IN THE MIDDLE OF 1985 both Argentina and Israel launched frontal attacks on inflation. This was not surprising since both countries were approaching hyperinflation. What was somewhat surprising was that, despite basic differences in the structure of these economies, the programs were extremely similar in their design and in their effects.1 Since both countries succeeded in reducing inflation drastically during the first year without incurring very significant costs in terms of employment and output, it is important to analyze the characteristics of the plans and to consider to what extent they also led to the establishment of a path of sustained price stability.

Abstract

IN THE MIDDLE OF 1985 both Argentina and Israel launched frontal attacks on inflation. This was not surprising since both countries were approaching hyperinflation. What was somewhat surprising was that, despite basic differences in the structure of these economies, the programs were extremely similar in their design and in their effects.1 Since both countries succeeded in reducing inflation drastically during the first year without incurring very significant costs in terms of employment and output, it is important to analyze the characteristics of the plans and to consider to what extent they also led to the establishment of a path of sustained price stability.

IN THE MIDDLE OF 1985 both Argentina and Israel launched frontal attacks on inflation. This was not surprising since both countries were approaching hyperinflation. What was somewhat surprising was that, despite basic differences in the structure of these economies, the programs were extremely similar in their design and in their effects.1 Since both countries succeeded in reducing inflation drastically during the first year without incurring very significant costs in terms of employment and output, it is important to analyze the characteristics of the plans and to consider to what extent they also led to the establishment of a path of sustained price stability.

The first part of this paper describes some of the similarities in the design, implementation, and results of the stabilization programs. In general, both were based on a mixture of broad traditional contractionary measures with relatively tight incomes policies. The nature and the initial results of these measures are discussed in this section.

The large degree of similarity between the two plans, both at the conceptual and at the operational levels, do not seem to be accidental but rather the consequence of a perception common to both countries about the nature of the inflationary process. The second part of the paper analyzes the rationale for this perception which led to the adoption of the specific type of shock policies used. This section emphasizes the argument that there is a connection between the use of price freezes and demand policies which led to some temporary overadjustment of real variables and, in particular, to an “overshooting” of wages and interest rates. It is argued that such transitional overadjustment was required to facilitate the implementation of the price freeze by reducing the need for administrative supervision. The need for the price freeze itself arose from the perception that inflation in these countries contains a large inertial component that is difficult to reduce gradually. The connection and the implicit trade-off between incomes policies and the degree of the adjustment needed has not been noted in previous literature, and is central to an understanding of the short-term dynamics of stabilization programs.

One of the striking similarities in the short-term dynamics of stabilization in the two countries is related to the time profile of the overshooting of real wages and interest rates. In both countries the overshooting was equally severe and lasted a similar time span. It is also remarkable that the real effects of the programs on employment and output followed similar patterns. Some explanations are given for these similarities.

While most of the paper concentrates on the strategy for making the transition from high to low inflation, it also examines the ability of the programs to establish sustained stability. It is emphasized that, although the initial shock usually reduces inflation significantly, the transition to longer-term price stability cannot be assured without tackling appropriately the fundamental causes of hyperinflation. These fundamentals include a balanced government budget and a sustainable position of the current account of the balance of payments. The paper also develops an analytical framework which formalizes the concept of overshooting and describes the theoretical underpinnings of the approach used.

I. Common Features of Programs and Actual Developments

Basic Common Features

The most important common feature of the two programs was the aim of the authorities in both countries to bring down inflation drastically and immediately. This must be seen in the context of the historical record—the failure of recent attempts in both countries to slow inflation gradually by decelerating the rate of devaluation.2 Renewed attempts in Argentina to apply the gradualist approach in late 1983 were also ineffective. It can be seen from Figure 1 that the programs adopted by both countries in mid-1985 were, however, extremely successful. This success was achieved with the aid of extensive wage-price controls. As the controls were gradually lifted, or became less effective, the programs were confronted with the first moment of truth: could lower inflation be maintained without controls?3 While Israel passed this stage successfully, Argentina seems to have encountered difficulties as it shifted from a fixed exchange rate policy to a crawling peg in April 1986 and inflation reaccelerated later in the year.

A second common feature of the programs was their comprehensiveness; both encompassed policies directed at all the basic macroeconomic variables. For Israel this was an entirely new approach to dealing with inflation; fiscal and monetary restraint had been attempted in 1980, exchange rate policy in 1982-83, and incomes policies from November 1984 until the initiation of the 1985 stabilization plan, but a fully comprehensive and coordinated strategy had never been tried.4 There is no doubt that this approach had a favorable public reception, which, in turn, enhanced the credibility and effectiveness of the program.

In Argentina, on the other hand, a number of comprehensive programs had been tried before (in particular that implemented between 1967-70 which had some similarities with the Austral Plan), and the 1985 program was therefore less innovative. This led to a tendency by the Argentine Government to make stronger commitments—such as introducing a new currency, the Austral, at a fixed exchange rate with the U.S. dollar, and making an explicit promise not to print money to finance the budget deficit. The eventual failure of fairly comprehensive plans in the past also created some skepticism and lack of confidence in economic policy, which strengthened the need to reinforce public credibility at the outset of the Austral Plan.5

Figure 1.
Figure 1.

Monthly Rate of Inflation-Consumer Price Index

(January 1985-August 1986)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Note: Observations for Israel are led by one period.
Figure 2.
Figure 2.

Public Sector Deficit as Percentage of GDP

(Quarterly, 1984.III-1986.II)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

The need for a comprehensive plan is largely dictated by the objective of bringing down inflation immediately. A partial approach may result in serious distortions which could endanger the entire plan; for example, if wage policy were excluded, a rise in real wages (in terms of tradable goods) would lead to a balance of payments crisis. However, a comprehensive plan that does not cover only nominal policies, but includes adjustments in unsustainable levels of the key macrovariables (such as the current account or public debt ratios) is also needed to deal with inertial inflation. Unlike earlier attempts in both countries, the 1985 programs were based on the conviction that one cannot, and should not, deal only with inflationary inertia when the basic variables (the “fundamentals”) are unsustainable—primarily because the public will rationally suspect that the government will eventually have to adopt corrective measures which usually have price effects. For example, a persistent current account deficit is bound to result eventually in devaluations and, consequently, price rises, while a growing public debtincome ratio will result in eventual inflationary pressures. Inflationary inertia, either institutionalized in backward-looking indexation or based on an entrenched pessimism concerning the government’s ability to reduce inflation, cannot be overcome if the public expects that other variables will dictate a continuation of inflation.

The third common feature of the two programs is the synchronization of the policies affecting nominal variables, especially the synchronization of the exchange rate and wage policies with the price freeze. Such synchronization is essential when accelerating inflation is to be checked to avoid serious distortions in relative prices in the initial stages of the program.

Specific Policies

Budget Deficit

A sharp reduction in the domestic fiscal deficit as a ratio of GDP was one of the centerpieces of the programs in both countries; initially the deficits in both countries were reduced by more than half, or by some 7-8 percent of GDP (Figure 2). Was this budget cut sufficient to deal with the situation?

In Israel the government deficit continued to decline after the initial fall and was eliminated completely in the course of 1986. The domestic deficit of the public sector as a whole turns out to be slightly positive but, with the surplus in the external accounts of government, the overall budget seems to be in balance.6 In Argentina the deficit rose after the initial decline, increasing recently to around 6-7 percent of GDP, with no offsetting unilateral receipts. The rule of “not printing money to finance the deficit” is by itself not strictly sufficient for long-term price stability; if the deficit is financed by the monetization of foreign loans, demand pressure will persist and, even if sustainable in the short run, the growing foreign debt will complicate future fiscal management. If such a result is expected, it would tend to reduce the current credibility of the plan.

While a complete and full elimination of the fiscal deficit at the outset of the program is not a necessary condition for stabilization—there have been successful stabilizations which did not start with an immediate balancing of the budget—it is important that the expected future path of deficits should be compatible with stability.7

The importance of fiscal policy is a forceful reminder of the need for persistence of the policy. This problem will be addressed in the later discussion of some aspects of long-term sustainability. The point here is that the actual budget cut undertaken in both countries could have been interpreted as a major step toward balancing the budget. However, the fact that the cut was achieved mainly by increasing net taxation in both countries and by taking some temporary steps may have raised some doubts about their long-term sustainability.8

Exchange Rate Policy

A fixed exchange rate policy has been used in many successful stabilization efforts but has special significance when the inflationary process is thought to be dominated by inertia. The basic strategy of overcoming such inertia is to create a reality of price stability within the framework of a sustainable path for the real sector. A fixed exchange rate policy is a basic ingredient of this strategy. A slowly crawling peg could, in principle, serve the same purpose but is, in practice, much more vulnerable to the resumption of inflation.

The sharp transition to a fixed exchange rate regime vis-á-vis the U.S. dollar (see Figure 3) signifies, more than anything else, the beginning of the stabilization policy. The Israeli plan was more successful in maintaining exchange rate stability through the first year while Argentina shifted to a crawling peg in the second quarter of 1985. Both countries started the programs with a large initial devaluation followed by a subsequent peg. This entailed a real devaluation intended to provide some cushion against inevitable price increases despite the price controls. However, in both countries the erosion of the real exchange rate against the dollar in 1986 was far greater than the initial devaluation (see Figure 4).

Figure 3.
Figure 3.

Nominal Exchange Rates Against U.S. Dollar

(Monthly, January 1985-August 1986; 1985= 100)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Figure 4.
Figure 4.

Real Exchange Rates Against U.S. Dollar

(Monthly, January 1985-June 1986; 1985= 100)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

It is important to note that the movements in the real exchange rate against a currency basket (of main trading partners) (see Figure 5) were much more stable for Israel, due to the weakening of the dollar vis-á-vis the European currencies. Because of Argentina’s different trade patterns, this stability was much less evident for the Austral.9 This larger deterioration of the real exchange rate in Argentina compared with that of Israel may be one of the reasons for Israel’s lower and more stable black market premium (Figure 6).10

The strategy of maintaining a fixed exchange rate while implementing rapid disinflation is closely related to the use of price controls. When a price freeze is in effect, an independently active exchange rate policy cannot be sustained, as was the case in the period preceding the Israeli stabilization plan.

Overshooting in Wages and Interest Rates

The policies followed in both countries led to extremely high real interest rates on commercial credit during the first three quarters of the stabilization programs and to the maintenance, over a number of months, of low real wages. This phenomenon is defined here as “overshooting”in the sense that the levels of these real variables depart, during the adjustment period, from their longer-term equilibrium value consistent with stability.

In Israel the slow reduction in nominal interest rates after the initiation of the program led to a plateau of real interest rates at about 5 percent a month around mid-1985, while in Argentina this rate was somewhat lower and more stable due to the introduction of special conversion tables (see Figures 7 and 8).11 The calculation ex ante of “real” interest rates is, of course, arbitrary because of the lack of direct data on inflationary expectations. It is clear, however, that during the first three quarters of the stabilization programs the price freeze considerably reduced short-run uncertainty about inflation so that the estimates of real interest rates, based on realized inflation rates, cannot be very far off the mark.12

Figure 5.
Figure 5.

Real Exchange Rales Against Basket

(Monthly, January 1985-June 1986; 1985 = 100)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Note: Observations for Israel are led by one period.
Figure 6.
Figure 6.

Parallel Exchange Market Spread

(Monthly. January 1985-June 1986; in percent)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Note: Observations for Israel are led by one period.
Figure 7.
Figure 7.

Real Interest Rates

(Monthly. January 1985-July 1986)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Note: Real rates are nominal interest rates deflated by consumer price inflation index in the previous month. Observations for Israel are led by one period.
Figure 8.
Figure 8.

Nominal Interest Rates

(Monthly, January 1985-July 1986)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Note: Observations for Israel are led by one period.

The emergence of high real interest rates in the early stages of disinflation has been observed in other countries as well (see Dornbusch and Fischer (1986)). One of the arguments advanced to explain this phenomenon is that the monetary authorities are slow to raise the money supply to meet the increased demand for money resulting from lower inflationary expectations (Dornbusch (1986)). This explanation does not seem to be appropriate for Israel where interest rates rather than the money supply are the policy instruments, and the money supply (base money as well as M1 and M2) is determined endogenously by the private sector who can exchange it on fixed terms for certain other liquid assets. In addition, banks can borrow funds from the Bank of Israel, at a fixed interest schedule, to cover reserve deficits. Moreover, the possibility was open in both countries for capital inflows, particularly from the repatri ation of capital.

In general, therefore, in a system where the money supply is determined endogenously, insufficient monetization is unlikely to exist and high real interest rates during disinflation could more probably be interpreted as government strategy (see Figure 9).13 The monetary authorities may, for instance, wish to maintain high real interest rates to compensate for an insufficient reduction in the budget deficit and to protect a fixed exchange rate from the effects of possible speculation over devaluation. In the following analysis, another important reason is adduced for maintaining very high real interest rates in the early stages of disinflation. This has to do with the incentive to freeze prices effectively with minimal administrative intervention.14

The level of real interest rates prevailing in the early stages of disinflation in Argentina and Israel was clearly untenable. Indeed, after the first quarter of 1986 there was a marked downward tendency in these rates, especially in Argentina. In Israel, the estimated real rate on free com mercial credit (in domestic currency) dropped in mid-1986 to more than 2 percent a month, a still unsustainable level, but much lower than earlier rates.15

A second variable subject to short-term overshooting in the two countries was real wages. Within the atmosphere of a national emergency and price freezes, the two governments were able to maintain a U-shaped path for real wages during the first stage of the programs (Figure 10). In Israel the severe, but temporary, cut in real wages that preceded the program was hammered out within the framework of an implicit “social contract,” while in Argentina more coercion seems to have been applied.

Unlike a rise in real interest rates, a reduction in real wages has not necessarily always been connected closely with disinflationary programs, but is to be expected when the government has the political power to impose wage controls.16 Low real wages at the outset of a program reduce expectations that inflationary methods will be used to erode them further, and may also signal the seriousness of the government’s con tractionary policies. There is also the traditional argument that a reduc tion in real wages helps to contain consumer demand (although the temporary nature of the wage cut and the redistribution effects of low-wage policies weaken this argument), and may also reduce the impact of the monetary and fiscal tightness on unemployment.

These considerations may encourage the government to enforce a temporary wage cut to establish a firm foundation for disinflation. However, another reason for cutting real wages—and one that is related to the use of price controls to reduce inflation drastically—is that they facilitate the management of price controls and reduce the necessary administrative intervention.

The actual movements of wages and interest rates over time in both countries reveal an almost identical shape, both in amplitude and in time span. The severe implementation of tight money and real wage cuts lasted for around eight to nine months in both countries.

A possible explanation of the similarity of the overshooting in both real interest rates and real wages in both countries is their similar initial level of inflation, which seems to have been in the range of 15-20 percent a month. In Israel the actual rate of inflation during the eight months preceding the program was considerably less, but this was due to a system of price and wage controls that suppressed the basic rate. Toward mid-1985, inflation was heading higher again. There seems to be a con nection between the basic rate of inflation and the time it takes to overcome the “institutional” component of inertia, which is the com ponent that captures the effect of a “catch-up” agreement (such as cost of living allowances). The higher the level of inflation, the shorter are the the time lags and, therefore, the shorter the time needed to extricate this type of inertia from the system. The similarity between the inflation rates in the two countries may therefore explain the similarity in the time span of these severe measures.

Figure 9.
Figure 9.

Real M1 Balances in Logarithms and at 1985 Constant Prices

(Monthly, January 1985-June 1986)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Note: Observations for Israel are led by one period.
Figure 10.
Figure 10.

Real Wage Indices

(Monthly, January 1985-July 1986)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Note: Observations for Israel are led by one period.
Disinflation, Employment, and the Trade Balance

One of the important results of the programs was the relatively small effect they had on employment and production compared with their dramatic effect on inflation. Indices of employment and real GDP (Figures 11 and 12) indicate that the effects of the programs on employ ment in Israel were small and they were transitory on GDP in both countries, although they also exhibited the U-shapes of the other real variables. Thus, the cost of disinflation in terms of lost output and employment appears to be small so far and, in fact, some acceleration in real growth was evident in 1986.

The behavior of the trade balance during the disinflationary period exhibits some degree of deterioration in both countries (Figure 13) and no visible improvement in the external accounts of either country could be attributed to the programs.

This less-than-satisfactory behavior of the trade balance is rather puzzling, since it is not, in principle, consistent with the severe budget cut that took place. In addition to the behavior of the terms of trade, one plausible explanation for this apparent contradiction is that the increase in taxation may have been construed as temporary; another (more popular in Israel) is that the increased stability of prices reduced the public’s precautionary demand for assets which led to an increase in consumption demand. This effect, however, can only have been temporary. One way or another, the fact that the programs did not improve the trade account may have serious implications for the sustainability of the level of stability initially achieved.

In the Israeli case, external support for the program also seems to have played an important role. In particular, there was the special U.S. grant of US$1.5 billion (over two years) specifically made to support the program. Consequently, no inflationary influences on expectations were exerted from the balance of payments. The presence of external support led, therefore, to an increase in confidence in the overall program.

Figure 11.
Figure 11.

Employment Indices

(Quarterly, 1984, I-1986, I; 1985 = 100)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Figure 12.
Figure 12.

Real GDP Indices

(Quarterly. 1984,[-1986. II 1985 = 100)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Figure 13.
Figure 13.

Trade Balance

(Quarterly. 1984.I-1986.II)

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

Note: Index numbers, with relevant quarters of 1983 = 100, seasonally adjusted.

II. Credibility, Inertia, and Controls

Credibility and Inertia

A well-known argument of the rational expectations school is that inflation can be stopped abruptly, with low transition costs, if the government implements drastic fiscal and monetary policies consistent with price stability.17 However, these policies must be credible—the public must be convinced that the policies will be maintained over a long time.

Unfortunately, there is no simple way in which the government can convince the public of its determination to maintain its policies. More over, even if the fiscal policy is credible, monetary policy may not be, so that there is no guarantee that monetary policy will not accommodate inflation. This is a major difficulty since it is well known that inflation can evolve independently of the real budget deficit when monetary instruments are accommodating.18

Suppose that the public does not believe that the government will enforce monetary tightness in the face of persistent inflationary expectations (for fear of large unemployment). In this situation, the public’s inflationary expectations will tend to be maintained. Faced with this situation, the government will not be inclined to adopt tight monetary policies at all, and the public’s expectations will have proved to be self-justifying or “rational.”

Among the reasons the public may not accept the government’s an nounced long-term disinflationary targets are that countries with a long history of inflation and failures of disinflationary policies exhibit deep-rooted pessimism about the government’s ability to control inflation. This causes considerable downward rigidity in basic, longer-term, infla tionary expectations. (An upward adjustment in basic inflationary ex pectations is obviously easier since the public is more easily convinced that the government’s position on inflation could weaken.) These basic inflationary expectations are, in addition to institutional mechanisms, the forces behind the concept of “expectational inflationary inertia” and make the use of a monetary squeeze as the basic disinflationary instru ment very costly. In contrast, the use of temporary incomes policies and initial price freezes may avoid the large costs in employment that arise from the combination of monetary restraint and inflationary inertia.

The question is, of course, what ensures that the expectational inertia will indeed be affected by the package of wage-price-exchange-rate freezes? Certainly, there is no guarantee that incomes policy will reduce the basic level of inflationary expectations, but there are two considera tions which indicate that it could. First, provided that the fundamental variables (such as the current account and the budget deficit) have adjusted properly, the imposition of a price (and exchange rate) freeze can help the government demonstrate that the economy can effectively func tion without inflation and without any stresses in the strategic areas— such as the current account of the balance of payments or the debt-income ratio. As a result of this demonstration effect, individual agents may come to accept the state of low inflation as a realistic possibility.

Second, as long as the government combines incomes policy with a fixed exchange rate, there is an explicit or implicit commitment not to resort to inflation to achieve short-term gains (in terms, for instance, of an erosion of the real wage to make the economy more competitive). When the fundamentals are consistent, this commitment may be credible and may carry over to the period when controls are lifted.19 The credibility of this commitment also depends on the political stability of the government. In this respect, both Argentina and Israel changed regimes with the establishment of the wide coalition government in Israel and of the popular democratic government of Argentina.

It seems, therefore, that a combination of incomes policies with a proper adjustment of the fundamentals does provide, in an appropriate political environment, an opportunity to break the vicious circle associated with credibility and inflationary inertia.

Incomes Policies, Stabilization, and Overadjustment

In addition to their role in reducing inflationary expectations, incomes policies can play another part in quick stabilization, one associated with the need to perform an immediate and drastic cut in the budget deficit. This cut must be based, in the short run, on an increase in taxation and on a reduction in subsidies, which may create an initial, sudden price rise. With the system of catch-up agreements and the inherent pessimistic expectations, such a price shock can become a longer-term wage-price spiral which will make the government’s policy untenable. The imme diate role of incomes policies is, therefore, to contain this initial shock to the price level arising from the adjustment measures themselves.

However, controls raise problems of their own. These include some well-known problems—such as the economic costs caused by inter ference with the market mechanism and the allocation of resources to direct supervision—and a less well-known problem—that the need to use a pervasive and very visible enforcement mechanism is by itself an indi cation that any stability achieved is artificial.20 The demonstration that the economy can function without inflation requires that controls be enforced with minimal coercion.

It is a common, and somewhat surprising, feature of both programs that they employed minimal enforcement in administering the controls.21 It seems, therefore, that the authorities made great efforts to reduce administrative intervention as much as possible. This involved the use of various methods of persuasion and propaganda. However, perhaps of greater significance was the strategy of creating conditions of excess supply during the critical phase of the price freeze.

It is at this point that the short-term overshooting in real interest and wage rates steps in. High real interest rates induce reduction of inventories and, thus, excess supply, which will be amplified by the negative effect of high interest rates and lower real wages on consumer demand.22 In general, excess supplies are not incompatible with inflationary expectations when the absolute price level is rigid downward. For example, in an uncontrolled economy with (short-term) downward rigidity in the price level and potential excess supply in the commodities and labor markets, prices cannot go down by assumption. But if the money supply is increasing, and inflationary expectations prevail, prices may increase proportionately. Real wages will not rise (and hence the potential supply of commodities will not diminish) because of excess supply in the labor market. A combination of excess supply and inflation therefore arises which is closely related to the phenomenon of stagflation.

In an economy subject to effective price controls the coexistence of basic inflationary expectations (driven by inertia) and excess supply may again arise. However, the effect of these expectations on the actual inflation rate is drastically reduced by the direct controls. But, as noted before, the effective implementation of controls is costly and, to reduce that cost, the government is motivated to generate excess supplies which reduce the need for administrative intervention.23

III. Formal Model

It may be useful to formulate the foregoing ideas more precisely. The purpose of this section is, first, to analyze within the framework of a macroeconomic model the creation of excess supply during the price freeze. Second, the concept of inflationary inertia is incorporated and the trade-off between adjustment and controls is formally presented.

Consider an economy which produces a single tradable final good (used for consumption, investment, exports and so on). This output (Y) is produced using domestic labor, an imported intermediate good (there is no direct import of consumption goods), and a constant initial stock of capital (including inventories). The demand function facing the country’s output abroad is downward sloping, depending only on the reciprocal of the real exchange rate (in units of domestic currency per foreign currency unit).

Consider a solution to the IS-LM model, where the (demand-determined) level of output is given by:

YD=YD(+)()(+)(+)(m,t,q,π),(1)

where m denotes real balances, t denotes tax parameters, q is the real exchange rate, and π denotes short-term inflationary expectations (which are also assumed to equal the actual rate of inflation). The signs above the variables are those of the partial derivatives. In a similar manner the solution for the real interest (i) in the IS-LM model is given by:

iD=iD()()(+)()(m,t,q,π),(2)

In this formulation we assume that real government expenditures are constant (which was roughly the case in practice), that the momentary price level is given, and that short-run inflationary expectations (during the official price freeze) are zero or some low fixed value. The aggregate supply function is given by:

YS=YS()()()()()(q,w,t,i,C),(3)

where the sign of t reflects the negative effect of taxes on supply and where C denotes price controls which affect aggregate production negatively. An increase in the interest rate reduces supply by increasing the cost of working capital and the rate of bankruptcies. An increase in import prices (through q) and an increase in the real wage (w) also reduce supply.

It is assumed throughout that during the price freeze Ys ≥ YD = Y which equals actual output (therefore, i is determined by equation (2)). Although price controls take the form of maximum prices, the price level cannot fall in the short run because of the assumed price rigidity. Substituting equation (2) in equation (3):

YS=YS()(+)()()()(+)(q,m,t,w,C,π),(4)

where the signs of equation (3) are assumed to dominate. The trade balance is given by:

B=B()(+)(Y,q),(5)

where Y = YD. It is assumed that imports are positively related to output and that an increase in the real exchange rate improves the balance of trade. Export demand of the country’s final good is fully determined by q.

Finally, combining equations (1) and (4), excess supply is:

E=YsYD=E()()(?)()()(?)(q,m,t,w,c,π).(6)

The tightening of monetary policy, through a reduction in m, reduces both demand and supply. However, in the short run the liquidation of inventories following a sharp rise in 1 is assumed to dominate—producing excess supply. This state can be maintained in the short run because of downward rigidity in prices. At the same time, the excess supply is not exported when q is held constant because the demand curve for exports is downward sloping.

The original position, prior to stabilization, can be described by the point A in Figure 14. where the curves YDYD and ysys (corresponding to equations (1) and (4), respectively) intersect. An iso-B curve corre sponding to equation (5) passes through this point, reflecting the original level of the balance of trade. The relatively flatter slope of BB reflects the usual conditions of trade balance stability.

The initial fiscal contraction associated with stabilization is reflected in an increase in t. This shifts both YDYD and ysys to the left, which does not ensure the creation of excess supply. Moreover, the imposition of controls tends to reduce excess supply. However, two factors do tend to work in the direction of creating excess supply: first, the monetary contraction connected with stabilization which creates excess supply through the liquidation of inventories; second, the reduction in real wages has the same effect. Both shift the YsYs curve to the right. Assume, therefore, that excess supply of BC (Figure 14) is created at the original level of q, with output at point B.

Note that by increasing the real exchange rate to q1 the government may reduce unemployment and improve further the trade balance. However, in doing so, it will reduce the excess supply (see equation (6)) and, thus, make the supervision of the price freeze more difficult.24

Equation (6) shows that by manipulating the policy instruments (q, m, t, w) the government may increase E, which helps maintain the controls.25 However, the use of each policy instrument for this purpose involves other types of costs. For example, a reduction in q worsens the trade balance, which may undermine the fixed exchange rate strategy. The reduction of m raises interest rates and thus deepens the recession and impedes growth. The government’s pressure to reduce w may be undesirable socially and costly politically. This is mitigated to some extent by the reduction in unemployment. The advantages in creating excess supply have to be weighed against the foregoing costs.

To turn to a more explicit formulation of the role of E in the price control regime, in the spirit of the “inflationary-inertia” theory, assume that in the uncontrolled economy inflation (which may also equal short-term expectations) is determined by:

π=πi+f(E),f(0)=0,f<0,(7)
Figure 14.
Figure 14.
Note: Yf denotes full employment output. It is not essential for Ys’ to be to the right of Ys, but B must be to the left of C.

where πI denotes the component associated with inflationary inertia (πI may be considered as a function of all past variables, including past inflation rates, budget deficits, and import surpluses), and E denotes current excess supply that reduces inflation.26 If prices are rigid down ward and flexible upward, we shall have π = πI, except when there exists excess supply.

We assume that by using controls (C), measured (say) by the input of supervisional effort, the government may reduce the effect of πI on π Thus, let:27

π=φ(c)πI+f(E),φ(0)=1,φ<0.(8)

Solving for c:

()()(+)C=U(E,π,π1),(9)

where the negative partial derivative of π means that when the target level of π is higher (less stringent), the government may do with less controls.

During the price freeze there is clearly a trade-off between adjustment and controls. This is depicted in Figure 15 which shows, along the P0 curve, the relationship between C and E for given levels of πI, and π. Using much C involves economic and other costs which the government may reduce by using its policy instruments to create excess supply. The optimal combination of C and E is at Q0. The more ambitious the government’s target π (denoted π¯*

article image
), the further out the P curve will lie.

If the government succeeds in maintaining a low π without a deterioration in the budget deficit or in the trade balance, then one may assume that πI will be gradually reduced. This will shift the P curve leftward to P1 where the chosen point is represented by Q1. Thus, as πl is reduced, less of C and E is required to attain the same inflation target. A success of the disinflationary policy means a gradual abolition of controls along 0Q0.

Substituting equation (6) into equation (9) and solving for C obtains:28

(+)(+)(?)(+)()(+)C=C(q,m,t,w,ππ1).(10)

Equation (10) shows how the various policy parameters can be used to reduce the burden of controls.29 However, each of these measures in volves a specific cost, as we explained earlier.

The data on the stabilization programs, which we described earlier, revealed similar “tastes” by the government by choosing a monetary squeeze and a wage cut of similar form in order to reduce the burden of controls in the early stages of stabilization. This particular mix resulted in a small effect on unemployment but seems to have slowed down economic growth for a while. Neither government exercised in the early stages of the programs the option of reducing the real exchange rate in order to reduce the need for controls.

The foregoing model predicts that the overshooting should be accompanied by an improvement in the trade balance. This prediction is confirmed only partially by the data. While it is true that there was some improvement in the trade balance during the first quarter (as we have seen earlier), the trade balance returned to its original state in the second quarter when the overshooting in wages and interest rates was still con siderable. This is related to an increase in spending, the reasons for which are still unclear.

Figure 15.
Figure 15.

Trade-off Between Price Controls and Excess Supply

Citation: IMF Staff Papers 1987, 002; 10.5089/9781451972931.024.A001

IV. Adjustment of the Fundamentals

It is clear that long-term stability can be achieved and maintained only if it is based on a balanced government budget and on a sustainable position of the current account of the balance of payments. It was seen earlier that the programs of both countries included a significant reduction in the budget deficit. It is extremely important, in terms of breaking past inflationary trends and, therefore, expectations, that these steps should be interpreted as part of a permanent policy of long-run budgetary equilibrium.

In this regard, some doubts remain about the longer-term nature of the fiscal improvement in both countries, where the bulk of the gains came from higher revenues without corresponding declines in public sector expenditures. Moreover, a large portion of the growth in revenues arose from temporary taxes and other contingency measures. Also, par ticularly in Argentina, some of the fiscal measures, such as import surcharges and export taxes, have negative structural effects and may hinder the long-term performance of the balance of payments.

Some aspects of the problems arising from the specific form of the budget cut can be illustrated by the model illustrated in Figure 14. The termination of the short-term overshooting leaves the economy with both yDyD and ySyS shifted to the left as a result of the increased level of taxation designed to close the budget deficit. Since in the longer run price controls are removed, the economy will actually be at the inter section of yDyD and ysys to the left of the point A in a state of unemployment.30

The return to full employment could be achieved by a reduction in real wages. However, this cannot be expected realistically because of social considerations or political constraints. Besides, the high level of taxation would still impede normal economic growth. The required long-term solution should take the form of a simultaneous reduction in taxes and public consumption. The released resources will be diverted to private investment which will be accommodated by further easing of monetary policy. This will enable an economically more efficient form of full employment.

The basic task of reducing the public sector is, therefore, the main test the programs must face in the longer run. As long as this aspect remains unsolved, the pressures from the absence of growth will induce the government to ease the burden of taxation through increased budget deficits which will undermine price stability.

Turning to the comparison between the two countries under review, it seems that the basic difference between them in the area of “fundamentals” is related not to the fiscal but rather to the external position. We know from past experiences in both countries, particularly those arising from the collapse of exchange-rate-based stabilization programs, that a current account deficit raises strong expectations of devaluation that lead eventually to inflationary consequences. If stability is to be sustained, it is therefore imperative to attain a level of external adjust ment that is necessary to remove this threat from the system.

In this respect, developments in Israel have been very encouraging. The import surplus showed a marked reduction in the period preceding the plan and the capital account was revolutionized as a result of a shift in the U.S. aid from loans to grants which, in itself, was sufficient to balance the current account.

The external situation in Argentina is quite different. In recent years Argentina has been running a current account deficit of over US$2 billion, which is caused by the large interest payments component that outweighs the trade account surplus. Unlike in Israel, most of the foreign debt is not contracted on concessional terms. Although there was some favorable response of the balance of payments to the program, many of the gains appear to be based on temporary developments with out a structural improvement. Under these circumstances, additional measures to improve fundamental variables should have been incorpo rated in the program to achieve a significant improvement in the trade account. The short-term nature of the improvements that were made during the program leads to the conclusion that the basic external difficulties of Argentina were not alleviated by the plan.

It is, therefore, quite possible that the difficulties that Argentina encountered at the end of the first year of the stabilization effort are connected not only to the intractability of the public sector situation but to the absence of a basic and structural solution to the external problem. The shift to a crawling-peg system in April 1986 is, therefore, unlikely to prevent, by itself, external sector tensions, and may rekindle the uncertainties which in the past led to the re-emergence of inflation.

Therefore, the transition to longer-term stability following the initial positive effects of the shock, and the overall sustainability of the achievements, cannot be assured without dealing with the fundamentals in the proper manner.

V. Concluding Remarks

The comparison of the recent stabilization experiences of Israel and Argentina show some common perceptions of the nature of the problem and in policy implementation, which give rise to three basic observa tions. In the first place, they illustrate that there are considerable differ ences in the strategy to disinflate in cases of very high and persistent inflation (or hyperinflation) compared with situations when inflation is moderate and occasional. In particular, it is important to recognize that, when inflation becomes ingrained, it acquires its own dynamics. This requires comprehensive stabilization packages containing specific transitional mechanisms to deal with the inertial forces which are certain to develop as inflation rises. The transitional mechanisms should combine and coordinate the correction of fundamental imbalances with the temporary implementation of direct price-signalling policies which serve both to assure individual economic agents that a clear-cut reversal in previous trends has taken place and to indicate the commitment of the government to implement and maintain policies consistent with the price signals.

Second, transitional incomes policies and direct price interventions may accompany, but never substitute for, the appropriate correction of the fundamental imbalances of the economy, including, particularly, budgetary adjustments and external sector sustainability. Moreover, the controls and the signalling policies which are designed to address the problem of lack of credibility and to deal with inertial inflation will not be successful if demand is not restrained over and above the short-run balancing requirements. To be effective in breaking inflationary inertia, an overadjustment that leads to some degree of excess supply in key markets should be implemented during the transition. The overshooting observed in both countries in the instruments of adjustment policies, such as real wages and real interest rates, are consistent with this observation. The implicit trade-off between excess supply and the strength of the controls for different degrees of inertial inflation indicates that over-adjustment also provides the mechanism for the eventual smooth elimination of direct intervention.

Third, the initial success in breaking the inflationary trend could be translated into sustainable stability only to the extent that the adjust ment of the fundamental variables can be maintained. In the longer run, not only a balanced budget but also a reduction in the size of the public sector are required for sustained stability. The most significant indicator for the state of the fundamentals seems to be the external balance. A failure to make this balance sustainable undermines anti-inflationary efforts and may well be a major cause of the problems with the Austral Plan in Argentina. In this respect, the situation in Israel is more satis factory, but maintaining the external balance promises to be a hard struggle.

APPENDIX Data Sources and Definitions

Sources

For Israel, data are obtained from the Bank of Israel’s Research Department’s data bases, and the Central Bureau of Statistics. For Argentina, Fund staff estimates have been used, except for data on nomirtal and effective exchange rates, employment, trade balance, and government deficit, which are taken from Novedades Económicas, Fundación Mediterránea, various issues. Data for real wages are obtained from the Department of Commerce.

Definitions

Inflation rates are measured from the CPI.

Nominal interest rates: for Israel, the interest on free commercial short-term bank credit denominated in new shegalim; for Argentina, free market rates on bill brokerage operations. Real rates are obtained by deflating nominal rates by the previous month’s inflation rate.

Real interest rates for the period 1967-72 in Argentina are calculated as:

RI=(QINF+1QIN+T)1,

where:

RI = real interest rate in percent per quarter;

QINFI= (QCPItQCPIt4)/4

article image
is the inflation rate;

QINT = quarterly nominal interest rates; and

QCPl - quarterly average of monthly CPI.

Real exchange rates against the U.S. dollar and the basket are defined as nominal exchange rates against the U.S. dollar and the basket, respectively, deflated by CPI.

Real money balances are obtained by deflation of nominal balances using the CPI.

Real wages are equal to nominal wage index divided by CPI.

Government deficits: for Israel, this is the deficit of the domestic public sector (including the Government and the Jewish Agency), it is on a cash basis, and includes an imputed real interest rate on public debt. For Argentina, the deficit is calculated from total public sector debt.

Monthly indices have been rebased to 1985 by setting the average of January-June 1985 = 100. Quarterly indices have been rebased by setting the average of the first two quarters of 1985 = 100.

REFERENCES

  • Barro, Robert J., and David B. Gordon, “Rules, Discretion, and Reputation in a Model of Monetary Policy,” Journal of Monetary Economics (Amsterdam), Vol. 12 (July 1983), pp. 10121.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bruno, Michael, “Generating a Sharp Disinflation: Israel 1985,” NBER Working Paper 1822 (Cambridge, Massachusetts: National Bureau of Economic Research, 01 1986).

    • Search Google Scholar
    • Export Citation
  • Calvo, Guillermo A., and Roque B. Fernandez, “Competitive Banks and the Inflation Tax,” Economics Letters (Amsterdam), Vol. 12, Nos. 3-4 (1983), pp. 31317.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dadone, A., and J.A. Ingaramo, “Controles de Precios en Argentina, 1960-1980” (Cordoba: Fundacion Mediterranea, 1986).

  • Dornbusch, Rudiger, “Inflation, Exchange Rates, and Stabilization,” NBER Working Paper 1739 (Cambridge, Massachusetts: National Bureau of Economic Research, October 1985).

    • Search Google Scholar
    • Export Citation
  • Dornbusch, Rudiger, “Tight Fiscal Policy and Easy Money” (unpublished; July 1986).

  • Dornbusch, Rudiger, Stanley Fischer, “Stopping Hyperinflation: Past and Present,” Welt wirtschaftliches Archiv (Kiel, Germany), 122 (April 1986).

  • Fundación Mediterránea, “The Austral Plan: Ten Months Later,” Newsletter (Cordoba), Vol. 1, No. 2 (April-June 1986).

  • Liviatan, Nissan (1986a). “Inflation and Stabilization in Israel: Conceptual Issues and Interpretation of Developments,” International Monetary Fund Working Paper (Washington, November 1986).

    • Search Google Scholar
    • Export Citation
  • Liviatan, Nissan (1986b), “The Evolution of Disinflationary Policies in Israel,” prepared for the Conference on “Economic Issues and Policy in Israel, 1985-86” held in Jerusalem, Israel in June 1986.

    • Search Google Scholar
    • Export Citation
  • Sargent, Thomas J., “Ends of Four Big Inflations,” in Inflation, Causes and Effects, ed. by Robert Ernst Hall (Chicago: University of Chicago Press, 1982).

    • Search Google Scholar
    • Export Citation
*

Mr. Blejer is Division Chief of the Special Fiscal Studies Division in the Fiscal Affairs Department of the Fund. He holds a Ph.D. from the University of Chicago and has taught at the Hebrew University of Jerusalem, Boston University, and New York University.

Mr. Liviatan is a Professor of Economics at the Hebrew University of Jerusalem, where he also received his Ph.D. He has also taught, among others, at the Massachusetts Institute of Technology; at the University of California, Berkeley; and at the University of Chicago. This paper was written while he was a visiting scholar in the Research Department of the Fund.

1

See, for example, Dornbusch and Fischer (1986), Dornbusch (1986), Fundación Mediterránea (1986), Bruno (1986), and Liviatan (1986a).

2

Reference could be made to the Martinez de Hoz period in Argentina (1976-80) and the Aridor experiment in Israel (1981-83).

3

In Israel most of the official controls were removed in the first half of 1986.

4

For a discussion of these policies, see Liviatan (1986a).

5

In other words, given the credibility problems and past experience, it seems that the policy measures used to affect expectations as well as the fundamental adjustment will have to be stronger than they might otherwise be. This may have implications for the flexibility that the Government has during the course of the stabilization program.

7

See Dornbusch and Fischer (1986) for the case of the Austrian stabilization.

8

A significant portion of increased revenues arose from emergency taxes, once-and-for-all forced saving schemes, and the shortening of collection lags.

9

In August 1986 Israel shifted to a fixed exchange rate policy against the basket of currencies. It should be noticed, however, that when measured from the beginning of 1985, the real exchange rate in Argentina improved considerably compared with that in Israel.

10

Notice that following the introduction of the crawling peg, the black market premium fell sharply in Argentina. However, this fall was reversed in August 1986 when the loss of credibility in the program became more widespread.

11

Real interest rates are calculated by deflating the nominal rate by consumer price inflation in the previous month.

12

This, of course, does not apply to the transitional period, that is the month during which the stabilization plan is actually started.

13

The remarkable similarity in the growth of real M, in the two countries (Figure 9) suggests that the authorities in Argentina were as liberal about the increase in M1 as tn Israel. Consequently, the hypothesis of insufficient moneti zation seems to be questionable for Argentina as well.

14

It should be noticed, however, that high levels of real interest rates may reflect uncertainty and the low level of credibility in the whole or in parts of the program, particularly regarding the fixed exchange rule. If such element of risk is present, observed nominal rates may not necessarily represent high ex ante real rates.

15

With the erosion of credibility and the re acceleration of inflation in Argen tina in the third quarter of 1986, the real interest rate rose substantially, to 4.5 percent a month on average.

16

See Dornbusch and Fischer (1986) for the analysis of real wages in stopping the German hyperinflation.

17

See, for example, Sargent (1982).

18

See, for example, Calvo and Fernandez (1983) for the possibility of varying the steady-state level of inflation, at a given budget deficit, by means of monetary management, mainly changes in the reserve ratio of commercial banks.

19

This is in the spirit of recent literature on rules and discretion as, for exam ple, in Barro and Gordon (1983).

20

For a comprehensive study of the real costs associated with price controls in the last 25 years in Argentina, see Dadone and Ingaramo (1986).

21

It is estimated that during the austerity period in Israel in the early 1950s, thousands of people were employed in the price controls system while in 1985 the number of people actively engaged in this area was not even in the hundreds.

22

These effects may, however, be partially offset by the adverse effect of high interest on current production.

23

The creation of excess supply clearly involves costs of its own. The govern ment must, therefore, eventually strike some balance between these and the administrative and economic costs of the price controls.

24

One way of reducing the burden of controls is by giving up the fixed ex change rate policy and switching to a crawling peg, as Argentina did in April 1986, This, however, is undoubtedly a serious blow to the credibility of the program.

25

In the short run the government is assumed to have sufficient influence on wages within the framework of its incomes policies.

26

One may add to equation (7) a component which depends on new informa tion obtained in the current period, like an announcement of a new government plan, but this is not essential for the presentation.

27

More generally, π=F()(+)()(C,πI,E).

article image

28

The effect of π in equation (9) is assumed to dominate.

29

This derivation is based on the (stability) assumption that (∂U/∂E)/(∂E/∂C) < 1, meaning that the positive effect of C on U through E (that is, an increase in C reduces Ys and, therefore, reduces E which increases the need for controls) is less than unity.

30

This unemployment will enable some improvement in the trade balance as imports fall.

IMF Staff papers: Volume 34 No. 3
Author: International Monetary Fund. Research Dept.