rate. Indeed, based on history, it could have been expected that Argentina would have needed to keep interest rates high to avoid hyperinflation, although this is not what happened when the currency eventually floated.

1 See Duguay (1994).

Summary

With the economy operating above potential in the first half of 1998, the initial downturn occurred as consumer confidence was sapped by external financial shocks and domestic political uncertainties, compounded by tighter monetary conditions and trade-related shocks; thereafter, the structural weaknesses came into play. The structure and low share of exports, arguably partly related to an appreciated real exchange rate, meant that the economy could not export its way out of recession; labor market rigidities limited the adjustment of real wages and contributed to a sharp increase in unemployment, which further eroded consumer confidence; debt dynamics ruled out expansionary fiscal policy; and while the scope for monetary policy in a small and financially open emerging-market economy is generally limited, the currency board regime precluded even a modest monetary stimulus.

Thus, once the depression was under way, there was no easy way out, as the authorities had no policy instruments that could have enabled them to stimulate the economy without compromising debt sustainability. Exiting the currency board via dollarization would not have solved this dilemma, while a float—which could have helped, in principle, to jump-start the economy through a large depreciation—would have had major adverse repercussions via domestic balance sheets (including the public sector’s large dollar-denominated liabilities). A possible alternative approach, at this stage, would have been a preemptive sovereign debt restructuring, which would have provided both sufficient liquidity and net present value relief, combined with a strategy to limit the adverse repercussions on banks’ balance sheets. Taking this step could have also provided an opportunity to simultaneously exit from the currency board in favor of a more flexible arrangement that would have been more suitable for Argentina. However, at that stage, the authorities were understandably reluctant to adopt such a drastic approach, as a sovereign default was regarded as a last resort that should be taken only after all other policy options were exhausted. In these circumstances, the authorities hoped instead that an economic recovery would put a brake on the public debt dynamics as well as boost investors’ confidence.

Was There a Credit Crunch?

A key question in assessing the various explanations for the economic downturn is whether there was a “credit crunch,” perhaps because of the strictures on monetary easing implied by the currency board arrangement. To examine this possibility, a credit supply and credit demand function is estimated in a framework that allows explicitly for quantity rationing or disequilibrium in the credit markets. Credit supply is assumed to depend upon “lending capacity”—total banking system assets (excluding repos) less cash-in-vault, deposits at the central bank, and equity—and upon the real lending rate. Credit demand is assumed to depend upon real GDP and on the real interest rate (both variables are instrumented using their lagged value, to reduce problems of endogeneity). The model is estimated using quarterly data over the period 1993–2001. The estimated parameters are of the expected signs and are generally statistically significant.

Demand and Supply of Credit to Private Sector

article image

The top panel of the figure shows the estimated supply, demand, and actual credit to the private sector (in billions of 1993 pesos). The overall fit of the model may be gauged by how closely the minimum of supply and demand (at any instant) tracks the actual level of credit; the model seems to perform better in the second half of the sample.

Although actual credit falls precipitously beginning in early 1999, the model suggests that this mostly reflects falling credit demand rather than a contraction of credit supply—as such, there is little evidence of a quantity-rationing credit crunch. Indeed, as depicted in the lower panel, there seemed to be excess supply of credit as the economy contracted.

In part, however, credit demand was being choked off by rising real interest rates. The “counterfactual” excess credit demand line therefore shows how much excess demand there would have been had real lending interest rates remained at their 1994 level during the Tequila crisis, and at their 1998H1 level during the period 1998: H2-2001. Even in this counterfactual simulation, there is little evidence of excess demand for credit, belying the hypothesis that a “credit crunch” caused the recession.

uch03fig04

Credits to Private Sector

Sources: Central Bank of Argentina; and IMF staff estimates.1 Counterfactual is the excess demand that would have prevailed if, during 1995, the real lending interest rate had remained at its average 1994 level, and during 1998: H2–2001, the real lending rate had remained at its average 1998: H1level.

Empirical Estimates of the Effects of a Depreciation

Although the real exchange rate and the terms of trade had diverged sharply at the beginning of 1999, by mid-2000, the real exchange rate had stabilized and the terms of trade had improved. In volume terms, merchandise exports rose 14 percent between 1998 and 2000, and in value terms, despite the terms of trade deterioration, the average level of exports in 2000 was roughly the same as it had been in 1998. Many commentators have argued that Argentina should have devalued its currency and allowed it to float at this time to protect it against Brazil’s devaluation. However, given the structure of Argentina’s exports, in particular the large share of primary products, the price responsiveness of exports is low. Moreover, the small share of exports and imports in the economy implies that very substantial real depreciations would have been required to have an appreciable impact on growth.

To determine the magnitude of exchange rate adjustment that would be needed to stabilize the external debt in relation to GDP, a quarterly trade model for Argentina was estimated. The model consists of separate equations for the exports of commodities and other products, but treats all imports as one aggregate commodity. Since Argentina is a price taker for commodities, it takes the world commodity price as given. The volume of commodity exports is associated with the import volume of Argentina’s trading partners and with the relative price of the world commodity price and foreign domestic deflators. The price of manufactured and services products is a weighted average of domestic unit labor costs and wholesale prices in Brazil (the main trading partner). The volume of manufactured goods is associated with the volume of Brazilian imports, and the price of Argentina’s manufactured exports relative to Brazil’s wholesale price index. Finally, import prices are associated with the world export price and import volumes are associated with domestic output and the relative price of imports versus that of domestic goods (measured by the GDP deflator). Since the volume and price terms are nonstationary variables, the analysis was conducted in first differences with a cointegrating relationship between the volume and price terms also included in the specifications.

The estimates revealed that real exports of commodities has a unit elasticity with respect to the world import volume and that the price effect is insignificant, implying that the demand for Argentine commodities does not react to price changes. Similarly, real exports of other goods and services has a long-run unit elasticity with respect to Brazilian import volumes but in this case the price effect is significant, although the coefficient is significantly below unity signifying an inelastic demand for these products. The price of exports of goods and services (excluding commodities) is a weighted average of unit labor costs and the Brazilian wholesale price index in U.S. dollars with relative weights in the ratio of 1:3. The long-run elasticity of real imports with respect to domestic output is 2.76, considerably higher than the corresponding activity elasticity for exports. Finally, there is no identifiable price effect on the demand for imports.

The change in the balance on goods and services was calculated based on a 20 percent depreciation, and since Argentina is assumed to be a price taker for commodities, shocking the system by a depreciation of 20 percent has no effect on the world price of commodities. Moreover, since the world demand for commodities has not changed, revenues rise by a comparable amount. Taking advantage of the depreciation, exporters raise their domestic manufactures price by 6 percent, generating a 14 percent competitive improvement in the relative price of their products since competitor prices are now 20 percent higher. However, the demand for Argentine manufactures is price inelastic so that manufacturing revenues decline by 12 percent in foreign currency. On the imports side, foreign exporters lower their prices by 12 percent to maintain competitiveness generating an 8 percent increase in the import price in domestic currency. The demand for imports is inelastic so that import volumes are lowered only by 3 percent and expenditures in foreign currency are lowered by 15 percent.

Based on this model, a 60 percent depreciation (in line with the actual fall in Argentina’s real effective exchange rate in 2002) would have improved the balance on goods and services by about US$24 billion (8½ percent of 1999 GDP), with the positive impact waning over time as the import price rises to fully reflect the depreciation. This improvement would have been sufficient to cover about half of Argentina’s gross external financing needs, but would have resulted in a sharp increase in the debt-to-GDP ratio due to valuation effects. Indeed, the actual response in export volumes in 2002 was more sluggish, with an improvement in the trade and services balance equivalent to only half of the model’s prediction, while the external debt ratio increased by about 80 percentage points of GDP.

  • Alesina, Alberto, and Roberto Perotti, 1995, “Taxation and Redistribution in an Open Economy,European Economic Review, Vol. 39 (May), pp. 96179.

    • Search Google Scholar
    • Export Citation
  • Alesina, Alberto, Ricardo Hausmann, Rudolf Hommes, and Ernesto Stein, 1996, “Budget Institutions and Fiscal Performance in Latin America,NBER Working Paper No. 5586 (Cambridge, MA: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Blanchard, Olivier, and Augustin Landier, 2001, “The Perverse Effects of Partial Labor Market Reform: Fixed Duration Contracts in France,NBER Working Paper No. 8219 (Cambridge, MA: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Caballero, Ricardo, and Arvind Krishnamurthy, 2002, “A Dual Liquidity Model for Emerging Markets,American Economic Review: Papers and Proceedings, Vol. 92, No. 2 (May), pp. 3337.

    • Search Google Scholar
    • Export Citation
  • Calmfors, Lars, and John Drifill, 1988, “Bargaining Structure, Corporatism, and Macroeconomic Performance,Economic Policy, Vol. 3, No. 1.

    • Search Google Scholar
    • Export Citation
  • Calvo, Guillermo, 2002, “Sudden Stops, the Real Exchange Rate and Fiscal Sustainability” (unpublished; Washington: Inter-American Development Bank).

    • Search Google Scholar
    • Export Citation
  • Canzoneri, Matthew, Robert Cumby, and Behzad Diba, 1998, “Fiscal Discipline and Exchange Rate Regimes,CEPR Discussion Paper No. 1899 (London: Centre for Economic Policy Research).

    • Search Google Scholar
    • Export Citation
  • Card, David, and Richard Freeman, 2002, “What Have Two Decades of British Economic Reform Delivered?NBER Working Paper No. 8801 (Cambridge, MA: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Chudnovsky, Daniel, Andres Lopez, and Fernando Porta, 1996, “Intra-Industry Trade and Regional Integration: The Case of the Auto Industry in Argentina,” in Latin American Trade in Manufactures: A Handbook of Regional and Country Cases, ed. M. Lord (Ann Arbor: University of Michigan Press).

    • Search Google Scholar
    • Export Citation
  • Collyns, Charles, and Russell Kincaid, eds. 2003, Managing Financial Crises: Recent Experience and Lessons for Latin America, IMF Occasional Paper No. 217 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Cuevas, Alfredo, 2003, “Reforming Intergovernmental Fiscal Relations in Argentina,IMF Working Paper No. 03/90 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • De la Torre, Augusto, Eduardo Yeyati, and Sergio Schmukler, 2002, “Argentina’s Financial Crisis: Floating Money, Sinking Banking” (unpublished; Washington: World Bank).

    • Search Google Scholar
    • Export Citation
  • Di Tella, Rafael, and Robert MacCulloch, 1998, “The Consequences of Labor Market Flexibility: Panel Evidence Based on Survey Data,ZEI Working Paper.

    • Search Google Scholar
    • Export Citation
  • Duguay, Pierre, 1994, “Empirical Evidence on the Strength of the Monetary Transmission Mechanism in Canada: An Aggregate Approach,Journal of Monetary Economics, Vol. 33 (February), pp. 3961.

    • Search Google Scholar
    • Export Citation
  • Edwards, Sebastian, and Alejandra Cox Edwards, 2000, “Economic Reforms and Labor Markets: Policy Issues and Lessons from Chile,NBER Working Paper No. 7646 (Cambridge, MA: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Feldstein, Martin, 2002, “Lessons from Argentina” (unpublished; Cambridge, MA: Harvard University).

  • Galiani, Sebastian, 2001, “Labor Market Reform in Argentina: Where Do We Stand?” (unpublished; Buenos Aires, Universidad de San Andres, November).

    • Search Google Scholar
    • Export Citation
  • Galiani, Sebastian, and Hugo Hopenhayn, 2003, “Duration and Risk of Unemployment in Argentina,Journal of Development Economics, Vol. 71, pp. 199212.

    • Search Google Scholar
    • Export Citation
  • Ghosh, Atish, Anne Marie Gulde, and Holger Wolf, 2002, Exchange Rate Regimes: Choices and Consequences (Cambridge, MA: MIT Press).

  • Ghosh, Atish, Timothy Lane, Marianne Schulze-Ghattas, Aleš Bulíř, Javier Hamann, and Alex Mourmouras, 2002, IMF-Supported Programs in Capital Account Crises, IMF Occasional Paper No. 210 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Ghosh, Atish, and Steven Phillips, 1998, “Warning: Inflation May Be Harmful to Your Growth,IMF Staff Papers, International Monetary Fund, Vol. 45 (December), pp. 672710

    • Search Google Scholar
    • Export Citation
  • Gruber, Jonathan, 1995, “The Incidence of Payroll Taxation: Evidence from Chile,Journal of Labor Economics, Vol. 15 (July), pp. 72101.

    • Search Google Scholar
    • Export Citation
  • Hausmann, Ricardo, and Andrés Velasco, 2002, “The Argentine Collapse: Hard Money’s Soft Underbelly” (unpublished; Cambridge, MA: Harvard University).

    • Search Google Scholar
    • Export Citation
  • Heckman, James, and Carmen Pagés, 2000, “The Cost of Job Security Regulation: Evidence from Latin American Labor Markets,NBER Working Paper No. 7773 (Cambridge, MA: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Hirschman, Albert O., 1985, “Reflections on the Latin American Experience,” in The Politics of Inflation and Economic Stagnation, ed. Leon Lindberg and Charles Maier (Washington: Brookings Institution).

    • Search Google Scholar
    • Export Citation
  • Hopenhayn, Hugo, 2001, “Labor Market Policies and Employment Duration: the Effects of Labor Market Reform in Argentina” (unpublished; Washington: Inter-American Development Bank, February).

    • Search Google Scholar
    • Export Citation
  • IMF, 2002a, “Assessing Sustainability,available via the Internet at www.imf.org/external/np/pdr/sus/2002/eng/052802.htm.

  • IMF, 2002b, “IMF Executive Board Reviews the Fund’s Surveillance,Public Information Notice No. 02/44. Available via the Internet at www.imf.org/external/np/sec/pn/2002/pn0244.htm.

    • Search Google Scholar
    • Export Citation
  • Lane, Timothy, Atish Ghosh, Javier Hamann, Steven Phillips, Marianne Schulze-Ghattas, and Tsidi Tsikata, 1999, IMF-Supported Programs in Indonesia, Korea, and Thailand: A Preliminary Assessment, IMF Occasional Paper No. 178 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Levine, David, and Laura D’Andrea Tyson, 1990, “Participation, Productivity and the Firm’s Environment,in Paying for Productivity: A Look at the Evidence, ed. Alan S. Blinder (Washington: Brookings Institution).

    • Search Google Scholar
    • Export Citation
  • Mussa, Michael, 2002, Argentina and the Fund: from Triumph to Tragedy, Policy Analysis in International Economics 67 (Washington: Institute for International Economics).

    • Search Google Scholar
    • Export Citation
  • Nickel, Stephen, and Kevin Denny, 1990, “Unions and Investment in British Industry,University of Oxford Discussion Paper (Oxford, U.K.).

    • Search Google Scholar
    • Export Citation
  • Perry, Guillermo, and Luis Serven, 2002, “The Anatomy and Physiology of a Multiple Crisis: Why Was Argentina Special and What Can We Learn from It?” (unpublished; Washington: World Bank).

    • Search Google Scholar
    • Export Citation
  • Roubini, Nouriel 2001, “Should Argentina Dollarize or Float? The Pros and Cons of Alternative Exchange Rate Regimes and Their Implications for Domestic and Foreign Debt Restructuring/Reduction,Stern School of Business draft Working Paper (New York: New York University).

    • Search Google Scholar
    • Export Citation
  • Stiglitz, Joseph, 2002, “Argentina, Shortchanged,The Washington Post, May 12.

  • Thomas, Alun, 2002, “The Costs and Benefits of Various Wage Bargaining Structures: An Empirical Exploration,IMF Working Paper No. 02/71 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Yeats, Alexander, 1998, “Does Mercosur’s Trade Performance Raise Concerns About the Effects of Regional Trade Arrangements?The World Bank Economic Review, Vol. 12, No. 1.

    • Search Google Scholar
    • Export Citation
  • Woodford, Michael, 1994, “Monetary Policy and Price Level Determinacy in a Cash-in-Advance Economy,Economic Theory, Vol. 4, No. 3, pp. 34580.

    • Search Google Scholar
    • Export Citation
  • Woodford, Michael, 1995, “Price Level Determinacy Without Control of a Monetary Aggregate,Carnegie-Rochester Conference Series on Public Policy, Vol. 43, pp. 153.

    • Search Google Scholar
    • Export Citation