Since the advent of floating exchange rates in the early 1970s, visible disparities have emerged in the balance of payments (BOP) of countries with weak and strong currencies. By 1976, the Federal Republic of Germany and Japan were experiencing large current account surpluses, low inflation rates, and appreciating exchange rates. Italy and the United Kingdom, on the other hand, had current account deficits, high inflation rates, and depreciating exchange rates. The experiences of these countries have been interpreted in terms of, and have led to a revival of interest in, the hypothesis that economies become trapped in either a vicious or a virtuous circle.
The vicious circle hypothesis takes many forms. The simplest one is that an initial disturbance which causes a depreciation of the exchange rate can set in motion an unstable process: exchange depreciation rapidly leads to higher domestic prices and costs, which in turn lead to further depreciation in the exchange rate and the cycle is recreated. (The virtuous circle is the antithesis of the vicious circle: an initial shock leads to exchange appreciation which slows down domestic inflation and in turn leads to further exchange appreciation.)
Most versions of the vicious circle hypothesis recognize that a country can move into an endless circle of inflation and exchange depreciation only if monetary policy is expansionary and accommodates wage and price increases. In the long run, this kind of policy will hardly increase output but will push prices up as wage and price adjustment becomes rapid. Many recognize that the authorities may have little choice but to adopt monetary accommodation because the inevitable high unemployment during a certain phase of adjustment may become politically unacceptable.
Much of the present interest in vicious and virtuous circles centers around finding why they occur, establishing whether certain countries are more vulnerable to them than others—either because of the structure of their economies or the political pressures faced by their governments—and reviewing what a country might do in order to avoid or escape from a vicious circle. This article, which is based on a wider theoretical and empirical analysis of the vicious circle phenomenon prepared by the author, examines the adjustment process within a national economy following an initial disturbance, in an attempt to throw light on some of these questions.
The circle begins
A vicious circle normally begins as a phase of a country’s adjustment process following a shock. This shock may be set off by domestic monetary policy—such as an expansion in the money supply made in an attempt to reduce the unemployment rate—or it may originate as a sudden increase in foreign prices, such as the oil price increase of late 1973. The following discussion will trace a country’s theoretical adjustment to an initial shock through its various phases.
In the first phase of the adjustment to a monetary shock, an expansion in the domestic money supply, at a given level of prices, will have two effects: to induce a fall in domestic interest rates and, because the yield on domestic assets has fallen below the yield on foreign assets, to set in motion anticipations of an exchange depreciation. Because domestic assets will have become less competitive, an actual (or incipient) capital outflow will take place, and the exchange rate will be pushed down. This exchange depreciation leads to a rise in the domestic price of imported goods and a fall in the foreign price of exported goods. These price effects will induce an increase in domestic output for two reasons: consumers go to local substitutes for imported products, and foreign demand for the domestic product also rises. Because the initial impact is to reduce unemployment and to strengthen the trade account, developments in this phase can be favorable.
In the second phase, called the stagflation phase because of the presence of inflation and lack of growth in the economy, the process of wage and price adjustment begins. The rate of wage inflation accelerates because the increase in the domestic price of imported goods pushes up the consumer price level and lowers the domestic real wage. It is to be expected that some unions will attempt to maintain their real wages. In addition, these inflationary expectations will be exacerbated by the accelerating domestic inflation caused by the increase in aggregate demand for domestic goods, now cheaper than their imported counterparts, and by the tendency for the costs of imported inputs and domestic wages to rise. Rising domestic prices and costs reduce the earlier competitive advantage of the country’s exports abroad; as this margin diminishes, domestic output will fall. It will also be accompanied by a deteriorating trade account and exchange depreciation.
If monetary policies are not relaxed in an attempt to reverse the decline in output, conventional classical theory suggests that the economy should move to a new equilibrium in a third phase. The fall in real domestic income and the rise in unemployment slows down wage and price inflation; as these forces abate, the trade account should improve and the exchange rate appreciate.
The initial impact of the adjustment to a foreign price shock, such as the 1973-74 rise in oil prices, can usually be found in the goods market. The first phase of adjustment is characterized by increasing unemployment, a deteriorating current account, and a downward tendency in the exchange rate. In the second phase, rises in import prices produce increases in domestic prices and wages and, in turn, a tendency for unemployment to continue to rise and domestic output to fall; this trend may continue until the advances in domestic prices and wages abate. In this phase, there is high inflation, unemployment, current account deficits, capital inflows (to cover the deficits and as a result of a combination of higher domestic interest rates and currency depreciation), and exchange depreciation. As the inflation process really slows down, unemployment should begin to stop rising, particularly as equilibrium is approached. Equilibrium will be reached when domestic inflation has fallen enough to offset the foreign inflationary impact and all real variables are at their original level. In the third phase, the trade account will begin to improve and domestic output will increase. In turn, unemployment and domestic inflation will continue to decrease, and the domestic inflation will soon fall enough to offset the inflationary impact of the improved trade balance.
If an economy is to make a smooth and stable adjustment to exogenous shocks along the lines we have described, several preconditions are likely to be necessary. First, capital has to be highly mobile and expectations about exchange rate changes have to be rational (and predictable). That is, market participants behave as though they are able to predict the path to equilibrium and act on these expectations. Thus, an expansionary monetary policy induces expectations of a fall in the exchange rate which will eliminate any expansionary effect because of the domestic inflationary effects. Second, domestic production has to be able to respond quickly to relative price changes; in addition, wages and prices must be able to fall, or at least to rise more slowly than the inflation rate in response to excess supply of goods and labor. Third, there must be a clear understanding that certain macroeconomic policies can contribute to a stable adjustment, and that other policies will hinder it. In the context of our assumptions, it would be an inappropriate macropolicy for the authorities to initiate a continuous monetary expansion to meet rising costs and wages; such a policy would have little effect on output but a major impact on prices. Such a policy choice could also worsen expectations and lead to unpredictable exchange rate changes. On the other hand, the authorities could introduce nonmonetary policies designed to increase output, a policy choice likely to lead to favorable exchange rate effects and to speed up the adjustment process.
Smooth adjustment has often not taken place in the real world for five main reasons. First, downward exchange rate movements have often led to sales of the currency at a price below the rate justified by current underlying economic variables. Domestic price rises resulting from these excessive exchange rate changes will thus be higher than they would otherwise be, adding to the domestic inflation.
Second, in small open economies, an increase in the price of imported goods may not mean that domestic substitutes will be cheaper and available to local consumers. Small open economies have a small share in world markets, a high proportion of imported goods in domestic output, and a narrow range of products. The impact of a depreciating exchange rate may thus be felt fairly rapidly in higher domestic costs and prices, which can depress output at home. If this occurs, there may be more, and longer periods of, unemployment than are likely to occur in larger and more diversified economies.
Third, relative price changes in many countries, even when they occur, are in fact slow to affect the volume of imports and exports. This sluggish response of foreign trade flows to changes in relative prices has been shown in the World Trade Model of the International Monetary Fund and is well documented in the current literature. In the short run, a sluggish response of exports may cause the current account to deteriorate after an exchange depreciation.
Fourth, a failure of real wages to fall can delay the rate at which the increasing numbers of unemployed can be profitably absorbed. This may lead to political pressures for the authorities to implement policies to reduce unemployment.
The choice of inappropriate macropolicies is the fifth main reason why a country may not adjust its BOP position smoothly to a long-run equilibrium. Indexation policies, for instance, can speed up the domestic inflationary process. Similarily, an increase in the domestic money supply may solve unemployment in the short run, but, unless the real wage falls, it will not produce an increase in real domestic output in the long run. Further, an increase in money supply during the period of rising prices and falling output will prevent the fall in real domestic income that is necessary before wage and price inflation can be expected to abate. Injecting more money into the economy will only lead the country eventually into another round of domestic spending, exchange depreciation, further inflation, and, as output is affected, to unemployment pressures. Thus, it can ensure that the cycle will begin all over again.
On the basis of this analysis, it can be suggested that three conditions seem to be necessary for a vicious circle to be set in motion: (1) the existence of a flexible exchange rate system, since without this a country would not be able to adopt an independent monetary policy; (2) some rigidity in the labor market which prevents the real wage from falling as domestic prices rise and, instead, helps to stimulate a more rapid rate of rise in nominal wage rates; and (3) political conditions which lead the monetary authorities to increase the money supply as wage and price inflation proceeds in an attempt to ensure that unemployment never rises sufficiently high for prices and wages to fall.
Why the circle occurs
Various attempts have been made in the past few years to explain why a vicious circle occurs. One simple answer which is sometimes advanced is that floating exchange rates alone can produce a depreciation-domestic inflation cycle. The argument against this view rests on the fact that both exchange depreciation and inflation respond to the same initial disturbance—a disequilibrating exogenous shock. Exchange rates are simply adjusting more rapidly than domestic prices in this situation, a sequence which gives the illusion that the exchange rate changes cause the domestic price changes.
There is a more sophisticated version of the hypothesis, however, which stresses excessive exchange rate movements. In this view, it is argued that an “overshooting” of the exchange rate pushes domestic prices higher than they would otherwise have been. Economic models can be formulated to show that some external shocks can indeed lead to a change in expectations about exchange rates which cause excessive exchange rate depreciation.
Both theoretical and empirical research shows that expectations can exert downward pressures on the exchange rate for some countries in ways which push the market rate below levels consistent with long-run domestic cost-price relationships; but this phenomenon by itself may not be held responsible for causing a vicious circle. Exchange rates are determined in the asset market over the short run. In order to assess whether these short-run effects are responsible for generating a vicious circle, it is necessary to consider whether the macroeconomic policies adopted by the country support, or run counter to, the necessary domestic economic adjustments.
Small country vulnerability
It has also been suggested that small open economies are more vulnerable to vicious circles than larger economies because their domestic prices rise faster following exchange rate depreciation. There is little doubt that in the short run small countries are less able to absorb exchange depreciation than large ones. The evidence in the table suggests that small countries may experience higher import and foreign export prices following depreciation (and therefore higher domestic prices) than large countries that exert some power over their import or export prices. Column 2 of the table shows that the Federal Republic of Germany, Japan, and the United States exert considerable power over the prices of their imports—a 1 per cent depreciation of the exchange rate leads to rises of only 0.6, 0.8, and 0.5 per cent in the domestic import prices of these countries after two quarters. The medium-sized and small countries, on the other hand, seem to experience almost the full impact of an exchange rate depreciation on domestic import prices after two quarters.
On the export side, the table shows that changes in exchange rates provoke completely offsetting changes in export prices within one or two years for small countries (for example, Belgium and the Netherlands—whose exports are concentrated in products for which there is perfect competition). The export prices of medium-sized countries (such as France, the Federal Republic of Germany, Italy, and the United Kingdom) reflect an exchange movement of between 70 per cent and 93 per cent within two years. Export prices of larger countries, such as Japan and the United States, in contrast, show less than 70 per cent of the exchange rate change. The results also show that the effect of import prices on domestic prices will be greater for small than for large countries as domestic wages and prices adjust to an exchange depreciation.
|Effect of 1 per cent change in||Mean estimate of first year effect on wage inflation of a 25 per cent rise in unemployment|
|Traded goods sector relative to GNP (1972)||Exchange rate on import prices over two quarters||Local currency import prices on consumer prices within one year||Exchange rate on export prices over two quarters|
|Germany, Fed. Rep. of||21||0.60||0.08||0.93||-0.91|
By itself, however, this tendency of small countries to experience larger short-run price effects following a depreciation is not sufficient to prove that small countries are more vulnerable than large ones to vicious circles. They may well also experience higher unemployment as the economy slows down and prices rise in the stagflation phase, but there are other factors that are more likely to be responsible for putting a country into a vicious circle following an exogenous shock. These include the resistance of labor to reduction in real wages, particularly over the period when unemployment is rising, and domestic policies that inject more money or aggregate demand into the economy to absorb the unemployed. Where a small country has a weak currency and a poorly functioning labor market, the evidence suggests it will be more vulnerable than a larger one to vicious circles: its exchange rate is more liable to excessive movements and there will be a larger impact of import prices on domestic prices. If prices and wages do not slow down as unemployment rises, these vulnerable countries will experience high unemployment over a prolonged period unless domestic policies are used at this stage to reduce unemployment. None of these factors alone is sufficient to sustain a vicious circle, although any one may lead a country to the initial stages of such a process.
There are two theories which have important implications for the role of the labor market in a vicious circle situation. One theory suggests that workers base their resistance to a reduction in the real wage on the concept of a “fair wage” derived from both their own past wage and the wages of others. Since people do riot easily change the concepts on which they base their behavior, it follows under this hypothesis that any attempt to reduce inflation by increasing unemployment is bound to be slow and painful. A second theory contrasts with the first: it proposes that resistance to reducing real wages will continue only as long as the authorities continue to expand the money supply and achieve additional output to absorb the unemployed. A restrictive monetary policy—announced by the government—will lead workers to accept short-run reductions in their real wage because behavior will change when a decrease in aggregate demand is anticipated and workers realize they are overestimating the domestic rate of inflation. This theory implies, therefore, that expectations about the inflation rate can be altered relatively quickly and at little unemployment cost by a government which decides to implement a restrictive monetary policy. Al-though higher inflation in small countries may generate more unemployment, it follows that, if this second theory is true, both inflation and unemployment can be reduced relatively quickly. If the alternative view is true, then small countries will suffer more than large ones.
Information on the short-run response of wages to unemployment has been obtained from a number of econometric studies, and the computed mean estimates from these studies for four countries is shown in the table. A cross-country comparison of these results shows that a hypothetical 25 per cent increase in the unemployment rate results in a large percentage reduction in the rate of change of wages in Canada, France, and Japan, with a lesser but still very pronounced response in the Federal Republic of Germany. Wages seem to be much less sensitive to variations in the unemployment rate in the United Kingdom, and least sensitive of all in Italy and the United States. These results suggest that for Canada, France, and Japan the second theory may be more relevant; for Italy, the United Kingdom, and the United States in contrast, the first theory may be more relevant.
These differing results may be due to several factors. First, the wage-unemployment relationship in a country is directly affected by productivity growth rates which determine the rate of increase in money wages that an economy can absorb without pushing up consumer prices. Such growth rates differ between countries. Second, institutional factors in the labor markets differ between countries—particularly the power of the trade unions, which are more radical in countries like the United Kingdom, for instance, than in the Federal Republic of Germany. Third, and most important, the basic philosophy and institutional traditions of the society determine monetary policy which in turn affects the wage-unemployment relationships. For instance, there is concern in Italy and the United Kingdom when unemployment rises; but in Japan or the Federal Republic of Germany there is concern when the inflation rate rises. When a country has a history of accommodating monetary policy, workers will tend not to moderate wage demands in the face of unemployment.
Expansionary monetary policy by itself will, in general, not pull a country out of a vicious circle because it is not designed to promote external or internal stability. One of the main consequences of expanding the money supply to meet increases in prices and wages is counterproductive: it means that the mechanism of reducing expenditure through lowering the exchange rate is prevented from taking place. The competitive advantage of an exchange rate change is quickly canceled out because in vicious circle countries wages and prices rise promptly as the exchange rate declines.
Expansionary monetary policies, combined with schemes indexing wages to inflation rates, in Italy and the United Kingdom in the mid-1970s were largely responsible for resistance to reducing real wages and for the continuation of the vicious circle in these countries. These two countries also experienced the biggest fall in productivity growth between 1972 and 1976. The chart shows how the oil crisis affected various countries’ unemployment rates as the process of adjustment began. A comparison of these unemployment rates also reflects the variance in monetary policy over the period.
Unemployment in four industrial countries, 1972-76
Sources: U.S. Department of Labor, except for Italy. Data for Italy from IMF Current Studies Division.
The evidence shows three major characteristics common to the role of the monetary authorities of the seven major industrial countries. First, a pattern of monetary restraint seems to appear in virtuous circle countries, while an accommodating pattern is much more likely for vicious circle countries. Second, in both vicious and virtuous circle countries, it appears that internal targets (for unemployment and inflation) took precedence over external targets (regarding exchange rates and current accounts); however, vicious circle countries showed more concern in reducing unemployment and virtuous circle countries more concern in reducing inflation. Neither took policy steps to alter the current account through devaluation or revaluation. In vicious circle countries, the authorities thought the real wage would fall and thus would alter the terms of trade and the volume of unemployment. When it did not, they increased the money supply to meet increases in wages and prices in order to prevent unemployment rising still further. Third, the lack of concern over external balances in both vicious and virtuous circle countries led to changes in expectations which became erratic and helped to induce destabilizing capital flows and exchange rate movements.
What policies enable a country to avoid or escape from a vicious circle? Ideally, for both vicious and virtuous circle countries, monetary, fiscal, and exchange rate policies should all be utilized in a coordinated combination to establish both external and internal stability. Monetary policy must allow the expenditure-reducing mechanism of an exchange depreciation (or expenditure-increasing mechanism of an exchange appreciation) to take place so that current account adjustment may occur. To the extent that this policy may produce an unacceptable level of unemployment in a vicious circle country, other policies should be used in order to increase real output. If fiscal policy is used to increase aggregate demand, it should not be financed by an expansion in the money supply. The major problem with using fiscal policy is that unstable adjustment may result if agents anticipate that fiscal policy will continue to be expansionary, A better policy would be to try to increase aggregate supply by using incentives for investors such as investment tax credits or employment subsidies. This policy is more attractive because in the long run it reduces the relative price of domestic goods and consequently is more likely to lead to an improvement in the current account.
At the same time, both vicious and virtuous circle countries must be aware of what monetary policy can and cannot achieve in order that monetary policy may aim at promoting stable and predictable exchange rate adjustment. A country where exchange depreciation produces a noticeable and long-run improvement in the current account and a long-run increase in domestic output will have fewer problems than a country where monetary policy and exchange depreciation lead to little improvement in the current account or domestic output. To the extent that monetary policy leads to unpredictable exchange rate adjustment, official intervention may be necessary to facilitate smooth adjustment until market participants are convinced that policy changes are permanent. An intervention policy should not be utilized to prevent the necessary exchange rate adjustment from taking place; nor should it weaken domestic monetary management through its effect on reserve movements.
It is quite clear, both theoretically and empirically, that without monetary accommodation a vicious circle cannot continue for any length of time. Either the economy will adjust to a new equilibrium or, if wages and prices do not adjust downward, it will adjust to a higher unemployment rate than that previously sustained. Why do the authorities not use the recognized combination of macropolicies to break out of the vicious circle at the point where unemployment is increasing? The main reason is political. If it appears that a long and deep recession is necessary before adjustment takes place, the authorities often feel they have no alternative but to accommodate wage and price changes through an expansionary monetary policy.
J.Bilson,“The ‘Vicious Circle’ Hypothesis,”IMF Staff Papers,Vol. 26 (March1979).
M.Bond,“Exchange Rates, Inflation and the Vicious Circle,”IMF, unpublished.
R.Dornbusch, and P.Krugman,“Flexible Exchange Rates in the Short Run,”Brookings Papers on Economic Activity,1976:3.
M.Kreinin,“The Effect of Exchange Rate Changes on the Prices and Volume of Foreign Trade,”IMF Staff Papers, Vol. 24 (July1977).
P.Robinson, T.Webb, M.Townsend,“The Influence of Exchange Rate Changes on Prices; A Study of 18 Industrial Countries,”Economica (February1979).
J.Sachs,“Wage Indexation, Flexible Exchange Rates and Macroeconomic Policy,”Quarterly journal of Economics, forthcoming.
W.Salant,“International Transmission of Inflation,” in L.Krause and W.Salant,eds., Worldwide Inflation, Brookings (Washington, D.C., 1977).
To our readers
Finance & Development is distributed without charge by the International Monetary Fund and the World Bank to qualified readers. Once your name is on our mailing list, you will continue receiving our publication only if you return the poll card that will be sent to you every three years. When you receive your poll card, you must return it promptly. Otherwise, your name will be dropped from our list.