Lessons from Experiences with High Inflation PDF

Lessons from Experiences with High Inflation Author(s): Rudiger Dornbusch Source: The World Bank Economic Review , Jan., 1992, Vol. 6, No. 1 (Jan., 1992), pp. 1331 Published by: Oxford University Press Stable URL: https://www.jstor.org/stable/3989792 REFERENCES Linked references are available on JST...

Lessons from Experiences with High Inflation Author(s): Rudiger Dornbusch Source: The World Bank Economic Review , Jan., 1992, Vol. 6, No. 1 (Jan., 1992), pp. 1331 Published by: Oxford University Press Stable URL: https://www.jstor.org/stable/3989792 REFERENCES Linked references are available on JSTOR for this article: https://www.jstor.org/stable/3989792?seq=1&cid=pdfreference#references_tab_contents You may need to log in to JSTOR to access the linked references. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to The World Bank Economic Review This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. 1: 13-31 Lessons from Experiences with High Inflation Rudiger Dornbusch In economies where price control has been the rule, the most serious concern may be recognition of the inflation problem. Beyond the initial correction of subsidies there is the broader issue of the risk of a serious inflation. This article looks at the problem of high inflation in developing countries in Europe and Latin America and draws lessons from historical experience. It analyzes the dynamics of the interaction among deficit finance, institutional innovation in financial markets, dollarization, and the s'hortening of wage contracts in high-inflation situations. When stabilization is undertaken, there is neither immediate, spontaneous resumption of longer adjustment periods for wages and prices nor instant increase of real money demand to noninflationary levels. Incomes policy-freezing exchange rates, wages, and prices-is advocated as an effective supplement to the inevitable budget cut to make up for institutional inertia and facilitate the start of the stabilization process. In early 1991 Argentina, Brazil, Peru, Poland, and Yugoslavia were in the midst of extreme instability or at best in the early stages of stabilization. Another group of countries, including the Soviet Union, Romania, and Bulgaria, were on the verge of slipping into high or even extreme inflation. A third group had already run the course and stabilized, as did Bolivia and Israel, or had avoided the extreme experience and opted for stabilization early and decidedly, as Mexico did. The evidence from some 20 experiences with high inflation establishes that the similarities between the experiences of various countries become sharper and clearer and the differences less significant as the inflation rate rises. The particu- lar mechanism by which monetary expansion occurs may differ-say, deficits of state enterprises rather than of a particular ministry-but the general pattern that runs from deficits to an expansion of money and credit is broadly the same, as are the dynamics of inflation. Of course the experience in the post-communist economies is special in that it starts from repressed inflation, but even that is not very. different from experiences in Argentina or Brazil, where cycles of price controls and hyperinflation are now common (Cardoso 1991; Dornbusch, Sturzenegger, and Wolf 1990). The experience of Poland and Yugoslavia, and Rudiger Dornbusch is Ford International Professor of Economics at the Massachusetts Institute of Technology. ? 1992 The International Bank for Reconstruction and Development/THE WORLD BANK 13 This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms 14 THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. 1 the extraordinary problems already apparent in the Soviet Union, Bu Romania, suggest that discussion of high inflation is timely (Commander and Coricelli 1992; Coricelli and Rocha 1990). I. LESSONS FROM HISTORY Should hyperinflation make policymakers opt for zero inflation at any price? Or is there room in between, with cost-benefit analysis and with the lessons from a rich inflation experience across time and space? Three questions are of special interest in the context of inflationary instability and stabilization. First, how does a country fall into hyperinflation? Second, what is necessary to stop high inflation and return to nornal growth? And, third, will Eastern Europe and the Soviet Union soon resemble Latin America, with some success stories, a few countries on the verge of high or extreme inflation, and some countries experiencing hyperinflation? The statistics are, of course, open to question, but the size of the Soviet Union's budget deficits in the past few years indicate the problem ahead (table 1). The data show persistent and increasing deficits. Inflation is still negligible in official markets, but the stage is set for a dramatic inflation unless both the overhang and the deficit are addressed at the outset of any attempt to restructure the economy. The answers to these questions can to a large extent be discerned from histori- cal experiences of high inflation (Bruno and others 1988; Yeager 1981; Dornbusch and Fischer 1986; Dornbusch, Sturzenegger, and Wolf 1990; Dornbusch and Wolf 1990). Some of the lessons from the past are described below. The Similarity of Inflationary Experiences across Countries It is a mistake to believe that the problems of a particular country are unique. But it is common that policymakers in Brazil, Peru, or the Soviet Union cannot accept that the experience of their country is not unique in the essential facts of inflation. Yet their situation is not substantially different from that of the 20 Table 1. The Soviet Budget Deficit, 1985-90 1985 1986 1987 1988 1989 1990" Percentage of GNP Revenues 47.3 45.8 43.6 41.7 41.0 42.8 Spending 49.7 52.0 52.0 51.0 49.5 50.6 Budget deficit 2.4 6.2 8.4 9.2 8.5 7.9 Percentage per year Retail inflation 2.0 1.3 0.6 2.0 4.8 Money incomes 3.6 3.9 9.2 13.1 14.5 Not available. a. Plan and estimates. Source: IMF, World Bank, OECD, and EBRD (1990). This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms Dornbusch 15 other countries where policymakers also thought that their unique situation ordinary economics aside (Dornbusch and Edwards 1991). The Danger of Complacency Hyperinflation is not around the corner whenever there is a budget deficit. But inflation can easily become a habit-and from there an unstable process. Com- placency comes at a disastrous price: society falls apart as the middle class disappears and society is divided between those who know how to get ahead with inflation and those that fall behind (see Guttman and Meehan 1976 and Fergusson 1975 for dramatic descriptions of the German experience). Pauperization of the middle class rapidly corrodes social institutions. Public administration, the tax system, and all social relations become undermined by corruption and fraud. The middle class revolts against the state, and the poor revolt against property. Participatory Democracy and Instability Political change toward a more participatory democracy has not been the traditional vehicle for stability. Political change may carry with it the expectation of an improvement in opportunities and living standards. An inflationary response to raised expectations occurred in Europe in the 1920s, in the Soviet Union in 1919-21, during the Allende period in Chile, and with the growth of Solidarity in Poland. Destructive inflation may accomplish many things that had been considered politically impossible. The destruction that takes place calls for particularly stern measures to rebuild confidence and stability. Because democratic institutions do not facilitate hard choices, democratic countries have almost invariably implemented special procedures to adopt and implement the hard measures necessary for stabilization. The arrangements differ from national unity governments (Israel in 1985) to restricted special powers for the executive (Poincare in France in 1926) or special parliamentary committees charged to interact expeditiously with the executive (Germany in 1923). In the end the politically impos- sible gets done because the destruction brought about by uncontrolled inflation is so devastating that it forces cooperation. Fiscal Austerity Stabilization does not imply that zero inflation must be achieved at any cost. Some policies incorporate a moderate rate of inflation because there is a steeply rising cost of disinflation. Policies that do not set a limit on inflation or policies that merely repress it for a while do not encourage confidence and stability. Policymakers need to have a good grasp of the role of control of the budget and incomes policy in stabilization. Without fiscal austerity stabilization cannot last; without incomes policy it cannot start. This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms 16 THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. 1 Table 2. Recent European Experience with High Inflation, 1986-90 (percentage per year) Year Hungary Turkey Poland Yugoslavia Soviet Union 1986 5.2 34.6 17.7 89.8 2.0 1987 8.7 38.8 25.2 120.8 1.3 1988 15.6 75.4 60.0 194.1 0.6 1989 16.9 69.6 251.1 1,239.9 2.0 1990 28.3 60.3 S85.8 583.1 4.8 Source: IMF, International Financial Statistics (1990). Structural Reform and External Support A separate set of issues concerns the transition from stabilization to growth What kind of policies at home and abroad can help to decrease the risk o protracted stagnation? Structural reform and external support play a role reassuring potential investors and thus moving the economy to growth (Dornbusch 1991). II. VARIETIES OF INFLATIONARY EXPERIENCE Table 2 shows inflation rates for several European countries in the range from moderate to acute and extreme inflation. Table 3 shows average annual inflatio rates in industrial and developing countries. High inflation is a problem of the developing countries of Europe and Latin America; it is not endemic to Africa, Asia, or industrial countries. Extreme inflation (hyperinflation) is rare. The generally accepted operationa definition of hyperinflation proposed by Cagan (1956) sets the benchmark at a inflation rate of 50 percent a month (12,875 percent at an annual rate): The term hyperinflation must be properly defined. I define hyperinflation as beginning in the month the rise in process exceeds 50 percent and as ending in the month before the monthly rise im pnces drops below that amount and stays below for at least a year. The definition does not rule out Table 3. Inflation around the World, 1970-89 (percentage per year) Developing countries Latin America Industrial Middle and the Year countries Africa Asia Europe East Caribbean 1970 5.6 5.4 6.5 3.0 12.4 1979 9.2 15.4 7.5 20.9 10.2 50.2 1988-89 4.7 23.3 11.8 126.0 17.0 208.0 - Not available. Source: uws, International Financial Statistics (various This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms Dornbusch 1 7 a rise in prices at a rate below 50 percent per month for the intervening months, and many of these months have rates below that figure. Until recently there were few cases of hyperinflation in modem history. Bu now, with fresh cases emerging in Latin America and possibly in Eastern Eu rope, the phenomenon is more pervasive. The distinction between hyperinflation and cases of lower and yet extrem inflation is somewhat arbitrary. Whether the inflation rate is really 50 percen month or only 20 does not make too much difference because in either case inflation will be the dominant factor in the economy and will overshadow most other issues. Countries experiencing inflation rates of 10 or 15 percent a month for any length of time are moving toward hyperinflation. Table 4 shows the movement into high inflation in Argentina, Bolivia, Brazil, Mexico, Peru, and Israel in the 1980s. Unlike the inflation experiences associated with war dislocation or civil war, high inflation in these six countries is rooted in domestic mismanagement and, to some extent, in external shocks. In each case the transition to extreme high inflation started off with quite moderate rates, then suddenly gathered speed and became extreme. Only Mexico cut the process short and stabilized before all the mechanisms of instability could gather force. Stabilization cannot afford to be weak. "Soft measures do not create hard currencies" the German authorities said in 1948 when a drastic monetary reform had to be administered (Dornbusch and Wolf 1990). The lack of thorough reform in Argentina, Brazil, and Peru shows up in the continuation for more than five years of off-and-on-again extreme inflation, which of course is accompanied by a dramatic decline in economic activity and the standard of living. III. THE SOURCES AND DYNAMICS OF HIGH INFLATION This section discusses the interaction of financing requirements and the financial structure by assuming full wage-price flexibility. It focuses on the role of contracts in the inflation process and the dynamics of the interaction among Table 4. Recent Experiences with High Inflation, 1981-89 (percentage per year, December to December) Year Argentina Bolivia Brazil Mexico Peru Israel 1981 105 29 106 28 75 117 1982 165 133 98 59 64 120 1983 344 269 142 102 111 146 1984 627 1,281 197 66 110 374 1985 672 11,748 227 58 163 305 1986 91 276 145 86 78 48 1987 132 15 230 132 86 20 1988 343 16 682 114 10,205 16 1989 3,079 15 1,287 20 3,390 20 1990 2,314 Source: 17 IMF, 2,938 27 7,482 International This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms 16 Financial S 18 THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. 1 deficit finance, institutional innovation in financial markets, dollari shortening contracts. Explosive inflation arises from the disintegratio ing of several institutions. A framework of the main determinants highlights the roles of budget finance, tax and financial institutions tracts in creating high inflation. The following analysis not only ide determinants of inflation but explains the mechanics of the very shar tion that has been witnessed on several occasions. Deficit Finance There is considerable controversy in high-inflation countries about the exact, or even the approximate, size of budget deficits. Reliable public data, covering an extended period of time in a comparable fashion, are simply unavailable. Various series differ in their coverage of the public sector, in the distinction between budget and cash bases, and in the inclusion of certain expenditure items, especially with respect to the quasi-fiscal deficit of the central bank. Both the adjustment of velocity and the presence of alternative means of financing the deficit (foreign borrowing, use of reserves, and domestic debt finance) help explain the lack of a tight link between inflation and the deficit. Controversies arise about the reason for the budget deficits, their endogeneity as a result of inflation, and their amplification by financial adaptation. That the actual outburst of inflation is often triggered by a foreign exchange crisis does not alter the fact that high inflation is a fiscal phenomenon. The most common view asserts that high inflation is the result of budget deficits. If the government spends more than it receives in tax collection, the remainder is financed by creating money. That means more money-too much money-chasing too few goods with the predictable outcome of inflation. This view needs considerable refinement to be entirely correct. Three directions of correction are essential. First, there is some room for noninflationary deficit finance. Second, deficits can be financed by debt. Third, there is a channel of causation that runs from inflation to deficits, as well as the other way around. A model of these important qualifications is given by equation 1. The deficit can be financed with high-powered money, with domestic debt, or with foreign debt: (1) gY =M/P + B/P + B* e/P where g is the deficit ratio, Y is real gross d domestic base money, B and B* are domestic a rate (domestic currency over dollars), and P variable denotes the rate of change.) It is clear borrowing from abroad or at home, thus entir money stock, at least for the time being. Focus entire deficit is financed by money will show h ated by such a system. This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms Dornbusch 19 Inflationary finance. Financing deficits by money creation mean money that is not demanded at the current level of prices must be for public by inflation. In a growing economy some extra real money bal demanded in order to finance the growing level of transactions. B that, the demand for nominal money expands only to the extent that erodes the purchasing power of existing real balances. To restore their real balances (at least partially), the public has to add to nominal money holdings. Thus inflationary finance automatically creates a demand for the money issue that finances the deficit. Keynes (1923, p. 37), in his splendid description of the inflation tax, noted the scope for inflationary finance even in a country with the poorest economic and political conditions: A government can live for a long time, even the German government or the Russian government, by printing paper money. That is to say, it can by this means secure the command over real resources, resources just as real as those obtained by taxation. The method is condemned, but its efficacy, up to a point, must be admitted.... so long as the public use money at all, the government can continue to raise resources by inflation... a government can get resources by a continuous practice of inflation, even when this is foreseen by the public generally, unless the sums they seek to raise in this way are very grossly excessive.... What is raised by printing notes is just as much taken from the public as is a beer duty or an income tax. What a government spends the public pays for. There is no such thing as an uncovered deficit. But, as Keynes has noted, significant inflation reduces the amount of money people choose to hold, because they will substitute toward assets that are more inflation proof. Thus, just as high taxation erodes the tax base, high inflation leads to a reduction in real balances and hence to an increase in the rate of inflation necessary to finance a given deficit. Moreover there may be a maximum amount of resources the government can extract. The long-run relation between the money-financed budget deficit and the rate of inflation is shown in equation 2 (for a derivation, see Dornbusch 1985): (2) r = (cag - y)/(1 - g), 1 >i ,g where gr and y are the rate of inflation and the grow a represents the noninflationary level of velocity velocity to the rate of inflation. This equation sh financed by money creation, there is inflation. B tionary impact of a given deficit can differ wide structure and the growth rate of output. There are three key points of this relation. First, higher the growth rate of output is. When outpu This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms 20 THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. 1 demand for real money. Accordingly there is room for some extra issued without introducing the risk of inflation. Second, inflation larger the budget deficit is. Moreover this relation is nonlinear. A ment tries to finance a larger deficit, the required rate of inflati steeply. Depending on the particular form of the money demand eq may even be a maximum deficit that can be financed by money. G that range implies hyperinflation. Third, the inflation rate depend ity parameters in equation 2. The higher is the level of noninflati (that is, because of dollarization) the higher the rate of inflation as any given deficit. A high degree of responsiveness of velocity to implies a larger rate of inflation. The increase in inflation brought about by a one percentage poin the deficit is higher, the higher are inflation and the budget defic one starts. Inflationary finance thus exerts a very powerful impact it is used in large doses or in an environment where a high level of strong responsiveness of velocity to inflation, leaves little scope fo tax. Likewise, dollarization or a drop in growth bring about large the inflation rate, more so the higher the initial extent of deficit finan The Olivera-Tanzi effect. One of the striking effects of inflation i of the real value of taxation. If there is any delay between accrual of taxes, the inflaton in the interim will mean that the real value o is lower the higher the rate of inflation. With moderate inflation difference that 1987 taxes are paid in 1988. But when inflation is effect wreaks havoc with the real value of tax collection. Keynes, on the impact of inflation on the budget noted this point, as d Turroni (1937). Tanzi (1978) and others have recognized this effect cific context of Latin American inflation. The empirical importance is large whenever inflation is high and tax collection lags are long there is no provision for tax indexation. External shocks and inflation. Suppose, as is the case in Argentin public sector has a large external debt and an external debt sho Specifically, assume that before the disturbance any existing exter rolled over with interest fully capitalized through automatic "new that there is no domestic debt. Let d* be the flow of external (measured as a percent of gross national product [GNP]), and thus deficit ratio that is financed by money creation. Thus, (3) g= (r) + d From equation 3 reduced ac ments implies that external d money creation. The country else finance the purchase of This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms Dornbusch 21 government will issue more money to finance the purchase of foreign for interest payments (assuming, of course, that there are no expenditu tax increases). Second, there will typically be a real depreciation in improve the external balance. The increase in inflation resulting from an external financing disru larger, the larger is the debt service shock and the real depreciation, b depends on the responsiveness of velocity to inflation and on the degre increased inflation erodes real tax collection. Each of these factors wil the inflationary impact of the debt shock significantly. The "balance of payments school" would argue that external balan lems and the resulting depreciation of the exchange rate are the prim of the deficit. By contrast, the "quantity theory school" would point t deficits and their financing by money creation as the reason for inflatio money is the essential ingredient in reconciling the quantity schoo balance of payments doctrine. Not surprisingly, suspension of reparat ments in Germany and of debt service in Bolivia in 1985 were essentia the stabilization of inflation. In Argentina involuntary external debt ser 1982 became an important source of inflation in exactly the manner t of payments school emphasizes. Deteriorating terms of trade further ag the external debt shock by forcing real depreciation and hence an incre real value (in terms of GDP or the tax base) of the existing external debt Endogenous financial innovation and liberalization. The interest th tional depository institutions can pay is typically controlled. There ma outright limitation on interest rates, or else institutions may be require reserves or government debt at controlled rates. These restrictions mak tions unable to compete in financial markets where nominal interest r nearly reflect the ongoing inflation. New, unregulated financial institut offer depositors higher interest rates spring up and thus draw custom from traditional depository institutions. There is a fall in the ratio of tional money tO GDP. The government loses part of its inflation tax b hence equilibrium inflation increases. The government may aggravate m when it responds to the increasing inflation by raising reserve require forcing traditional banks to hold government debt. Governments often promote this process, most obviously under the financial liberalization. Since inflation is a tax on money (or commercia non-interest-bearing reserves), financial liberalization means that the p avoid the tax on money. Financial liberalization may take the form of bearing deposits or formal dollarization, each of which reduces the dem high-powered money; velocity rises and so does the inflation rate a with the financing of a given deficit by money creation. Thus, from an point of view, financial repression, not liberalization, is appropriate. Fi liberalization requires that extra tax revenue be available to avoid the in ary impact of a reduction in the captive inflation tax base. Governmen This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms 22 THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. I condone dollarization likewise promote inflation. Dollarization is cap equation 2 by both the coefficients a and ,B. The shift from the dom tary base into dollars reduces the base for the inflation tax and hence must increase inflation. One is tempted to explain inflation experiences in some countries by dollarization and new financial intermediaries. Thus countries with stronger dollarization have higher inflation. A government that experiences some inflation and makes dollarization easier will experience even more inflation. However, dollarization is also a response to inflation. The financial adaptation to inflation intensifies the inflationary process. In response to inflation there is a flight from money into interest-bearing financial assets, to the extent that they exist at all, or into dollars. But there is also an institutional adaptation: financial institutions spring up that offer protection against inflation. The better the protection they offer, the more substantial the flight from money or the larger the increase in velocity. Timing. Inflation controls and managed exchange rates can slow down the buildup of inflation. The loss of resources, or forced saving, is an alternative mode of financing. However, deficits do imply money creation and inflation. Moreover the longer the delay the more dramatic the inflationary explosion. This is especially the case when a managed exchange rate and reserve losses have financed the deficit in a relatively noninflationary manner. When these mechanisms are no longer possible, there will be a sudden shift toward the inflation tax at the same time that a real depreciation is required. The Role of Contracts in the Inflation Process As inflation accelerates, contracts shorten, and that shortening of contracts is itself a factor that causes inflation to accelerate. Institutional wage-setting mechanisms often rely on a fixed contract length, with wage adjustments occurring at specified intervals. The adjustments are based on the cumulated increase in prices since the last adjustment. For example, earners might receive full compen- sation for past actual price increases at regular intervals, say yearly. Now suppose there is a shift to six-month intervals. There are two interesting questions. The first concerns the dynamics of shifting to shorter contacts. What is the threshold for inflationary erosion of wages that causes the shift, and what makes it economywide rather than just for a particular firm? The other interesting question is what happens when the frequency of adjustment increases yet further. This point has been developed especially by Pazos (1972). It is of interest here because contract deterioration is one of the important charactenrstics of an accelerating inflation and because exchange depreciation often plays an important role in setting off the process. If nominal wages are adjusted only periodically, the real wage follows a sawtooth pattern. On each adjustment date the nominal wage is increased by the cumulated inflation since the preceding adjustment. Until the next adjustnent This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms Dornbusch 23 date the real wage declines as the ongoing inflation erodes the purchasin of the constant nominal payments. By the end of the adjustment interv wage has declined below its period average. The higher the rate of inf moreover, the lower the average real wage, given the interval of adjust In a system of full, but lagged, indexation, the real wage can be cut o moving to a higher rate of inflation. Thus, once-and-for-all depreciatio currency immediately raises the rate of inflation and erodes existing But wage indexation ensures that inflation must be pushed to an even h so that there is always some group of wage earners whose wages are sti the increasing rates of price increases. The same principle applies to th of subsidies undertaken to correct the budget. Measures undertaken to competitiveness or the budget can be effective only if they achieve a real wage, but because of full indexation that cut can take place only if is allowed to run at a higher rate. This mechanism often sets the stage for inflation explosions. Consider a country that requires adjustments in the budget and external com- petitiveness. Suppose that the government lacks the political force to suspend full indexation, so that the removal of subsidies or a real exchange depreciation will speed up the inflation rate. Workers in the middle of their contracts, for example, will find that their real wages fall below what they consider a minimum standard of living. They cannot borrow, even in perfect capital markets. Hence they will call for a shorter interval between wage adjustments in order to recover the real wage losses imposed by inflation. They will ask for an advance of what they think is due. If the economy does, in fact, shift from, say, sixmonth to three-month indexation intervals, the inflation rate will simply double (Simonsen 1986). But once the contract structure has moved to a three-month scheme, it is unlikely that the indexation structure will return spontaneously to a longer interval, even if shocks are favorable. And there is nothing to make the three-month interval more stable than the six-month interval that was just aban- doned. New shocks will shift the economy to even more frequent adjustments and hence to correspondingly higher rates of inflation. At this stage the exchange rate becomes critical. The dramatic escalation of inflation, seemingly out of proportion to the disturbances, arises from the endogeneity of the adjustment interval. This is due not so much to the direct impact on inflation of corrective exchange rate or price policies. It occurs because increases in inflation, which may be minor but highly visible (such as a 10 percent devaluation over and above a purchasing power parity rule or a removal of bread subsidies), lead to an increase in the frequency of wage adjustments, which brings on a much higher inflation rate. The endogeneity of adjustment intervals is the mechanism that connects small inflation disturbances with a shift from 50 to 100 percent inflation or beyond to hyperinflation. As long as full indexation remains, even seemingly small corrections are a dramatic threat to the stability of the inflation rate and hence may not be worth undertaking. Incomes policy designed to avoid inflationary explosion must avoid accelerating the frequency of adjustments. This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms 24 THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. 1 Dynamics The actual dynamics of the economy emerge from the interact inflationary aspects of deficit finance and the contracting proc equilibrium may not actually exist. When inflation rises significantly nently, institutions adapt. In doing so, they help to increase infl conditions of extreme inflation, institutions break down. There is a nearabandonment of domestic money, which means the government must continue to increase inflation to get any seigniorage. Contracts are set for a shorter duration and are more likely to be dollar-based. In the analysis of inflationary experiences, it is common to assume adaptive inflationary expectations (Cagan 1956). There appears to be a significant sluggishness in the initial phases of high inflation as well as a subsequent acceleration, which suggests exactly such an expectations mechanism. Adaptive inflationary expectations are often the key model device to slowing the impact of money on inflation. An alternative and perhaps more accurate model focuses much more on the dynamics of deterioration in contracts, both in the goods and labor markets, and on the inflationary adaptation of financial institutions. Insti- tutional dynamics seem to offer a more suitable framework for studying high inflation. As economic institutions break down and time intervals for contracts and adaptation to inflation become shorter, the inflation process becomes explosive. The economic time horizon shrinks along with contracts and maturities of finan- cial assets until, when the economy converges to a spot market with dollar pricing, the budget or external balance deficit leads to hyperinflation. Hyperinflation is inevitable because the inflation tax, with sufficient financial adapta- tion, can be almost totally evaded, and hence the budget deficit cannot be financed. The Olivera-Tanzi effect, the shortening of contracts, and financial adaptation all react in a perverse way (from the perspective of stabilization) in that they widen the deficit and accelerate explosively the inflation process. IV. STABILIZATION The preceding discussion helps to explain why stabilization is difficult and often takes more than one attempt to succeed. In the process of high inflation all institutions break down. When stabilization is tndertaken, there is neither inmediate, spontaneous resumption of longer adjustment periods for wages and prices nor an instant increase of real money demand to noninflationary levels. As a result more sizable adjustments in the budget are required, and more dramatic measures are necessary to create the confidence that stabilization will, in fact, last. Because the fiscal measures have to be particularly large they are also particularly difficult and hence often can not be sustained. When they fail, inflation returns instantly at exceptionally high levels because institutional inertia had not recovered. This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms Dornbusch 25 Incomes policy-freezing exchange rates, wages, and prices-can be an effective supplement to the inevitable budget cut. It makes up for institutional inertia and, to that extent, gives a government a better chance to start stabilization. But, as is clear from the experiences of Argentina, Brazil, and Peru, failure to correct the budget implies that high inflation will soon return. The decline in the ratio of Ml to GDP is not typically fully reversed in the initial stabilization. As a result financing even a moderate deficit is much more inflationary than it was before the experience of extremely high inflation. This hysteresis effect of high inflation (similarly apparent in contracts, pricing, and tax collection) sharply reduces the chances of stopping inflation with anything short of a dramatic budget cut. The task of stabilizing inflation involves stopping inflation quickly and avoiding the resurgence of inflationary pressures. To end inflation by incomes policy is relatively easy, but to keep it down requires fiscal support. The chief mistake in stabilization policy is to rely too much on incomes policy-fixed exchange rates and wage and price freezes-and too little on fiscal austerity. Such programs quickly lead to repressed inflation and overvaluation, in which tight monetary policy is introduced to sustain the imbalances. Ultimately that does not work, and another inflationary explosion offers the starting point for yet another stabilization. Argentina offers a clear example of this process with its successive failed stabilization programs in the past five years. Budget Balancing Budget deficits are the ultimate source of inflation. When external financing or the domestic capital market cannot finance deficits, then the deficits must be adjusted. Two questions immediately emerge. The first is how large a deficit is consistent with stability; the second is how to cut deficits down to the required size. Argentina, Brazil, and Peru failed to adjust fiscal deficits in the aftermath of their 1985 heterodox stabilization. Wage-price controls and fixed exchange rates quickly stopped inflation and raised the political popularity of the president. The resulting possibility for fiscal stabilization was, however, not used. Instead the deficits persisted and were financed by creating money. As a result, inflation continues. Quasi-fiscal deficits. The starting point for budget balancing is the need for a transparent accounting of the consolidated government. Because the issue is control of monetary emission, it is essential that the central bank's "quasi-fiscal" deficit be part of the accounting. An accounting framework is needed for the consolidated government sector, including not only the central govermnent and the central bank, but also state enterprises and local government. Extreme inflation invariably reflects deficits financed by writing checks on the central bank, whether it be by provincial authorities, as in China; by state enterprises, as in Yugoslavia; or by a government bank, as in Brazil. The deficits may have as This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms 26 THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. 1 a counterpart purchases of foreign exchange, payments of wages, deficits railroads, external debt service, election spending by a governor, or corruption. In any case a deficit leads to money creation. Quasi-fiscal deficits arise from loans by the central bank at subsidized losses on foreign exchange operations in the form of guarantees, forwar tracts, or simply purchases at a high rate (under multiple exchange rate sales at a low one. In Peru in 1986-87, for example, exchange losses accou for 2.3 percent of GDP. But central bank losses also arise from credit ope Subsidized credit is no different from any other subsidy; in fact credit s have long ceased being investment subsidies and have become simply a p tion subsidy that finances wages when prices are not allowed to reflect co Revenues. The second point on the reform agenda is to achieve a produ tax system. The reform must raise revenue on a substantially larger scale more efficiently. Increasing the yield of the tax system is dictated by the eliminate deficits. Inflation stabilization makes an immediate contribution cause the inflationary erosion of revenues ceases. But that is only a sma perhaps as much as 2 percent of GNP in revenues. The major effort mus reconstructing the tax system, including stopping the corruption and e that now undermine the collection of taxes as well as introducing and d strating mechanisms to increase compliance. The complacent acceptance of vasive tax evasion is the most regressive aspect of the Latin American t tem, and it must be carefully watched as Eastern Europe moves to taxat the chief source of government revenue. The revenue effort must concentrate both on collecting taxes and on el ing subsidies in public sector enterprises. Many countries now have perv systems to manage public sector prices, both to control inflation and to t prevent a decline in real wages. The implied revenue losses are extraordi large and cannot be justified by any of the objectives. For example, in Peru controlled telephone rates have reduced the real price of the service to one-tenth the 1985 level. It is difficult to argue that telephone rates have an important incidence either on inflation or on welfare of the poor, but they do contribute to deficits. Governments should therefore eliminate totally all subsidies. The resulting revenue gains must be applied to eliminate inflation, which in itself raises the welfare of the poor since inflation is a highly regressive tax. Part of the revenues should also be used for targeted food and employment programs for the poorest groups. Beyond cutting all subsidies and raising revenue under the existing governments should use the crisis to institute a more efficient tax system should produce more revenue with fewer distortions, which inating the pervasive exemptions from direct taxes and raising the ra levels. A comprehensive value added tax of 15 percent, with a 5 per charge for luxuries, might be the starting point for discussion. This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms Dornbusch 27 Government spending. For many observers the right direction for a is to cut government spending, not to raise taxes. Inefficiency in gove pervasive, and public sector employment in many countries is unjustifi But there is no presumption that the fiscal problem can be solved co massive firing of public sector employees and privatization. There is restructure public sector spending, from consumption to investment a ductive services. But as for the level of spending, it certainly is not ex More of the spending absolutely and relatively should fall on infrast health, and social services for the poorer groups. The current compos spending is not only unproductive but probably also regressive. Most of infrastructure investment could be done by the private sect certainly the case, for example, for telephone services, but also for pu port and even the road system. Mexico is now exploring such options substantial success. But the fact remains that infrastructure spending be the priority in balancing the budget, certainly not at the expense of education. Incomes Policy Fiscal austerity is the essential aspect of stabilization, but incomes policy is an important, desirable component. Incomes policy is designed to bring about a rapid, coordinated end to inflation. In a hyperinflation, incomes policy amounts to fixing the exchange rate. Because price setting is geared to the movements of the dollar, the move to a fixed exchange rate is enough to break the inflation an the expectation of inflation. But when annual inflation is only 100 or 200 percent, incomes policy is both more essential and more complicated. Without incomes policy ending inflation by demand management alone would create an extraordinary depression. The current inflation will be a weighted average of cost increases that are equal to past inflation, which enters costs by explicit or implicit indexation, and the current rate of exchange depreciation, e, plus a cyclical component, which is denoted by "GAP": (4) r = r-1 + (1-T) e+ *GAP. Because of the inertia represented by th or implicit indexation, current inflation unless the government breaks the pro means-fixing the exchange rate and stop will have to intervene in loan contracts to reduce real interest burdens that otherwise would result from the unanticipated decline in inflation, and intervention will be required in wage contracts. Because these contracts have periodic adjustments for inflationary erosion, a sudden ending of inflation requires intervention. Some wage contracts have to be rolled back, and others need to be adjusted upward. This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms 28 THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. 1 Exchange rate policy. Exchange rate policy assumes a strategi lization, as does pricing in the public sector. The starting point invariably a fixed exchange rate. But if inflation does not end c or later adjustments in the exchange rate and public sector pri The decision to abandon the fixed rate is a difficult one becaus government's acceptance of inflation as something inevitable. As a temptation to postpone exchange rate adjustment until a sign tion has developed. Overvaluation in turn creates an expectation of a devaluation real interest rates become necessary to stop speculation. High r in turn increase domestic debt service and worsen the budget. exchange rate adjustment does have to come, but often the ove gone so far that an outright exchange crisis and collapse are the end of the abortive attempt to practice a fixed exchange rate. The pragmatic answer is to move after two or three months to a crawling peg, depreciating the exchange rate at a pace that maintains external competitiveness. The risk of an overvaluation maintains short economic horizons and stands in the way of recovery. The right time for a crawling peg is very early because the government should try to preserve maximum competitiveness. Holding onto an exchange rate too long may yield an extra month of low inflation, but it also sacrifices competitiveness and therefore prejudices the return of growth. Indexation. A major stabilization decision regards indexation. The common view is that indexation is responsible for the inflation and that accordingly it should be abolished. Moreover governments should declare a zero inflation target rather than create mechanisms that make it easier to live with inflation. However, it is not necessarily true that without indexation there is inflation stability because inflationary shocks, such as public sector price increases and depreciation, are not fully absorbed into lower real wages. Without explicit indexation the government becomes the judge of what wage increases to grant. The wage becomes politicized, which means invariably larger rather than smaller wage increases, and wage increases come sooner rather than later. In fact in economies in which a government seeks to avoid explicit indexation, as in Brazil after 1985, inflation soon becomes more unstable and susceptible to a far more rapid escalation than had ever been experienced under indexation. Indexation is a mechanism that creates inertia and also preserves inertia. Reintroducing half-yearly indexation may therefore be a key step in establishing the expectation of low inflation. Once the wage is locked away, a very rapid resumption of inflation will not be expected. As a result horizons can lengthen far more effectively than under threshold provisions or in the absence of any kind of formal indexation. Monetary policy. Monetary policy does not play an independent role in stabilization; it is dictated by the budget and the exchange rate policy. Following This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms Dornbusch 29 stabilization real interest rates are too high. One could argue that to growth the economy needs reliquification. There is very limited room f uification, but that is best done by monetizing reserve inflows rather t deficit finance or domestic credit creation. The alternative is an overly firm commitment to a zero inflation targ policymaker might be tempted to make monetary policy (and the exchan do what fiscal policy has not achieved. The risk is a long period of extra narily high real interest rates and possibly an exchange overvaluatio might stop inflation, but they also will destroy the real economy. The only lasting way to bring about low real interest rates and to achieve m inflation is by a balanced budget (including state enterprises, except wh are financed in the capital market) and a very competitive real exchange V. CONCLUDING REMARKS: PRIORITIES Without financial stability, economic reconstruction and growth will simply not occur. If inflation is high and variable, then it will be the most important issue. It will take up policymakers' precious time. It will tempt them into superficial remedies, which help in the short run but set back economic activity because they create uncertainty. It will also lead the private sector to focus on protection against inflation and government's arbitrary interventions. The first priority then must be to reduce inflation. A lastingly balanced budget is required to achieve financial stability. In economies in which pnrce control has been the rule, the recognition of the problem of inflation may be the most serious issue. There has been no experi- ence with inflation, and most attention focuses on the popular revolt against removing some subsidies. But beyond the initial correction of subsidies there is the broader issue of the risk of a serious inflation. Serious inflation can emerge either because there is an initial monetary overhang or because the subsidy correction does not go far enough. Deficits remain, and money creation starts to interact with corrective inflation. (The Soviet Union is an obvious case in point.) In countries with major fiscal problems it is not politically impossible to make the necessary adjustments; it is politically difficult, but the adjustments will ultimately be made. The only question is how large the loss in the standard of living has to be before it is done and how much time and political capital will be lost (Blanchard and others, 1991; Fischer and Gelb 1990, and Dornbusch 1990). REFERENCES The word "processed" describes informally reproduced works that m monly available through library systems. Bresciani-Turroni, C. 1937. The Economics of Inflation. London: Allen & This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms 30 THE WORLD BANK ECONOMIC REVIEW, VOL. 6, NO. I Blanchard, Oliver, Rudiger Dornbusch, Richard Layard, and Lawrence Su Economic Reform in the East. Cambridge, Mass.: MIT Press. Bruno, Michael, Guido Di Tella, Rudiger Dornbusch, and Stanley Fischer, eds. 1988. Stopping High Inflation. Cambridge, Mass.: MIT Press. Cagan, Phillip. 1956. "The Monetary Dynamics of Hyperinflation." In Milton Friedman, ed., Studies in the Quantity Theory of Money. University of Chicago Press. Cardoso, Eliana. 1991. "From Inertia to Megainflation: Brazil's Macroeconomic Policies in the 1980s." In Michael Bruno and Stanley Fischer, eds., Lessons of Economic Stabilization and Its Aftermath. Cambridge, Mass.: MIT Press. Commander, Simon, and Fabrizio Coricelli. 1992. "Price-Wage Dynamics and Inflation in Socialist Economies: Empirical Models for Hungary and Poland." The World Bank Economic Review, this issue. Coricelli, Fabrizio, and Roberto Rocha. 1990. "Stabilization Programs in Eastern Europe: A Comparative Analysis of the Polish and Yugoslav Programs of 1990." World Bank, Washington, D.C. Processed. Dornbusch, Rudiger. 1985. "Stopping Hyperinflation: Lessons from the German Experience in the 1920s." In Rudiger Dornbusch, Stanley Fischer, and John Bossons, eds., Macroeconomics and Finance: Essays in Honor of Franco Modigliani. Cambridge, Mass.: MIT Press.. 1990. "Priorities of Economic Reform in Eastern Europe and the Soviet Union." Massachusetts Institute of Technology, Department of Economics, Cambridge, Mass. Processed.. 1991. "Policies to Move from Stabilization to Growth." Proceedings of the World Bank Annual Conference on Development Economics- 1990. Washington, D.C.: World Bank. Dornbusch, Rudiger, and Sebastian Edwards. 1990. "The Macroeconomics of Populism in Latin America." Trimestre Economico 57 (January-March): 121-62. , eds. 1991. The Macroeconomics of Populism in Latin America. University of Chicago Press. Dornbusch, Rudiger, and Stanley Fischer. 1986. "Stopping Hyperinflations: Past and Present." Weltwirtschaftliches Archiv 122 (1, April): 1-14. Dornbusch, Rudiger, Federico Sturzenegger, and Holger Wolf. 1990. "Extreme Inflation: Dynamics and Stabilization." Brookings Papers on Economic Activity 2: 1-64. Dornbusch, Rudiger, and Holger Wolf. 1990. "Monetary Overhang and Reforms in the 1940s." Massachusetts Institute of Technology, Department of Economics, Cambridge, Mass. Processed. Fergusson, A. 1975. When Money Dies. London: William Kimber. Fischer, Stanley, and Alan Gelb. 1990. "Issues in Socialist Economy Reform." Massachusetts Institute of Technology, Department of Economics, Cambridge, Mass. Processed. Guttman, William, and P. Meehan. 1976. The Great Inflation. London: Gordon and Cremonesi. International Monetary Fund. Various issues. International Financial Statistics. Washington, D.C. This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms Dornbusch 31 Intemational Monetary Fund, World Bank, Organisation for Economic Co-operation and Development, and European Bank for Reconstruction and Development. 1990. The Economy of the Soviet Union. Washington D.C. Keynes, John M. 1923. A Tract on Monetary Reform. Reprinted by the Royal Economic Society, London, 1971. Pazos, Felipe. 1972. Chronic Inflation in Latin America. New York: Praeger. Simonsen, Mario. 1986. "Indexation: Current Theory and the Brazilian Experience'" In Rudiger Dornbusch and Mario Simonsen, eds., Inflation, Debt and Indexation. Cambridge, Mass.: MIT Press. Tanzi, Vito. 1978. "Inflation, Real Tax Revenue, and the Case for Inflationary Finance: Theory with an Application to Argentina." IMF Staff Papers 25 (September): 417-51. Yeager, Leland. 1981. Experiences with Stopping Inflation. Washington D.C.: American Enterprise Institute. This content downloaded from 92.202.12.4 on Thu, 02 May 2024 05:36:55 +00:00 All use subject to https://about.jstor.org/terms